Citation
The effects of positive behavior interventions and supports on academic achievement and behavior

Material Information

Title:
The effects of positive behavior interventions and supports on academic achievement and behavior
Creator:
Taylor, Jessica Marie ( author )
Place of Publication:
Denver, Colo.
Publisher:
University of Colorado Denver
Publication Date:
Language:
English
Physical Description:
1 electronic file (91 pages) : ;

Thesis/Dissertation Information

Degree:
Doctorate ( Doctor of education)
Degree Grantor:
University of Colorado Denver
Degree Divisions:
School of Education and Human Development, CU Denver
Degree Disciplines:
Leadership for educational equity

Subjects

Subjects / Keywords:
Academic achievement ( lcsh )
Behavior modification ( lcsh )
Academic achievement ( fast )
Behavior modification ( fast )
Genre:
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )

Notes

Review:
The purpose of this study was to determine if a positive behavior and intervention supports model curriculum (PBIS) had an effect on student academic achievement tests score outcomes and school wide number of behavioral incidents over a two-year period. The data collected were based on the first and second year outcomes of academic achievement tests scores and number of behavioral incidents by type of PBIS interventions offered. The following questions were explored: Is there an overall difference between year one and year two student academic post test scores in reading and math? Is there a difference in the total number of behavioral incidents between year one and year two? Is there a difference between pre and post student academic scores and low or high number of incidents? Is there an interaction between pre and post student academic scores and low or high number of behavioral incidents?
Review:
For this study a quasi-experimental one- group, pretest-posttest design was used. The statistical analysis utilized in this study was a dependent/paired samples t test for questions one and two and a mixed ANOVA for question three. These different analyses were conducted in SPSS. The overall findings indicated that there was no statistically significant difference with the reading or math scores between year one and year two but that there was a statistically significant difference between year one documented behavior incidents and year two documented behavior incidents.
Review:
Additionally, the year one and year two findings indicated there was no statistically significant interaction between having a high or low number of behavior incidents on reading pre and posttest scores. The findings for the math pre and posttest scores were similar, there was no statistically significant interaction between having a high or low number of behavior incidents on math pre and posttest scores for year one or year two.
Bibliography:
Includes bibliographical references.
System Details:
System requirements: Adobe Reader.
Statement of Responsibility:
by Jessica Marie Taylor.

Record Information

Source Institution:
University of Colorado Denver Collections
Holding Location:
Auraria Library
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
987242986 ( OCLC )
ocn987242986
Classification:
LD1193.E37 2016d Y49 ( lcc )

Downloads

This item has the following downloads:


Full Text
THE EFFECTS OF POSITIVE BEHAVIOR INTERVENTIONS AND SUPPORTS ON
ACADEMIC ACHIEVEMENT AND BEHAVIOR
by
JESSICA MARIE TAYLOR
M.A., Curriculum and Instruction of Reading and Writing, University of Colorado, Denver B.A., Behavioral Science and Psychology, Metropolitan State University
A dissertation submitted to the Faculty of the Graduate School of the University of Colorado, Denver in partial fulfillment of the requirements for the degree of Doctorate of Education Leadership and Educational Equity Program
2016


This dissertation for the Doctor of Education degree by
Jessica Marie Taylor has been approved for the Leadership and Educational Equity Program by
Kara Viesca, Chair Nancy Leech Molly Ramirez
December 17, 2016


Taylor, Jessica Ed.D., (Educational Equity and Leadership)
The Effects of Positive Behavior Interventions and Supports on Academic Achievement and Behavior
Dissertation directed by Assistant Professor, Kara Viesca
ABSTRACT
The purpose of this study was to detennine if a positive behavior and intervention supports model curriculum (PBIS) had an effect on student academic achievement tests score outcomes and school wide number of behavioral incidents over a two-year period. The data collected were based on the first and second year outcomes of academic achievement tests scores and number of behavioral incidents by type of PBIS interventions offered. The following questions were explored: Is there an overall difference between year one and year two student academic post test scores in reading and math? Is there a difference in the total number of behavioral incidents between year one and year two? Is there a difference between pre and post student academic scores and low or high number of incidents? Is there an interaction between pre and post student academic scores and low or high number of behavioral incidents?
For this study a quasi-experimental one- group, pretest-posttest design was used. The statistical analysis utilized in this study was a dependent/paired samples t test for questions one and two and a mixed ANOVA for question three. These different analyses were conducted in SPSS. The overall findings indicated that there was no statistically significant difference with the reading or math scores between year one and year two but that there was a statistically significant difference between year one documented behavior incidents and year two documented behavior incidents.
Additionally, the year one and year two findings indicated there was no statistically significant interaction between having a high or low number of behavior incidents on reading pre and posttest scores. The findings for the math pre and posttest scores were similar, there was no statistically significant


interaction between having a high or low number of behavior incidents on math pre and posttest scores for year one or year two.
The Form and content of this abstract are approved. I recommend its publication.
Approved: Kara Viesca
IV


TABLE OF CONTENTS
CHAPTER
I. INTRODUCTION.................................................................1
Background and Significance..................................................1
Statement of the Problem.................................................2
Review of the Literature.....................................................2
Academic Achievement and Problem Behavior...............................3
School-wide Positive Behavior Supports and Academic Achievement.........4
School-Wide Positive Behavior Supports and School Culture................5
Positive Behavior Supports and At-Risk Youth.............................6
II. THEORETICAL FRAMEWORK.......................................................9
PBIS.........................................................................9
Operant Conditioning.................................................. 10
Reinforcements and Rewards..............................................11
Critical Evaluation of Operant Conditioning Theory......................11
III. METHODLOGICAL APPROACH....................................................14
Measures....................................................................14
Variables...............................................................14
Participants and Site...................................................15
Data Collection and Procedure...........................................16
IV. FINDINGS...................................................................18
Research Question One and Two............................................. 19
Research Question One.......................................................19
Reading Posttest Scores.................................................19
Math Posttest Scores................................................ 20
v


Research Question Two..........................................................20
Documented Behavioral Incidents.........................................20
Research Question Three........................................................21
Reading Year One........................................................22
Reading Year Two........................................................23
Year One Math...........................................................23
Year Two Math...........................................................24
V. DISCUSSION.................................................................27
Summary of Findings.........................................................27
Implications............................................................28
Limitations and Suggestions for Future Research.........................29
REFERENCES.....................................................................32
APPENDIX
A. Detailed Methodology.......................................................38
B. Sample of analyses.........................................................52
C. Prospectus/Proposal........................................................61
VI


LIST OF TABLES
TABLE
1. PBIS Tiers and Core Elements......................................................9
2. Artifact and data collection......................................................15
3. Pre/post test scores and documented behaviors over a two-year period..............16
4. Differences between posttest scores and behavioral incidents for year one and two.19
5. Interactions between test scores and high-low behaviors...........................21
vii


ABSTRACT
The purpose of this study was to determine if a positive behavior and intervention supports model curriculum (PBIS) had an effect on student academic achievement tests score outcomes and school wide number of behavioral incidents over a two-year period. The data collected were based on the first and second year outcomes of academic achievement tests scores and number of behavioral incidents by type of PBIS interventions offered. The following questions were explored: Is there an overall difference between year one and year two student academic post test scores in reading and math? Is there a difference in the total number of behavioral incidents between year one and year two? Is there a difference between pre and post student academic scores and low or high number of incidents? Is there an interaction between pre and post student academic scores and low or high number of behavioral incidents?
For this study a quasi-experimental one- group, pretest-posttest design was used. The statistical analysis utilized in this study was a dependent/paired samples t test for questions one and two and a mixed ANOVA for question three. These different analyses were conducted in SPSS. The overall findings indicated that there was no statistically significant difference with the reading or math scores between year one and year two but that there was a statistically significant difference between year one documented behavior incidents and year two documented behavior incidents.
Additionally, the year one and year two findings indicated there was no statistically significant interaction between having a high or low number of behavior incidents on reading pre and posttest scores. The findings for the math pre and posttest scores were similar, there was no statistically significant interaction between having a high or low number of behavior incidents on math pre and posttest scores for year one or year two.


The Form and content of this abstract are approved. I recommend its publication.
Approved: Kara Viesca


CHAPTER 1
INTRODUCTION
The purpose of this study was to determine if a positive behavior and intervention supports model curriculum (PB1S) had an effect on student academic achievement test score outcomes and school wide number of behavioral incidents over a two-year period. The data collected were based on the first and second year outcomes of academic achievement tests scores and number of behavioral incidents by type of PBIS interventions offered. PBIS curriculum framework provides the systems and tools for establishing a continuum of evidence-based practices in any school setting with the goal of assisting in meeting the particular academic and/or behavior support needs of students (Simonsen & Sugai, 2013). In the first year, all school personnel at the study site utilized universal tier one supports. In the second year, after receiving training, all school personnel utilized individualized and targeted tier two and three supports. The following questions were explored: : Is there an overall difference between year one and year two student academic post test scores in reading and math? Is there a difference in the total number of behavioral incidents between year one and year two? Is there a difference between pre and post student academic scores and low or high number of incidents? Is there an interaction between pre and post student academic scores and low or high number of behavioral incidents?
Background and Significance
This study focused on the documented behaviors and academic test scores of day treatment students attending a Colorado approved and eligible facility school to explore questions about PBIS curriculum. Eligible and approved facility schools serve a unique purpose in the Colorado school systems and are designed to provide educational services to at-risk youth. The mission of an Eligible Facility School is to assure that all students in placement receive a
1


quality education and to improve educational outcomes for those students (Colorado Department of Education, 2014). The term Eligible Facility is defined as a day treatment center, residential childcare facility or a hospital licensed by either the Department of Human Services or the Department of Public Health and Environment. Approved Facility School refers to an educational program that is operated by a licensed agency and that has been approved to receive reimbursement for education services for students placed in the program. In the spring of 2008, legislation was passed that addressed approved facility schools in a number of ways. Commonly referred to as house bill 1204 (22-2-401 C.R.S.), it established a board with rule making authority to set graduation requirements and the approval process for facility schools (http://www.cde.state.co.us/facilityschools/hb081204act). Day treatment programs, which are viewed as a form of alternative education options, have targeted students with behavioral problems who have been unsuccessful in traditional educational settings (Yi, 2012).
Statement of the Problem
The literature indicates that there is an increase in the number of students who are in need or enrolled in alternative educational treatment programs, such as day treatment programs due to a variety of behavior problems adjustment issues (Garrison, 1987; Cox, 1999). Although, there is a rise in the need of alternative school placements and programs focusing on behavioral supports continues (Hughes & Adera, 2006), there is limited research regarding important outcome variables for students enrolled in such programs.
Review of the Literature
For the purposes of this study, the literature reviewed examined research pertaining to student problem behaviors and their relation to academic achievement, as well as the relationship between PBIS and academic achievement, and finally how behavioral outcomes and school
2


climate relate to at-risk youth. The parameters for this literature review were set for research conducted between 1997 to 2016. These dates were chosen because PBIS was established in 1997.
During the 1990s the term positive behavior support became popular in school systems. It referred to behavior interventions or strategies that could theoretically be used to reduce problem behavior and assist in promoting desirable behavior (Dunlap et al., 2000).
Academic Achievement and Problem Behavior
There have been several studies conducted linking academic performance and problem behaviors by Tobin and Sugai (1999), Morrison et al (2001), Roeser et al (2000), and Nelson et al (2004). This relationship between academic performance and problem behaviors has also been studied at the middle school and high school levels by Tobin and Sugai (1999), Roeser et al (2000), and Nelson et al (2004) Luiselli, Putnam, and Sunderland (2002) and Luiselli, Putnam, Handler, and Feinberg (2005). Tobin and Sugai (1999) found that individual student academic failure in high school was linked with three or more suspensions in ninth grade. They also found links between grade point average (GPAs) and specific types of office discipline referral (ODR) behaviors, such as fighting, harassing and threats of violence, nonviolent misbehavior. Morrison, Anthony, Storino, and Dillon (2001) reviewed the records of students who were referred to an inschool suspension program. Those students who had no previous ODRs had higher GPAs than the students who had ODRs. Roeser, Eccles, and Sameroff (2000) found the relationship between academic performance and behavior strengthened in middle school by reviewing ODR documents and suspensions.
Other research conducted by Nelson, Benner, Lane, & Smith (2004), found that students with documented problem behavior, such as ODRs experienced large academic deficits in
3


reading and math when compared to same age peers. Also that externalizing or acting out behaviors were strongly related to academic performance deficits when compared to internalizing behaviors. Another study by Lee, Sugai and Homer (1999) found improvements in escape-maintained or acting out problem behavior when students received academic support that made them effective with academic tasks. Research pertaining to the link between academic outcomes and problem behaviors suggests a need for continued research and exploration on the affect PB1S has on academic outcomes.
School-wide Positive Behavior Support and Academic Achievement
Luiselli, Putnam, and Sunderland (2002) discovered that after the implementation of school-wide behavior support (PBIS) in a suburban middle school, detentions for disruptive behavior and ODRs decreased over a four-year period. School attendance also increased over the same period. The reward for meeting predetermined academic criteria, such as maintaining a specific grade point average, receiving passing grades for all subjects on the report card, and having no more than two homework detentions, as well as, behavioral, attendance, detentions, expulsions was a lottery drawing that was conducted each quarter. The percent of students who were eligible for the lottery increased from 40% of the schools population to 55% of the schools' population over the course of four years.
In another study, Luiselli, Putnam, Handler, and Feinberg (2005) implemented a schoolwide behavior support plan at an urban school and found decreases from baseline to intervention to follow-up in documented behaviors and suspensions. Reading comprehension and mathematics percentile ranks on standardized tests improved from the first to the second test dates, with an increase of 18 and 25 percentage points. In another study, Putnam, Handler, & OLeary-Zonarich (2003) discovered that reading and math scores improved on standardized
4


testing following the implementation of behavior support intervention at an urban elementary school.
An analysis of academic performance of schools that are implementing school wide positive behavior support compared to schools not implementing such programs was conducted in Illinois (Homer, Sugai, Eber, & Lewandowski, 2004). Schools that were implementing PBIS had scored 80% on the School Evaluation Tool, a tool utilized by PBIS to measure the success of implementation in school systems and assist in developing improvement plans (Sugai, Lewis-Palmer, Todd & Homer, 2001). Schools implementing PBIS also had 80% of their students able to discuss their school wide expectations and rules (Sugai, et al., 2001).
Homer, Sugai, Todd, and Lewis-Palmer (2005) found similar findings with another school district with nineteen elementary schools. Between the 1997-98 and 2001-2002 academic years, thirteen of the schools implemented school-wide positive behavior support and six schools did not. They compared the percentage of 3rd graders who met state wide reading standards in the academic year 1997-98 with the percentage in the academic year 2001-2002. Their findings concluded that ten out of the thirteen schools that implemented PBIS practices had improved outcomes. The overall increase in percentage of students meeting standards ranged from 2% to over 15% in these schools. Only one of the six schools that not did implement school-wide positive behavior support showed improvement.
School-wide Positive Behavior Support and School Culture
School culture has been defined as the belief system that directly influences school climate (Sugai, 2013). School culture refers to the manner in which teachers and staff members work together, while school climate refers more toward the school's effects on students. ODRs have been primarily used in PBIS research to track behavioral and school climate outcomes
5


(Sugai, 2013). Anecdotal evidence would suggest that PB1S schools with reduced referrals and discipline issues have better teacher retention and higher satisfaction thus leading to a positive school climate (Sugai, 2013).
Oliver, Wehby, and Reschly (2011) found that teachers who experience difficulty controlling classroom behavior have higher stress and burnout making it more difficult for them to meet the instructional demands of the classroom. Research has suggested that schools can support classroom teachers with PB1S by focusing on prevention; using multiple data sources to develop strategies for screening, identification and treatment, and taking a coordinated, schoolwide approach to reducing problem behaviors among students (Oliver et al., 2011).
McClure (2011) reviewed studies supporting PB1S and found several studies which associated PBIS with decreases in office discipline referrals (ODRs) as well as increased consistency and positive interactions among school staff. There have been many studies conducted that have indicated that schools implementing PBIS have significant reductions in ODRs data (Nelson, 1997; Sprague, et ah, 2001). Luiselli, Putnam, and Handler (2001) indicated a 69% reduction in ODRs after the implementation of PBIS in their study. Similarly, Todd, Haugen, Anderson, and Spriggs (2002) indicated an 80% reduction in ODRs in the first year of PBIS implementation and a 76% reduction in the second year in their review of behavior documentation. More recently, Bradshaw and Leaf (2008) indicated reduced ODRs data as well as improved perceptions of school safety among teachers and staff in the Maryland school system.
Positive Behavior Supports and At-Risk Youth
There are a large number of youth educated in restrictive or alternative education (AE) settings. AE schools and programing include those housed in juvenile detention centers (Carver,
6


Lewis, & Tice, 2010). Estimates suggest that between 12% and 50% of these youth have disabilities, and most youth are placed in restrictive settings as a result of significant behavior challenges (Carver, et al., 2010). Public school districts report transferring youth to AE settings for a variety of reasons, including physical aggression (61% of districts); "disruptive verbal behavior" (57%); "possession, distribution, or use" of controlled substances (57%); chronic academic failure (57%) or truancy (53%); possession or use of firearms (42%) or other weapons (51%); "arrests or involvement with the criminal justice system" (42%); teen parenthood (31%); and/or mental health needs 27%; (Carver et al., 2010, p. 11). Therefore, AE settings need to be able to support youth with a variety of behavioral needs and challenges, as well as meet individual academic and behavioral needs.
Empirical research on the presence and effectiveness of behavior support practices/PBIS in AE settings is limited (Flower, McDaniel, & Jolivette, 2011; Lehr, 2004). Youth that are unresponsive to tier 1 practices such as universal supports that are offered daily to all students may require additional tier 2 practices such as including an individualized goal on a youth's school-wide point card, providing additional adult mentoring and support to enhance social skills instruction, and developing a menu of more individualized reinforcements. For youth whose behaviors are unresponsive to tier 2, individualized tier 3 practices may be added. Tier 3 practices should be based on a full functional behavioral planning (Eber, Sugai, Smith, & Scott, 2002). Following these implementation practices in an AE setting, suggests that a PB1S framework may result in positive outcomes for youth educated within AE settings, including increases in appropriate behavior, decreases in problem behaviors, and decreases in use of crisis-emergency responses, such as restraint (Simonsen, Young, & Britton, 2010).
Summary
7


Tobin and Sugai (1999), Morrison et al (2001), Roeser et al (2000), and Nelson et al (2004) presented a relationship between academic performance and problem behaviors that have also been studied at the middle school and high school levels. These connections were linked to office referrals, documented behaviors and academic performance. An improvement in school climate and academic outcomes has been linked to PBIS practices Luiselli, Putnam, and Sunderland (2002) discovered that after the implementation of PBIS in a suburban middle school, detentions for disruptive behavior and ODRs decreased and attendance increased. Additionally, Luiselli, Putnam, Handler, and Feinberg (2005) implemented a school-wide behavior support plan at an urban school and found reading comprehension and mathematics percentile ranks on standardized tests improved from the first to the second test dates, with an increase of 18 and 25 percentage points. There was also evidence that suggested that PBIS schools with reduced referrals and discipline issues have better teacher retention and higher satisfaction thus leading to a positive school climate (Sugai, 2013).
There are a large number of youth educated in restrictive or alternative education (AE) settings. AE schools and programing include those housed in juvenile detention centers (Carver, Lewis, & Tice, 2010). Empirical research on the presence and effectiveness of behavior support practices/PBIS in AE settings is limited (Flower, McDaniel, & Jolivette, 2011; Lehr, 2004). The research suggests that there is a need for further research to examine the academic and behavioral outcomes from PBIS use of at-risk youth in alternative settings.
8


CHAPTER II
THEORETICAL FRAMEWORK
This study focused on B.F. Skinner's (1950) operant conditioning theory as a basis of measurement due to the similarities with the PB1S programming. Operant conditioning and PB1S similarities in theory and practice will be discussed below.
PBIS
School-wide Positive Behavior Interventions and Support (PBIS) is a systems approach to establishing the social culture and behavioral supports needed for all students in a school to achieve both social and academic success (Homer at el., 2004). PBIS is an approach that defines core elements of behavioral and student struggles that can be addressed through a variety of tiered supports and strategies.
The table outlined below sums up the idea of PBIS and the areas a school team needs to collect data to analyze behavioral and academic trends. The team plans and implements group and individualized intervention based on assessment information focusing on (a) prevention of problem contexts, (b) instruction on functionally equivalent skills, and instruction on desired performance skills, (c) strategies for placing problem behavior on extinction, (d) strategies for enhancing contingence reward of desired behavior, and (e) use of negative or safety consequences if needed (Homer at el., 2004). B.F. Skinners Operant Conditioning Theory and PBIS interventions share a similar model and participation planning.
9


Table 1
PBIS Tiers and Core Elements
Prevention Tier Core Elements
Primary Behavioral expectations are defined and taught Reward system is established for appropriate behaviors Consequence continuum is established for problem behavior. Continuous data collection and planning takes place
Secondary Universal screening Progress monitoring for at risk students System for increasing structure and predictability System for increasing contingent adult feedback System for linking academic and behavioral performance System for increasing home/school communication Collection and use of data for decision-making
Tertiary Functional Behavioral Assessment Team-based comprehensive assessment Linking of academic and behavior supports Individualized intervention based on assessment information focusing on (a) prevention of problem contexts, (b) instruction on functionally equivalent skills, and instruction on desired performance skills, (c) strategies for placing problem behavior on extinction, (d) strategies for enhancing contingence reward of desired behavior, and (e) use of negative or safety consequences if needed. Collection and use of data for decision-making
(Homer at el., 2004)
Operant Conditioning
Operant conditioning has been widely applied in clinical settings for behavior modification as well as, for classroom management and instructional development (Skinner, 1950). The operant conditioning theory of B.F. Skinner is based upon the idea that learning is a function of change in overt behavior (Skinner, 1950). Changes in behavior are the result of an individuals response to events (stimuli) that occur in the environment. When a particular Stimulus-Response (S-R) pattern is reinforced or rewarded, the individual is conditioned to respond (Skinner, 1954).
10


Stimulus response in Skinners S-R pattern theory and the PBIS prescribed group and individual interventions, strategies, and supports follow the same processes. The reinforcements and rewards of the S-R pattern and PBIS are discussed further below.
Reinforcement and Rewards
Reinforcement is the key element in Skinner's S-R theory (Skinner, 1968). A reinforcer is anything that strengthens the desired response. It could be verbal praise, a good grade or a feeling of increased accomplishment or satisfaction (Skinner, 1968). The theory also covers negative reinforcers, which are any stimulus that results in the increased frequency of a response when it is withdrawn (Skinner, 1968).
In past years, school-wide discipline has focused mainly on reacting to specific student misbehavior by implementing punishment-based strategies including reprimands, loss of privileges, office referrals, suspensions, and expulsions (Sugai & Homer, 1999). Meaning that punishment utilized to promote desired behaviors, especially when it is used inconsistently is ineffective. Introducing, modeling, and reinforcing positive social behavior is an important aspect of a student's educational experience (Sugai & Homer, 1999).
The purpose of school-wide PBIS is to establish a climate in which appropriate behavior is the norm (Barrett at el., 2008). The PBIS strategy focuses on identifying problem behaviors and the environments in which they occur in and then providing programming that involves teaching and modeling desired behaviors (Barrett at el., 2008). Then when students demonstrate the desired behaviors they receive positive reinforcement through a variety of stimuli. Teaching behavioral expectations and rewarding students for following them is a much more positive approach than waiting for misbehavior to occur before responding thus creating a school environment geared towards success (Barrett at el., 2008).
11


Critical Evaluation of Operant Conditioning Theory
Operant conditioning can be used to explain a wide variety of behaviors, from the process of learning, to language acquisition. Operant conditioning also has practical applications, such as token economies, which can be applied in classrooms, prisons and psychiatric hospitals, and are also utilized among PBIS programming plans (McLeod, 2015). Despite this, operant conditioning has been disputed and critiqued over the years. Some of the best-known critics include Noam Chompsky (1959) and J.E.R. Staddon (1995) who both question the methods used to develop this theory. Additionally, later critiques asserted that the theory fails to take into account the role of inherited and cognitive factors in learning, thus making the theory an incomplete explanation of the learning process (McLeod, 2015). Lastly, the use of animal research in operant conditioning studies also raised the issue of extrapolation (McLeod, 2015). Some psychologists argue that generalizations cannot be made from studies on animals to humans because their anatomy and physiology are different from humans (McLeod, 2015).
There are visible similarities between operant conditioning Theory and PBIS programming, which despite the critiques makes this an appropriate perspective for this study. There are a variety of different positive reinforcements prescribed in both operant conditioning theory and PBIS planning used to increase the likelihood of desired behavior in the classroom. Some of the most widely used reinforcements include, consumable rewards, positive social interactions, and earned activities.
According to Landrum and Kauffman (2006), Despite a rich history and extensive empirical underpinnings, the behavioral perspective on teaching and management is not highly regarded in the education community (2006, p. 47). Critics argue that operant conditioning is an unfeeling approach more suited to animals than to humans (Landrum & Kauffman, 2006).
12


Regardless of the critiques, operant conditioning is commonly used in classrooms and is viewed by many teachers as an effective approach to improving classroom practice. It provides teachers with a set of tools for improving classroom management and student learning (Landrum & Kauffman, 2006). In addition, the underlying purpose of PBIS programming follows operant conditioning theory closely, but also applies the element of data collection. The data collected in PBIS programmed schools measures student behaviors while demanding tiered intervention programming changes if desired behaviors are not exhibited, which made the connection between operant conditioning and PBIS programming relevant to this study.
13


CHAPTER III
METHODOLOGICAL APPROACH
A quasi-experimental approach was utilized to investigate the research questions. Quantitative research approaches are viewed to be objective in nature (Gliner at el., 2009). The term objective implies that behaviors are easily classified and that data are usually collected with some sort of instrument (Gliner at el., 2009). Quasi-experimental research designs contain both independent and dependent variables (Gliner at el., 2009). Active independent variables are defined as a treatment, such as a workshop, new curriculum, or other intervention for a specific amount of time throughout the duration of the study. Dependent variables are the measure or the outcome, such as test scores, behavioral changes, or other effect related outcomes (Gliner at el., 2009). Additionally, quasi-experimental designs can be divided into four different categories based on the basic design of the study. For the purposes of this study a quasi-experimental one-group, pretest-posttest design was used. This type of design requires that an initial observation (pretest) is given, then the intervention or independent variable is applied, and finally a second observation (posttest) is applied (Gliner at el., 2009). The statistical analysis utilized in this study was a mixed ANOVA for research question one and a dependent t test for research question two and three. This analysis was conducted in SPSS.
Measures
This study consisted of two dependent variables and one active independent variable. The researcher relied on previously collected data from two databases.
Variables
In quantitative research variables are the key elements in research questions. Variables are defined as the characteristics of situations or participants in any given study having different
14


values (Gliner at el., 2009). This study measured the characteristics of two dependent variables. Dependent variables are defined as the assumed measure or effect of the independent variable (Gliner at el., 2009). Independent and dependent variables can be divided into two types; active or attribute (Gliner at el., 2009). For the purposes of this study the active independent variable was the PBIS curriculum. There were three dependent variables in this study. Two of the dependent variables are academic test scores measured at student intake and discharge creating the pretest and posttest scores in the content areas of reading comprehension and math. The independent variable for this study was the implementation of PBIS curriculum and supports over a two-year period. During year one, the 2013-2014 school year tier one supports were put into place throughout the school environment for all students. In year two, the 2014-2015 school year tier two and three supports were put into place throughout the school environment. Participants and Site
Day treatment students are commonly referred to as at risk youth. At-risk students are typically defined as students who are likely to fail or drop out of school before a high school diploma or GED is attained (Kaufman at el., 1992). Research has indicated that students with disruptive behavior disorders are at risk for numerous adjustment problems during adolescence, which include violence and juvenile delinquency (Yi, 2012). These various adjustment problems can result in academic failure, substance abuse, risky sexual behavior, and antisocial behavior (Broidy et al., 2003; Yi 2012). Thus, resulting in the need for an alternative educational option, such as a day treatment program (Yi, 2012).
The day treatment students in this sample are all male students between the ages of nine to seventeen who have been diagnosed with various behavior disorders including, attention deficit/ hyperactivity disorder (ADHD), oppositional defiant disorder (ODD) and conduct
15


disorder (CD), and a variety of additional behavior disorders, which are typically associated with behavioral, social emotional, academic impairments. As defined by the American Psychiatric Association's Diagnostic and Statistical Manual of Mental Disorders, fourth edition, text revision (DSM-IV-TR; APA, 2000), Attention-Deficit/Hyperactivity Disorder (ADHD) is one of the most common and pervasively impairing childhood psychiatric disorders. ADHD and CD are characterized by the primary features of inattention, hyperactivity, and impulsivity, and are associated with other disruptive behaviors. ODD is characterized by patterns of negativistic, disobedient, hostile, and defiant behavior toward authority figures that persists for at least 6 months and is characterized by the frequent behaviors of temper loss, arguments with adults, defiance or refusal to comply with adult instructions, and deliberate actions of annoying others (DSM-IV-TR, APA, 2000).
Data Collection Procedure
All schools receiving any state or federal funds are required to collect data for reporting purposes. This study will utilize two data sources collected from an alternative school setting. The two samples are pre and post academic test scores and documented behavior incidents. The sections below will discuss the specifics about how each sample is collected at the site. Additionally, the chart below depicts the data being collected from the sample, as well as, the duration and frequency of collection.
16


Table 2
Artifact and data collection
Artifacts/ data Time Period Anticipated Sample Size
Year 1 number of individual student documented incidents Aug. 2013- May 2014 n=53 collected from
Year 2 number of individual student documented incidents Aug. 2014- May 2015 n=54 collected from
Year 1 pretest scores Rdg Aug. 2013-April 2014 n=53
Year 1 pretest scores Math Aug. 2013- April 2014 II G
Year 1 posttest scores Rdg Sep. 2013- May 2014 n=53
Year l posttest scores Math Sep. 2013- May 2014 n-53
Year 2 pretest scores Rdg Aug. 2014- April 2015 II a
Year 2 pretest scores Math Aug. 2014- April 2015 n=54
Year 2 posttest scores Rdg Sep. 2014- May 2015 n=54
Year 2 posttest scores Math Sep. 2014- May 2015 n=54
Year 1 total number of documented behavioral incident reports Aug. 2013-May 2014 n=808 collected from
Year 2 total number of documented Aug. 2014- May 2015 behavioral incident reports n=l,126 collected from
17


CHAPTER IV
FINDINGS
A descriptive statistic analysis was conducted on the 2013-2015 data for the following variables; year one reading pre and posttest scores, year one math pre and posttest scores, year two reading pre and posttest scores, year two math pre and posttest scores, year one documented behavior incidents and year two documented behavioral incidents. The results are as follows: in year one there were 53 student reports with a minimum grade level of 4 and a maximum grade level range of 12. In year two there were 54 student reports with a minimum grade level of 4 and a maximum grade level of 11. In both years the most common grade level was grade 8.
Table 3
Pre/post Test Scores and Documented Behaviors Over a Two-Year Period
Years Mean Median Mode skewness Min. Max
Y1 rdg pre 6.0 5.5 5.7 .68 .2 12.0
Y1 rdg post 7.0 6.4 5.7 .55 2.8 12.9
Y1 math pre 5.5 5.1 4.0 1.3 1.9 12.9
Y1 math post 6.2 5.4 4.3 1.1 1.4 12.9
Y2 rdg pre 6.6 5.8 4.3 .63 1.4 12.9
Y2 rdg post 7.4 6.9 10.8 .25 2.6 12.9
Y2 math pre 5.4 4.9 3.5 1.3 1.4 12.9
Y2 math post 5.7 4.8 4.8 1.0 .8 12.9
Y1 IR behavior 15.2 9 1 2.1 0 94
Y2 IR behavior 20.8 13.0 2 1.6 0 103
The mean of the pre reading scores was 6.05, which was similar to the math pretest scores 5.56. The mean of the posttest reading scores for year one was 7.02 and the mean for the
18


posttest math scores was 6.28 which would indicate that there was growth between the pre and posttest scores in reading and math in year one. The most common pretest reading scores for year one was 5.7 the most common pretest math scores for year one was 4.0, meaning that in year one and year two when students were taking their pretests in both reading and math they were scoring between and forth and fifth grade level, despite their actual grade level, which was most commonly an eighth grade level. The most common posttest reading posttest score for year one was 5.7 and the most common posttest score for math in year one was 4.3, meaning that students were scoring below the most common grade level 8 on their reading and math posttests.
The results for year two were similar to year one. The year two pretest reading scores was (M= 6.67) and the math pretest scores were (M= 5.45) and the posttest reading scores were (M= 7.46) and the posttest math scores was (M= 5.78), indicating that there was also growth in year two in reading and math between the pre and posttest scores. Additionally, there were similar pre and posttest common scores in year two. The most common pre test scores for reading was 4.3 and for math was 3.5, meaning that in year one and two when students were taking their pretests in both reading and math they were scoring between and third and fifth grade level, despite their actual grade level, which was most commonly an eighth grade level. The most common math was 4.8, indicating that students were scoring below the most common grade level 8 on math posttests. The math posttest scores for year one and two were skewed. In year two the reading posttest scores mean had a dramatic increase to 10.8 from the pretest mean 4.3, indicating that there was a large increase in the reading scores from pre to posttest.
The mean of year ones documented behavior incidents was 15.25 and the mean for year two's documented behavior incidents was 20.85 indicating that there was an increase in the overall number of incident reports written documenting behavioral incidents from year one to
19


year two. The range for documented behavior incidents in year one was 94, with a minimum range of 0 and maximum range of 94. The range for year two documented behavioral incidents was 103, with a minimum range of 0 and a maximum range of 103. The middle amount of incident reports written for a student in year one was 9 and the middle amount of incident reports written for a student in year two was 13. The most common number of incidents in year one was 1, and the most common number of incidents in year two was 2. This indicates that individual student(s) were generating more incident reports in year two.
Research Questions One and Two
This study examined if the PBIS curriculum had a significant difference between the year one and year two interventions by measuring if there was an overall difference between year one and year two student academic post test scores and behavioral incidents. To assess these differences three t tests were conducted in (1) reading posttest scores for year one and two, (2) math posttest scores for year one and two for research question one, and (3) the documented behavioral incidents for year one and two for research question two. The findings for each question are depicted in table four and discussed below.
Table 4
Differences Between Posttest Scores and Behavior Incidents for Year one and Two
Tests and Years df V Variance Transformed sk
Reading 52 .46
Reading posttest Y1 9.0 .55
Reading posttest Y2 10.0 .25
Math 52 .37
Math posttest Y1 .04 -.04
Math posttest Y2 .05 -.77
Behavioral incident 52 <.001
Research Question One
Is there an overall difference between the year one and year two student academic post test scores? To determine if there was an overall difference between year one and year two
20


student academic posttest scores two t tests were conducted. The findings are discussed below. Reading Posttest scores
To assess if there was a significant difference between the reading posttest scores over a two-year period a t test was conducted. There were (n=53) scores from year one and (N=54) scores from year two. The assumptions were checked and have been met, the data is normal and the variables are independent.
There was no significant difference between year one reading posttest scores and the year two posttest scores, t{52) = -.746, p .46. The effect size, d was not reported because there was no statistically significant difference found in these results.
Math Posttest Scores
To assess if there was a significant difference between the math posttest scores over a two-year period a / test was conducted. There were (;i=53) scores from year one and (=54) scores from year two. The assumptions were been checked, and one was not been met. The data was not normally distributed. The math posttest scores from year one had a skewness score of 1.1 and the posttest scores from year two had a skewness score of 1.08. For this reason, a Logl 0 transfer was executed in SPSS to transform the variable. Thus, transforming the skewness score for math year one -.049 and for math year two -.77.
There was no statistically significant difference between year one math posttest scores and year two math posttest scores, t(52) = .901 p = .37. The effect size, d was not reported because there was no statistically significant difference found in these results.
Research Question Two
21


Is there a difference in the total number of behavioral incidents between year one and year two? To determine if there was an overall difference between year one and year two documented behavior incidents two / tests were conducted. The findings are discussed below. Documented Behavioral Incidents
To assess if there was a significant difference between the number of documented behavioral incidents over a two-year period a t test was conducted. There were (n=53) students generating documented incidents from year one and (=54) students from year two. The assumptions have been checked and one has not been met, year one behavioral incidents has a skewness score of 2.1 and year two has a skewness score of 1.6. For this reason a Logl 0 transfer was executed in SPSS to transform the variable.
There was a statistically significant difference between year one documented behavior incidents and year two documented behavior incidents, t(52) = -89.23,p < .001. The effect size,
(d = -0.27) is close to a medium effect size according to Cohen (1988).
Research Question Three
Is there a difference between pre and post student academic scores and low or high number of incidents? Is there an interaction between pre and post student academic scores and low or high number of behavioral incidents? To answer this question, four mixed ANOVAs were conducted to assess the following: (1) whether there was an interaction between pre and posttest scores in reading for year one and the level of behavior documented incidents (i.e., low vs. high), (2) whether there was an interaction between pre and posttest scores in math for year one and the level of behavior documented incidents (i.e., low vs. high), (3) whether there was an interaction between pre and posttest scores in reading for year two and the level of behavior documented incidents (i.e., low vs. high), and (4) whether there was an interaction between pre and posttest
22


scores in math for year two and the level of behavior documented incidents (i.e., low vs. high). The following assumptions were tested (a) independence of observations (b) normality, and (c) sphericity. All assumptions were met. The findings for each test are depicted in Table 5 and discussed below.
Table 5
Interactions between test scores and hish-low behaviors
Year df F P Partial eta2 Power
Reading year 1 1 .115 .73 .02 .063
Reading year 2 1 .336 .56 .006 .088
Math year 1 1 2.47 .12 .046 .339
Math year 2 1 .129 .72 .002 .064
Reading Year One
To assess whether there was an interaction between pre and posttest scores in reading for year one and the level of behavior documented incidents (i.e., low vs. high) a Mixed ANOVA was conducted. The results depicted in the table above indicated no statistically significant interaction between the reading year one test scores and high-low behaviors, F{ 1,1 )= 115, p =.73. The results indicated there was a significant main effect for tests scores, F(l,l)= 19.28,p <.001, partial eta2 = .27 which is a small effect size according to Cohen (1988), but there was not a statistically significant effect for the high-low behavior incidents, F( 1,1 )= 1.20, p=.28.
This indicates that even though there was a statistically significant effect between the reading pre and posttest scores for year one there was no statistically significant interaction between the pre and posttest scores and having a high or low number of behavior incidents.
The power for the pre and posttest scores finding of statistically significant main effect was .99, indicating there was an adequate amount of power to determine the statistically significant main effect. The power for the high-low behavior incidents was 19, indicating that there was not enough power to determine if there was a statistically significant effect. The power
23


for the interaction of the pre/posttest scores and behavior incidents was .063, indicating that we did not have enough power to find a statistically significant interaction if one had existed. Reading Year Two
To assess whether there was a statistically significant interaction between pre and posttest scores in reading for year two and the level of behavior documented incidents (i.e., low vs. high) a Mixed ANOVA was conducted. The results, depicted in the table above indicated no statistically significant interaction between the reading year two test scores and high-low behaviors, F(l,l) = .336,p = .56. The results indicated there was a statistically significant main effect for tests scores, F(l,l)= 16.57,/? = .017, partial eta2 = .105 which is a large effect size according to Cohen (1988), but there was not a statistically significant effect for the high-low behavior incidents, F(1,1) = .011, p = .91. This indicates that even thought there was a statistically significant effect between the reading pre and posttest scores for year two there was no statistically significant interaction between the pre and posttest scores and having a high or low number of behavior incidents.
The power for the pre and posttest scores finding of statistically significant main effect was .67, indicating there was not an adequate amount of power to determine the significant main effect.
The power for the high-low behavior incidents was .051, indicating that there was not enough power to determine if there was a statistically significant effect. The power for the interaction of the pre/posttest scores and behavior incidents was .088, indicating that there not was enough power to determine the statistically significant interaction, if one had existed.
Year One Math
24


To assess whether there was an interaction between pre and posttest scores in math for year one and the level of behavior documented incidents (i.e., low vs. high) a Mixed ANOVA was conducted. The results depicted in the table above indicated no statistically significant interaction between the math year one test scores and high-low behaviors, F(1,1) = 2.47, p 12. The results indicated there was a statistically significant main effect for tests scores, F( 1,1) = 9.09, p = .004, partial eta2 = .15 which is a small effect size according to Cohen (1988), but there was not a statistically significant effect for the high-low behavior incidents, F(1,1)= 143, /?=.70. This indicates that even though there was a statistically significant effect between the math pre and posttest scores for year one there was no statistically significant interaction between the pre and posttest scores and having a high or low number of behavior incidents.
The power for the pre and posttest scores finding of statistically significant main effect was .84, indicating there was an adequate amount of power to determine the statistically significant main effect. The power for the high-low behavior incidents was .06, indicating that there was not enough power to determine if there was a statistically significant effect. The power for the interaction of the pre/posttest scores and behavior incidents was .33, indicating that there was not enough power to determine if there was a statistically significant interaction.
Year Two Math
To assess whether there was an interaction between pre and posttest scores in math for year two and the level of behavior documented incidents (i.e., low vs. high) a Mixed ANOVA was conducted. The results depicted in the table above indicated no statistically significant interaction between the math year two test scores and high-low behaviors, F( 1,1) = 129,p- .72. The results indicated there was no statistically significant main effect for tests scores, F(l,l) = 3.29, p = .075, there was also no statistically significant effect for the high-low behavior
25


incidents, F(31,l) = .126,/? = .72. This indicates that there was no statistically significant effect between the math pre and posttest scores or high-low behavior incidents for year two.
Additionally, there was no statistically significant interaction between the pre and posttest scores and having a high or low number of behavior incidents. The power for the pre and posttest scores finding of statistically significant main effect was .42, indicating there was not an adequate amount of power to determine if there was a statistically significant main effect. The power for the high-low behavior incidents was .064, indicating that there was not enough power to determine a statistically significant effect. The power for the interaction of the pre/posttest scores and behavior incidents was .064, indicating that there was not enough power to determine if there was a statistically significant interaction.
26


CHAPTER V
DISCUSSION
This study set out to explore if a positive behavior and intervention supports model curriculum (PBIS) had an effect on student academic achievement test score outcomes and school wide number of behavioral incidents over a two-year period. The data collected were based on the first and second year outcomes of academic achievement tests scores and number of behavioral incidents by type of PBIS interventions offered. The following questions were answered: : Is there an overall difference between year one and year two student academic post test scores in reading and math? Is there a difference in the total number of behavioral incidents between year one and year two? Is there a difference between pre and post student academic scores and low or high number of incidents? Is there an interaction between pre and post student academic scores and low or high number of behavioral incidents?
Summary' of Findings
The first research questions explored was: Is there an overall difference between year one and year two student academic post test scores? The findings indicated that there was no significant difference with the reading or math scores between year one and year two.
The second research question explored was: Is there a difference in the total number of behavioral incidents between year one and year two? The findings indicated there was a significant difference between year one documented behavior incidents (A/=5) and year two documented behavior incidents (A/=4). The effect size, d= -0.27, which is close to a medium effect size.
Additionally, four t tests were conducted to determine if there was a difference between the pre and posttest scores in reading and math for both years. The assumptions were checked
27


and one was not met, the math scores for year one and two were skewed. A LoglO transformation of these variables was completed.
The results for year one were as follows; there was a significant difference between the pre and posttest scores in reading t(52) = -.461 p< ,00\,d= -0.35, which is a medium effect size according to Cohen (1988). There was also statistically significant difference between pre and posttest math scores r(52) = -2.45 p = .017, d= -0.23, which is a small effect size according to Cohen (1988).
For year two the results were as follows; there was a statistically significant difference between the pre and posttest scores in reading t(53) = -2.51 p = .015, d = -0.25, which is a small effect size according to Cohen (1988). There was no statistically significant difference between the pre and posttest scores in math /(53 )= -1.68 p = .097. The effect size, d was not reported because there was no statistically significant difference found in these results.
The third research questions explored was: Is there a difference between pre and post student academic test scores in each year, when adjusting for low or high number of behavioral incidents? The year one and year two findings indicated there was no statistically significant interaction between having a high or low number of behavior incidents on reading pre and posttest scores. Additionally, in year one and two indicated that there was also no statistically significant interaction between having a high or low number of behavior incidents on the pre and posttest math scores.
Implications
When comparing year one and year twos academic outcomes in reading and math there was lack of significance growth from year one to year two. However, when comparing the academic growth within each year from pretest to posttest there was a significant amount of
28


growth for both years indicating that the students did demonstrate academic growth during their time of enrollment at the school. The findings when comparing all academic scores indicate that the school was able to maintain consistent academic growth for students each year, but not enough growth to demonstrate an increase in scores from year one to year two. This lack of growth could be attributed to the novelty the PBIS system in the school and the implementation of additional tiered supports for struggling students. Additional training or professional development pertaining to PBIS practices and time to monitor interventions and academic growth could yield different academic results.
Comparing the documented behavior incidents from year one to year two there was a significant difference between year one and year two. Looking at the descriptive analysis there were on average more incidents documented per student in year two. This could be attributed to different school, student and staff related factors. The first possible factor being the students on average could have demonstrated more incident related behaviors in year two. Another possible factor being the education staff may have documented more incidents in the second year due to the nature of PBIS programming. One of the major focuses of PBIS is collecting data in regards to tracking behaviors and identifying behaviors requiring interventions, meaning that there is a possibility that more behaviors were documented in the second year for the purpose of tracking to identify where interventions are needed.
When examining if there was an interaction between a high or low number if incidents and academic scores there were no significant interactions found. There are a variety of circumstances or situations that could attribute to these outcomes. Previous research Tobin and Sugai (1999), Morrison et al (2001), Roeser et al (2000), and Nelson et al (2004) have found that there is a link between behavior and academic perfonnance, so perhaps with larger effect sizes or
29


a longer period of time over years to measure these outcomes a link could be made, if one exists in this situation.
When discussing these results with the possible situational circumstances that could have attributed to the findings of this study have been speculated by this researcher. Below this studys limitations and recommendations for additional research, including school actions are discussed.
Limitations and Suggestions for Future Research
This study was theoretically grounded in operant conditioning theory, which has been widely applied in clinical settings for behavior modification as well as, for classroom management and instructional development (Skinner, 1950). The operant conditioning theory of B.F. Skinner is based upon the idea that learning is a function of change in overt behavior (Skinner, 1950). Changes in behavior are the result of an individuals response to events (stimuli) that occur in the environment. When a particular Stimulus-Response (S-R) pattern is reinforced or rewarded, the individual is conditioned to respond (Skinner, 1954). This theory holds the same basis of the underlying ideology of PBIS strategy focuses on identifying problem behaviors and the environments in which they occur in and then providing programming that involves teaching and modeling desired behaviors (Barrett at el., 2008).
While the participant school reported that tier one and tier two interventions were applied to meet the individual needs of the students this study did not focus on the specific interventions being utilized. This study focused on the academic and behavioral outcomes for each student. There was no measurement of tier two and three interventions being provided to specific students. Suggestions for future research would include pairing specific interventions administered to various students in relation to their academic achievement and behavioral
30


outcomes. Additionally, tracking specific times behaviors are occurring and focusing on targeting patterns in times of day or classes where behaviors are occurring will provide additional information regarding when interventions are needed on a class or school wide scale. Collecting this data could provide additional information pertaining to the interactions between behaviors and academic achievement.
Further more, tracking academic achievement results for years prior to the PB1S implementation and then comparing the findings from the implementation years could shed some light on the school perfonnance before the interventions were applied to the school and students. This information would assist in determining the effectiveness of the PBIS curriculum over time.
There are additional assessment factors that could be taken into consideration to examining academic growth within each year and between years. The use of different academic assessments with smaller units of measurement could open up additional information pertaining to academic growth in specific areas. Completing assessments on a progress-monitoring basis could also assist in the overall development of tracking patterns in academic difficulties and growth of the students, which would allow for additional academics to be put into place. Implementing additional practices with assessments could allow for additional information and understanding of academic growth both within and between school years.
31


Running Head: REFERENCES
REFERENCES
American Psychiatric Association (2000). Diagnostic and statistical manual of mental disorders (4th ed., text rev.). Washington, DC
Bradshaw, C., & Leaf, P. (2008). Project target: Update on key Findings from project target.
Barrett, S., Bradshaw, C., & Lewis-Palmer, T. (2008). Maryland state-wide PB1S initiative. Journal of Positive Behavior Interventions, 10, 1005-114.
Carver, P. R., Lewis, L., & Tice, P. (2010). Alternative schools and programs for public school students at risk of educational failure: 2007-08 (NCES 2010-026). U.S. Department of Education, National Center for Education Statistics. Washington, DC: Government Printing Office.
Chomsky, N. (1959). "Reviews: Verbal behavior by B. F. Skinner". Language 35 (1): 26-58. retrieved from: JSTOR 411334.
Colorado Department of Education. (2008). H.B. 1204 (22-2-401 C.R.S.) http://www.cde.state.co.us/facilityschools/hb081204act
Colorado Department of Education. (2014). Facility schools, Colorado. Retrieved from: http://www.cde.state.co.us/facilityschools
Dell, C.A., Harrold, B., & Dell, T. (2008). Test review: Wide range achievement test (4lh ed.). Lutz, FL: Psychological Assessment Resources.
Eber, L., Sugai, G., Smith, C. R., & Scott, T. M. (2002). Wraparound and positive behavioral
interventions and supports in the schools. Journal of Emotional and Behavioral Disorders, 10, 171-181.
Ennis, R. P., Jolivette, K., Swoszowski, N. C., & Johnson M. L. (2012): Secondar prevention efforts at a residential facility for students with emotional and behavioral disorders: Function-based check-in, check-out, Residential Treatment for Children & Youth, 29, 79-102.
Fleming, C. B., Harachi, T. W., Cortes, R. C., Abbott, R. D. & Catalano, R. F. (2004). Level and change in reading scores and attention problems during elementary school as predictors of problem behavior in middle school. Journal of Emotional and Behavioral Disorders, 12(3), 130-144.
Garrison, R. W. (1987). Alternative schools for disruptive youth: NSSC resource paper. Malibu, CA: National Safety Center, Pepperdine University, Office of Juvenile Justice and Delinquency Prevention.
32


Running Head: REFERENCES
Gliner, J. A., Morgan, G. A., Leech, N. L. (2009). Research methods in applied settings: An integrated approach to design and analysis. (2nd ed.). New York, NY: Taylor and Francis.
Homer, R. H., Todd, A., Lewis-Palmer, T., Irvin, L., Sugai, G., & Boland, J. (2004). The school-wide evaluation tool (SET): A research instrument for assessing school-wide positive behavior support. Journal of Positive Behavior Intervention (5(1) 3-12.
Homer, R. H., Sugai, G., Todd, A. W., & Lewis-Palmer, T. (2005) School-wide positive
behavior support: An alternative approach to discipline in schools. In L. M. Bambara & L L. Kern (Eds.), Individualized supports for students with problem behaviors.
(pp. 359- 90). New York: Guilford Press.
Homer, R., Sugai, G., Eber, L., & Lewandowski, H. (2004). Illinois Positive behavior
interventions and support project: 2003-2004 Progress Report. University of Oregon: Center on Positive Behavior Interventions and Support & Illinois State Board of Education.
Hughes, A. F., & Adera, B. (2006). Education and day treatment opportunities in schools: Strategies that work. Preventing School Failure, 51, 26-30.
Irwin, L, K., Tobin, T. J., Sprague, J. R., Sugai, G. & Vincent, C. G. (2004) Validity of office discipline referral measures as indices of school-wide behavioral status and effects of school-wide behavioral interventions. Journal of Positive Behavior Interventions, 6(3) 131-147.
Kaufman, P., Bradbury, D., & Owings, J. (1992). Characteristics of at-risk students in NELS:88. Nathional eduation longitudinal study of1988. Contractor report. Berkely, CA.
Kaufman, A., & Kaufman, F. (2004). Kaufman test of educational achievement (2nd ed.) Minneapolis, MN: AGS Publishing.
Kaufman, A., & Kaufman, F. (2005). Kaufman test of educational achievement-Brief Edition. Minneapolis, MN: AGS Publishing.
Landrum, T. J., & Kauffman, J. M. (2006). Behavioral approaches to classroom management. In C. M. Evertson & C. S. Weinstein (Eds.), Handbook of classroom management: Research, practice and contemporary issues. Mahwah, NJ: Erlbaum.
Lee, Y., Sugai, G. & Homer, R. H. (1999). Using an instructional intervention to reduce problem and off- task behaviors. Journal of Positive Behavior Inten>entions, 1(4), 195-204.
Lee, J. & Ware, B. (2002). Open source development with LAMP-.Using linux, apache, mySQL, perl, and PHP. Boston, MA: Addison-Wesley Professional
33


Running Head: REFERENCES
Leech, N. L., Barrett, K., & Morgan, G. A. (2011). SPSS for intermediate statistics: Use and interpretation (4,h ed.). New York, NY: Taylor and Francis
Lehr, C. A., & Lange, C. M. (2003). Alternative schools and the stu-dents they serve:
Perceptions of state directors of special edu-cation. Policy Research Brief (University of Minnesota: Minne-apolis, Institute on Community Integration), 14(1).
Lewis-Palmer, T., Sugai, G., & Larson, S. (1999). Using data to guide decisions about program implementation and effectiveness. Effective School Practices, 17(4), 47- 53.
Luiselli, J., Putnam, R., & Handler, M. (2001). Improving discipline practices in public schools: Description of a whole school and district wide model of behavior analysis consultation. The Behavior Analyst Today, 2(1), 18-26.
Luiselli, J. K., Putnam, R. F., & Sunderland, M. (2002). Longitudinal evaluation of behavior support intervention in a public middle school. Journal of Positive Behavior Interventions, 6(3), 182-188.
Luiselli, J. K., Putnam, R. F., Handler, M. W., & Feinberg A. B. (2005). Whole-school positive behavior support: Effects on student discipline problems and academic performance. Educational Psychology, 25(2-3), 183-198. 81
Lewis-Palmer, T., Homer, R. H., Sugai, G., Eber, L., & Phillips, D. (2002). Illinois Positive
Behavior Interventions and Support Project: 2001-2002 Progress Report. University of Oregon: OSEP Center on Positive Behavior Support.
McClure, C. T. (2009). Behavior needs school wide effort. National staff development council: Teachers teaching teachers. Research Brief. May 2009, pp. 8-9. Retrieved from www.nsdc.org
McIntosh, K. (2005, March). Use ofDIBELS ORF trajectories to predict office discipline referrals. Paper presented at D1BELS Summit 2005, Ratin, N. M.
McKinney, J. D. (1989). Longitudinal research on the behavioral characteristics of children with learning disabilities. Journal of Learning Disabilities, 22(3), 141-150, 165.
McLeod, S. A. (2015). Skinner Operant Conditioning. Retrieved from www.simplypsychology.org/operant-conditioning.html
Muscott, H. (2006). Implementing PBIS with fidelity in PBIS NH schools. [PowerPoint Slides], Retrieved from www.nhcebis.seresc.net
Nelson, J.R., Benner, G. J., Lane, K. & Smith, B. W. (2004). Academic achievement of K-12 students with emotional and behavioral disorders. Exceptional Children, 71(1), 59-73.
34


Running Head: REFERENCES
Nelson, C. M., Sprague, J. R., Jolivette, K., Smith, C. R., & Tobin, T. J. (2009). Positive
behavior support in alternative education, community-based mental health and juvenile justice settings. In G. Sugai, R. Homer, G. Dunlap, and W. Sailor (Eds.), Hand-book of positive behavior support (pp. 465-496). New York, NY: Springer.
Nelson, J. (1996). Designing schools to meet the needs of students who exhibit disruptive behavior. Journal of Emotional and Behavioral Disorders, 4, 147-161.
No Child Left Behind (NCLB) Act of 2001, Pub. L. No. 107-110, § 115, Stat. 1425 (2002). National School Climate Center (NSCC). (2013). How do we define school climate? Retrieved from http://www.schoolclimate.org/climate/
Oliver, R. M., Wehby, J. H., Reschly, D. J. (2011). Teacher classroom management practices Effects on disruptive or aggressive student behavior. Campbell Systematic Reviews, 2011.4. doi 10.4073/esr.2011.4
Putnam, R. F, Handler, M., & O'Leary-Zonarich, C. (2003). Improving academic achievement using school-wide behavioral support interventions. Paper presented at the Annual Conference of the Association of Behavior Analysis. San Francisco, CA.
Putnam. R. F., Handler, M., Rey, J., & O'Leary-Zonarich, C. (2002). Classwide behavior support interventions: Using functional assessment practices to design effective interventions in general classroom settings. Paper presented at the Annual Conference of the Association of Behavior Analysis. Toronto, Canada.
Putnam, R. F., Homer, R. H., Algozzine, R. (2006). Academic achievement and the
implementation of school-wide behavior support. Positive Behavioral Supports Newsletter, 3(1), 1-2.
Reynolds, R., & Kamphaus, C. (2003). Reynolds Intellectual Assessment Scales and Reynolds Intellectual Screening Test Professional Manual. Psychological Assessment
Roid, G. (2003). Stanford-Binet Intelligence Scales, Fifth Edition.
Roeser, R. W., Eccles, J. S. & Sameroff, A. J. (2000). School as a context of early adolescents academic and social-emotional development: a summary of research findings. The Elementary School Journal, 100(5), 443-471
Simonsen, B., Britton, L., & Young, D. (2010). School-wide posi-tive behavior support in a non-public school setting: A case study. Journal of Positive Behavior Interventions, 12, 180-191. doi: 10.1177/1098300708330495
Simonsen, B., & Sugai, G. (2013). PBIS in alternative education settings: Positive support for youth and high-risk behavior. Education and Treatment of Children. 36(3), 3-14. doi: 10.1353/etc.2013.0030
35


Running Head: REFERENCES
Skinner, B.F. (1950). Are theories of learning necessary? Psychological Review, 57(4), 193-216.
Skinner, B.F. (1954). The science of learning and the art of teaching. Harvard Educational Review, 24(2), 86-97.
Skinner, B.F. (1968). The Technology of Teaching. New York: Appleton-Century-Crofts.
Sprague, J., Walker, H., Golly, A., White, K., Myers, D., & Shannon, T. (2001). Translating
research into effective practice: The effects of universal staff and student intervention on indicators of discipline and school safety. Education &Treatment of Children, 24, 495 511.
Stevens, J. (2007). Intermediate statisitcs: A modem approach (3rd ed.). New York, NY: Taylor & Francis
Staddon, J. (1995) On responsibility and punishment. The Atlantic Monthly, Feb., 88-94
Sugai, G., Lewis-Palmer, T., Todd, A. & Homer, R. (2001). School-wide evaluation tool. University of Oregon.
Sugai, G., & Homer, R. H. (1999). Discipline and behavioral support: Preferred processes and practices. Effective School Practices, 17(4), 10-22.
Sugai, G., Homer, R. H., Dunlap, G., Hieneman, M., Lewis, T. J., Nelson, C., et al. (2000). Applying positive behavior support and functional behavioral assessment in schools. Journal of Positive Behavior Interventions, 2(3), 131-143.
Sugai, G., Homer, R. H., Dunlap, G. (2003). Effective behavioral support (EBS) survey version 2.0. Educational and Community Supports: University of Oregon.
Tobin, T., & Sugai, G. (1999). Predicting violence at school, chronic discipline problems, and high school outcomes from sixth graders' school records. Journal of Emotional Disorders. 7, 40-53.
Todd, A., Haugen, L., Anderson, K., & Spriggs, M. (2002). Teaching recess: Low cost efforts producing effective results. Journal of Positive Behavior Interventions, 2, 233-245.
Wilkinson, G. S., & Robertson, G. J. (2006). Wide range achievement test- fourth edition. Lutz, FL: Psychological Assessment Resources
Wechsler, D. (2005). Wechsler Individual Achievement Test 2nd Edition (WIAT II). London: The Psychological Corp.
36


Running Head: REFERENCES
Wechsler, D. (2004). The Wechsler intelligence scale for childrenfourth edition. London: Pearson Assessment.
Wechsler, D. (2008). The Wechsler abbrevaited scale of intelligence. London: Pearson Assessment.
Woodcock, R. & Johnson, M. (1977). Woodcock-Johnson tests of achaivment. Riverside Publishing.
Yi, M. S. (2012). Evaluation of therapeutic progress in at-risk youth in a behavioral day-treatment school program. Orange, CA. Chapman University.
37


Running head: APPENDIX A
APPENDIX A Research Design
The purpose of this study was to determine if a positive behavior and intervention supports model curriculum (PBIS) had an effect on student academic achievement test score outcomes and school wide number of behavioral incidents over a two-year period. The data collected was based on the first and second year outcomes of academic achievement tests scores and number of behavioral incidents varied by the types of PBIS interventions offered. PBIS curriculum framework provides the systems and tools for establishing a continuum of evidence-based practices in any school setting with the goal of assisting in meeting the particular academic and/or behavior support needs of students (Simonsen & Sugai, 2013). In the first year tier one, all school personal in the school utilized universal supports. In the second year, after receiving training, all school personal utilized individualized and targeted tier two and three supports. The following sections will describe this studys methodological approach, subjects and sampling, the types of data that was collected, the sources and instruments utilized for data collection, and the methods in which the data was analyzed to explore the following questions: Is there an overall difference between year one and year two student academic post test scores? Is there a difference in the total number of behavioral incidents between year one and year two? Is there a difference between pre and post student academic test scores in each year, when comparing a low or high number of behavioral incidents?
Methodological Approach
A quasi-experimental approach was be utilized to investigate the research questions. Quantitative research approaches are viewed to be objective in nature (Gliner at el., 2009). The term objective implies that behaviors are easily classified and that data are usually collected with
38


Running head: APPENDIX A
some sort of instrument (Gliner at el., 2009). Quasi-experimental research designs contain both independent and dependent variables (Gliner at el., 2009). Active independent variables are defined as a treatment, such as a workshop, new curriculum, or other intervention for a specific amount of time throughout the duration of the study. Dependent variables are the measure or the outcome, such as test scores, behavioral changes, or other effect related outcomes (Gliner at el., 2009). Additionally, quasi-experimental designs can be divided into four different categories based on the basic design of the study. For the purposes of this study a quasi-experimental one-group, pretest-posttest design was used. This type of design requires that an initial observation (pretest) is given, then the intervention or independent variable is applied, and finally a second observation (posttest) is applied (Gliner at el., 2009). The statistical analysis utilized in this study was a mixed ANOVA for research question one and a dependent/paired samples t test for research question two and three. This analysis was conducted in SPSS.
Participants and Site
This study focused on the documented behaviors and academic test scores of day treatment students who attended a Colorado approved and eligible facility school. This school is both an eligible site and approved site due to the populations being served and the multiple services being provided and billed for. Below the basic purpose of an eligible and approved facility school is discussed to describe environment the data was originally collected in.
Site
The mission of an Eligible Facility Schools is to assure that all students in placement receive a quality education and to improve educational outcomes for those students (Colorado Department of Education, 2014). The tenn Eligible Facility is defined as a day treatment center, residential childcare facility or a hospital licensed by either the Department of Human Services
39


Running head: APPENDIX A
or the Department of Public Health and Environment. Approved Facility School refers to an educational program that is operated by a licensed agency and that has been approved to receive reimbursement for education services for students placed in the program. In the spring of 2008, legislation was passed that addressed approved facility schools in a number of ways. Commonly referred to as house bill 1204 (22-2-401 C.R.S.), it established a board with rule making authority to set graduation requirements and the approval process for facility schools (http://www.cde.state.co.us/facilityschools/hb081204act). Day treatment programs, which are viewed as a fonn of alternative education options, have targeted students with behavioral problems who have been unsuccessful in traditional educational settings. (Yi, 2012). Participants
Day treatment students are commonly referred to as at risk youth. An at-risk student is generally defined as a student who is likely to fail at school (Kaufman at el., 1992). In this context, school failure is typically seen as dropping out of school before high school graduation (Kaufman at el., 1992). Research has indicated that students with disruptive behavior disorders are at risk for numerous adjustment problems during adolescence, which include violence and juvenile delinquency (Yi, 2012). These various adjustment problems can result in academic failure, substance abuse, risky sexual behavior, and antisocial behavior (Broidy et al., 2003; Yi, 2012). Thus, resulting in the need for an alternative educational option, such as a day treatment program (Yi, 2012).
The day treatment students in this sample are male students between the ages of nine to seventeen who have been diagnosed with various behavior disorders including, attention deficit/ hyperactivity disorder (ADHD), oppositional defiant disorder (ODD) and conduct disorder (CD), and a variety of additional behavior disorders, which are typically associated
40


Running head: APPENDIX A
with behavioral, social emotional, academic impainnents. As defined by the American Psychiatric Associations Diagnostic and Statistical Manual of Mental Disorders, fourth edition, text revision (DSM-IV-TR; APA, 2000), Attention-Deficit/Hyperactivity Disorder (ADHD) is one of the most common and pervasively impairing childhood psychiatric disorders. ADHD and CD are characterized by the primary features of inattention, hyperactivity, and impulsivity, and is associated with other disruptive behaviors. ODD is characterized by patterns of negativistic, disobedient, hostile, and defiant behavior toward authority figures that persists for at least 6 months and is characterized by the frequent behaviors of temper loss, arguments with adults, defiance or refusal to comply with adult instructions, and deliberate actions of annoying others (DSM-IV-TR, APA, 2000).
Measures
This study consisted of two dependent variables and one active independent variable. The researcher relied on previously collected data from two databases.
Variables
In quantitative research variables are the key elements in research questions. Variables are defined as the characteristics of situations or participants in any given study having different values (Gliner at el., 2009). This study measured the characteristics of two dependent variables. Dependent variables are defined as the assumed measure or effect of the independent variable (Gliner at el., 2009). Independent and dependent variables can be divided into two types; active or attribute (Gliner at el., 2009). For the purposes of this study an active independent variables was used. An active independent variable is defined as variable containing an intervention such as a new curriculum (Gliner at el., 2009).
Dependent Variables
41


Running head: APPENDIX A
There were three dependent variables in this study. Two of the dependent variables were academic test scores measured at student intake and discharge creating the pretest and posttest scores in the content areas of reading comprehension and math. The terms intake and discharge were used in this academic setting due to the untraditional nature of a student's length of stay in the school. In a more traditional setting students are tested at the beginning of the school year and end of the school year, in a day-treatment setting the students are tested when they arrive at the school (ie., intake) and when they leave the school (ie., discharge) this time frame is typically less then a school year. The scale of measurement for this variable was an interval measurement. An interval measurement was utilized when the difference between two values is meaningful (Stevens, 2007). These test scores were measured by the Wide Range Test of Achievement (WRAT-IV) assessment. The third dependent variable was documented behavioral incident reports collected by a Lamp Stack operating system. These variables are discussed in detail below.
Wide Range Test of Achievement (WRAT-IV)
The pretest-posttest scores are derived from the Wide Range Test of Achievement (WRAT-IV) assessment, which is administrated to students upon intake and again when students discharge. The scores provided by the WRAT4 are scaled scores indicating grade level equivalency scores in the content areas of reading comprehension and fluency, spelling, and math. For the purposes of this study the reading comprehension and math scores was examined. The WRAT4 is designed to provide "a quick, simple, psychometrically sound assessment of academic skills'5 (Wilkinson & Robertson, 2006, p. 3). Joseph F. Jastak first published the test in 1946 with the purpose of augmenting the cognitive performance measures of the Wechsler-Bellevue Scales, developed by David Wechsler (Dell, Harrold, & Dell, 2008). Jastak (1946)
42


Running head: APPENDIX A
believed that academic performance should also be considered during a cognitive assessment battery, since then the WRAT has been used as a nonn-referenced measure of basic academic skills in the areas of reading, spelling, and mathematical calculations (Dell, Harrold, & Dell, 2008). The WRAT4 development team utilized a stratified quota-based sampling procedure to standardize the WRAT4 these results were based on the 2001 U.S. Census to develop a nonn referenced assessment. The sample size was 3,007,with 100 to 150 participants matched across nineteen age groups (Wilkinson & Robertson, 2006). The samples were also matched to census data by gender, race/ethnicity, educational attainment, and geographic region. Educational attainment was used to indicate socioeconomic status, and it was based on years of school completion for individuals who were eighteen and older. The four census categories were employed: college graduate, some college (no degree), high school graduate only, and less than a high school diploma. Again, for those younger than age eighteen, parent educational attainment was used (Wilkinson & Robertson, 2006). For grade-based samples, which are the forniat pre and posttest scores for this study was reported, individuals with educational disability classifications were included in the norming process. The following disabilities were represented in the norming process: specific learning disabilities, speech and language impairments, developmentally disabled, emotional disturbance, physical impairments (including hearing, orthopedic, visual, and other health impairments), and attention-deficit/hyperactivity disorder (Wilkinson & Robertson, 2006). The representation of individuals with various types of disabilities were not adequately accomplished during the standardization process because visually, hearing, and physically disabled children had a lower percentage of representation when compared to that of the National Center for Education Statistics reports for 2002 (Dell, Harrold, & Dell, 2008).
43


Running head: APPENDIX A
WRAT-IV Validity and Reliability'
The assessment developers conducted an analysis in reliability for the WRAT4, which includes internal consistency and alternate-forms (Wilkinson & Robertson, 2006). The WRAT4 overall has high levels of internal consistency, ranging from .92 to .98. The data from the assessment was also found to have moderate levels of internal consistency within its subtests, with reliability coefficients ranging from .87 to .93. For example, the reading composite score coefficients are high, ranging from .95 to .96 on both the blue and green forms (Wilkinson & Robertson, 2006).
Test validity refers to how closely scores of a given test are related to those of another established test based on the same criteria (Gay, 1992; Wilkinson & Robertson, 2006). To explore the validity of the WRAT4 test results the subtests were compared to several tests of academic achievement. These comparison tests included the WRAT-Expanded (Wilkinson & Robertson, 2006), Wechsler Individual Achievement Test (Wechsler, 2005). Woodcock Johnson III, (Woodcock & Johnson, 1977), Kaufman Test of Educational Achievement-Comprehensive Fonn (Kaufman & Kaufman, 2004), Kaufman Test of Educational Achievement-Brief Form (Kaufman, & Kauffnan, 2005), Wide Range Intelligence Test, Wechsler Intelligence Scale for Children-Fourth Edition (Wechsler, 2004), Stanford-Binet Intelligence Scale-Fifth Edition (Roid, 2003), Wechsler Abbreviated Scale of Intelligence (Wechsler, 2008), and the Reynolds Intellectual Assessment Scales (Reynolds & Kamphaus, 2003). The comparison study found low to moderate relationships between the results/ scores from all assessments pertaining to full-scale IQ and subtest scores, concluding that the WRAT4 is a valid and reliable assessment (Wilkinson & Robertson, 2006).
Documented Behavioral Incident Reports (IR)
44


Running head: APPENDIX A
The documented behavioral incident reports (IR) are forms that teachers, paraprofessionals, and various adult-care takers in the school environment submit each time a student has a behavioral incident. This process is used to track behaviors, report individual behavior trends to a student's caregivers and containment team, as well as to evaluate the students safety level. The behavioral incidents that require documentation in this day-treatment setting include, physical threatening, verbal threatening, physical aggression, assault, police contact, suicidal ideation, absent without official leave (AWOL), police contact, in possession of contraband, and inappropriate talk including discussions about drugs, colluding, sexual comments, and other harmful statements.
The documented behavioral incident reports are stored and tracked in a privately and independently developed electronic tracking system developed by the school to store personal student information. The system adheres to all privacy rules and regulations. This system allows users to track, pull and group reports for individual students and school wide tracking purposes. When building the system, the private programmer utilized the LAMP stack database technology, which is basically the combination of open source technologies that together produce an application-serving platform. Lamp is an acronym for Linux, Apache Web server, MySQL database, and Perl, Python, or PHP (Lee & Ware, 2002). These four applications make up the LAMP stack. There have been government-funded analysis completed to detennine the reliability of the LAMP stack software, which established baseline for software quality and security in open source based on sophisticated analyses of more than 17.5 million lines of source code being used (Lee & Ware, 2002). The measurement utilized for this variable was nominal. A nominal measurement is used when categories are being numbers but not ordered (Stevens, 2007).
45


Running head: APPENDIX A
Documented Behavioral Incident Reports (IR) Validity and Reliability
Teachers and para-professional employed in the school receive a variety of trainings pertaining to documenting incidents and including pertinent information, how to access the system, and the importance of communicating incidents to the team members. In addition, employees also receive a variety of trainings pertaining to being an informed supervisor of "at-risk- students, spotting problematic behaviors, and crisis intervention strategies. While, behaviors are more subjective based on the individual observing and documenting the incident most documented reports will contain elements of the school defined behavioral categories discussed above. Additionally, all reports are read and approved by a school administrator thus creating reliable and valid infonnation to examine.
Independent Variables
The independent variable for this study was the implementation of PBIS curriculum and
supports over a two-year period. During year one. the 2013-2014 school year tier one supports
were put into place throughout the school environment for all students. In year two, the 2014-
2015 school year tier two and three supports were put into place throughout the school
environment. The table below briefly depicts PBIS tier one, two, and three interventions.
PBIS tier interventions_________________________________________________________________
Tier__________________Interventions_____________________________________________________
Tier 1 School-wide positive and proactive interventions implemented by all staff
to support student behavior across all settings, which may include:
Establishing, teaching, prompting, and monitoring student behavior with respect to a few positive setting- and class-wide expectations
Frequent and explicit school- and class-wide social skills instruction
School-and/or class-wide student recognition systems (e.g., point card or check-in/check-out intervention)
Continuum of responses for inappropriate behavior that include an instructional focus
Tier 2 Targeted or intensified positive and proactive interventions Implemented
by staff to support targeted-students' behavior across all settings, which may include:
46


Running head: APPENDIX A
Additional teaching, prompting, and monitoring with respect to positive setting-wide expectations
More frequent or explicit social skills instruction
Additional mentoring and structure provided within the setting-wide point or check-in/check-out intervention (e.g., individualized goals, additional check-ins)
Increased instructional support related to chronic social-behavior errors
Tier 3 Individualized and intensive positive and proactive interventions, based
on a functional-behavioral assessment and implemented to support individual students behavior, which may include:
Antecedent strategies that include environmental changes and added prompting for replacement and desired behavior
Instructional strategies to explicitly teach replacement behavior(s) and a plan for shaping toward desired behaviors
Consequence strategies that provide functionally appropriate reinforcement for replacement behavior(s), increase reinforcement for desired behaviors, and prevent or reduce
___________________reinforcement currently maintaining problem behaviorts)
(Simonsen & Sugai, 2013)
Data Collection Procedure
All schools receiving any state or federal funds are required to collect data for reporting purposes. This study utilized two data sources collected from an alternative school setting. The two samples were pre and post academic test scores and documented behavior incidents. The sections below will discuss the specifics about how each sample was collected at the site. Additionally, the chart below depicts the data being collected from the sample, as well as, the duration and frequency of collection.
Artifact and data collection
Artifacts/ data Time Period Anticipated Sample Size
Year 1 number of individual student documented incidents Aug. 2013- May 2014 n=53 collected from
Year 2 number of individual student documented incidents Aug. 2014- May 2015 n=54 collected from
Year 1 pretest scores Rdg Aug. 2013-April 2014 n=53
Year 1 pretest scores Math Aug. 2013- April 2014 n=53
47


Running head: APPENDIX A
Year 1 posttest scores Rdg Sep. 2013- May 2014 n=53
Year 1 posttest scores Math Sep. 2013- May 2014 n=53
Year 2 pretest scores Rdg Aug. 2014- April 2015 n=54
Year 2 pretest scores Math Aug. 2014- April 2015 n=54
Year 2 posttest scores Rdg Sep. 2014- May 2015 n=54
Year 2 posttest scores Math Sep. 2014- May 2015 n=54
Year ltotal number of documented Aug. 2013- May 2014 behavioral incident reports Year 2 total number of documented Aug. 2014- May 2015 n=808 collected from n=l, 126 collected from
behavioral incident reports
Test Scores
The pre and posttest scores are collected on site when a student comes to the school and then again when they leave the school. In alternative school settings the length of stay for a student is not a typical school year, as such, to measure academic growth students are given an assessment to track their progress for their length of stay. The instrument used to obtain the test scores is the WRAT- IV, which is a norm-referenced measure of basic academic skills in the areas of reading, spelling, and mathematical calculations (Dell, Harrold, & Dell, 2008).
The WRAT-IV tests are complied on a working Excel document by the school. This document is updated throughout the year and closed on the last day of the school calendar. Each year is tracked in the same format.
Documented Behavior Incidents
The documented behavioral incident reports (IR) data sample are electronic forms that teachers, paraprofessionals, and various adult-care takers in the school environment submit to the LAMP based tracking system (Lee & Ware, 2002) each time a student has a behavioral incident.
48


Running head: APPENDIX A
The purpose of the documentation is to track behaviors, report individual behavior trends to a students caregivers and containment team, as well as to evaluate the students safety level. The behavioral incidents that are documented and will be included in this study include physical threatening, verbal threatening, physical aggression, assault, police contact, suicidal ideation, absent with official leave (AWOL), police contact, in possession of contraband, and inappropriate talk including discussions about drugs, colluding, sexual comments, and other harmful statements. The electronic tracking system has the ability to sort and categorize these documents to meet the needs of this study.
Data Analysis
This study used a repeated measures approach to analyze the data. A repeated measures design is defined as a design that measures subjects several times, either on the same dependent variable or on different measures (Stevens, 2007). The data being utilized in this study was analyzed in two different ways as to answer all of the research questions. To better explain the analysis plan the following sections below discuss the analysis plan for each research question. A mixed ANOVA analysis was utilized to analyze research question one. A mixed ANOVA compares the mean differences between groups that have been split on two independent variables (Gliner at el., 2009). The primary purpose of a mixed ANOVA is to understand if there is an interaction between the independent variables on the dependent variable (Gliner at el., 2009). Additionally, a dependent/paired samples t test was used to analyze research questions two and three. A dependent or paired samples t-test is used to compare whether two groups have different average values (Gliner at el., 2009).
Research Questions One and Two
Research questions one and two both utilized a dependent/paired samples Mest to analyze
49


Running head: APPENDIX A
the data and answer the questions.
The first research question explored was: Is there an overall difference between year one and year two student academic post test scores? For this research questions the data utilized consisted of posttest scores from year one and year two.
The second research question explored was: Is there a difference in the total number of behavioral incidents between year one and year two? This research question utilized the total number of behavioral incidents from year one and two.
Research Question Three
The third research question explored was: Is there a difference between pre and post student academic test scores when factoring in a low or high number of behavioral incidents? This question utilized the individual pre and post test scores for each student in year one and two, as well as, individual high or low number of behavioral incidents in year one and year two. The high or low number of behavioral incidents was determined by identifying the median score of total behavioral incidents and categorizing high behaviors above the median and low behaviors below the median. For the purposes of entering the high-low behaviors into SPSS the low behaviors were labeled with a 0 and the high behaviors were labeled with a 1. This data was analyzed using mixed ANOVA.
Checking Assumptions
Research questions one and two had two assumptions that need to be checked. The first was checked upon the completion of the data collection process, the dependent variable should be approximately normally distributed for each group of the independent variable (Leech at el., 2011). This assumption was violated among the math pre and posttest scores for both years, as well as among the documented behavioral incidents for both years. Prior to running the t test
50


Running head: APPENDIX A
these variables were transformed using a LoglO transfer in SPSS. The second assumption was met, the independent variable was dichotomous, meaning they were paired, or matched, in some way (Leech at el., 2011).
For research question three the normality, homogeneity and sphericity assumptions need to be checked when conducting a mixed ANOVA analysis (Leech at el., 2011). These assumptions were tested using SPSS and all were met. Normality implies that there were no significant outliers in any group from the within-subjects independent variables or between-subjects independent variables (Leech at el., 2011). Outliers are simply single data points within your data that do not follow the usual pattern (Leech at el., 2011). Homogeneity of variances (Leech at el., 2011) was tested using the Levenes test for homogeneity of variances (Leech at el., 2011). Sphericity, was the variances of the differences between the related groups of the within-subject independent variables for all groups of the between-subjects independent variables and dependence between pairs of groups is roughly equal (Leech at el., 2011). Additional Analysis conducted
An additional t test was conducted after view the descriptive statistic findings to detennine if there was a significant difference between the pre and posttest scores for both year one and year two. For the analysis the assumption of approximately normally distributed data for each group of the independent variable (Leech at el., 2011) was checked and violated among the math pre and posttest scores for both years, as well as among the documented behavioral incidents for both years. The transfonned variables from the LoglO transfer in SPSS for research questions one and two were utilized for this t test. The second assumption was met, the independent variable was dichotomous, meaning they were paired, or matched, in some way (Leech at el., 2011).
51


Running Head: APPENDIX B
APPENDIX B
T-TEST PAIRS=yearlreadingposttestmathltransform behaviorltransformWITH year2 readingposttest
math2transform behavior2transorra(PAIRED)
/CRITERIA=CI(.9500)
/MISSING=ANALYSIS.
T-Test
Notes
Output Created
28-JUL-2016 22:22:03
Comments
input Data /Users/jmtaylor/disserta tion.sav
Active Dataset DataSetl
Filter
Weight
Split File
N of Rows in Working Data File 58
Missing Value Handling Definition of Missing User defined missing values are treated as missing.
Cases Used Statistics for each analysis are based on the cases with no missing or out-of-range data for any variable in the analysis.
Syntax T-TEST PAIRS=year1 readingpos ttest mathltransform behaviorl transform WITH year2readingposttest math2transform behavior2transorm (PAIRED) /CRITERIA=CI(.9500) /MISSING=ANALYSIS.
Resources Processor Time 00:00:00.01
Elapsed Time 00:00:00.00
52


Running Head: APPENDIX B
Paired Samples Statistics
Std. Error
Mean N Std. Deviation Mean
Pair 1 yearl rdg posttest 7.023 53 2.9849 .4100
year 2 rdg posttest 7.496 53 3.1914 .4384
Pair 2 mathltransform .7522 53 .19961 .02742
math2transform .7105 53 .22600 .03104
Pair 3 behaviorl transform 1.9734 53 .00234 .00032
behavior2transorm 2.0124 53 .00216 .00030
Paired Samples Correlations
N Correlation Sig.
Pair 1 yearl rdg posttest & year 53 2 rdg posttest -.119 .394
Pair 2 mathltransform & math2transform 53 -.255 .065
Pair 3 behaviorl transform behavior2transorm & 53 -.004 .980
Paired Samples Test
Paired Differences
95% Confidence... Std. Error Mean Std. Deviation Mean Lower
Pair 1 yearl rdg posttest -2 rdg posttest year -.4736 4.6229 .6350 -1.7478
Pair 2 mathltransform -math2transform .04178 .33752 .04636 -.05125
Pair 3 behaviorl transform behavior2transorm -.03905 .00319 .00044 -.03993
Paired Samples Test
Paired ...
95% Confidence Interval of the...
Upper t df Sig. (2-tailed)
Pair 1 yearl rdg posttest year .8006 2 rdg posttest -.746 52 .459
Pair 2 mathltransform -math2transform .13481 .901 52 .372
Pair 3 behaviorl transform behavior2transorm -.03817 -89.235 52 .000
53


Running Head: APPENDIX B
MEANS TABLES=yearlreadingposttestyear2readingposttestmathltransform math2tra nsform
behaviorltransformbehavior2transorm /CELLS=MEAN STDDEV SKEW VAR.
Means
Notes
Output Created 28-JUL-2016 22:20:10
Comments
input Data /Users/jmtaylor/disserta tion.sav
Active Dataset DataSetl
Filter
Weight
Split File
N of Rows in Working Data File 58
Missing Value Handling Definition of Missing For each dependent variable in a table, user-defined missing values for the dependent and all grouping variables are treated as missing.
Cases Used Cases used for each table have no missing values in any independent variable, and not all dependent variables have missing values.
Syntax MEANS TABLES=year1 readingp osttest year2readingposttest math 1 transform math2transform behaviorl transform behavior2transorm /CELLS=MEAN STDDEV SKEW VAR.
Resources Processor Time 00:00:00.01
Elapsed Time 00:00:00.00
54


Running Head: APPENDIX B
Case Processing Summary
Cases
Included Excluded Total
N Percent N Percent N Percent
yearl rdg posttest 53 91.4% 5 8.6% 58 100.0%
year 2 rdg posttest 54 93.1% 4 6.9% 58 100.0%
math 1 transform 53 91.4% 5 8.6% 58 100.0%
math2transform 54 93.1% 4 6.9% 58 100.0%
behaviorl transform 53 91.4% 5 8.6% 58 100.0%
behavior2transorm 54 93.1% 4 6.9% 58 100.0%
Report
yearl rdg posttest year 2 rdg posttest mathltransfor m math2transfor m behaviorltran sform
Mean 7.023 7.469 .7522 .7107 1.9734
Std. Deviation 2.9849 3.1677 .19961 .22387 .00234
Skewness .554 .251 -.049 -.772 .053
Variance 8.909 10.034 .040 .050 .000
Report
behavior2tran
sorm
Mean 2.0125
Std. Deviation .00219
Skewness .252
Variance .000
GLM yearlreadingpretestyearlreadingposttestBY yearlhighlow /WSFACTOR=testscores 2 Polynomial /MEASURE=highlow /METHOD=SSTYPE(3)
/PRINT=DESCRIPTIVE ETASQ OPOWER PARAMETER HOMOGENEITY /CRITERIA=ALPHA(.05)
/WSDESIGN=testscores /DESIGN=yearlhighlow.
General Linear Model
Notes
14-AUG-2016 16:16:...
Output Created Comments
55


Running Head: APPENDIX B
Input Data /Users/jmtaylor/disserta tion.sav
Active Dataset DataSetl
Filter
Weight
Split File
N of Rows in Working Data File 58
Missing Value Handling Definition of Missing User-defined missing values are treated as missing.
Cases Used Statistics are based on all cases with valid data for all variables in the model.
Syntax GLM yearl readingpretest yearlreadingposttest BY yearl highlow
/WSFACTOR=testscores 2 Polynomial /MEASURE=highlow /METHOD=SSTYPE(3) /PRINT=DESCRIPTIVE ETASQ OPOWER PARAMETER HOMOGENEITY /CRITERIA=ALPHA(.05) /WSDESIGN=testscores
/DESIGN=year1 highlow.
Resources Processor Time 00:00:00.04
Elapsed Time 00:00:00.00
Within-Subjects
Factors
Measure: highlow
Dependent testsco Variable res
56


Running Head: APPENDIX B
1 yearl reading
pretest
2 yearl reading posttest
year one high-low
0
1
Between-Subjects Factors
Value Label N low 36
high 17
Mauchly's Test of Sphericity a
Measure: highlow
Epsilon*
Approx. Chi- Greenhouse-
Within Subjects Effect Mauchl/s W Square df Sig. Geisser
testscores 1.000 .000 0 . 1.000
Mauchly's Test of Sphericity a
Measure:
hi9hlow Epsilon*
Within Subjects Effect Huynh-Feldt Lower-bound testscores 1.000 1.000
Tests the null hypothesis that the error covariance matrix of the orthonormalized transformed dependent variables is proportional to an identity matrix.
a. Design: Intercept + yearlhighlow Within Subjects Design: testscores
b. May be used to adjust the degrees of freedom for the averaged tests of significance. Corrected tests are displayed in the Tests of Within-Subjects Effects table.
Measure: highlow
Source Tests of Within-Subjects Effects Type III Sum of Squares df Mean Square
testscores Sphericity Assumed 23.064 1 23.064
Greenhouse-Geisser 23.064 1.000 23.064
Huynh-Feldt 23.064 1.000 23.064
Lower-bound 23.064 1.000 23.064
testscores yearihighlow Sphericity Assumed .137 1 .137
Greenhouse-Geisser .137 1.000 .137
Huynh-Feldt .137 1.000 .137
57


Lower-bound .137 1.000 .137
Error(testscores) Sphericity Assumed 61.006 51 1.196
Greenhouse-Geisser 61.006 51.000 1.196
Huynh-Feldt 61.006 51.000 1.196
Lower-bound 61.006 51.000 1.196
58


Measure: highlow
Tests of Within-Subjects Effects
Partial Eta
Source F Sig. Squared
testscores Sphericity Assumed 19.281 .000 .274
Greenhouse-Geisser 19.281 .000 .274
Huynh-Feldt 19.281 .000 .274
Lower-bound 19.281 .000 .274
testscores yearlhighlow Sphericity Assumed .115 .736 .002
Greenhouse-Geisser .115 .736 .002
Huynh-Feldt .115 .736 .002
Lower-bound .115 .736 .002
Error(testscores) Sphericity Assumed
Greenhouse-Geisser
Huynh-Feldt
Lower-bound
Measure: highlow
Source Tests of Within-Subjects Effects Noncent. Parameter Observed Powera
testscores Sphericity Assumed 19.281 .991
Greenhouse-Geisser 19.281 .991
Huynh-Feldt 19.281 .991
Lower-bound 19.281 .991
testscores yearl highlow Sphericity Assumed .115 .063
Greenhouse-Geisser .115 .063
Huynh-Feldt .115 .063
Lower-bound .115 .063
Error(testscores) Sphericity Assumed
Greenhouse-Geisser
Huynh-Feldt
Lower-bound
a. Computed using alpha = .05
59


Measure: highlow
Tests of Within-Subjects Contrasts
Type III Sum of
Source testscores Squares df Mean Square F
testscores Linear 23.064 1 23.064 19.281
testscores yearlhighlow Linear .137 1 .137 .115
Error(testscores) Linear 61.006 51 1.196
Measure: highlow
Tests of Within-Subjects Contrasts
Source testscores Partial Eta Sig. Squared Noncent. Parameter Observed Power a
testscores Linear .000 .274 19.281 .991
testscores yearlhighlow Linear .736 .002 .115 .063
Error(testscores) Linear
a. Computed using alpha = .05
Levene's Test of Equality of Error Variances a
F df1 df2 Sig.
year 1 rdg pretest .625 1 51 .433
yearl rdg posttest .851 1 51 .361
Tests the null hypothesis that the error variance of the dependent variable is equal across groups.
a. Design: Intercept + yearlhighlow Within Subjects Design: testscores
Measure: highlow
Tests of Between-Subjects Effects
Transformed Variable: Average
Source Type III Sum of Squares df Mean Square F Sig. Partial Eta Squared
Intercept 4131.664 1 4131.664 303.413 .000 .856
yearlhighlow 16.379 1 16.379 1.203 .278 .023
Error 694.482 51 13.617
60


Running Head: APPENDIX C
APPEXDIX C Introduction
This study will be an examination of the effects of the implementation of School wide Positive Behavior support systems for at-risk youth in a behavioral day treatment school setting. For he purposes of this study at-risk students are identified as students with disruptive behavior disorders that include Attention Deficit Hyperactive Disorder (ADHD), Conduct Disorder (CD), and Oppositional Defiant Disorder (ODD). School-wide Positive Behavior Support (PBIS) is a systems approach to establishing the social culture and behavioral supports needed for all children in a school to achieve both social and academic success. PBIS is not a packaged curriculum, but an approach that defines core elements that can be achieved through a variety of strategies and supports.
The literature reviewed pertaining to PBIS examined its effectiveness on student achievement, the behavioral outcomes and climate of a school, and the at-risk population of students being served. The research indicated that there is a need for further research to examine the academic and behavioral outcomes of at-risk youth in alternative settings. Additionally, considerations should be taken when reviewing outcome results of PBIS when taking into account teachers' satisfaction level and opinions of PBIS.
A quasi-experimental approach will be utilized to investigate the research questions being explored in this study. Quantitative research approaches are viewed to be objective in nature (Gliner at el., 2009). For the purposes of this study a quasi-experimental one- group, pretest-posttest design will be used. This type of design requires that an initial observation (pretest) is given, then the intervention or independent variable is applied, and finally a second observation
61


Running head: APPENDIX C
(posttest) is applied (Gliner at el., 2009). There will be two statistical analyses utilized in this study. They will be a mixed ANOVA and a dependent t test. This analysis will be conducted in SPSS.
Conceptual Framework
This study aims to determine if a positive behavior and intervention supports model curriculum (PBIS) has an effect on student academic achievement test score outcomes and school wide number of behavioral incidents over a two-year period. This study will focus on B.F. Skinners (1950) operant conditioning theory as a basis of measurement due to the similarities with the PBIS programming. Operant conditioning and PBIS similarities in theory and practice will be discussed below.
PBIS
School-wide Positive Behavior Interventions and Support (PBIS) is a systems approach to establishing the social culture and behavioral supports needed for all students in a school to achieve both social and academic success (Homer at el., 2004). PBIS is an approach that defines core elements of behavioral and student struggles that can be addressed through a variety of tiered supports and strategies. The core elements of the three tiers in the prevention model are outlined below:
PBIS Tiers and Core Elements___________________________________________________________
Prevention Tier______Core Elements_____________________________________________________
Primary Behavioral expectations are defined and taught
Reward system is established for appropriate behaviors Consequence continuum is established for problem behavior.
Continuous data collection and planning takes place Secondary Universal screening
Progress monitoring for at risk students System for increasing structure and predictability System for increasing contingent adult feedback System for linking academic and behavioral performance System for increasing home/school communication
62


Running Head: APPENDIX C
Collection and use of data for decision-making Tertiary Functional Behavioral Assessment
Team-based comprehensive assessment Linking of academic and behavior supports
Individualized intervention based on assessment information focusing on (a) prevention of problem contexts, (b) instruction on functionally equivalent skills, and instruction on desired performance skills, (c) strategies for placing problem behavior on extinction, (d) strategies for enhancing contingence reward of desired behavior, and (e) use of negative or safety consequences if needed.
____________________Collection and use of data for decision-making_______________
(Homer at el., 2004)
To sum up the table outlined above the idea of PBIS is for a school team to collect data to analyze behavioral and academic trends. The team plans and implements group and individualized intervention based on assessment information focusing on (a) prevention of problem contexts, (b) instruction on functionally equivalent skills, and instruction on desired performance skills, (c) strategies for placing problem behavior on extinction, (d) strategies for enhancing contingence reward of desired behavior, and (e) use of negative or safety consequences if needed (Homer at el., 2004). B.F. Skinners Operant Conditioning Theory and PBIS interventions share a similar model and participation planning.
Operant Conditioning
Operant conditioning has been widely applied in clinical settings for behavior modification as well as, teaching for classroom management and instructional development for programmed instruction (Skinner, 1950). The operant conditioning theory of B.F. Skinner is based upon the idea that learning is a function of change in overt behavior (Skinner, 1950). Changes in behavior are the result of an individual's response to events (stimuli) that occur in the environment. When a particular Stimulus-Response (S-R) pattern is reinforced or rewarded, the individual is conditioned to respond (Skinner, 1954).
63


Running head: APPENDIX C
Stimulus response in Skinner's S-R pattern theory and the PBIS prescribed group and individual interventions, strategies, and supports follow the same processes. The reinforcements and rewards of the S-R pattern and PBIS are discussed further below.
Reinforcement and Rewards
Reinforcement is the key element in Skinner's S-R theory (Skinner, 1968). A reinforcer is anything that strengthens the desired response. It could be verbal praise, a good grade or a feeling of increased accomplishment or satisfaction (Skinner, 1968). The theory also covers negative reinforcers, which are any stimulus that results in the increased frequency of a response when it is withdrawn (Skinner, 1968). A great deal of attention was given to schedules of reinforcement and their effects on establishing and maintaining behavior.
In past years, school-wide discipline has focused mainly on reacting to specific student misbehavior by implementing punishment-based strategies including reprimands, loss of privileges, office referrals, suspensions, and expulsions (Sugai & Homer, 1999). Meaning that punishment utilized to promote desired behaviors, especially when it is used inconsistently is ineffective. Introducing, modeling, and reinforcing positive social behavior is an important step of a student's educational experience (Sugai & Homer, 1999).
The purpose of school-wide PBIS is to establish a climate in which appropriate behavior is the norm (Barrett at el., 2008). The PBIS strategy focuses on identifying problem behaviors and the environments in which they occur in and then providing programming that involves teaching and modeling desired behaviors (Barrett at el., 2008). Then when students demonstrate the desired behaviors they receive positive reinforcement through a variety of stimuli. Teaching behavioral expectations and rewarding students for following them is a much more positive
64


Running Head: APPENDIX C
approach than waiting for misbehavior to occur before responding thus creating a school environment geared towards success (Barrett at el., 2008).
Critical Evaluation of Operant Conditioning Theory
Operant conditioning can be used to explain a wide variety of behaviors, from the process of learning, to language acquisition. Operant conditioning also has practical applications, such as token economies, which can be applied in classrooms, prisons and psychiatric hospitals, and are also utilized among PBIS programming plans (McLeod, 2015).
Operant conditioning has been disputed and critiqued over the years. Some of the best known critics include Noam Chompsky (1959) and J.E.R. Staddon (1995) who both questioning the methods used to develop this theory. Additionally, later critiques discuss that the theory fails to take into account the role of inherited and cognitive factors in learning, thus making the theory an incomplete explanation of the learning process (McLeod, 2015). Lastly, the use of animal research in operant conditioning studies also raises the issue of extrapolation (McLeod, 2015). Some psychologists argue that generalizations cannot be made from studies on animals to humans because their anatomy and physiology are different from humans (McLeod, 2015).
There are visible similarities between operant conditioning Theory and PBIS programming, which despite the critiques makes this an appropriate perspective for this study. There are a variety of different positive reinforcements prescribed in both operant conditioning Theory and PBIS planning used to increase the likelihood of desired behavior in the classroom. Some of the most widely used reinforcements include, consumable rewards, positive social interactions, and earned activities.
According to Landrum and Kauffman (2006), Despite a rich history and extensive empirical underpinnings, the behavioral perspective on teaching and management is not highly
65


Running head: APPENDIX C
regarded in the education community (2006. p. 47). Critics argue that operant conditioning is an unfeeling approach more suited to animals than to humans (Landrum & Kauffman, 2006). Regardless of the critiques, operant conditioning is commonly used in classrooms and is viewed by many teachers as an effective approach to improving classroom practice. It provides teachers with a set of tools for improving classroom management and student learning (Landrum & Kauffman, 2006). In addition, the underlying purpose of PBIS programming follows operant conditioning theory closely, but also applies the element of data collection. The data collected in PBIS programmed schools measures student behaviors while demanding tiered intervention programming changes if desired behaviors are not exhibited, making the connection between operant conditioning and PBIS programming relevant to this study.
Literature Review
The purpose of this study is to determine if a positive behavior and intervention supports model curriculum (PBIS) has an effect on student academic achievement test score outcomes and school wide number of behavioral incidents over a two-year period. The data collected will be based on the first and second year outcomes of academic achievement tests scores and number of behavioral incidents varied by the types of PBIS interventions offered. PBIS curriculum framework provides the systems and tools for establishing a continuum of evidence-based practices in any school setting with the goal of assisting in meeting the particular academic and/or behavior support needs of students (Simonsen & Sugai, 2013).
For the purposes of this study, this literature review will examine research pertaining to problem behaviors and how they relate to academic achievement, PBIS related to academic achievement, behavioral outcomes and school climate, and how it relates to at-risk youth. The parameters for this research were set for research conducted in 1997 to 2015. These dates were
66


Running Head: APPENDIX C
chosen because PBIS was established in 1997.
PBIS Development and Implementation
In order to describe the development and implementation of PBIS, the historical context around the program will be briefly discussed. During the 1990s the term positive behavior support became popular in school systems. It referred to behavior interventions or strategies that could theoretically be used to reduce problem behavior and assist in promoting desirable behavior (Dunlap et al., 2000). In 1997, there were amendments made to the Individuals with Disabilities Education Act (IDEA). One important aspect of these amendments was the concept of positive behavior support (PBS) for students whose behaviors violated school rules or was outside personal or interpersonal norms of acceptable social behavior (Sugai et al, 2000, p. 131). Meaning, legally, if a student who has been identified as having a disability displays behaviors that affect his or her learning or others learning, then that childs Individualized Education Program (IEP) team, including teachers, para-professionals, and those interacting with the student must include positive behavior interventions or supports to address the behavior (Sugai et al, 2000). The implementation of IDEA was the main contributing factor in the developments on PBIS, but another factor also contributed to the development of PBIS years later in 2001. President George W. Bush signed the No Child Left Behind Act (NCLB, 2002). This act changed the federal governments involvement with public education and included a component of PBIS implementation for students (Sugai et al, 2000).
Schools have had practices in place to deal with problem behaviors for many years (Sugai et al., 2000). Legal policies and amendments have contributed and affected the development of many behavior programs. Research pertaining to the link between academic achievement and
67


Running head: APPENDIX C
problem behaviors has also sparked the continued development and implementation of PBIS (Sugai et al., 2000)
Academic Achievement and Problem Behavior
Higher rates of office discipline referrals (ODRs) are associated with problematic behavioral climates in schools (Irwin, at el., 2004). McIntosh (2005) found relationships between academic performance and problem behavior across grade levels. McIntosh (2005) investigated how early screening measures pertaining to assessments in kindergarten targeting behavior and reading outcomes predicted if a student would have two or more discipline contacts in the third and fifth grade. He found that office referrals (ODRs) in first and second grade were predictors of ODRs in third grade (McIntosh, 2005). His results also indicated that reading competence in kindergarten, measured by the Dynamic Indicators of Basic Early Literacy Skills (DIBELS) were also linked to ODRs in 3rd grade. McIntosh (2005) found that overall predictors of students receiving two or more discipline contacts in fifth grade were linked to their fourth grade ODRs and low DIBELS Oral Reading Fluency scores.
The overall findings provided by McIntoshs study were that children start their kindergarten school year with varying reading skills. However, if they do not respond to literacy instruction during kindergarten, and fall behind, a negative spiral of achievement and behavior becomes more likely (McIntosh, 2005). If a students literacy skills do not keep pace with peers, academic tasks become more difficult, and problem behaviors that lead to escape from these academic tasks become more likely.
There have been several studies conducted on the topic of linking academic performance and problem behaviors. The following studies discussed briefly discuss the similar findings over the years. This relationship between academic performance and problem behaviors has also been
68


Running Head: APPENDIX C
studied at the middle school and high school levels. Tobin and Sugai (1999) found that individual student academic failure in high school was linked with three or more suspensions in ninth grade. They also found links between grade point average (GPAs) and specific types of office discipline referral (ODR) behaviors, such as fighting, harassing and threats of violence, nonviolent misbehavior. Morrison, Anthony, Storino, and Dillon (2001) reviewed the records of students who were referred to an in-school suspension program. Those students who had no previous ODRs had higher GPAs than the students who had ODRs. Roeser, Eccles, and Sameroff (2000) found the relationship between academic performance and behavior strengthened in middle school by reviewing ODR documents and suspensions.
Other research conducted by Nelson, Benner, Lane, & Smith (2004), found that students with documented problem behavior, such as ODRs experienced large academic deficits in reading and math when compared to same age peers. Also that externalizing or acting out behaviors were strongly related to academic performance deficits when compared to internalizing behaviors. In a similar study, Harachi, Cortes, Abbott, and Catalano (2004) found that students with higher reading scores in elementary school and students' whose scores increased between third and sixth grade demonstrated significantly less problem behavior in seventh grade. Another study by Lee, Sugai & Horner (1999) found improvements in escape-maintained or acting out problem behavior when students received academic support that made them effective with academic tasks. Research pertaining to the link between academic outcomes and problem behaviors has created a need for continued research and exploration on the affect PBIS has on academic outcomes.
School-wide Positive Behavior Support and Academic Achievement
69


Running head: APPENDIX C
There have been a number of studies conducted that explore how school-wide behavior supports, also known as positive behavior interventions and supports (PBIS) decrease problem behavior, increase time spent in academic instruction, and seem to be connected with improved academic outcomes. The following sections below will discuss these studies.
Improving academic outcomes
Luiselli, Putnam, and Sunderland (2002) discovered that after the implementation of school-wide behavior support (PBIS) in a suburban middle school, detentions for disruptive behavior and ODRs decreased over a four-year period. School attendance also increased over the four year period. The reward for meeting predetermined academic criteria, such as maintaining a specific grade point average, receiving passing grades for all subjects on the report card, and having no more than two homework detentions, as well as, behavioral, attendance, detentions, expulsions was a lottery drawing that was conducted each quarter. The percent of students who were eligible for the lottery increased from 40% of the schools' population to 55% of the schools' population over the course of four years.
In another study, Luiselli, Putnam, Handler, and Feinberg (2005) implemented a schoolwide behavior support plan at an urban school and found decreases from baseline to intervention to follow-up in documented behaviors and suspensions. Reading comprehension and mathematics percentile ranks on standardized tests improved from the first to the second test dates, with an increase of 18 and 25 percentage points. In another study, Putnam. Handler, & O'Leary-Zonarich (2003) discovered that reading and math scores improved on standardized testing following the implementation of behavior support intervention at an urban elementary school.
70


Running Head: APPENDIX C
A recent analysis of academic performance of schools that are implementing school wide positive behavior support compared to schools not implementing such programs was conducted in Illinois (Homer, Sugai, Eber, & Lewandowski, 2004). Schools that were implementing PB1S had scored 80% on the School Evaluation Tool, a tool utilized by PBIS to measure the success of implementation in school systems and assist in developing improvement plans (Sugai, Lewis-Palmer, Todd & Homer, 2001). Schools implementing PBIS also had 80% of their students being able to discuss their school wide expectations and rules (Sugai, et al., 2001). In addition, the schools (n=52) in which school-wide positive behavior supports were implemented had 62% of their 3rd grade students meeting the Illinois State Achievement Test Reading Standard (Sugai, et al., 2001). In comparison, only 47% of students met the Illinois State Achievement Reading Test Standard in schools (n=69) that had not fully implemented positive behavior support (Sugai, et al., 2001).
Homer, Sugai, Todd, and Lewis-Palmer (2005) discovered similar findings with another school district with nineteen elementary schools. Between the 1997-98 and 2001-2002 academic years, thirteen of the schools implemented school-wide positive behavior support and six schools did not. They compared the percentage of 3rd graders who met state wide reading standards in the academic year 1997-98 with the percentage in the academic year 2001-2002. Their findings concluded that ten out of the thirteen schools that implemented PBIS practices had improved outcomes. The overall increase in percentage of students meeting standards ranged from 2% to over 15% in these schools. Only one of the six schools that not did implement school-wide positive behavior support showed improvement.
School Wide Positive Behavior Support and School Culture
71


Running head: APPENDIX C
School culture has been defined as the belief system that directly influences school climate (Sugai, 2013). In a positive school climate students feel comfortable, valued, accepted, wanted and secure in a positive environment where they can interact with people whom they trust. Climate reflects the positive or negative feelings toward the school environment. School culture refers to the manner in which teachers and staff members work together, while school climate refers more toward the schools effects on students. ODRs have been primarily used in PB1S research to track behavioral and school climate outcomes (Sugai, 2013). Teachers are key stakeholders in implementing PBIS. If the school staff do not support or buy in" to the program, the effectiveness of PBIS will be compromised. Research has shown that PBIS can be an effective behavioral intervention program; however, there is limited research on how teachers perceive PBIS and its impacts on teacher motivation and satisfaction (Sugai, 2013). Anecdotal evidence would suggest that PBIS schools with reduced referrals and discipline issues have better teacher retention and higher satisfaction thus leading to a positive school climate (Sugai, 2013).
Oliver, Wehby, & Reschly (2011) found that teachers who experience difficulty controlling classroom behavior have higher stress and burnout making it more difficult for them to meet the instructional demands of the classroom. Effective approaches to managing the classroom environment are necessary to establish environments that support student behavior and the learning process as well as to reduce teacher stress and burnout. Research has suggested that schools can support classroom teachers with PBIS by focusing on prevention; using multiple data sources to develop strategies for screening, identification and treatment, and taking a coordinated, school-wide approach to reducing problem behaviors among students (Oliver et al., 2011). The purpose of PBIS was to establish a climate in which appropriate behavior is the norm.
72


Running Head: APPENDIX C
In the past, school-wide discipline has focused on reacting to implementing punishment-based strategies, including reprimands, loss of privileges, office referrals, suspensions, and expulsions for misbehavior.
McClure (2011) reviewed studies supporting PBIS and found several studies which associated PBIS with decreases in office discipline referrals (ODRs) as well as increased consistency and positive interactions among school staff. There have been many studies conducted that have indicated that schools implementing PBIS have significant reductions in ODRs data (Nelson, 1996; Sprague, et al., 2001). Luiselli, Putnam, and Handler (2001) indicated a 69% reduction in ODRs after the implementation of PBIS in their study. Similarly, Todd, Haugen, Anderson, and Spriggs (2002) indicated an 80% reduction in ODRs in the first year of PBIS implementation and a 76% reduction in the second year in their review of behavior documentation. More recently, Bradshaw and Leaf (2008) indicated reduced ODRs data as well as improved perceptions of school safety among teachers and staff in the Maryland school system.
Positive Behavior Supports and At-Risk Youth
There are a large number of youth educated in restrictive or alternative education (AE) settings. AE schools and programing include those housed in juvenile detention centers (Carver, Lewis, & Tice, 2010). Estimates suggest that between 12% and 50% of these youth have disabilities, and most youth are placed in restrictive settings as a result of significant behavior challenges (Carver, et ah, 2010). Public school districts report transferring youth to AE settings for a variety of reasons, including physical aggression (61% of districts); "disruptive verbal behavior" (57%); "possession, distribution, or use" of controlled substances (57%); chronic academic failure (57%) or truancy (53%); possession or use of firearms (42%) or other weapons
73


Running head: APPENDIX C
(51%); "arrests or involvement with the criminal justice system" (42%); teen parenthood (31%); and/or mental health needs 27%; (Carver et ah, 2010, p. 11). Therefore, AE settings need to be able to support youth with a variety of behavioral needs and challenges, as well as meet individual academic and behavioral needs.
The overall empirical research on the presence and effectiveness of behavior support practices/PBIS in AE settings is limited (Flower, McDaniel, & Jolivette, 2011; Lehr, 2004). There is, however, initial research that suggests behavior management practices in these settings may be more punitive than positive (Lehr & Lange, 2003).
When youth displaying high-risk behaviors are educated together in an AE setting, it is a common misconception by teachers and school staff that all of the students will require tier three supports. Instead, PBIS experts suggest that all three tiers are necessary elements for the successful implementation of PBIS (Nelson et al., 2009). Unfortunately, the current existing research on PBIS and at-risk youth focuses on suggestions for implementation and data collection to provide tiered supports rather than outcomes.
Youth that are unresponsive to tier 1 practices, may require additional tier 2 practices to be added, such as including an individualized goal on a youths school-wide point card, providing additional adult mentoring and support to enhance social skills instruction, and developing a menu of more individualized reinforcements. For youth whose behaviors are unresponsive to tier 2, individualized tier 3 practices may be added. Tier 3 practices should be based on a full functional behavioral planning (Eber, Sugai, Smith, & Scott, 2002). Following these implementation practices in an AE setting, suggests that a PBIS framework may result in positive outcomes for youth educated within AE settings, including increases in appropriate behavior, decreases in problem behaviors, and decreases in use of crisis-emergency responses,
74


Running Head: APPENDIX C
such as restraint (Simonsen, Young, & Britton, 2010). In addition, single-case design studies have demonstrated that targeted-group interventions, such as check-in/check-out, have promise in AE settings (Ennis, Jolivette, Swoszowski, & Johnson, 2012) Thus, emerging evidence supports the implementation of intensified proactive and positive practices within a PBIS framework to support youth in AE settings.
Conclusion
In summary, this literature review has discussed research pertaining to PBIS and examined its effectiveness on student achievement, the behavioral outcomes and climate of a school, and the at-risk population of students being served. The research indicates that there is a need for further research to examine the academic and behavioral outcomes from PBIS use of at-risk youth in alternative settings. Additionally, considerations should be taken when reviewing outcome results of PBIS when taking into account teachers satisfaction level and opinions of PBIS.
Research Design
The purpose of this study is to determine if a positive behavior and intervention supports model curriculum (PBIS) has an effect on student academic achievement test score outcomes and school wide number of behavioral incidents over a two-year period. The data collected will be based on the first and second year outcomes of academic achievement tests scores and number of behavioral incidents varied by the types of PBIS interventions offered. PBIS curriculum framework provides the systems and tools for establishing a continuum of evidence-based practices in any school setting with the goal of assisting in meeting the particular academic and/or behavior support needs of students (Simonsen & Sugai, 2013). In the first year tier one, all school personal in the school utilized universal supports. In the second year, after receiving
75


Running head: APPENDIX C
training, all school personal utilized individualized and targeted tier two and three supports. The following sections will describe this studys methodological approach, subjects and sampling, the types of data that will be collected, the sources and instruments utilized for data collection, and the methods in which the data will be analyzed to answer the following questions: Is there an overall difference between year one and year two student academic post test scores? Is there a difference in the total number of behavioral incidents between year one and year two? Is there a difference between pre and post student academic test scores in each year, when comparing a low or high number of behavioral incidents?
76


Running Head: APPENDIX C
References
American Psychiatric Association (2000). Diagnostic and statistical manual of mental disorders (4th ed., text rev.). Washington, DC
Bradshaw, C., & Leaf, P. (2008). Project Target: Update on Key Findings from Project Target.
Barrett, S., Bradshaw, C., & Lewis-Palmer, T. (2008). Maryland state-wide PBIS initiative. Journal of Positive Behavior Interventions, 10, 1005-114.
Carver, P. R., Lewis, L., & Tice, P. (2010). Alternative Schools and Pro-grams for Public School Students At Risk of Educational Failure: 2007-08 (NCES 2010-026). U.S. Department of Education, Na-tional Center for Education Statistics. Washington, DC: Government Printing Office.
Chomsky, Noam (1959). "Reviews: Verbal behavior by B. F. Skinner". Language 35 (1): 26-58. JSTOR 411334.
Colorado Department of Education. (2008). H.B. 1204 (22-2-401 C.R.S.) http://www.cde.state.co.us/facilityschools/hb081204act
Colorado Department of Education. (2014). Retrieved from: http://www.cde.state.co.us/facilityschools
Dell, C.A., Harrold, B., & Dell, T. (2008). Test review: Wide range achievement test (4th ed.). Lutz, FL: Psychological Assessment Resources.
Eber, L., Sugai, G., Smith, C. R., & Scott, T. M. (2002). Wraparound and positive behavioral
interventions and supports in the schools. Journal of Emotional and Behavioral Disorders, 10, 171-181.
Ennis, R. P., Jolivette, K., Swoszowski, N. C., & Johnson M. L. (2012): Secondar prevention efforts at a residential facility for stu-dents with emotional and behavioral disorders: Function-based check-in, check-out, Residential Treatment for Children & Youth, 29, 79-102.
Fleming, C. B., Harachi, T. W., Cortes, R. C., Abbott, R. D. & Catalano, R. F. (2004). Level and change in reading scores and attention problems during elementary school as predictors of problem behavior in middle school. Journal of Emotional and Behavioral Disorders, 12(3), 130-144.
Garrison, R. W. (1987). Alternative schools for disruptive youth: NSSC resource paper. Malibu,
CA: National Safety Center, Pepperdine University, Office of Juvenile Justice and Delinquency
Prevention.
Gliner, J. A., Morgan, G. A., Leech, N. L. (2009). Research methods in applied settings: An
77


Running head: APPENDIX C
integrated approach to design and analysis. (2nd ed.). New York, NY: Taylor and Francis.
Homer, R. H Todd, A., Lewis-Palmer, T., Irvin, L., Sugai, G., & Boland, J. (2004). The school-wide evaluation tool (SET): A research instrument for assessing school-wide positive behavior support. Journal of Positive Behavior Intervention <5(1) 3-12.
Homer, R. H., Sugai, G., Todd, A. W., & Lewis-Palmer, T. (2005) School-wide positive
behavior support: An alternative approach to discipline in schools. In L. M. Bambara & L L. Kern (Eds.), Individualized supports for students with problem behaviors.
(pp. 359- 90). New York: Guilford Press.
Homer, R., Sugai, G., Eber, L., & Lewandowski, H. (2004). Illinois Positive Behavior
Interventions and Support Project: 2003-2004 Progress Report. University of Oregon: Center on Positive Behavior Interventions and Support & Illinois State Board of Education.
Hughes, A. F., & Adera, B. (2006). Education and day treatment opportunities in schools:
Strategies that work. Preventing School Failure, 51, 26-30.
Irwin, L. K Tobin, T. J., Sprague, J. R., Sugai, G. & Vincent, C. G. (2004) Validity of office discipline referral measures as indices of school-wide behavioral status and effects of school-wide behavioral interventions. Journal of Positive Behavior Interventions, 6(3) 131-147.
Kaufman, P., Bradbury, D., & Owings, J. (1992). Characteristics of at-risk students in NELS:88. Nathional eduation longitudinal study of 1988. Contractor report. Berkely, CA.
Kaufman, A., & Kaufman, F. (2004). Kaufman Test of Educational Achievement (2nd ed.) Minneapolis, MN: AGS Publishing.
Kaufman, A., & Kaufman, F. (2005). Kaufman Test of Educational Achievement-Brief Edition. Minneapolis, MN: AGS Publishing.
Landrum, T. J., & Kauffman, J. M. (2006). Behavioral approaches to classroom management. In C. M. Evertson & C. S. Weinstein (Eds.), Handbook of classroom management:
Research, practice and contemporary issues. Mahwah, NJ: Erlbaum.
Lee, Y., Sugai, G. & Homer, R. H. (1999). Using an instructional intervention to reduce problem and off- task behaviors. Journal of Positive Behavior Interventions, 1(4), 195-204.
Lee, J. & Ware, B. (2002). Open source development with LAMP-.Using linux, apache, mySQL, perl, and PHP. Boston, MA: Addison-Wesley Professional
Leech, N. L., Barnett, K., & Morgan, G. A. (2011). SPSS for intermediate statistics: Use and interpretation (4th ed.). New York, NY: Taylor and Francis
78


Running Head: APPENDIX C
Lehr, C. A., & Lange, C. M. (2003). Alternative schools and the stu-dents they serve:
Perceptions of state directors of special edu-cation. Policy Research Brief (University of Minnesota: Minne-apolis, Institute on Community Integration), 14(1).
Lewis-Palmer, T., Sugai, G., & Larson, S. (1999). Using data to guide decisions about program implementation and effectiveness. Effective School Practices, 17(4), 47- 53.
Luiselli, J., Putnam, R., & Handler, M. (2001). Improving discipline practices in public schools: Description of a whole school and district wide model of behavior analysis consultation. The Behavior Analyst Today, 2(1), 18-26.
Luiselli, J. K., Putnam, R. F., & Sunderland, M. (2002). Longitudinal evaluation ofbehavior support intervention in a public middle school. Journal of Positive Behavior Interventions, 6(3), 182-188.
Luiselli, J. K., Putnam, R. F., Handler, M. W., & Feinberg A. B. (2005). Whole-school positive behavior support: Effects on student discipline problems and academic performance. Educational Psychology, 25(2-3), 183-198. 81
Lewis-Palmer, T., Homer, R. H., Sugai, G., Eber, L., & Phillips, D. (2002). Illinois Positive
Behavior Interventions and Support Project: 2001-2002 Progress Report. University of Oregon: OSEP Center on Positive Behavior Support.
McClure, C. T. (2009). Behavior needs school wide effort. National staff development council: Teachers teaching teachers. Research Brief. May 2009, pp. 8-9. Retrieved from www.nsdc.org
McIntosh, K. (2005, March). Use ofDIBELS ORF trajectories to predict office discipline referrals. Paper presented at DIBELS Summit 2005, Ratin, N. M.
McKinney, J. D. (1989). Longitudinal research on the behavioral characteristics of children with learning disabilities. Journal of Learning Disabilities, 22(3), 141-150, 165.
McLeod, S. A. (2015). Skinner Operant Conditioning. Retrieved from www.simplypsychology.org/operant-conditioning.html
Muscott, H. (2006). Implementing PBIS with fidelity in PBIS NH schools. [PowerPoint Slides]. Retrieved from www.nhcebis.seresc.net
Nelson, J.R., Benner, G. J., Lane, K. & Smith, B. W. (2004). Academic achievement of K-12 students with emotional and behavioral disorders. Exceptional Children, 71(1), 59-73.
Nelson, C. M., Sprague, J. R., Jolivette, K., Smith, C. R., & Tobin, T. J. (2009). Positive
behavior support in alternative education, community-based mental health and juvenile justice settings. In G. Sugai, R. Homer, G. Dunlap, and W. Sailor (Eds.), Hand-book of positive behavior support (pp. 465-496). New York, NY: Springer.
79


Running head: APPENDIX C
Nelson, J. (1996). Designing schools to meet the needs of students who exhibit disruptive behavior. Journal of Emotional and Behavioral Disorders, 4, 147-161.
No Child Left Behind (NCLB) Act of 2001, Pub. L. No. 107-110, § 115, Stat. 1425 (2002). National School Climate Center (NSCC). (2013). How do we define school climate? Retrieved from http://www.schoolclimate.org/climate/
Oliver, R. M., Wehby, J. H., Reschly, D. J. (2011). Teacher classroom management practices Effects on disruptive or aggressive student behavior. Campbell Systematic Reviews, 2011.4. doi 10.4073/esr,2011.4
Putnam. R. F. Handler, M.. & O'Leary-Zonarich, C. (2003). Improving academic achievement using school-wide behavioral support interventions. Paper presented at the Annual Conference of the Association of Behavior Analysis. San Francisco, CA.
Putnam, R. F.. Handler. M., Rey, J., & OLeary-Zonarich, C. (2002). Classwide behavior support inter\>entions: Using functional assessment practices to design effective interventions in general classroom settings. Paper presented at the Annual Conference of the Association of Behavior Analysis. Toronto, Canada.
Putnam, R. F., Homer, R. H., Algozzine, R. (2006). Academic achievement and the
implementation of school-wide behavior support. Positive Behavioral Supports Newsletter, 3(1), 1-2.
Reynolds, R., & Kamphaus, C. (2003). Reynolds Intellectual Assessment Scales and Reynolds Intellectual Screening Test Professional Manual. Psychological Assessment
Roid, G. (2003). Stanford-Binet Intelligence Scales, Fifth Edition.
Roeser, R. W., Eccles, J. S. & Sameroff, A. J. (2000). School as a context of early adolescents academic and social-emotional development: a summary of research findings. The Elementary School Journal, 100(5), 443-471
Simonsen, B., Britton, L., & Young, D. (2010). School-wide posi-tive behavior support in a non-public school setting: A case study. Journal of Positive Behavior Interventions, 12,
180-191. doi: 10.1177/1098300708330495
Simonsen, B., & Sugai, G. (2013). PBIS in alternative education settings: Positive support for youth and high-risk behavior. Education and Treatment of Children, 36(3), 3-14. doi: 10.1353/etc.2013.0030
Skinner, B.F. (1950). Are theories of learning necessary? Psychological Review, 57(4),
193-216.
Skinner, B.F. (1954). The science of learning and the art of teaching. Harvard Educational Review, 24(2), 86-97.
80


Running Head: APPENDIX C
Skinner, B.F. (1968). The Technology of Teaching. New York: Appleton-Century-Crofts.
Sprague, J., Walker, H., Golly, A., White, K., Myers, D., & Shannon, T. (2001). Translating
research into effective practice: The effects of universal staff and student intervention on indicators of discipline and school safety. Education &Treatment of Children, 24, 495 511.
Stevens, J. (2007). Intermediate statisitcs: A modern approach (3rd ed.). New York, NY: Taylor & Francis
Staddon, J. (1995) On responsibility and punishment. The Atlantic Monthly, Feb., 88-94
Sugai, G., Lewis-Palmer, T., Todd, A. & Flomer, R. (2001). School-wide evaluation tool. University of Oregon.
Sugai, G., & Homer, R. H. (1999). Discipline and behavioral support: Preferred processes and practices. Effective School Practices, 17(4), 10-22.
Sugai, G., Homer, R. H., Dunlap, G., Hieneman, M., Lewis, T. J., Nelson, C., et al. (2000). Applying positive behavior support and functional behavioral assessment in schools. Journal of Positive Behavior Interventions, 2(3), 131-143.
Sugai, G., Homer, R. H., Dunlap, G. (2003). Effective behavioral support (EBS) survey version 2.0. Educational and Community Supports: University of Oregon.
Tobin, T., & Sugai, G. (1999). Predicting violence at school, chronic discipline problems, and high school outcomes from sixth graders' school records. Journal of Emotional Disorders. 7, 40-53.
Todd, A., Haugen, L., Anderson, K & Spriggs, M. (2002). Teaching recess: Low cost efforts producing effective results. Journal of Positive Behavior Interventions, 2, 233-245.
Wilkinson, G. S., & Robertson, G. J. (2006). Wide range achievement test- fourth edition. Lutz, FL: Psychological Assessment Resources
Wechsler, D. (2005). Wechsler Individual Achievement Test 2nd Edition (WIAT II). London: The Psychological Coip.
Wechsler, D. (2004). The Wechsler intelligence scale for childrenfourth edition. London: Pearson Assessment.
Wechsler, D. (2008). The Wechsler abbrevaited scale of intelligence. London: Pearson Assessment.
81


Running head: APPENDIX C
Woodcock, R. & Johnson, M. (1977). Woodcock-Johnson tests of achaivment. Riverside Publishing.
Yi, M. S. (2012). Evaluation of therapeutic progress in at-risk youth in a behavioral day-treatment school program. Orange, CA. Chapman University.
82


Full Text

PAGE 1

THE EFFECTS OF POSITIVE BEHAVIOR INTERVENTIONS AND SUPPORTS ON ACADEMIC ACHIEVEMENT AND BEHAVIOR by JESSICA MARIE TAYLOR M.A. Curriculum and Instruction of Reading and Writing, University of Colorado, Denver B.A., Behavioral Science and Psychology Metropolitan State University A dissertation submitted to the Faculty of the Graduate School of the University of Colorado Denver in partial fulfillment of the requirements for the degree of Doctorate of Education Leadership and Educational Equity Program 2016

PAGE 2

This dissertation for the Doctor of Education degree b y Je ssica Marie Taylor ha s been approved for the Leader s hip an d Educational Equity Program by Kara Viesca, C hair Na ncy Leech Molly Ramirez De ce mb er 17, 2016 ii

PAGE 3

Taylor, Jessica Ed.D., (Educational Equity and Leadership) The Effects of Positive Behavior Intervention s and Supports on Academic Achievement and Behavior Dissertation directed by Assistant Professor, Kara Viesca ABSTRACT The purpose of this study was to detennine if a positive behavior and intervention s upport s model curriculum (PBIS) had an effect on s tud en t academic achievement tests score outcomes and sc hool wide number of behavioral incidents over a two -yea r period. The data collected were based on the first and second year outcomes of academic achievement tests sco res and number of behavioral incidents by type ofPBIS interventions offered. The following questions were explored: I s there an overall difference between year one and year two student academic post test scores in reading and math? I s there a di f ference in the total number of behavioral incidents between year one and year two ? Is there a difference between pre and post student academic sco res and low or high number of incident s? Is there an interaction between pre and po s t st udent acade mic sco res and low or high number of behavioral incidents? For this st udy a quasi-experimental onegroup, pretest-postlest design was u se d The statistical analysis utilized in this study was a dependen t/pa ired sam ple s I test for question s one and two and a mix e d ANOY A for question three. These different analyses were conducted in SPSS. The overall findings indicated that there was no statistically sig nificant difference with the reading or math scores between year one and year two but that there was a sta tistically significant difference between year one documented behavior incidents and year tw o documented behavior incidents. Additionally, the year one and year two findings indicated there was no s tati stica lly s ignificant interaction between having a high or low number of behavior incident s on reading pre and postlest scores. The findings for the math pre and postlest scores were simi lar there was no statistically sig nificant iii

PAGE 4

interaction between having a high or low number of behavior incidents on math pre and posttest scores for year one or year two. The Form and content of this abstract are approved. I recommend its publication. Approved: Kara Viesca iv

PAGE 5

T A B LE OF CONTENTS C H A PTER I. INT R O DUCTION .................. ....... .... ........ ...... .. .......... .... ... ..... .... ............. .... ... . ... ... ..... I Background and Si gnificance ... ... ... ... ... ............... ........ .. ...... ............... .... ....... .............. .... I Stat e m ent of the Pro bl e m ............ ....... ....... ...... .............. ........ ... ........... . .. ... ........... 2 Rev iew of t h e Lit era tu r e .......................... ................................................. .......... .... ............. 2 Acad emic Achieve m e n t and Proble m Beh avior. .......... ... ....... ... ... ... .............................. 3 Sc h ool-wide P ositive Be h avior Supports a n d Acad emic Ach ieve m e nt... .... ................. .4 Sc h ool-Wide Pos i t i ve Behavi o r Supports a n d School Cu ltu re .... ... ...... ...... ....... ............ 5 Positive Behavi o r Supports and At-Risk youth ......... ...................... ........ ... ...... ...... 6 II. THEORETI CAL FRAMEWORK ... ..... ... .... .. ........ ......... ...... ............... .... ... ... ... .... ........ 9 PBIS .............. ....... .... ............. ........ . ...... .......... .. ....... .... ... ............. ...... .... .... ........... ...... 9 Operan t Condit io n ing .................................... ... ............................ ....... ... ....................... 1 0 R einforce ment s a n d R ewards .... .... ...... .... .......... .... ..... .... ... ... ......... ....... ...... ........ ........ I I Critical Evaluatio n of Operant Condit i o n i n g T h eory ............................................ .. .. ...... I I III. M ET HOD LOG I C A L APPROACH .... .... ... ................. ... ............ ......... .............. ... ..... .... ... 14 Meas u res .... ...................... .... ............... ...... .... .... ...... .... ............ ................. .... ... ................. 14 Variab l es ... ... .... ... .... ...... ... .............. . ................. ...... .... ..... .......... ..... ... .. ... ....... ...... 1 4 Parti cipa nts and S it e ........................ ... ... .... ......... ..... ........ .... ........... .................... ......... 1 5 D a t a Collectio n a n d Procedure ................... .. ............. ... ......................................... ........ 1 6 IV. FINDINGS ....... ................... ............................. ....... .............. ............... .... .... .............. ... 1 8 Research Q uestio n One and Two ...... ...... ................. .................. ............... ...... ................ ...... 1 9 Research Questio n One ....... .... ....... .. .... .. ... ....... .................. ..... .. .............. .................. 1 9 Readin g Posttes t Scores ......... ... ....... ....... ......... ... ................ ........ ... .... .... .... ... .... .... ..... 1 9 Math P osttest Sco r es ........... .... ........ ...... ....... ........ ... ...... ....... ..... .... ..... ... ... .... ...... ... 20 v

PAGE 6

Research Question Two ...................................... ............. .............................................. .. ...... .... 20 Documented Beha viora l Incidents .............. .... ... ............... ... ....................................... ..... 20 Research Question Three ...................... ....... .................... ............................ ..................... ........... 2 1 Reading Year One ......... ........ .......... ............... ............... .... ............ ........ ................. .......... 22 Reading Year Two ............ ......................................... ................... ...................... ........ .. .... 23 Year One Math ... ......... ............... .... .............. .... ................................................................ 23 Year Two Math .......... ... ........... .... ..... ............................................... .... ..... ...... ................. 24 V. DiSCUSSiON ................................................. ....................................... .................... .............. 27 S umm ary of Findings .......... ........... ........................... ............................. ..... ....... ........... ......... 27 Impli ca tion s ............................................................................. .............. .............. ............. 28 Limitations and Sugge st ion s for Future Research ........................................................... 29 REFERENCES ...... .......... ......... ............ .......... .. .......... .......... ........... ...... ..... ................... ... .. .......... 32 APPENDIX A. Detailed Methodology ....... ................... .. ..... .. .. ..... ...... ......... ............... ...... ............................. 38 B. Sa mpl e of analyses ............................. .......... ....................... ............. .............. ..................... 52 C. Prospectus / Propo sal.. .................. .... ...... ..... ......... .... .............................. .. ................................ 6 1 vi

PAGE 7

LIST OF TAB LES TABLE I. PBIS Tiers and Core E lem ents .... ..... .......... .. .. ...... ..... ...... ......... ............... ...................... .. .. 9 2. Artifact and data collection ...... .............. .. ..... ........ ................... .................... .............. ..... ...... 1 5 3. Pre / post t est scores and documented b e h avio rs over a two-year period ...... ......................... 1 6 4. Diff erences between posttest sco re s a nd beha v i ora l incide nt s for year one and two .. .......... 1 9 5 Int eraction s between te s t scores and high-low beh av i o r s ............................ ............. ............ 2 1 vii

PAGE 8

ABSTRACT The purpose of this study was to detennine if a positive behavior and intervention supports model curriculum (PBIS) had an effect on student academic achievement test s sco re outcomes and school wide number of behavioral incidents over a two-year period The data collected were based on the first and second year outcomes of academic achievement tests scores and number of behavioral incidents by type ofPBIS interventions offered. The following questions were explored: Is there an overall difference between year one and year two student academic post test scores in reading and math? Is there a difference in the total number of behavioral incidents between year one and year two ? Is there a difference between pre and post student academic scores and low or high number of incidents? Is there an interaction between pre and post student academic scores and low or high number of behavioral incidents? For this study a quasi-experimental onegroup, pretest-postlest design was used. The statistical analysis utili ze d in this study was a dependent / paired samples t test for questions one and two and a mixed ANOYA for question three These different analyses were conducted in SPSS. The overall findings indicated that there was no statistically significant difference with the reading or math scores between year one and year two but that there was a statistically significant difference between year one documented behavior incidents and year two documented behavior incidents. Additionally, the year one and year two findings indicated there was no statistically significant interaction between having a high or low number of behavior incidents on reading pre and postlest scores. The findings for the math pre and postlest scores were s imilar, there was no statistically significant interaction between having a high or low number of behavior incidents on math pre and posttest scores for year one or year two.

PAGE 9

The Form and content of this abstract are approved. I recommend its publication. Approved: Kara Viesca

PAGE 10

CHAPTER I INTRODUCTION The purpose of this study was to detennine if a positive behavior and intervention supports model curriculum (PBIS) had an effect on student academic achievement test score outcomes and school wide number of behavioral incidents over a two-year period. The data collected were based on the first and second year outcomes of academic achievement tests scores and number of behavioral incidents by type ofPBIS interventions offered. PBlS curriculum framework provides the systems and tools for establishing a continuum of evidence-based practices in any school setting with the goal of assisting in meeting the particular academic and / or behavior support needs of students (Simonsen & Sugai, 2013). In the first year, all school personnel at the study site utilized universal tier one supports. In the second year, after receiving training all school personnel utilized individualized and targeted tier two and three supports. The following questions were explored: : Is there an overall di fference between year one and year two student academic post test scores in reading and math? Is there a difference in the total number of behavioral incidents between year one and year two? Is there a difference between pre and post student academic scores and low or high number of incidents? Is there an interaction between pre and post student academic scores and low or high number of behavioral incidents ? Background and Significance This study focused on the documented behaviors and academic test scores of day treatment students attending a Colorado approved and eligible facility school to explore questions about PBlS curriculum. Eligible and approved facility schools serve a unique purpose in the Colorado school systems and are designed to provide educational services to at-risk youth. The mission of an Eligible Facility School is to assure that all students in placement receive a

PAGE 11

quality education and to improve educational outcomes for those students (Colorado Department of Education, 2014). The term Eligible Facility i s defined as a day treatment center residential childcare facility or a hospital licensed by either the Department of Human Services or the Department of Public Health and Environment. Approved Facility School refers to an educational program that is operated by a licensed agency and that has been approved to receive reimbursement for education services for s tudents placed in the program. In the spring of 2008, legi s lation was passed that addressed approved facility schools in a number of ways. Commo nly referr e d to as hou se bill 1204 (22-2-401 C .R.S.), it established a board with rule m a king authority to set graduation requirements and the approval process for facility sc hool s (http :// www.cde.state.co.us / facility sc h oo l slh b0 8 1204act). Day treatment programs which are viewed as a form of alternative education options, have targeted students with behavioral problem s who have been unsuccessful in traditional educational settings (Yi, 2012). Statement of the Problem The literature indicates that th e re is an incr ease in the number of students who are in need or e nrolled in alternative educational treatment programs such as da y treatment program s due to a variety of behavior problems adjustment i ss ues (Garrison, 1987 ; Cox 1999). Although there i s a rise in the need of alternative school placements and programs focu s ing on b e h av ior a l s upports continues (Hughes & Adera 2006) there is limited research regarding important outcome variables for s tudents enrolled in such programs. Review of the Literature For the purposes of this s tudy the lit e rature reviewed examined research p erta ining to s tudent pr o blem behaviors and their relati o n to academic achievement as well as the relationship between PSIS and academic achievement, and finall y how behavioral outcomes and sc hool 2

PAGE 12

climate relate to at-risk youth. The parameters for this literature review were set for research conducted between 1997 to 2016. These dates were chosen because PBIS was established in 1997. During the I 990s the term positive behavior support became popular in school systems. It referred to behavior interventions or strategies that could theoretically be used to reduce problem behavior and assist in promoting desirable behavior (Dunlap et aI., 2000). Academic Achievement and Problem Behavior There have been several studies conducted linking academic perfonnance and problem behaviors by Tobin and Sugai (1999) Morrison et al (2001) Roeser et al (2000) and Nelson et al (2004). This relationship between academic performance and problem behaviors has also been studied at the middle school and high school levels by Tobin and Sugai (1999), Roeser et al (2000), and Nelson et al (2004) Luiselli Putnam, and Sunderland (2002) and, Luiselli Putnam, Handler and Feinberg (2005). Tobin and Sugai (1999) found that individual student academic failure in high school was linked with three or more suspensions in ninth grade. They also found links between grade point average (GPAs) and specific types of office discipline referral (ODR) behaviors, such as fighting, harassing and threats of violence nonviolent misbehavior. Morrison Anthony, Storino and Dillon (200 I) reviewed the records of students who were referred to an in school suspension program. Those students who had no previous ODRs had higher GPAs than the students who had ODRs. Roeser Eccles and Sameroff (2000) found the relationship between academic performance and behavior strengthened in middle school by reviewing ODR documents and suspensions. Other research conducted by Nelson Benner Lane & Smith (2004), found that students with documented problem behavior such as ODRs experienced large academic deficits in 3

PAGE 13

reading and math when compared to same age peers. Also that externalizing or acting out behaviors were strongly related to academic performance deficits when compared to internalizing behaviors. Another study by Lee, Sugai and Homer (1999) found improvements in escape-maintained or acting out problem behavior when students received academic support that made them effective with academic tasks. Research pertaining to the link between academic outcomes and problem behaviors suggests a need for continned research and exploration on the affect PBIS has on academic outcomes. School-wide Positive Behavior Support and Academic Achievement Luiselli, Putnam, and Sunderland (2002) discovered that after the implementation of schoo l-wide behavior support (PBIS) in a suburban middle school, detentions for disruptive behavior and ODRs decreased over a four-year period. School attendance also increased over the same period. The reward for meeting predetermined academic criteria, such as maintaining a specific grade point average, receiving passing grades for all subjects on the report card, and having no more than two homework detentions, as well as, behavioral, attendance, detentions, expulsions was a lottery drawing that was conducted each quarter. The percent of students who were eligible for the l ottery increased from 40% of the schools' population to 55% of the schools' population over the course of four years [n another study, Luiselli, Putnam, Handler, and Feinberg (2005) implemented a schoolwide behavior support plan at an urban school and found decreases from baseline to intervention to follow-up in documented behaviors and suspensions. Reading comprehension and mathematics percentile ranks on standardized tests improved from the first to the second test dates, with an increase of 18 and 25 percentage points. In another study, Putnam, Handler, & O Leary-Zonarich (2003) discovered that reading and math scores improved on standardized 4

PAGE 14

testing following the implementation of behavior s upport intervention at an urban elementary school. An analysis of academic performance of schools that are implementing school wide positive behavior su pport compared to sc hools not implementing such programs was conducted in Illinoi s (Homer, Sugai, Eber, & Lewandowski, 2004). Schools that were implementing PBIS had scored 80% on the School Evaluation Tool, a tool utili ze d by PBIS to measure the s uccess of implement atio n in school systems and assist in developing improvement plans (Sugai Lewis Palmer Todd & Homer, 2001). Schools implementing PBIS also had 80% of their students able to discu ss their school wide expectations and rule s (Sugai, et aI., 2001). Homer, Sugai, Todd, and Lewis-Palmer (2005) found similar findings with another school district with nineteen elementary schools. Between the 1997-98 and 2001-2002 academic years, thirteen of the schools implemented school-wide positive behavior support and six schools did not. They compared the percentage of 3rd graders who met state wide reading s tandard s in the academic year 1997-98 with the percentage in the academic year 2001-2002. Their findings concluded that ten out of the thirteen schools that implemented PBIS practices had improved outcomes. The overall increase in percentage of students meeting standards ranged from 2% to over 15% in the se schools. Only one of the six schools that not did implement school-wide positive behavior s upport showed improvement. School-wide Positive Behavior Support and School Culture School culture has been defined as the belief sys tem that directly influences school climate (Sugai, 2013). School culture refers to the manner in which teachers and staff members work together while school climate refers more toward the school's effects on students. ODRs have been primarily used in PBIS research to track beha v ioral and sc hool climate outcomes 5

PAGE 15

(Sugai, 2013). Anecdotal evidence would suggest that PBIS schools with reduced referrals and discipline issues have better teacher retention and higher satisfaction thus leading to a positive school climate (Sugai 2013). Oliver Wehby and Reschly (2011) found that teachers who experience difficult y controlling classroom behavior have higher stress and burnout making it more difficult for them to meet the in s tructional demand s of the classroom. Research has suggested that schools can support classroom teachers with PBIS by focusing on prevention; using multiple data sources to develop strategies for screening, identification and treatment, and taking a coordinated, school wide approach to reducing problem behaviors among students (Oliver et ai., 2011). McClure (2011) reviewed studies supporting PBIS and found several studies which associated PBIS with decrease s in office discipline referrals (OORs) as well as increased consistency and positive interactions among school staff. There have been many studies conducted that have indicated that schools implementing PBIS have significant reductions in OORs data (Nelson 1997 ; Sprague, et ai., 2001). Luiselli, Putnam, and Handler (2001) indicated a 69% reduction in OORs after the implementation of PBIS in their study. Similarly, Todd, Haugen Anderson and Spriggs (2002) indicated an 80% reduction in OORs in the first year of PBIS implementation and a 76% reduction in the second y ear in their review of behavior documentation. More recently Bradshaw and Leaf (2008) indicated reduced OORs data as well as improved perceptions of school safety among teachers and staff in the Mar y land school system. Positive Behavior Supports and At-Risk Youth There are a large number of youth educated in restrictive or alternative education (AE) settings. AE sc hools and programing include those housed in juvenile detention centers (Carver, 6

PAGE 16

Lewis, & Tice, 2010). Estimates suggest that between 12% and 50% of these youth have disabilities, and most youth are placed in restrictive settings as a result of significant behavior challenges (Carver et a!., 2010). Public school districts report transferring youth to AE settings for a variety of reasons, including physical aggression (61 % of districts) ; "disruptive verbal behavior" (57%); "possession distribution, or use" of controlled substances (57%); chronic academic failure (57%) or truancy (53%); possession or use offireanns (42%) or other weapons (51 %); "arrests or involvement with the criminal justice system" (42%); teen parenthood (31 %); and/or mental health needs 27% ; (Carver et a!., 2010, p. II). Therefore, AE settings need to be able to support youth with a variety of behavioral needs and challenges, as well as meet individual academic and behavioral needs. Empirical research on the presence and effectiveness of behavior support practices / PBIS in AE settings is limited (Flower McDaniel, & Jolivette 2011 ; Lehr, 2004). Youth that are unresponsive to tier 1 practices such as universal supports that are offered daily to all students may require additional tier 2 practices such as including an individualized goal on a youth s school-wide point card, providing additional adult mentoring and support to enhance social skills instruction, and developing a menu of more individuali z ed reinforcements For youth whose behaviors are unresponsive to tier 2, individualized tier 3 practices may be added. Tier 3 practices should be based on a full functional behavioral planning (Eber Sugai, Smith, & Scott 2002). Following these implementation practices in an AE setting, suggests that a PBIS framework may result in positive outcomes for youth educated within AE settings, induding increases in appropriate behavior decreases in problem behaviors and decreases in use of crisis emergency responses such as restraint (Simonsen Young & Britton 20 10). Summary 7

PAGE 17

Tobin and Sugai (1999), Morrison et al (2001), Roeser et al (2000), and Nelson et al (2004) presented a relationship between academic performance and problem behaviors that have also been studied at the middle school and high school levels. These connections were linked to office referrals, documented behaviors and academic perfonnance. An improvement in school climate and academic outcomes has been linked to PBIS practices Luiselli, Putnam, and Sunderland (2002) discovered that after the implementation of PBIS in a suburban middle school detentions for disruptive behavior and ODRs decreased and attendance increased. Additionally, Luiselli, Putnam, Handler, and Feinberg (2005) implemented a school-wide behavior support plan at an urban school and found reading comprehension and mathematics percentile ranks on standardized tests improved from the first to the second test dates, with an increase of 18 and 25 percentage points. There was also evidence that suggested that PBIS schools with reduced referrals and discipline issues have better teacher retention and higher satisfaction thus leading to a positive school climate (Sugai 20 I 3). There are a large number of youth educated in restrictive or alternative education (AE) settings. AE schools and programing include those housed in juvenile detention centers (Carver Lewis & Tice 2010). Empirical research on the presence and effectiveness of behavior support practices / PSIS in AE settings is limited (Flower, McDaniel, & Jolivette, 2011; Lehr, 2004). The research suggests that there is a need for further research to examine the academic and behavioral outcomes from PBIS use of at-risk youth in alternative settings. 8

PAGE 18

CHAPTER II THEORETICAL FRAMEWORK This stud y focused on B.F. Skinner's (1950) operant conditioning theory as a ba s i s of measurement due to the similarities with the PBIS programming. Operant conditioning and PBIS similarities in theory and practice will be discussed below. PBIS School-wide Positive Behavior Interventions and Support (PBIS) is a systems approach to establishing the social culture and behavioral supports needed for all students in a school to achieve both social and academic success (Homer at el., 2004). PBIS is an approach that defines core element s of behavioral and student struggles that can be addressed through a variety of tiered supports and strategies. The table outlined below sums up the idea of PBIS and the areas a school team needs to collect data to analyze behavioral and academic trend s The team plan s and implements group and individuali zed intervention based on assessment information focusing on (a) prevention of problem contexts, (b) instruction on functionally equivalent skills and instruction on desired performance skills (c) strategies for placing problem behavior on extinction (d) strategie s for enhancing contingence reward of desired behavior and (e) use of negative or safety consequences if needed (Homer at el., 2004). B.F. Skinner s Operant C onditioning Theory and PBIS interventions share a similar model and participation planning 9

PAGE 19

Table I PBIS Tiers and Core Elements Prevention Tier Primary Secondary Tertiary Core Elements Behavioral expectations are defined and taught Reward sys tem is established for appropriate behaviors Consequence continuum is established for problem beha v ior. Continuous data collection and planning takes place Universal screening Progre ss monitoring for at risk students System for increasing structure and predictability System for increasing contingent adult feedback System for linking academic and behavioral performance System for increasing home / school communication Collection and us e of data for decision-making Functional Behavioral Assessment Team-based comprehensive assessment Linking of academic and behavior supports Individualized intervention based on assessment information focusing on (a) prevention of problem contexts (b) instruction on functionally equivalent skills, and instruction on desired performance skills, (c) strategies for placing problem behavior on extinction, (d) strategies for enhancing contingence reward of desired behavior and (e) use of negative or safety consequences if needed. Collection and use of data for decision-making (Homer at eJ., 2004) Operant Conditioning Operant conditioning ha s been widely applied in clinical settings for behavior modification as well as, for cla ssroo m management and instructional development (Skinner 1950) The operant conditioning theory of B.F. Skinner is based upon the idea th at learning is a function of change in overt behavior (Skinner 1950) C hanges in behavior are the result of an individual's respon se to events (stimuli) that occur in the environment. When a particular Stimulus-Response (S-R) pattern is reinforced or rewarded, the individual is conditioned to respond (Skinner, 1954). 10

PAGE 20

Stimulus response in Skinner s S-R pattern theory and the PBIS prescribed group and individual interventions strategies and supports follow the same processes. The reinforcements and rewards of the S-R pattern and PBIS are discussed further below. Reinforcement and Rewards Reinforcement is the key element in Skinner's S-R theory (Skinner 1968). A reinforcer is anything that strengthens the desired response. It could be verbal praise a good grade or a feeling of increased accomplishment or satisfaction (Skinner 1968). The theory also covers negative reinforcers which are any stimulus that results in the increased frequency of a response when it is withdrawn (Skinner, 1968). In past years school wide discipline has focused mainly on reacting to specific student misbehavior by implementing punishment-based strategies including reprimands loss of privileges, office referrals suspensions and expulsions (Sugai & Homer, 1999). Meaning that punishment utili zed to promote desired behaviors especially when it is used inconsistently is ineffective. Introducing modeling, and reinforcing positive social behavior is an important aspect of a student's educational experience (Sugai & Homer 1999). The purpose of school-wide PBIS is to establish a climate in which appropriate behavior is the norm (Barrett at el., 2008) The PBIS strategy focuses on identifying problem behaviors and the environments in which they occur in and then providing programming that involves teaching and modeling desired behaviors (Barrett at el., 2008). Then when students demonstrate the desired behaviors they receive positive reinforcement through a variety of stimuli Teaching behavioral expectations and rewarding students for following them is a much more positive approach than waiting for misbehavior to occur before responding thus creating a school environment geared towards success (Barrett at el., 2008). I 1

PAGE 21

Critical Evaluation of Operant Conditioning Theory Operant conditioning can be used to explain a wide variety of behaviors from the process ofleaming, to language acquisition. Operant conditioning also has practical applications such as token economies which can be applied in classrooms prisons and psychiatric hospitals, and are also utilized among PBIS programming plans (McLeod, 2015). Despite this operant conditioning has been disputed and critiqued over the years. Some of the best-known critic s include Noam Chompsky (1959) and J.E.R. Staddon (1995) who both question the methods used to develop this theory. Additionally, later critiques asserted that the theory fails to take into account the role of inherited and cognitive factors in learning, thus making the theory an incomplete explanation of the learning process (McLeod 20 \5). Lastly the use of animal research in operant conditioning studies also raised the issue of extrapolation (McLeod, 2015). Some psychologists argue that generalizations cannot be made from studies on animals to humans because their anatomy and physiology are different from humans (McLeod, 20 \5). There are visible similarities between operant conditioning Theory and PSIS programming which despite the critiques makes this an appropriate perspective for this study. There are a variety of different positive reinforcements prescribed in both operant conditioning theory and PSIS planning u se d to increase the likelihood of desired behavior in the classroom. Some of the most widely used reinforcements include consumable rewards, positive social interactions and earned activities. According to Landrum and Kauffinan (2006) Despite a rich history and extensive empirical underpinnings the behavioral perspective on teaching and management is not highly regarded in the education community" (2006 p. 47). C ritics argue that operant conditioning is an unfeeling approach more suited to animals than to humans (Landrum & Kauffman 2006). 12

PAGE 22

Regardless of the critiques, operant conditioning is commonly used in classrooms and is viewed by many teachers as an effective approach to improving classroom practice. It provides teachers with a set of tools for improving classroom management and student l earning (Landrum & Kauffman 2006). In addition, the underlying purpose of PBlS programming follows operant conditioning theory closely but also applies the e l ement of data collection. The data collected in PSIS programmed schools measures student behaviors while demanding tiered intervention programming changes if desired behaviors are not exhibited which made the connection between operant conditioning and PSIS programming relevant to this study. 13

PAGE 23

CHAPTER III METHODOLOGICAL APPROACH A quasi-experimental approach was utilized to investigate the research questions. Quantitative research approaches are viewed to be objective in nature (Gliner at el., 2009). The tenn objective implies that behaviors are easily c1assi fied and that data are usually collected with some sort of instrument (Gliner at el. 2009). Quasi-experimental research designs contain both independent and dependent variables (Gliner at el., 2009) Active independent variables are defined as a treatment, such as a workshop, new curriculum, or other intervention for a specific amount of time throughout the duration of the study. Dependent variables are the measure or the outcome, such as test scores behavioral changes or other effect related outcomes (Gliner at el., 2009). Additionally quasi-experimental designs can be divided into four different categories based on the basic design of the study For the purposes of this study a quasi-experimental one group pretest-posttest design was used. This type of design requires that an initial observation (pretest) is given then the intervention or independent variable is applied and finally a second observation (posttest) is applied (Gliner at el., 2009). The statistical analysis utilized in this study was a mixed ANOV A for research question one and a dependent t test for research question two and three. This analysis was conducted in SPSS. Measures This study consisted of two dependent variables and one active independent variable. The researcher relied on previously collected data from two databases. Variables In quantitative research variables are the key elements in research questions. Variables are defined as the characteristic s of situations or participants in any given study having different 14

PAGE 24

values (Gliner at el. 2009). This stu d y me as ured the characteristics of two dependent variables. D e pendent variables are defined as the assumed measure or effect of the independent variable (Gliner at el., 2009). Independent and dependent variables can be divided into two types; active or attribute (Gliner at el., 2009). For the purpo ses of this study the active independent va riable was the PBIS cuniculum. There were three dependent variables in this study. Two of th e dependent variables are academic te s t sco res measured at student intake and discharge creating the pretest and po s ttest scores in the content a rea s of reading comprehension a nd math The independent varia ble for this s tud y was the implementation ofPBIS cuniculum and s upports over a twoyea r period. During yea r one the 2013-2014 school y ear tier one s upport s were put into place throughout the school environment f o r all students. In year two th e 20 14-2015 sc hool year tier two and three support s were put into place throughout the school environment. Participants and Site Day treatment students are commonly refe rred to a s at risk yo uth. At-risk s tud e nt s a re typically defined as students who are lik e ly to fail or drop out of school before a high sc h ool diploma or GED is attained (Kaufman at el., 1 992). Research ha s indicated that s tud ents with disrupti ve behavior disorders are a t ris k for numerous adjustment problem s during adolescence which include v iolence and juvenile delinquency (Yi, 2012). These various adjustment problem s can result in academic failure substance abuse, risky sexual behavior and antisocial beh av ior (Broidy et aI., 2003 ; Yi 2012). Thus, resulting in the need for an alternative e ducational option such a s a da y tr eat ment pr ogra m (Yi 2012). The day treatment students in thi s sa mple are all male s tudents between th e ages of nine t o seve nteen who ha ve been diagno se d with vario u s beha v ior diso rder s including attention deficit! hyperactivity di so rder (ADHD), o ppo si tional defiant disorder (ODD) and conduct 15

PAGE 25

disor der (CD) and a variety of additional behavior disorders which are typicall y associated with behavioral so cial emotional, academic impairments. As defined by the American P sy chiatric Association's Diagnostic and Statistical Manual of Mental Disorders fourth edition, text revision (DSM-IV-TR ; APA 2000), Attention-Defici t/ Hyperactivity Disorder (ADHD) is one of the most common and pervasively impairing childhood psychiatric disorders. ADHD and CD are characterized by the primary features of inattention, hyperactivity and impulsivity and are associated with other disruptive behaviors. ODD is characterized by patterns of negativi s tic disobedient hostile and defiant behavior toward authority figures that persists for at least 6 months and is characterized by the frequent behaviors of temper loss arguments with adults defiance or refusal to comply with adult instructions, and deliberate actions of annoying others (DSM-IV-TR, APA, 2000). Data Collection Procedure All schools receiving any state or federal funds are required to collect data for reporting purposes. This study will utili ze two data sources collected from an alternative school setting. The two samples are pre and post academic test scores and documented behavior incidents. The sections below will discuss the specifics about how each sample is collected at the site. Additionally, the chart below depicts the data being collected from the sample, as well as, the duration and frequency of collection. 16

PAGE 26

Table 2 Artifa c t and data collection Arti facts / data Year I number of individual student documented incidents Year 2 number of individual student documented incidents Year I pretest scores Rdg Year 1 pretest scores Math Year I posttest scores Rdg Year 1 posttest scores Math Year 2 pretest scores Rdg Year 2 pretest scores Math Year 2 posHest scores Rdg Year 2 posttest scores Math Time Period Aug. 2013May 2014 Aug. 2014May 2015 Aug 2013Apri12014 Aug. 2013April 2014 Sep 2013May 2014 Sep. 2013May 2014 Aug. 2014April 2015 Aug. 2014April 2015 Sep. 2014May 2015 Sep. 2014May 2015 Year I total number of documented Aug. 2013Ma y 2014 Anticipated Sample Size n =53 collected from n = 54 collected from n =53 n =53 n =53 n =53 n = 54 n = 54 n = 54 n = 54 behavioral incident reports n = 808 collected from Year 2 total number of documented Aug 2014May 2015 behavioral incident reports 17 n = I 126 collected from

PAGE 27

CHAPTER IV FINDINGS A de sc riptive statistic analysis was conducted on the 2013-2015 data for the following variables; year one reading pre and posttest scores year one math pre and posttest scores, year two reading pre and posttest scores year two math pre and posttest scores year one documented behavior incident s and year two documented behavioral incidents. The results are as follows: in year one there were 53 student reports with a minimum grade level of 4 and a maximum grade level range of 12. In year two there were 54 student reports with a minimum grade l eve l of 4 and a maximum grade level of II In both years the most common grade lev e l was grade 8. Table 3 Prei12pst Test Scores and Documented Behaviors Over a Two-Year P e riod Years Mean Median Mode skewness Min Max YI rdg pre 6.0 5.5 5.7 .68 .2 12. 0 YI rdg po st 7.0 6.4 5.7 .55 2.8 12.9 YI math pre 5.5 5.1 4.0 1.3 1.9 12. 9 YI math post 6.2 5.4 4.3 1.1 1.4 12.9 Y2 rdg pre 6.6 5 8 4.3 .63 1.4 12. 9 Y2 rdg post 7.4 6.9 10.8 .25 2.6 12.9 Y2 math pre 5.4 4.9 3.5 1.3 1.4 12.9 Y2 math p ost 5.7 4.8 4.8 1.0 .8 12.9 YI IR behavior 15. 2 9 2.1 0 94 Y2 IR behavior 20.8 13.0 2 1.6 0 103 The mean of the pre reading scores was 6.05, which was s imilar to the math prete st scores 5.56. The mean of the posttest reading scores for yea r one was 7.02 and the mean for the 18

PAGE 28

posttest math scores was 6.28 which would indicate that there was growth between the pre and posttest scores in reading and math in year one. The most common pretest reading scores for year one was 5.7 the most common pretest math scores for year one was 4.0, meaning that in year one and year two when students were taking their pretests in both reading and math they were scoring between and forth and fifth grade level despite their actual grade level, which was most commonly an eighth grade level. The most common posttest reading posttest score for year one was 5.7 and the most common posttest score for math in year one was 4.3, meaning that students were scoring below the most common grade level 8 on their reading and math posttests. The results for year two were similar to year one. The year two pretest reading scores was (M= 6.67) and the math pretest scores were (M= 5.45) and the posttest reading scores were (M= 7.46) and the posttest math scores was (M= 5.78), indicating that there was also growth in year two in reading and math between the pre and posttest scores Additionally, there were similar pre and posttest common scores in year two. The most common pre test scores for reading was 4.3 and for math was 3.5 meaning that in year one and two when students were taking their pretests in both reading and math they were scoring between and third and fifth grade level, despite their actual grade level, which was most commonly an eighth grade level. The most common math was 4.8, indicating that students were scoring below the most common grade level 8 on math posttests. The math posttest scores for year one and two were skewed. In year two the reading posttest scores mean had a dramatic increase to 10.8 from the pretest mean 4.3, indicating that there was a large increase in the reading scores from pre to posttest. The mean of year one's documented behavior incidents was 15.25 and the mean for year two's documented behavior incidents was 20.85 indicating that there was an increase in the overall number of incident reports written documenting behavioral incidents from year one to 19

PAGE 29

year two. The range for documented behavior incidents in year one was 94, with a minimum range of 0 and maximum range of 94. The range for year two documented behavioral incidents was 103, with a minimum range of 0 and a maximum range of 103. The middle amount of incident reports written for a student in year one was 9 and the middle amount of incident reports written for a student in year two was 13. The most common number of incidents in year one was I, and the most common number of incidents in year two was 2. This indicates that individual student(s) were generating more incident reports in year two. Research Questions One and Two This study examined if the PBIS curriculum had a significant difference between the year one and year two interventions by measuring if there was an overall difference between year one and year two student academic post test scores and behavioral incidents. To assess these differences three t tests were conducted in (I) reading posttest scores for year one and two (2) math posttest scores for year one and two for research question one, and (3) the documented behavioral incidents for year one and two for research question two. The findings for each question are depicted in table four and discussed below. Table 4 Differences Between Posttest Scores and Behavior Incidents {or Year one and Two Tests and Years df p Variance Trans{ormed sk Reading 52.46 Reading posttest Y 1 Reading posttest Y2 Math Math posttest Y 1 Math posttest Y2 Behavioral incident 52 52 .37 < .001 9.0 10.0 .04 .05 Research Question One .55 .25 -.04 -.77 Is there an overall difference between the year one and year two student academic post test scores? To determine if there was an overall difference between year one and year two 20

PAGE 30

student academic posttest scores two 1 tests were conducted. The findings are discussed below Reading Posttest scores To assess if there was a significant difference between the reading posttest scores over a two-year period a 1 test was conducted. There were (/1= 53) scores from year one and (N= 54) scores from year two. The assumptions were checked and have been met the data is nonnal and the variables are independent. There was no significant difference between year one reading posttest scores and the year two posttest scores, 1(52) = -.746,p = .46. The effec t size, d was not reported because there was no statistically significant difference found in these results Math Posttest Scores To assess if there was a significant difference between the math postte s t scores over a two-year period a 1 test was conducted. There were (11= 53) sco r es from year one and (/1= 54) scores from year two. The assumpt i ons were been checked, and one was not been met. The data was not nonnally distributed The math postte s t scores from year one had a skewness sco re of 1 1 and the postte s t scores from year two h ad a skewness score of 1.08. For this reason a Logl 0 transfer was executed in SPSS to transfonn the variable. Thus transfonning the skewness score for math year one .0 49 and for math year two -.77 There was no statistically significant difference between year one math posttest scores and year two math postte s t sco res 1(52) = .901 P = .37. The effect size, d was not reported because there was no statistically significant difference found in these results Research Question Two 21

PAGE 31

Is there a difference in the total number of behavioral incidents between year one and year two? To detennine if there was an overall difference between year one and year two documented behavior incidents two 1 tests were conducted. The findings are discussed below. Documented Behavioral Incidents To assess if there was a significant difference between the number of documented behavioral incidents over a two-year period a 1 test was conducted. There were (n = 53) students generating documented incidents from year one and (n = 54) students from year two. The assumptions have been checked and one has not been met, year one behavioral incidents has a skewness score of2.1 and year two has a skewness score of 1.6. For this reason a LoglO transfer was executed in SPSS to transfonn the variable. There was a statistically significant difference between year one documented behavior incidents and year two documented behavior incidents 1(52) = -89.23 p < .001. The effect size (d = -0.27) is close to a medium effect size according to Cohen (1988). Research Question Three Is there a difference between pre and post student academic scores and low or high number of incidents? Is there an interaction between pre and post student academic scores and low or high number of behavioral incidents? To answer this question, four mixed ANOVAs were conducted to assess the following: (I) whether there was an interaction between pre and posttest scores in reading for year one and the level of behavior documented incidents (i.e. low vs. high) (2) whether there was an interaction between pre and posttest scores in math for year one and the level of behavior documented incidents (i.e., low vs. high) (3) whether there was an interaction between pre and posttest scores in reading for year two and the level of behavior documented incidents (i.e., low vs. high), and (4) whether there was an interaction between pre and posttest 22

PAGE 32

scores in math for year two and the level of behavior documented incidents (i.e., low vs. high). The following assumptions were tested (a) independence of observations (b) normality, and (c) sphericity. All assumptions were met. The findings for each test are depicted in Table 5 and discussed below. Table 5 interactions between test scores and high-low behaviors Year d( F p Reading year I .1 15 .73 Reading year 2 .336 .56 Math year I 2.47 .12 Math year 2 .129.72 Reading Year One Partial etaZ .02 .006 .046 .002 Power .063 .088 .339 .064 To assess whether there was an interaction between pre and posttest scores in reading for year one and the level of behavior documented incidents (i.e., low vs. high) a Mixed ANOYA was conducted. The results depicted in the table above indicated no statistically significant interaction between the reading year one test scores and high-low behaviors F( I 1)= .115 P = .73. The results indicated there was a significant main effect for tests scores F(I, 1)= 19.28, P <.001, partial eta2 = .27 which is a small effect size according to Cohen (1988), but there was not a statistically significant effect for the high-low behavior incidents, F(I 1)= 1.20 p = .28. This indicates that even though there was a statistically significant effect between the reading pre and posttest scores for year one there was no statistically significant interaction between the pre and posttest scores and having a high or low number of behavior incidents. The power for the pre and posttest scores finding of statistically significant main effect was 99, indicating there was an adequate amount of power to detennine the statistically significant main effect. The power for the high-low behavior incidents was .19, indicating that there was not enough power to detennine if there was a statistically significant effect. The power 23

PAGE 33

for the interaction of the pre / posttest scores and behavior incidents was .063, indicating that we did not have enough power to find a statistically significant interaction if one had existed. Reading Year Two To assess whether there was a statistically significant interaction between pre and postlest scores in reading for year two and the level of behavior documented incidents (i.e. low vs. high) a Mixed ANOV A was conducted. The results depicted in the table above indicated no statistically significant interaction between the reading year two test scores and high-low behaviors F(1, I) = .336 p = .56. The results indicated there was a statistically significant main effect for tests scores F( I 1)= 16.57 P = .017 partial eta2 = .105 which is a large effect size according to Cohen (1988) but there was not a statistically significant effect for the high-low behavior incidents, F(I, I) = .011, p = .91. This indicates that even thought there was a statistically significant effect between the reading pre and posttest scores for year two there was no statistically significant interaction between the pre and postlest sco res and having a high or low number of behavior incidents. The power for the pre and postlest scores finding of statistically significant main effect was .67, indicating there was not an adequate amount of power to determine the significant main effect. The power for the high-low behavior incidents was .051, indicating that there was not enough power to determine if there was a statistically significant effect. The power for the interaction of the pre / postlest scores and behavior incidents was .088 indicating that there not was enough power to determine the statistically significant interaction if one had existed. Year One Math 24

PAGE 34

To assess whether there was an interaction between pre and pos ttest scores in math for year one and the level of behavior documented incidents (i .e low v s high) a Mixed ANOY A was conducted. The results depicted in the table above indicated no statistically significant interaction between the math year one test scores and high-low behaviors F(l, 1) = 2.47 p = .12. The results indicated there was a statistically significant main effect for tests scores, F(I, I) = 9.09,p = .004 partial eta2 = .15 which is a small effect size according to Cohen (1988) but there was not a statistically significant effect for the high-low behavior incidents, F(1, 1)= .143 p = .70. This indicates that even though there was a statistically significant effect between the math pre and posttest scores for year one there was no statistically significant interaction between the pre and posttest scores and having a high or low number of behavior incidents. The power for the pre and posttest scores finding of statistically significant main effect was .84 indicating there was an adequate amount of power to determine the statistically significant main effect. The power for the high-low behavior incidents was .06 indicating that there was not enough power to detennine if there was a statistically significant effect. The power for the interaction of the pre / posttest scores and behavior incidents was .33, indicating that there was not enough power to detennine ifthere was a statistically significant interaction. Year Two Math To assess whether there was an interaction between pre and posttest scores in math for year two and the level of behavior documented incidents (i.e., low vs. high) a Mixed ANOYA was conducted. The results depicted in the table above indicated no statistically significant interaction between the math year two test scores and high-low behaviors, F(1, 1) = 129 P = 72. The results indicated there was no statistically significant main effect for tests scores F( 1 I) = 3.29 p = 075 there was also no s tatistically significant effect for the high-low behavior 25

PAGE 35

incidents F(31, I) = .126 p = .72. This indicates that there was no statistically significant effect between the math pre and posttest scores or high-low behavior incidents for year two. Additionally there was no statistically significant interaction between the pre and posttest scores and having a high or low number of behavior incidents. The power for the pre and posttest scores finding of statistically significant main effect was .42, indicating there was not an adequate amount of power to detennine if there was a statistically significant main effect. The power for the high-low behavior incidents was .064 indicating that there was not enough power to detennine a statistically significant effect. The power for the interaction of the pre / posttest scores and behavior incidents was .064 indicating that there was not enough power to detennine ifthere was a statistically significant interaction. 26

PAGE 36

CHAPTER V DISCUSSION This study set out to explore if a positive behavior and intervention supports model curriculum (PBIS) had an effect on student academic achievement test score outcomes and school wide number of behavioral incidents over a two-year period. The data collected were based on the fir st and second year outcomes of academic achievement tests scores and number of behavioral incidents by type of PBIS interventions offered. The following questions were answered: : Is there an overall difference between year one and year two student academic post test scores in reading and math? Is there a difference in the total number of behavioral incidents between year one and year two ? Is there a difference between pre an d post student academic scores and low or high number of incidents? Is there an interaction between pre and post student academic scores and low or high number of behavioral incidents? Summary of Findings The first research questions explored was: I s there an overall difference between year one and year two student academic post test scores? The findings indicated that there was no s ignificant difference with the reading or math scores between year one and year two. The second research question explored was: Is there a difference in the total number of behavioral incidents between year one and year two? The findings indicated there was a s ignificant difference between year one documented behavior incidents (M=5) and year two documented behavior incident s (M=4). The effect s i ze, d = -0.27 which is close to a medium effect size. Additionally four t tests were conducted to determine ifthere was a difference between the pre and posttest scores in readin g and math for both years. The assumptions were checked 27

PAGE 37

and one was not met, the math scores for year one and two were skewed. A Log 10 transformation of these variables was completed The results for year one were as follows ; there was a significant difference between the pre and postlest scores in reading 1(52) = -.461 P < .001, d = -0.35, which is a medium effect size according to Cohen (1988). There was also statistically significant difference between pre and postlest math scores 1(52) = -2.45 P = .017 d = -0.23 which is a small effect size according to Cohen (1988). For year two the results were as follows; there was a statistically significant difference between the pre and postlest scores in reading 1(53) = -2.51 P = .015 d = -0.25 which is a small effect size according to Cohen (1988). There was no statistically significant difference between the pre and postlest scores in math 1(53 ) = -1.68 p = .097. The effect size d was not reported because there was no statistically significant difference found in these results. The third research questions explored was: Is there a difference between pre and post student academic test scores in each year, when adjusting for low or high number of behavioral incidents? The year one and year two findings indicated there was no statistically significant interaction between having a high or low number of behavior incidents on reading pre and posttest scores. Additionally in year one and two indicated that there was also no statistically significant interaction between having a high or low number of behavior incidents on the pre and postlest math scores. Implications When comparing year one and year two's academic outcomes in reading and math there was lack of significance growth from year one to year two. However when comparing the academic growth within each y ear from pretest to postlest there was a significant amount of 28

PAGE 38

growth for both years indicating that the students did demonstrate academic growth during their time of enrollment at the school. The findings when comparing all academic scores indicate that the school was able to maintain consistent academic growth for students each year but not enough growth to demonstrate an increase in scores from year one to year two. This lack of growth could be attributed to the novelty the PBIS system in the school and the implementation of additional tiered supports for struggling students. Additional training or professional development pertaining to PBIS practices and time to monitor interventions and academic growth could yield different academic results. Comparing the documented behavior incidents from year one to year two there was a significant difference between year one and year two. Looking at the descriptive analysis there were on average more incident s documented per student in year two. This could be attributed to different school, student and staff related factors. The first possible factor being the students on average could have demonstrated more incident related behaviors in year two Another possible factor being the education staff may have documented more incidents in the second year due to the nature of PBlS programming. One of the major focuses ofPBIS is collecting data in regards to tracking behaviors and identifying behaviors requiring interventions, meaning that there is a possibility that more behaviors were documented in the second year for the purpose of tracking to identify where interventions are needed. When examining if there was an interaction between a high or low number if incidents and academic scores there were no significant interactions found. There are a variety of circumstances or situations that could attribute to these outcomes. Previous research Tobin and Sugai (1999) Morrison et al (200 I), Roeser et al (2000) and Nelson et al (2004) have found that there is a link between behavior and academic performance so perhaps with larger effect sizes or 29

PAGE 39

a longer period of time over years to measure these outcomes a link could be made, if one exists in this situation. When discussing these results with the possible situationa l circumstances that could have attributed to the findings of this study have been speculated by this researcher. Below this study's limitations and recommendations for add ition a l research including school actions are discussed. Limitations and Suggestions for Future Research This study was theoretically grou nd ed in operant condit i oning th eory which has been widely app lied in clinical settings for behavior modification as well as, for classroom management and instructional development (Skinner 1950). The operant conditioning theory of B.F. Skil1l1er is based upon the idea that le arning is a function of change in overt behavior (Skinner 1950). Changes in behavior are th e result of an individual's response to events (stimuli) that occur in the environment. When a particular Stimulus-Response (S-R) pattern is reinforced or rewarded the individual is cond iti oned to respond (Skinner, 1 954). This theory holds the same basis of the und erlying ideology ofPBIS strategy focuses on identifying problem behaviors and the environments in whic h they occur in and then providing programming that involves teaching and modeling desired behaviors (Barrett at el. 2008). While the participant schoo l reported that tier o n e and tier two interventions were applied to meet the individual needs of the students this study did not focus on the specific interventions being utilized. This study focused on the academic and behavioral outcomes for each st udent. There was no measurement of tier two and three interventions being provided to specific students. Suggest i ons for future research would include pairing specific interventions administered to various students in relation to their academic achievement and behavioral 30

PAGE 40

outcomes. Additionally, tracking specific times behaviors are occurring and focusing on targeting pattems in times of day or classes where behaviors are occurring will provide additional information regarding when interventions are needed on a class or school wide scale. Collecting this data could provide additional infonnation pertaining to the interactions between behaviors and academic achievement. Further more tracking academic achievement results for years prior to the PBlS implementation and then comparing the findings from the implementation years could shed some light on the school performance before the interventions were applied to the school and students. This information would assist in determining the effectiveness of the PBlS curriculum over time. There are additional assessment factors that could be taken into consideration to examining academic growth within each year and between years. The use of different academic assessments with smaller units of measurement could open up additional information pertaining to academic growth in specific areas. Completing assessments on a progress-monitoring basis could also assist in the overall development of tracking pattems in academic difficulties and growth of the students, which would allow for additional academics to be put into place. Implementing additional practices with assessments could allow for additional information and understanding of academic growth both within and between school years. 31

PAGE 41

Running Head: REFEREN CES REFERENCES American Psychiatric Association (2000). Dia gnost i c and s tati stical manual of m e ntal disorders (4th ed. text rev.). Washington, DC Bradshaw, C., & Leaf, P. (2008). Project target: Update on key Findings from project target. Barrett, S., Bradshaw C., & Lewis-Palmer, T. (2008). Maryland s tate-wide PBlS initiative. Journal of Positive Behavior Interventions, 10 1005-114 Carver, P. R. Lewis L., & Tice, P. (2010). Alternative sc hools and programs for public school students at risk of educational failure: 2007-08 (NCES 2010-026). U.S. Department of Education, National Center for Education Statistics. Washington DC: Government Printing Office. Chomsky, N. (1959) "Reviews: V e rbal behavior by B. F. Skinner". Language 35 (1): 26-58. retrieved from: JSTOR 411334. Colorado Department of Education. (2008). H.B. 1204 (22-2-401 C.R.S.) http ://www.cde.state.co.us / facilityschool slh b081204act Colorado Department of Education. (2014). Facility schools, Colorado. Retrieved from: http://www.cde.state.co.us / facilityschools Dell, C.A., Harrold 8., & Dell, T. (2008). Test r e view: Wide rang e achievement test (4th ed.). Lutz, FL: Psychological Assessment Re so urces Eber, L. Sugai, G., Smith C. R., & Scott, T. M. (2002). Wraparound and positive behavioral interventions and supports in the schools. Journal of Emotional and Behavioral Disorders 10 171-181. Ennis, R. P., Jolivette, K., Swoszowski, N c., & Johnson M. L. (2012): Secondar prevention efforts at a residential facility for students with emotional and behavioral disorders: Function-based check-in, check-out, Residential Treatment for Children & Youth, 29, 79-102. Fleming, C. B., Harachi T. W. Cortes, R. c., Abbott, R. D. & Ca talano R. F. (2004). Level and change in reading scores and attention problems during elementary school as predictors of problem behavior in middle school. Journal of Emotional and Behavioral Disorders 12(3),130-144. Garrison, R. W. (1987). Alternative schools for disruptive youth: NSSC resource paper. Malibu CA: National Safety Center, Pepperdine University, Office of Juvenile Justice and Delinquency Prevention 32

PAGE 42

Running Head: REFERENCES Gliner J. A., Morgan, G. A., Leech N. L. (2009). Research methods in appli e d s e ttings: An integrat e d approach to d e sign and analysis. (2nd ed.). New York NY: Taylor and Francis. Horner, R. H., Todd A Lewis-Palmer, T. Irvin 1., Sugai, G. & Boland J (2004). The school-wide evaluation tool (SET): A research instrument for assessing school-wide positive behavior support. Journal of Positive Behavior Intervention 6(1) 3-12. Horner, R. H., Sugai, G., Todd, A. W., & Lewis-Palmer T. (2005) School-wide positive behavior support: An alternative approach to discipline in schoo l s. In L. M. Bambara & L 1. Kern (Eds.) Individualized supports for students with problem behaviors. (pp.35990). New York: Guilford Press. Horner, R. Sugai, G., Eber, 1., & Lewandowski H. (2004). Illinois Positive behavior interventions and support project: 2003-2004 Progr e ss Report. University of Oregon: Center on Positive Behavior Interventions and Support & Illinois State Board of Education. Hughes, A. F., & Adera, B. (2006). Education and day treatment opportunities in schools: Strategies that work. Preventing School Failure 51, 26-30. Irwin, 1. K. Tobin T. J Sprague J R., Sugai, G. & Vincent C G (2004) Validity of office discipline referral measures as indices of school-wide behavioral status and effects of school-wide behavioral interventions. Journal of Positive Behavior Interventions, 6(3) 131-147. Kaufinan P., Bradbury D., & Owings, J. (1992). Characteristics of at-risk students in NELS:88. Nathional eduation longitudinal study of 1988. Contractor report. Berkely, CA. Kaufinan A., & Kaufinan F. (2004). Kaufinan test of educational achievement (2nd ed.) Minneapolis, MN: AGS Publishing. Kaufinan, A., & Kaufinan, F. (2005). Kaufinan test of educational achievementBrief Edition. Minneapolis MN: AGS Publishing. Landrum, T. J., & Kauffinan, J. M. (2006) Behavioral approaches to classroom management. In C. M. Evertson & C. S. Weinstein (Eds ), Handbook of classroom management: Research practice and contemporary issues. Mahwah, NJ: Erlbaum. Lee, Y., Sugai, G. & Homer, R. H. (1999). Using an instructional intervention to reduce problem and offtask behaviors. Journal of Positive Behavior Interventions, 1(4), 195-204. Lee, J. & Ware, B. (2002). Open sourc e d e velopment with LAMP: Using linux, apache mySQL, perl, and PHP. Boston MA: Addison-Wesley Professional 33

PAGE 43

Running Head: REFERENCES Leech, N. L., Barrett K., & Morgan G. A. (2011). SPSSjor intermediate statistics: Use and intelpretation (4th ed.). New York, NY: Taylor and Franci s Lehr, C. A., & Lange C. M. (2003). Alternative schools and the stu-dents they serve: Perception s of state directors of special edu-cation. Policy Research Brief (University of Minnesota: Minne-apolis, Institute on Community Integration) 14(1). Lewis-Palmer, T. Sugai G., & Larson, S. (1999). Using data to guide decisions about program implementation and effectiveness. Effective School Practices 17(4) 47-53. Luiselli J., Putnam R., & Handler M. (2001). Improving discipline practices in public schools: Description ofa whole school and district wide model of behavior analysis consultation. The Behavior Analyst Today 2(1), 18-26. Luiselli, J. K., Putnam R. F., & Sunderland M. (2002). Longitudinal evaluation of behavior support intervention in a public middle school. Journal of Positive Behavior Interventions, 6(3), 182-188. Luiselli J. K., Putnam, R F., Handler, M. W., & Feinberg A. B. (2005). Whole-school positive behavior support: Effects on student discipline problems and academic perfonnance Educational Psychology 25(2-3), 183-198. 81 Lewis-Palmer T. Homer, R. H., Sugai, G., Eber, L. & Phillips D. (2002). Illinoi s Positiv e Behavior Int erventio n s and Support Project: 200/-2002 Progress R eport. University of Oregon: OSEP Center on Positive Behavior Support. McClure, C. T. (2009). Behavior needs school wide effort. National staff development council: Teachers teaching teachers. Research Brief. May 2009 pp. 8-9 Retrieved from www.nsdc.org McIntosh K. (2005, March). Use of D1BELS ORF traj ec tories to predict office discipline referrals. Paper presented at DIBELS Summit 2005 Ratin N. M. McKinney, J. D. (1989). Longitudinal resea rch on the behavioral characteristics of children with learning disabilities. Journal of Learning Disabilities, 22(3). 141-150, 165. McLeod S. A. (2015). Skinner Operant Conditioning. Retrieved from www.simplypsychology.orgloperant-conditioning.html Muscott, H. (2006). Implementing PBIS with fidelity in PBIS NH schools. [PowerPoint Slides). Retrieved from www.nhcebis.seresc.net Nelson, J.R., Benner, G. J., Lane, K. & Smith, B. W. (2004). Academic achievement ofK-12 students with emotional and behavioral disorders. Exceptional Children, 71(1),59-73. 34

PAGE 44

Running Head: REFERENCES Nelson, C. M., Sprague, J. R., Jolivette, K., Smith C. R., & Tobin, T. J. (2009). Positive behavior support in alternative education, community-based mental health and juvenile justice settings. In G. Sugai, R. Homer, G. Dunlap and W. Sailor (Eds.), Hand-book of positive behavior support (pp. 465-496). New York NY: Springer. Nelson J. (1996). Designing schools to meet the needs of students who exhibit disruptive behavior. Journal of Emotional and Behavioral Disorders 4,147161. No Child Left Behind (NCLB) Act of2001, Pub. L. No. 107-110 115, Stat. 1425 (2002). National School Climate Center (NSCC). (2013). How do we define school climate? Retrieved from http :// www.schoolclimate.orglclimate / Oliver R. M., Wehby J. H. Reschly, D. J. (2011) Teacher classroom management practices Effects on disruptive or aggressive student behavior. Campbell Systematic Reviews 201104. doi I OA073 / esr.20 I 104 Putnam R. F Handler M., & O Leary-Zonarich C. (2003). Improving academic achiev e m e nt using school-wide behavioral support interventions. Paper presented at the Annual Conference of the Association of Behavior Analysis. San Francisco CA. Putnam R. F Handler, M., Rey .I., & O'Leary-Zonarich, C. (2002). Classwide behavior support interventions: Using functional assessm e nt practices to design effective interv e ntions in general classroom settings. Paper presented at the Annual Conference of the Association of Behavior Analysis. Toronto Canada. Putnam, R. F Homer, R. H. Algo z zine R. (2006). Academic achievement and the implementation of school-wide behavior support. Positive Behavioral Supports Newsletter, 3(1), 1-2. Reynolds R., & Kamphaus C. (2003). Re y nold s Int e ll ec tual A s ses s m e nt S ca le s and R ey nold s In te ll ec tual S c re e nin g T es t Pr o f ess ional Manual. P sy chological A ss es s ment Roid G. (2003) Stanford Bin e t Intelligence Scale s, Fifth Edition. Roeser, R. W., Eccles, J. S. & Sameroff, A. J. (2000). School as a context of early adolescents' academic and social-emotional development: a summary of research findings. The El e m e ntm y School Journal, 100(5),443-471 Simonsen B., Britton, L., & Young D (2010). School-wide posi-tive behavior support in a non-public school setting: A case study. Journal of Positive Behavior Interventions 12, 180-191. doi: I 0.117711 098300708330495 Simonsen B. & Sugai, G. (2013). PBIS in alternative education settings: Positive support for youth and high-risk behavior. Education and Tre atment of Childr en. 36(3), 3-14. doi: 10.1353 / etc.2013.0030 35

PAGE 45

Running Head: REFERENCES Skinner, B.F. (\950). Are theories oflearning necessary? Psychological Review, 57(4) 193-216. Skinner, B.F. (\954). The science oflearning and the art of teaching. Harvard Educational Review, 24(2), 86-97. Skinner, B.F. (1968). The Technology of Teaching. New York: Appleton-Century-Crofts. Sprague, J., Walker, H., Golly, A., White, K., Myers, D., & Shannon, T. (2001). Translating research into effective practice: The effects of universal staff and student intervention on indicators of discipline and school safety. Education &Treatment of Children, 24, 495 511. Stevens, J. (2007). Intermediate statisitcs: A modern approach (3fd ed.). New York, NY: Taylor & Francis Staddon, J. (1995) On responsibility and punishment. The Atlantic Monthly, Feb., 88 94 Sugai, G., Lewis-Palmer, T., Todd, A. & Homer, R. (2001). School-wide evaluation tool. University of Oregon. Sugai, G., & Homer, R. H. (1999). Discipline and behavioral support: Preferred processes and practices. Effective School Practices, 17(4), 10-22. Sugai, G., Homer, R. H., Dunlap, G., Hieneman, M., Lewis, T. J., Nelson, c., et al. (2000). Applying positive behavior support and functional behavioral assessment in schools. Journal of Positive Behavior Interventions, 2(3), 131143. Sugai, G., Homer, R. H., Dunlap, G. (2003). Effective behavioral support (EBS) survey version 2.0. Educational and Community Supports: University of Oregon. Tobin, T., & Sugai, G. (1999). Predicting violence at school, chronic discipline problems, and high school outcomes from sixth graders' school records. Journal of Emotional Disorders. 7, 40-53. Todd, A., Haugen, 1., Anderson, K., & Spriggs, M. (2002). Teaching recess: Low cost efforts producing effective results. Journal of Positive Behavior Interventions, 2, 233-245. Wilkinson G. S., & Robertson, G. J. (2006). WIde range achievement test-fourth edition. Lutz, FL: Psychological Assessment Resources Wechsler, D. (2005). Wechsler Individual Achievement Test 2nd Edition (WIAT II). London: The Psychological Corp. 36

PAGE 46

Running Head: REFERENCES Wech s ler D. (2004). The W ec h s l er intelligence sca l e for children -foUl1h edi tion. L o nd o n : Pear so n Assessment. Wech s ler D. (2008). The Wechsler abbreva it ed sca l e of int e lli ge n ce. London: Pe a r so n Assess ment. Woodcock, R. & Johnson M. (1977). WoodcockJohnson te s t s of ac h aiv ment. Rive r s i de Publi s hing Y i M. S. (2012). Evaluati on of therapeutic progres s in a t-ri sk yout h in a behavioral da y treatment schoo l progr am. Orange, CA. C h apman Un i ve r s it y. 37

PAGE 47

Running head: APPENDIX A APPE NDIX A Research Design The purpose of this study was to determine if a positive behavior and intervention supports model curriculum (PBIS) had an effect on student academic achievement test score outcomes and school wide number of behavioral incidents over a two-year period. The data collected was based on the first and second year outcomes of academic achievement tests scores and number of behavioral incidents varied by the types ofPBIS interventions offered. PBIS curriculum framework provides the systems and tools for establishing a continuum of evidence based practices in any school setting with the goal of assisting in meeting the particular academic and / or behavior support needs of students (Simonsen & Sugai 2013). In the first year tier one, all school personal in the school utilized universal supports. In the second year after receiving training, all school personal utilized individualized and targeted tier two and three supports. The following sections will describe this study s methodological approach subjects and sampling the types of data that was collected, the sources and instruments utilized for data collection, and the methods in which the data was analyzed to explore the following questions: Is there an overall difference between year one and year two student academic post test scores? Is there a difference in the total number of behavioral incidents between year one and year two? Is there a difference between pre and post student academic test scores in each year, when comparing a low or high number of behavioral incidents? Methodological Approach A quasi-experimental approach was be utili zed to investigate the research questions. Quantitative research approaches are viewed to be objective in nature (Gliner at el., 2009). The term objective implies that behaviors are easily classified and that data are usually collected with 38

PAGE 48

Running head: APPENDIX A so me sort of instrument (Gliner at el., 2009). Quasi-experimental research de s ign s contain both indep e ndent and dependent variables (Gliner at el., 2009). Active independent variables are defined as a treatment such as a workshop new curriculum, or other intervention for a specific amount of time throughout the duration of the study. Dependent variables are the measure or the outcome, such as test scores, behavioral changes, or other effect related outcomes (Gliner at el. 2009). Additionally, quasi-experimental designs can be divided into four different categories ba se d on the basic design of the study. For the purposes of this study a quasi-experimental one group pretest-posttest design was used This type of design requires that an initial observation (pretest) is given then the intervention or independent variable is applied and finally a second observation (posttest) is applied (Gliner at el., 2009). The statistical analysis utili zed in this study was a mixed ANOV A for research question one and a dependent/paired samples I test for research question two and three. This analysis was conducted in SPSS. Participants and Site This study focused on the documented behaviors and academic test scores of day treatment students who attended a Colorado approved and eligible facility school. This school is both an eligible site and approved s ite due to the populations being served and the multiple services being provided and billed for. Below the basic purpose of an eligible and approved facility school is discussed to describe environment the data was originally collected in. Site The miss ion of an Eligible Facility Schools is to assure that all students in placement receive a qualit y education and to improve educational outcomes for tho se s tudents (Colorado Department of Education, 2014). The tenn Eligible Facility is defined as a day treatment center residential childcare facility or a hospital licensed by either the Department of Human Services 39

PAGE 49

Running head: APPENDIX A or the Department of Public Health and Environment. Approved Facility School refers to an educational program that is operated by a licensed agency and that has been approved to receive reimbursement for education services for students placed in the program. In the spring of 2008 legislation was passed that addressed approved facility schools in a number of ways. Commonly referred to as house bill 1204 (22-2-401 C R S.), it established a board with rule making authority to set graduation requirements and the approval process for facility schools (http: // www.cde state.co.us / facilityschoolslhb081204act). Day treatment programs, which are viewed as a form of alternative education options have targeted students with behavioral problems who have been unsuccessful in traditional educational settings. (Yi 2012). Participants Day treatment students are commonly referred to as at risk youth. An "at-risk student is generally defined as a student who is likely to fail at school (Kaufman at el., 1992). In this context school failure is typically seen as dropping out of school before high school graduation (Kaufman at el., 1992). Research has indicated that students with disruptive behavior disorders are at risk for numerous adjustment problems during adolescence which include violence and juvenile delinquency (Yi 2012). These various adjustment problems can result in academic failure, substance abuse, risky sexual behavior and antisocial behavior (Broidy et a!., 2003; Yi, 2012). Thus resulting in the need for an alternative educational option such as a day treatment program (Yi, 2012). The day treatment students in this sample are male students between the ages of nine to seventeen who have been diagnosed with various behavior disorders including, attention deficit / hyperactivity disorder (ADHD), oppositional defiant disorder (ODD) and conduct disorder (CD), and a variety of additional behavior disorders, which are typically associated 40

PAGE 50

Running head: APPENDIX A with behavioral social emotional, academic impainnents. As defined by the American Psychiatric Association's Diagnostic and Statistical Manual of Mental Disorders, fourth edition text revision (DSM-IV-TR; APA 2000), Attention-Deficit/Hyperactivity Disorder (ADHO) i s one of the most common and pervasively impairing childhood psychiatric disorders. ADHD and CD are characterized by the primary features of inattention, hyperactivity and impulsivity and is associated with other disruptive beha v iors ODD is characterized by patterns of negativistic, disobedient hostile and defiant behavior toward authority figures that persists for at least 6 months and is characterized by the frequent behaviors of temper loss argument s with adults defiance or refusal to comply with adult instructions, and deliberate actions of annoying others (DSM-IV-TR, APA 2000). Measures This study consisted of two dependent variables and one active independent variable. The researcher relied on previously collected data from two database s Variables In quantitative research variables are the key elements in research questions. Variables are defined as the characteristics of situation s or p a rticipants in any given study having different values (Gliner at el., 2009). This s tudy measured the characteristics of two dependent variables. Dependent variables are defined as the assumed measure or effect of the independent variable (Gliner at el., 2009). Independent and dependent variables can be divided into two t y pes ; ac tive or attribut e (Gliner at el., 2009). For the purpo ses of this study an active independent va riable s was used An active independent variable is defined as variable containing an intervention such as a new curriculum (Gliner at el., 2009). Dependent Variables 41

PAGE 51

Running head: APPENDIX A There were three dependent variables in this study. Two of the dependent variables were academic test scores measured at student intake and discharge creating the pretest and posHest scores in the content areas of reading comprehension and math. The tenns intake and dis charge were used in this academic setting due to the untraditional nature of a student's length of stay in the school. In a more traditional setting students are tested at the beginning of the school year and end of the school year in a day-treatment setting the students are tested when they arrive at the school (ie intake) and when they leave the school (ie., discharge) this time frame is typically less then a school year. The scale of measurement for this variable was an interval measurement. An interval measurement was utilized when the difference between two values is meaningful (Stevens 200 7 ) These test scores were measured by the Wide Range Test of Achievement (WRA T-IV) assessment. The third dependent variable was documented behavioral incident reports collected by a Lamp Stack operating system. These variables are discussed in detail below. Wide Range Test of Achievement (WRAT-IV) The pretest-posttest scores are derived from the Wide Range Test of Achievement (WRA T-IV) a s sessment, which is administrated to students upon intake and again when students discharge. The scores provided by the WRA T4 are scaled scores indicating grade level equivalency scores in the content areas of reading comprehension and fluency, spelling, and math. For the purposes of this study the reading comprehension and math scores was examined. The WRA T4 i s designed to provide a quick, simple psychometricall y sound assessment of academic skill s" (Wilkinson & Robertson, 2006 p. 3). Joseph F. Jastak first published the test in 1946 with the purpose of augmenting the cognitive perfonnance measures of the Wechsler Bellevue Scales developed by David Wechsler (Dell, Harrold & Dell 2008). Jastak (1946) 42

PAGE 52

Running head: APPENDIX A believed that academic perfonnance should also be considered during a cognitive assessment battery, since then the WRAT has been used as a nonn-referenced measure of basic academic skills in the areas of reading, spelling and mathematical calculation s (Dell, Harrold & Dell 2008) The WRA T4 development team utilized a stratified quota-based sampling procedure to standardize the WRA T4 these results were based on the 2001 U.S. Census to develop a nonn referenced assessment. The sample size was 3,007,with 100 to 150 participants matched across nineteen age groups (Wilkinson & Robertson, 2006). The samples were also matched to census data by gender, race / ethnicity, educational attainment and geographic region Educational attainment was used to indicate socioeconomic status and it was based on years of school completion for individuals who were eighteen and older. The four census categories were employed: college graduate some college (no degree) high school graduate only, and less than a high school diploma. Again for those younger than age eighteen, parent educational attainment was used (Wilkinson & Robertson 2006). For grade-based samples which are the fonnat pre and posttest scores for this study was reported, individuals with educational disability classifications were included in the nonning process. The following disabilities were represented in the nonning process: specific learning disabilities, speech and language impainnents developmentally disabled emotional disturbance physical impainnents (including hearing, orthopedic visual, and other health impainnents), and attention-deficitlhyperactivity disorder (Wilkinson & Robertson 2006). The representation of individuals with various types of disabilities were not adequately accomplished during the standardi z ation process because visually hearing, and physically disabled children had a lower percentage of representation when compared to that of the National Center for Education Statistics reports for 2002 (Dell Harrold & Dell 2008) 43

PAGE 53

Running head: APPENDIX A WRATI V Validity and Reliability The assessment developers conducted an analysis in reliability for the WRA T4 which includes internal consistency and alternate-fonns (Wilkinson & Robertson, 2006). The WRA T4 overall has high levels of internal consistency, ranging from 92 to .98. The data from the assessment was also found to have moderate levels of internal consistency within its subtests with reliability coefficients ranging from .87 to 93. For example, the reading composite score coefficients are high ranging from .95 to .96 on both the blue and green fonns (Wilkinson & Robertson 2006). Test validity refers to how closely scores of a given test are related to those of another established test based on the same criteria (Gay, 1992; Wilkinson & Robertson 2006). To explore the validity of the WRA T4 test results the subtests were compared to several tests of academic achievement These comparison tests included the WRAT-Expanded (Wilkinson & Robertson 2006) Wechsler Individual Achievement Test (Wechsler, 2005). Woodcock Johnson IJI, (Woodcock & Johnson 1977) Kaufman Test of Educational Achievement-Comprehensive Fonn (Kaufman & Kaufman 2004) Kaufman Test of Educational Achievement-BriefFonn (Kaufman & Kaufman, 2005) Wide Range Intelligence Test, Wechsler Intelligence Scale for Children-Fourth Edition (Wechsler 2004), Stanford-Binet Intelligence Scale-Fifth Edition (Roid, 2003) Wechsler Abbreviated Scale ofIntelligence (Wechsler 2008) and the Reynolds Intellectual A s sessment Scales (Reynold s & Kamphaus, 2003) The comparison study found low to moderate relationships between the results / scores from all assessments pertaining to full-scale IQ and subtest scores concluding that the WRA T4 is a valid and reliable assessment (Wilkinson & Robert s on, 2006). Documented Behavioral Inciden t Reports (IR) 44

PAGE 54

Running head: APPENDIX A The documented behavioral incident reports (lR) are forms that teachers paraprofessionals and various adult-care takers in the school environment submit each time a student has a behavioral incident. This process is used to track behaviors, report individual behavior trends to a student s caregivers and containment team a s well as to evaluate the student' s s afety level. The behavioral incident s that require documentation in this day-treatment setting include physical threatening, verbal threatening physical aggression, assault police contact suicidal ideation absent without official leave (AWOL) police contact in possession of contraband, and inappropriate talk including discussions about drugs colluding sexual comments and other harmful statements. The documented behavioral incident reports are stored and tracked in a privately and independently developed electronic tracking system developed by the school to store personal student information. The system adheres to all privacy rules and regulations. This system allows users to track pull and group reports for individual students and school wide tracking purposes. When building the system, the private programmer utilized the LAMP stack database technology which is basically the combination of open source technologies that together produce an application-serving platform. Lamp is an acronym for Linux, Apache Web server MySQL database and Perl Python or PHP (Lee & Ware 2002). These four applications make up the LAMP stack. There have been government-funded analysis completed to determine the reliability of the LAMP stack software which established baseline for software quality and security in open source based on sophisticated analyses of more than 17.5 million lines of source code being used (Lee & Ware 2002). The measurement utilized for this variable was nominal. A nominal measurement is used when categories are being numbers but not ordered (Stevens 2007). 4S

PAGE 55

Running head: APPENDIX A Documented Behavioral Incident Reports (IR) Validity and Reliability Teachers and para-professional employed in the school receive a va riet y of trainings pertaining to documenting incidents and including pertinent information how to access the system, and the importance of communicating incidents to the team members. In addition, employees also receive a variety of trainings pertaining to being an informed s upervisor of atriskstudents s potting problematic behaviors and crisis intervention strategies. While behaviors are mor e s ubjective based on the individual observing and documenting the incident most documented reports will contain elements of the school defined behavioral categories discu sse d above. Additionally, all reports are read and approved by a school administrator thus creating reliable and valid information to examine. Independent Variables The independent variable for this study was the implementation ofPBIS curriculum and supp0!1S over a twoyear period. During year one, the 2013-2014 school year tier one s upports were put into place throughout the sc hool environment for all students. In year two the 20142015 school year tier two and three s upports were put into place throughout the school environment. The table below briefl y depicts PBlS tier one, two, and three int erven tions. PBIS tier interventions Tier Tier I Tier 2 Interventions School-wide positive and proactive interventions implemented by all staff to support student behavior across all settings, which may include: Establishing, teaching prompting and monitoring student behavior with respect to a few positive settingand class-wide expectations Frequent and explicit schooland class-wide social skills instruction School-and / or class-wide student recognition systems (e.g. point card or check-inlcheck-out intervention) Continuum of responses for inappropriate behavior that include an instructional focus Targeted or intensified positive and proactive interventions Implemented by staff to support targeted-students' behavior across all settings which may include: 46

PAGE 56

Running head: APPENDIX A Tier 3 Additional t eac hing prompting and monitorin g with respec t to positive setting-wide expectations More frequent o r explicit social skills ins truction Additional mentoring and s tructure provided within the sett ing-wide point or check-inlcheck-out intervention (e.g., individuali ze d goals additional check-ins) Increased ins tructional support related to chronic social-behavior errors Individuali zed and intensive positive and proactive intervention s based on a functional-behavioral assessment and implement e d to support individual st udent s behavior which may include: Antecedent strategies that include environmental changes and added prompting for replacement and desired behavior Ins truction a l s trategie s to explicitly teach replacement b e h av ior( s) and a plan for s haping toward desired behaviors Consequence s trategie s that provid e functionall y appropriate reinforcement for replacement behavior(s) increase reinforcement for desired behaviors and prevent or reduce reinforcement currently maintaining problem behaviorC s ) (Simonsen & Sugai 2013) Data Collection Procedure All schools receiving any s tate o r federal funds are required to collect data for reporting purp oses. This s tudy utili z ed t wo data so urc es collected from an alternative school setting. The tw o samples were pre and po s t academic te s t sco re s and documented beha v i o r incidents. The sec tion s below will dis cu ss the s p ec ific s about h ow each sample was collected at th e s ite. Additionally the chart below depi c t s the data being collected from the sample, as well as, the duration and frequency of collecti o n Artifact and data collection Artifacts / data Year I number of individual student documented incidents Year 2 number of individual s tud e nt do c ument e d incidents Year I pret es t scor e s Rdg Year I pretest scores Math Time Period Anticipated Sample Size Aug. 2013May 2014 n =53 collected from Aug 2014May 201 5 n = 54 collected from Aug. 2013April 2014 n =53 Aug. 2013April 2014 n =53 47

PAGE 57

Running head: APPENDIX A Year 1 posttest scores Rdg Sep. 2013May 2014 n =53 Year 1 posttest scores Math Sep. 2013May 2014 n =53 Year 2 pretest scores Rdg Aug. 2014April 2015 n = 54 Year 2 pretest scores Math Aug. 2014April 2015 n = 54 Year 2 posttest scores Rdg Sep. 2014May 2015 n = 54 Year 2 posttest scores Math Sep. 2014May 2015 n = 54 Year !total number of documented Aug. 2013May 2014 behavioral incident reports n = 808 collected from Year 2 total number of documented Aug. 2014May 2015 behavioral incident reports Test Scores n = 1,126 collected from The pre and posttest scores are collected on site when a student comes to the school and then again when they leave the school. In alternative school settings the length of stay for a student is not a typical school year, as such, to measure academic growth students are given an assessment to track their progress for their length of stay. The instrument used to obtain the test scores is the WRAT-IV, which is a nonn-referenced measure of basic academic skills in the areas of reading spelling and mathematical calculations (Dell Harrold, & Dell 2008). The WRAT-JV tests are complied on a working Excel document by the school. This document is updated throughout the year and closed on the last day of the school calendar. Each year is tracked in the same fonnat. Documented Behavior Incidents The documented behavioral incident reports (IR) data sample are electronic fonns that teachers paraprofessionals and various adult-care takers in the school environment submit to the LAMP based tracking system (Lee & Ware, 2002) each time a student has a behavioral incident. 48

PAGE 58

Running head: APPENDIX A The purpose of the documentation is to track behaviors report individual behavior trends to a student s caregivers and containment team, a s well as to evaluate the student's safety level. The behavioral incidents that are documented and will be included in this study include physical threatening, verbal threatening, phy s ical aggression, assault police contact suicidal ideation, absent with official leave (AWOL) police contact, in possession of contraband, and inappropriate talk including discussions about drugs, colluding sexual comments, and other harmful statements. The electronic tracking system has the ability to sort and categorize these documents to meet the needs of this study. Data Analysis This study used a repeated measures approach to analyze the data. A repeated measures design is defined as a design that measures subjects several times, either on the same dependent variable or on different measures (Stevens, 2007). The data being utilized in this study was analyzed in two different ways as to answer all of the research questions To better explain the analysis plan the following sections below discuss the analysis plan for each research question. A mixed ANOV A analysis was utilized to analy z e research question one. A mixed ANOV A compares the mean differences between groups that have been split on two independent variables (Gliner at el., 2009). The primary purpose of a mixed ANOVA is to understand if there is an interaction between the independent variables on the dependent variable (Gliner at el., 2009). Additionally a dependent/paired samples I test was used to analy z e research questions two and three. A dependent or paired samples I-test is used to compare whether two groups have different average values (Gliner at el., 2009). Research Questions One and T wo Research questions one and two both utili zed a dependent / paired samples I-test to analyze 49

PAGE 59

Running head: APPENDIX A the data and answer the questions. The first research question explored was: Is there an overall difference between year one and year two student academic post test scores ? For this research questions the data utili z ed consisted of posttest scores from year one and year two. The second research question explored was: Is there a difference in the total number of behavioral incidents between year one and year two? This research question utilized the total number of behavioral incidents from year one and two. Research Question T hree The third research question explored was: Is there a difference between pre and post student academic test scores when factoring in a low or high number of behavioral incidents? This question utilized the individual pre and post test scores for each student in year one and two as well as individual high or low number of behavioral incidents in year one and year two. The high or low number of behavioral incidents was determined by identifying the median score of total behavioral incidents and categori z ing high behaviors above the median and low behaviors below the median. For the purposes of entering the high-low behaviors into SPSS the low behaviors were labeled with a 0 and the high behaviors were labeled with a I. This data was analyzed using mixed ANOV A. Checking Assumptions Research questions one and two had two assumptions that need to be checked. The first was checked upon the completion of the data collection process the dependent variable should be approximately normally distributed for each group of the independent variable (Leech at el. 2011). This assumption was violated among the math pre and posttest scores for both years as well as among the documented behavioral incidents for both years. Prior to running the t test 50

PAGE 60

Running head: APPENDIX A these variables were transfonned using a LoglO transfer in SPSS. The second assumption was met the independent variable was dichotomous, meaning they were paired or matched, in some way (Leech at el. 2011). For research question three the nonnality homogeneity and sphericity assumptions need to be checked when conducting a mixed ANOYA analysis (Leech at el. 2011). These assumptions were tested using SPSS and all were met. Nonnality implies that there were no significant outliers in any group from the within-subjects independent variables or between subjects independent variables (Leech at el. 2011). Outliers are simply single data points within your data that do not follow the usual pattern (Leech at el., 2011). Homogeneity of variances (Leech at el. 2011) was tested using the Levene s test for homogeneity of variances (Leech at el., 2011). Sphericity, was the variances of the differences between the related groups of the within-subject independent variables for all groups of the between-subjects independent variables and dependence between pairs of groups is roughly equal (Leech at el. 20 I I). Additional Analysis conducted An additional t test was conducted after view the descriptive statistic findings to detennine ifthere was a significant difference between the pre and posttest scores for both year one and year two. For the analysis the assumption of approximately nonnally distributed data for each group of the independent variable (Leech at el., 2011) was checked and violated among the math pre and posttest scores for both years, as well as among the documented behavioral incidents for both years. The transfonned variables from the LoglO transfer in SPSS for research questions one and two were utili ze d for this I test. The second assumption was met the independent variable was dichotomous meaning they were paired or matched, in so me way (Leech at el. 20 I I) 51

PAGE 61

Running H ead: APPENDIX B APPENDlXB T-TEST PAIRS=yearlreadingposttestmathltransform behaviorltransformWITH year2 readingposttest math2transform behavior2transorm(PAIRED) jCRITERIA=CI(.9500) j MISSING=ANALYSIS. T-Test Output Created Comments Input Missing Value Handling Syntax Resources Notes Data Active Dataset Filter Weight Split File N of Rows in Working Data File Definition of Missing Cases Used Processor Time Elapsed Time 52 28JUL-201622:22 :03 IUsers/jmtaylor/disserta tion.sav DataSet1 ------- User defined missing values a r e treated as missing 58 Statistics for each analysis are based on the cases with no missing or out-of-range data for any variable in the analysis. T-TEST PAIRS=year1 readingpos ttest math1transform behavior1 transform WITH year2readingposttest math2transform behavior2transorm (PAIRED) ICRITERIA=CI(.9500) IMISSING=ANAL YSIS. 00 : 00 :00.01 --00:00 :00.00

PAGE 62

Running Head: APPENDIX B Paired Samples Statistics Std Error Mean N Std. Deviation Mean Pair 1 year1 rdg posltest 7.023 53 2 .9849 .4100 year 2 rdg posltest 7.496 53 3 1914 .4384 Pair 2 math1 transform .7522 53 .19961 .02742 math2transform .7105 53 .22600 .03104 Pair 3 behavior1 transform 1.9734 53 .00234 .00032 behavior2transorm 2.0124 53 .00216 .00030 Paired Samp l es Correlations Pair 1 year1 rdg postlest & year 2 rdg posttest Pair 2 math1transform & math2transform --Pair 3 behavior1transform & behavior2transorm Pair 1 year1 rdg postlest -year 2 rdg posltest Pair 2 math1transform -math2transform N Correlation Sig 53 119 .394 53 255 .065 53 -.004 .980 Paired Samples Test Paired Differences Mean Std Deviation 4736 4.6229 .04178 .33752 Std Error Mean .6350 .04636 95% Confidence .. Lower -1.7478 -.05125 ---------------------Pair 3 behavior1transform behavior2transorm Pair 1 year1 rdg postlest year 2 rdg posltest Pa i r 2 math1transform rnath2transform Pair 3 behavior1transform-behavior2transorm -.03905 .00319 Paired Samples Test Paired ... 95% Confidence Interval of the ... Upper .8006 .13481 .03817 53 -.746 .901 -89 .235 .00044 -.03993 df Sig (2-tailed) 52 .459 52 .372 52 .000

PAGE 63

R u nn in g Head: APPENDIX B MEANS TABLES=yearlreadingposttestyear2readingposttestmathltransform math2tra nsform behaviorltransformbehavior2transorm /CELLS=MEAN STDDEV SKEW VAR. M ea n s Output Created Com ments Input Notes 28-JUL 201622:20 :10 --------Data IUsers/jmtaylor/disserta tion.sav .-... ---------Active Dataset Filter DataSet1 __ .. __ ._ .. ..:.=::.::.:.::c.-. ___ Weight _______ ......c.:.-'-"-'____ Split File N of Rows in W orking Data File 58 Missing Value Handling Definition of Missing For each depende n t variable in a table user-defined miss i ng values for the dependent and all grouping vari ab l es a r e treated as missing Cases Used Syntax Resources Processor Time Elapsed Time 54 Cases used for each table have n o miss i ng values in any independent variable and not all dependent variables have missing values MEANS TABLES=year1 readingp osttest year2readingposttest math1transform math 2transform behavior1transform behavior2transorm ICELLS=MEAN STDDEV SKEWVAR. 00:00: 00 .01 00:00:00.00

PAGE 64

R unning Head: APPEND I X B Case P r ocessing Summary Cases Included Excluded Total N year1 rdg pastiest 53 year 2 rdg pastiest 54 math1transform 53 math2transform 54 behavior1 transform 53 behavior2transorm 54 Mean Std. Deviation Skewness Variance Mean Std. Deviation Skewness Variance year1 rdg pastiest 7 .023 2 9849 55 4 8 909 behavior2tran sorm 2.0125 00219 252 .000 Percent 9 1. 4 % 93 1 % 91.4% 93 1 % 91. 4 % 93 1 % year 2 rdg pastiest 7.469 3 1677 .251 10 034 N Percent N Percent 5 8 6 % 58 100.0 % 4 6 9 % 58 100.0 % 5 8.6% 58 100.0% 4 6.9 % 58 100 0 % 5 8.6 % 58 100.0 % 4 6.9 % 58 100.0 % Rep ort math1transfor math2transfor behavior1 tran m m sform 7522 .7107 1 973 4 .19961 22387 0023 4 049 772 053 -------.040 .050 000 Repo rt GLM yearlreadingpretestyearlreadingposttestBYyearlhighlow /WSFACTOR=testscores 2 Polynomial /MEASURE=highlow /METHOD=SSTYPE(3) /PRINT=DESCRIPTIVE ETASQ OPOWER PARAMETER HOMOGENEITY / CRITERIA=ALPHA (.05) /WSDESIGN=testscores /DESIGN=yearlhighlow. G e n e r a l Lin ear M o del Output Created Comments Notes 1 4-AUG2016 16: 16 : ... 55

PAGE 65

R u nning Head : APPEND I X B inp u t -Data Active Dataset Filter Weight Split File N of Rows in Working Data File Missing Value Handling Definition of Missing Syntax Resources Within-Subjects Factors Cases Used Processor Time Elapsed Time Measure: highlow Dependent testsco Variable res 56 IUsers/jmtay l or/disserta tion.sav DataSet1 ------ 58 User-defined missing values are treated as missing Statistics are based on all cases with valid data for all variables in the model. GLM year1 readingpretest year1 readingposttest BY year1highlow IWSFACTOR=testscores 2 Polynomial IMEASURE=highlow IMETHOD=SSTYPE(3) IPRINT=DESCRIPTIVE ETASQ OPOWER PARAMETER HOMOGENEITY ICRITERIA=ALPHA(.05) IWSDESIGN=testscores IDESIGN=year1 high low. 00 :00:00. 04 00:00: 00.00

PAGE 66

Running He ad: AP P END IX B 2 year1 reading pre1est year1 reading posllest year one high -low 0 1 Measure: h ighlow Between-Subjects Factors Valu e Label N low high 36 17 Mauchly's Test of Sphericity a Approx. Chi-Within Subj ec t s Effect Mauchly's W Square df Sig testscores Measure: highlow Within Subjects Effect tests cores 1.000 000 o Mauchly's Test of Sphericity a Epsilonb Huynh-Feldt Lower-bound 1 000 1 000 Epsilonb G reen h ouse Geisser 1 000 Tests the null hypothesis that the error covariance matrix of the orthonormalized transformed dependent variables is proportional to an identity matrix. a Design : Intercept + year1 highlow Within Subjects Design:testscores b May be used to adjust the degrees of freedom for the averaged tests of significance. Corrected tests are displayed in the Tests of Within-Subjects Effects table Source te sts cores Measure, highlow Tests of Within-Subjects Effects Type III Sum of Squares df Sphericity Assumed GreenhouseGei sser Huynh-Feldt Lower-bound 23 064 1 23 064 1 000 23.064 1 000 tests cores year1 highlow Sphericity Assumed Greenhouse-Geisser 23. 064 137 .137 1.000 1 1 000 Huynh-Feldt 137 1.000 57 Mean Square 23 064 23 064 23 064 23. 064 .137 137 .137

PAGE 67

Lowerbound 137 1 000 137 Error(testscores) Sphericity Assumed 61. 006 51 1 196 Greenhouse-Geisser 61. 006 51.000 1 196 Huynh-Feldt 61. 006 51. 000 1.196 -_. --Lower-bound 61. 006 51. 000 1.196 5 8

PAGE 68

Measure: highlow T es t s o f Within-Subj ects E ffect s Partial Eta Source F Sig. Squared tests cores Sphericity Assumed 19 .281 000 .274 Greenhouse-Geisser 19 .281 000 .274 Huynh-Feldt 19 .281 000 274 19 .281 .000 274 -----. testscores year1 high low Sphericity Assumed .115 .736 .002 Greenhouse-Geisser 115 .736 .002 Huynh-Feldt 115 .736 .002 -----Lower-bound .115 .736 .002 ----_. Error(testscores) Sphericity Assumed Greenhouse-Geisser -----_. Huynh-Feldt Lower-bound Measure: highlow Tests of Within-Subjec t s Effects Source testscores Sphericity Assumed Greenhouse-Geisser Noncent. Pa rameter 19.281 19 .281 19.281 -. -----Huynh-Feldt Lower-bound tes t scores year1 high low Sphericity Assumed Error(testscores) Greenhouse-Geisser Huynh-Feldt Lower-bound Sphericity Assumed Greenhouse-Geisser Huynh-Feldt Lower-bound a. Computed using alpha = .05 59 19.281 .115 115 115 115 -------_. Observed Power a .991 .991 .991 .991 --. 063 063 063 .063

PAGE 69

Measure: highlow Tests of With i n Subjects Contrasts Type III Sum of Source tests cores Squares df Mean Square testscores Linear 23.064 1 23.064 ._------testscores year1 highlow Linear 137 1 137 _. __ ._-Error(testscores) Linear 61.006 51 1.196 Measure: highlow T ests of Wit hinS ub jects Contras t s Partial Eta Noncent Source tests cores Sig Squared Parameter testscores Linear 000 .274 19.281 testscores year1 highlow Linear 736 .002 115 ----------. --------Error(testscores) Linear a. Computed using alpha = .05 L evene's Test of Equality of Erro r Variances a year 1 rdg pretest yearl rdg postlest F 625 .851 df1 1 1 df2 Sig 51 .433 51 .361 Tests the null hypothesis that the error variance of the dependent variable is equal across groups. a. Design: Intercept + yearlhighlow Within Subjects Design: testscores Measure: highlow Tests of Between-S u b j ects Effects Transformed Variable: Average Type III Sum of Source Squares df Mean Square F Intercept 4 131 664 1 4131 66 4 303 .413 yearlhighlow 16.379 1 16 379 1 .203 Error 694.482 51 13 617 60 Sig 000 278 F 19.281 115 Observed Power a .991 .063 Partial Eta Squared .856 ----. 023

PAGE 70

Running Head: APPENDIX C APPEXDIXC Introduction This study will be an examination of the effects of the implementation of School wide Positive Behavior support systems for at-risk youth in a behavioral day treatment school setting. For he purposes of this study at-risk students are identified as students with disruptive behavior disorders that include Attention Deficit Hyperactive Disorder (ADHD), Conduct Disorder (CD), and Oppositional Defiant Disorder (ODD). School-wide Positive Behavior Support (PBIS) is a systems approach to establishing the social culture and behavioral supports needed for all children in a school to achieve both social and academic success. PBIS is not a packaged curriculum, but an approach that defines core elements that can be achieved through a variety of strategies and supports. The literature reviewed pertaining to PBIS examined its effectiveness on student achievement the behavioral outcomes and climate of a school, and the at-risk population of students being served. The research indicated that there is a need for further research to examine the academic and behavioral outcomes of at-risk youth in alternative settings. Additionally, considerations should be taken when reviewing outcome results of PBIS when taking into account teachers satisfaction level and opinions of PBIS. A quasi-experimental approach will be utilized to investigate the research questions being explored in this study. Quantitative research approaches are viewed to be objective in nature (Gliner at el., 2009). For the purposes of this study a quasi-experimental onegroup, pretest posttest design will be used. This type of design requires that an initial observation (pretest) is given, then the intervention or independent variable is applied, and finally a second observation 61

PAGE 71

Running head: APPENDIX C (posttest) is applied (Gliner at el. 2009). There will be two statistical analyses utilized in this study. They will be a mixed ANOV A and a dependent t test. TIns analysis will be conducted in SPSS. Conceptual Framework This study aims to detennine if a positive behavior and intervention supports model curriculum (PBlS) has an effect on student academic achievement test score outcomes and school wide number of behavioral incidents over a two-year period. This study will focus on B.F. Skinner's (1950) operant conditioning theory as a basis of measurement due to the similarities with the PBIS programming Operant conditioning and PBlS similarities in theory and practice will be discussed below. PBIS School-wide Positive Behavior Interventions and Support (PBIS) is a systems approach to establishing the social culture and behavioral supports needed for all students in a school to achieve both social and academic success (Homer at el., 2004). PBlS is an approach that defines core elements of behavioral and student struggles that can be addressed through a variety of tiered supports and strategies. The core elements of the three tiers in the prevention model are outlined below: PBIS Tiers and Core Elements Prevention Tier Primary Secondary Core Elements Behavioral expectations are defined and taught Reward system is established for appropriate behaviors Consequence continuum is established for problem behavior. Continuous data collection and planning takes place Universal screening Progress monitoring for at risk students System for increasing structure and predictability System for increasing contingent adult feedback System for linking academic and behavioral perfonnance System for increasing home / school communication 62

PAGE 72

Running Head: APPENDIX C Tertiary Collection and use of data for decision-making Functional Behavioral Assessment Team-based comprehensive assessment Linking of academic and behavior supports Individualized intervention based on assessment information focusing on (a) prevention of problem contexts (b) instruction on functionally equivalent skills and instruction on desired perfonnance skills (c) strategies for placing problem behavior on extinction, (d) strategies for enhancing contingence reward of desired behavior and (e) use of negative or safety consequences if needed. Collection and use of data for decision-making (Homer at el., 2004) To sum up the table outlined above the idea of PBIS is for a school team to collect data to analyze behavioral and academic trend s The team plans and implements group and individualized intervention based on assessment information focusing on (a) prevention of problem contexts, (b) instruction on functionally equivalent skills, and instruction on desired perfonnance skills (c) strategies for placing problem behavior on extinction, (d) strategies for enhancing contingence reward of desired behavior, and (e) use of negative or safety consequences if needed (Homer at el. 2004). B.F. Skinner s Operant Conditioning Theory and PBIS interventions share a similar model and participation planning. Operant Conditioning Operant conditioning has been widely applied in clinical settings for behavior modification as well as, teaching for classroom management and instructional development for programmed instruction (Skinner 1950). The operant conditioning theory ofB.F. Skinner is based upon the idea that learning is a function of change in overt behavior (Skinner, 1950). Changes in behavior are the result of an individual's response to events (stimuli) that occur in the environment. When a particular Stimulus-Response (S -R) pattern is reinforced or rewarded the individual is conditioned to respond (Skinner 1954). 63

PAGE 73

Running head: APPENDIX C Stimulus response in Skinner's S-R pattern theory and the PBIS prescribed group and individual interventions, strategies, and supports follow the same processes. The reinforcement s and rewards of the S-R pattern and PBIS are discussed further below. Reinforcement and Rewards Reinforcement is the key element in Skinner's S-R theory (Skilmer 1968) A reinforcer is anything that strengthens the desired re spo nse. It could be verbal praise, a good grade or a feeling of increased accomplishment or satisfaction (Skinner, 1968). The theory a lso covers negati ve reinforcers which are any stimulus that results in the increased frequency of a respon se when it is withdrawn (Skinner 1968). A great deal of attention was given to schedules of reinforcement and their effects on establishing and maintaining behavior. In past years, school-wide discipline has focused mainly on reacting to spec ific student misbehavior by implementing punishment-based strategies including reprimands, loss of privileges, office referrals, s u s pensions, and expulsions (Sugai & Homer, 1999). Meaning that puni s hment utili ze d to promote desired behaviors especially when it is used incon s istently is ineffective Introducing modeling and reinforcing positive social behavior is an important step of a s tudent's educational experience (Sugai & Homer 1999). The purpo se of sc hool-wide PBIS i s to establish a climate in which appropriate behavior i s the nonn (Barrett at el., 2008). The PBIS strategy focuses on identifying problem behaviors and the environments in which they occur in and then providing programming that involve s teaching and modeling desired behaviors (Barrett at el., 2008). Then when students demonstrate the desired behaviors they receive positive reinforcement through a variety of s timuli Teaching behavioral expectations and rewarding students for following them is a much more positive 64

PAGE 74

Running Head: APPENDIX C approach than waiting for misbehavior to occur before responding thus creating a schoo l environment geared towards success (Barrett at el., 2008). Critical Evaluation of Operant Conditioning T heo ry Operant condi tionin g can be used to explain a wide variety of behaviors from the process oflearning to l anguage acquisition. Operant conditioning also has practical applications such as token economies whic h can be applied in classrooms prisons and psychiatric hospitals and are also utilized among PBIS programming plans (McLeod, 2015). Operant conditio ning has been disputed and critiqued over the years. Some of the best known critics include Noam Chompsky (1959) and J.E .R. Staddon (1995) who both questioning the methods used to develop this theory. Additionally, later cri tiqu es discuss that the theory fails to take into account the role of inherited and cognitive factors in learning, thus making the theory an incomplete explanation of the l earning process (McLeod, 2015). Lastly the use of animal research in operant conditioning studies also raises the issue of extrapo l ation (McLeod 2015). Some psychologists argue that generalizations cannot be made from studies on animals to humans because their anatomy and physiology are different from humans (McLeod 2015). There are v isible simi l arities between operant conditioning Theory and PBlS programming which despite the critiques makes this an appropriate perspective for this study There are a variety of different positive reinforcements prescribed in both operant conditioning Theory and PBIS planning used to increase the likelihood of desired behavior in the classroom. Some of the most wide l y used reinforcements include co n sumab l e rewards positive social interactions, and earned activities. According to Landrum and Kauffman (2006) Despite a rich history and extensive empirica l underpirmings the behavioral perspective on teaching and management i s not highly 65

PAGE 75

Running head: APPENDIX C regarded in the education community" (2006, p. 47). Critics argue that operant conditioning is an unfeeling approach more suited to animals than to humans (Landrum & Kauffinan 2006). Regardless of the critiques, operant conditioning is commonly used in classrooms and is viewed by many teachers as an effective approach to improving classroom practice. It provides teachers with a set of tools for improving classroom management and student leaming (Landrum & Kauffman 2006). In addition the underlying purpose of PBlS programming follows operant conditioning theory closely, but also applies the element of data collection. The data collected in PBIS programmed schools measures student behaviors while demanding tiered intervention programming changes if desired behaviors are not exhibited, making the connection between operant conditioning and PBIS programming relevant to this study. Literature Review The purpose of this study is to determine if a positive behavior and intervention supports model curriculum (PBIS) has an effect on student academic achievement test score outcomes and school wide number of behavioral incidents over a two-year period. The data collected will be based on the first and second year outcomes of academic achievement tests scores and number of behavioral incidents varied by the types ofPBIS interventions offered. PBIS curriculum framework provides the systems and tools for establishing a continuum of evidence-based practices in any school setting with the goal of assisting in meeting the particular academic andlor behavior support needs of students (Simonsen & Sugai 2013). For the purposes of this study, this literature review will examine research pertaining to problem behaviors and how they relate to academic achievement PBlS related to academic achievement behavioral outcome s and school climate and how it relates to at-risk youth. The parameters for this research were set for research conducted in 1997 to 2015. These dates were 66

PAGE 76

Running Head: APPENDIX C chosen because PBIS was established in 1997 PBIS Development and Implementation In order to describe the development and implementation ofPBIS, the historical context around the program will be briefly discussed. During the 1990s the term positive behavior support became popular in school systems. It referred to behavior interventions or strategies that could theoretically be used to reduce problem behavior and assist in promoting desirable behavior (Dunlap et a!., 2000). In 1997 there were amendments made to the Individuals with Disabilities Education Act (IDEA). One important aspect of these amendments was the concept of positive behavior support (PBS) for students whose behaviors violated school rules or was "outside personal or interpersonal norms of acceptable social behavior" (Sugai et ai, 2000 p. 131). Meaning, legally, if a student who has been identified as having a disability displays behaviors that affect his or her learning or others' learning then that child's Individualized Education Program (JEP) team, including teachers, para-professionals, and those interacting with the student must include positive behavior interventions or supports to address the behavior (Sugai et aI, 2000). The implementation of IDEA was the main contributing factor in the developments on PBIS, but another factor also contributed to the development of PBIS years later in 2001. President George W. Bush signed the No Child Left Behind Act (NCLB, 2002). This act changed the federal government's involvement with public education and included a component of PBIS implementation for students (Sugai et ai, 2000). Schools have had practices in place to deal with problem behaviors for many years (Sugai et a!., 2000). Legal policies and amendments have contributed and affected the development of many behavior programs. Research pertaining to the link between academic achievement and 67

PAGE 77

Running head: APPENDIX C problem behaviors has also sparked the continued development and implementation of PBIS (Sugai et aI., 2000) Academic Achievement and Problem Behavior Higher rates of office discipline referrals (ODRs) are associated with problematic behavioral climates in schools (Irwin at el. 2004). Mcintosh (2005) found relationships between academic perfonnance and problem behavior across grade levels. Mcintosh (2005) investigated how early screening measures pertaining to assessments in kindergarten targeting behavior and reading outcomes predicted if a student would have two or more discipline contacts in the third and fifth grade. He found that office referrals (ODRs) in first and second grade were predictors ofODRs in third grade (Mcintosh 2005). His results also indicated that reading competence in kindergarten, measured by the Dynamic Indicators of Basic Early Literacy Skills (DIBELS) were also linked to ODRs in 3rd grade. Mcintosh (2005) found that overall predictors of students receiving two or more discipline contacts in fifth grade were linked to their fourth grade ODRs and low DIBELS Oral Reading Fluency scores. The overall findings provided by Mcintosh's study were that children start their kindergarten school year with varying reading skills. However, if they do not respond to literacy instruction during kindergarten and fall behind a negative spiral of achievement and behavior becomes more likely (McIntosh, 2005). If a student's literacy skills do not keep pace with peers, academic tasks become more difficult, and problem behaviors that lead to escape from these academic tasks become more likely. There have been several studies conducted on the topic oflinking academic perfonnance and problem behaviors. The following studies discussed briefly discuss the similar findings over the years. This relationship between academic performance and problem behaviors has also been 68

PAGE 78

Running Head: APPENDIX C studied a t the middle schoo l and high sc hool levels. Tobin and Sugai (1999) found that individual student academic failure in high school was linked with three or more suspensions in ninth grade. They also found links between grade point average (GPAs) and specific types of office discipline referral (ODR) behaviors, such as fighting, harassing and threats of violence, nonviolent misbehavior. Morrison, Anthony Storino and Dillon (200 I) reviewed the records of stu dents who were referred to an in-school s uspension program. Those students who had no previous ODRs had higher GPAs than the students who had ODRs. Roeser Eccles, and Sameroff (2000) found the relationship between academic performance and behavior strengthened in middle school by reviewing ODR documents and suspensions. Other research conducted by Nelson, Benner, Lane, & Smith (2004) found that students with documented problem behavior, such as ODRs experienced large academic deficits in reading and math when compared to same age peers. Also that externalizing or acting out behaviors were strongly related to academic performance deficits when compared to internali z ing behaviors. In a similar study, Harachi Cortes Abbott, and Catalano (2004) found that s tudent s with higher reading scores in elementary school and students whose sco res increased between third and sixth grade demonstrated significa ntl y less problem behavior in seventh grade. Another study b y Lee Sugai & Horner (1999) found improvements in escape maintained or acting out problem behavior when students received academic support that made them effective with academic tasks. Research pertaining to the link between academic outcomes and problem behaviors has created a need for continued research and exp l oration on the affect PBIS has on academic outcomes. School-wide Positive Behavior Support and Academic Achievement 69

PAGE 79

Running head: APPENDIX C There have been a number of s tudie s con ducted that explore how school-wide behavior s upports also known as positive behavior intervention s and supports (PBIS) decrea se problem behavior increase time spent in academic ins truction and seem to be connected with impro ve d academic outcomes. The following sections below will discuss these studies. Improving academic outcomes Luiselli Putnam and Sunderland (2002) discovered that after the implementation of sc hool-wide behavior support (PBIS) in a suburban middle school detention s for disruptive behavior and ODRs decreased over a four-year period School attendance also increased over the four year period. The reward for meeting predetermined academic criteria, such as maintaining a specific grade point average, receiving pa ss ing grades for all subjects on the report card and having no more than two homework detentions as well as, behavioral, attendance, detentions expulsions was a lottery drawing that was conducted each quarter. The percent of s tudents who were eligible for the lotter y increased from 40% of the sc hools population to 55% of the sc hools' population over the course of four yea rs. In another study Luiselli Putnam Handler and Feinberg (2005) implemented a sc hool wide beha vio r support plan at an urb a n school and found decreases from bas e line to intervention to follow-up in documented behavior s and suspensions. Reading comprehension and mathematics percentile ranks on standardized tests improved from the first to the second test d ates, with an increase of 18 and 25 percentage points. In another s tudy Putnam, Handler & O Leary-Zonarich (2003) discovered that reading and math scores improved on standardized te s ting following the implementation of behavior s upport intervention at an urban elementary sc hool. 70

PAGE 80

Running Head: APPENDIX C A recent analysis of academic performance of schools that are implementing school wide positive behavior support compared to schools not implementing such programs was conducted in Illinois (Homer Sugai Eber & Lewandowski 2004). Schools that were implementing PBIS had scored 80% on the School Evaluation Tool, a tool utilized by PBIS to measure the success of implementation in school systems and assist in developing improvement plans (Sugai Lewis Palmer Todd & Homer 2001). Schools implementing PBIS also had 80% of their students being able to discuss their school wide expectations and rules (Sugai, et aI., 2001). In addition the schools (n=52) in which school-wide positive behavior supports were implemented had 62% of their 3rd grade students meeting the Illinois State Achievement Test Reading Standard (Sugai et aI., 200 I). In comparison, only 47% of students met the Illinois State Achievement Reading Test Standard in schools (n = 69) that had not fully implemented positive behavior support (Sugai, et aI., 200 I) Homer Sugai Todd and Lewis-Palmer (2005) discovered similar findings with another school district with nineteen elementary schools. Between the 1997-98 and 2001-2002 academic years, thirteen of the schools implemented school-wide positive behavior support and six schools did not. They compared the percentage of 3rd grader s who met state wide reading standards in the academic year 1997-98 with the percentage in the academic year 2001 2002. Their findings concluded that ten out of the thil1een schools that implemented PBIS practices had improved outcomes. The overall increase in percentage of students meeting standards ranged from 2% to over 15% in these schools. Only one of the six schools that not did implement school-wide positive behavior support showed improvement. School Wide Positive Behavior Support and School Culture 71

PAGE 81

Running head: APPENDIX C School culture has been defined as the belief system that directly influences school climate (Sugai, 2013). In a positive school climate students feel comfortab l e, valued, accepted wanted and secure in a positive environment where they can interact with people whom they trust. Climate reflects the positive or negative feelings toward the school environment. School culture refers to the manner in which teachers and staff members work together, while school climate refers more toward the school s effects on students. ODRs have been primarily used in PBlS research to track behavioral and school climate outcomes (Sugai 2013). Teachers are key s takeholders in implementing PBlS. If the sc hool staff do not support or buy in" to the program, the effectiveness ofPBIS will be compromised. Research has shown that PBlS can be an effective behavioral intervention program ; however there is limited research on how teachers perceive PBlS and its impacts on teacher motivation and satisfaction (Sugai 2013). Anecdotal evidence would suggest that PBlS schools with reduced referrals and discipline issues have better teacher retention and higher satisfaction thus l eading to a positive school climate (Sugai 2013). Oliver Wehby & Reschly (20 II) found that teachers who experience difficulty controlling classroom behavior have higher stress and burnout making it more difficult for them to meet the instructional demands of the classroom. Effective approaches to managing the classroom environment are necessary to establish environments that support student behavior and the learning process as well as to reduce teacher stress and burnout. Research has suggested that schools can support classroom teachers with PBlS by focusing on prevention ; u s ing multiple data so urce s to develop strategies for screening identification and treatment and taking a coordinated schoo l -wide approach to reducing problem behaviors among students (Oliver et aI., 2011). The purpo se ofPBlS was to establish a climate in which appropriate behavior is the nonn. 72

PAGE 82

Running Head: APPENDIX C In the past, school-wide disci pline has focused on reacting to implementing punishment-based strategies including reprim a nd s, l oss of pri v ileg es, office referral s s uspen s ions and expulsions for misbehavior. McClure (2011) reviewed s tudies supporting PBIS and found s everal studies which associated PBIS with decreases in office discipline referrals (OORs) as well as incre ase d consistency and positive interaction s a mong school staff. There have been many studies conducted that have indic ate d that schools implementing PBlS have significant reductions in ODRs data (Nelson, 1996 ; Sprague, et aI., 2001). Luiselli Putnam and Handler (2001) indicated a 69% reduction in OORs after the implementation of PBIS in their study Similarly, Todd, Haugen Anderson and Spriggs (2002) indicated an 80% reduction in ODRs in the first year of PBIS implementation and a 76% reduction in the second year in their review of behavior documentation. More recently Bradshaw and Leaf (2008) indicated reduced OORs data as well as improved perceptions of sc hool safety among teachers and staff in the M ary land school sys tem. Positive Behavior Supports and At-Risk Youth There are a large number of youth educated in restrictive or alternative education (AE) se ttings. AE schools and programing include tho se housed in juvenile detention centers (Carver Lewis, & Tice, 2010). Estimates suggest that between 12% and 50% of the se youth have disabilities and most youth are placed in restrictive settings as a result of significant behavior challenges (Carver, et aI., 20 I 0). Public school district s report tran s ferring youth to AE settings for a variety of reasons, including physical aggression (61 % of districts) ; "disruptive verbal behavior (57%); "possession, distribution, or use of controlled substances (57%) ; chronic academic failure (57%) or truancy (53%); posse ss ion or u se offireanns (42%) or other weapons 73

PAGE 83

Running head: APPENDIX C (51 %); "arrests or involvement with the criminal justice system" (42%); teen parenthood (31 %); and /or mental health needs 27% ; (Carver et aI., 20 I 0 p. II). Therefore AE settings need to be able to support youth with a variety of behavioral needs and challenges as well as meet individual academic and behavioral needs. The overall empirical research on the presence and effectiveness of behavior support practices / PSIS in AE settings is limited (Flower, McDaniel & Jolivette 2011; Lehr, 2004) There is however initial research that suggests behavior management practices in these settings may be more punitive than positive (Lehr & Lange, 2003). When youth displaying high-risk behaviors are educated together in an AE setting it is a common misconception by teachers and school staff that all of the students will require tier three supports. Instead, PSIS experts suggest that all three tiers are necessary elements for the successful implementation of PSIS (Nelson et aI., 2009). Unfortunately, the current existing research on PSIS and at-risk youth focuses on suggestions for implementation and data collection to provide tiered supports rather than outcomes. Youth that are unresponsive to tier 1 practices may require additional tier 2 practices to be added such as including an individualized goal on a youth's school-wide point card providing additional adult mentoring and support to enhance social skills instruction and developing a menu of more individualized reinforcements. For youth whose behaviors are unresponsive to tier 2, individualized tier 3 practices may be added. Tier 3 practices should be based on a full functional behavioral planning (Eber Sugai Smith & Scott 2002). Following these implementation practices in an AE setting, suggests that a PSIS framework may result in positive outcomes for youth educated within AE settings including increases in appropriate behavior decreases in problem behaviors, and decreases in use of crisis-emergency re s ponses, 74

PAGE 84

Running Head : APPENDIX C such as re s traint (Simonsen Young, & Britton, 20 I 0). In addition s ingle-case design studies have demonstrated that targeted-group intervention s, s uch as check-inlch eck -out have promise in AE settings (Ennis Jolivette Swoszowski & John so n 2012) Thus emerging evidence supports the implementation of intensified proactive and positive practices within a PBIS framework to support youth in AE settings. Conclusion In summary, this literature review has discussed research pert a ining to PBIS and examined its effectiveness on student achievement, the behavioral outcomes and climate of a school and the at-risk population of students being served. The resear c h indicates that there is a need for further research to examine the academic and beha v ioral outcomes from PBIS u se of at risk youth in a lternative settings. Additionally, considerations should be taken when reviewing outcome results of PBIS when takin g into account teachers satisfaction level and opinions of PBIS Research Design The purpo se of this study is to determine if a po sitive behavior and intervention s upports model curriculum (PBIS) has an effec t on student acade mic achievement te s t sco re outcomes and school wide number of behavioral incidents over a two-year period The data collected will be based on the first and second year outcomes of academic achievement tests scores and number of behavioral incidents varied by the types of PBIS int erve ntion s offered. PBlS curriculum framework provides the systems and tools for establishing a continuum of evidence-based practices in any school setting with the goal of assisting in meeting the particular academic and/or behavior support need s of s tudents (Simonsen & Sugai 2013). In the first year tier one all school personal in the school utili ze d universal supports. In the second year after receiving 75

PAGE 85

Running head : APPENDIX C training all school personal utili zed indiv i duali zed and targeted tier two and three s upports. The following se ctions will de scr ibe thi s s tudy's methodological approach subjects and sampling the types of data that will be collected, th e sources and instruments utilized for dat a collection, and the methods in which the data will be analyzed to answer the following questions: Is there an overall difference between year one and year two student academic post test scores? Is there a difference in the total number of behavioral incidents between year one and year two ? I s there a difference between pre and post st udent academic test scores in each year when comparing a low or high number of behavioral incidents ? 76

PAGE 86

Running Head: APPENDIX C References American Psychiatric Association (2000). Diagnostic and s tati s ti cal manual of m e ntal disorders (4th ed., text rev.). Washington DC Bradshaw C., & Leaf, P. (2008). Project Target: Update on Key Findings from Project Target. Barrett S., Bradshaw, C., & Lewis-Palmer, T. (2008). Maryland state-wide PBIS initiative. Journal of Positive Behavior interve ntions. 10, 1005-1 14. Carver P. R., Lewis L., & Tice P (2010). Alternative Schools and Pro-grams for Public School Students At Risk of Educational Failure: 2007-08 (NCES 2010-026). U.S. Department of Education, Na-tional Center for Education Statistics. Washington DC: Goverrunent Printing Office. Chomsky Noa m (1959). "Reviews: V e rbal b e havior by B. F. Skinner". Language 35 (I): 26-58. JSTOR411334. Colorado Department of Education. (2008). H.B. 1204 (22 2-401 C.R.S.) http :// www.cde.state.co.us / facilityschoolslhb081204act Colorado Department of Education. (2014). Retrieved from: http: // www.cde.state.co.us / facilityschools Dell C.A., Harrold B., & DeB, T. (2008). Test r eview: Wide rang e achievement test (4th ed.). Lutz FL : Psychological Assessment Resources. Eber, L., Sugai G. Smith C. R., & Scott T M. (2002). Wraparound and positive behavioral interventions and supports in the schools. Journal of Emotional and Beh av ioral Disorders 10, 171-181. Ennis R P., Jolivette, K., Swos zo wski N. c., & Johnson M. L. (2012): Secondar prevention e fforts at a residential facility for stu-dents with emotional and behavioral disorders: Function-based check-in, check-out, Residential Treatment for Children & Youth 29, 79-102. Fleming C. B., Harachi, T. W. Cortes, R. C. Abbott R. D. & Catalano, R. F. (2004). Level and change in reading scores and attention problems during elementary school as predictors of problem behavior in middle school. Journal of Emotional and B e havior a l Disorders, 12(3) 130-144 Garrison R. W. (1987). Alternative sc h oo l s for disruptive yo uth: NSSC r eso ur ce paper. Malibu, CA: National Safety Center, Pepperdine University, Office of Juvenile Ju s ti ce and D e linqu e nc y Pr eve ntion Gliner, J. A., Morgan, G. A., Leech N. L. (2009). R esea rch methods in applied setti ngs: An 77

PAGE 87

Running head: APPENDIX C integrat e d approa c h t o de s ign and analysi s (2nd ed.). New York NY: Taylor and Francis. Horner, R. H., Todd, A., Lewis-Palmer, T., Irvin, L., Sugai, G & Boland, J. (2004). The school-wide evaluation tool (SET): A research instrument for assessing school-wide positive behavior support. Journal 0/ Positive B e havior Int e nlention 6( 1) 3-12. Horner, R. H., Sugai G., Todd A. W. & Lewis-Palmer T. (2005) School-wide positive behavior support: An alternative approach to discipline in schools. In L. M. Bambara & L L. Kern (Eds.) Individualized supports(or stud e nts with problem behaviors. (pp 35990). New York: Guilford Press. Horner, R Sugai, G., Eber L., & Lewandowski, H. (2004). Illinois Positive Behavior Interv e ntions and Support Project: 2003-2004 Progress Report. University of Oregon: Center on Positive Behavior Interventions and Support & Illinois State Board of Education. Hughes, A. F., & Adera, B. (2006). Education and day treatment opportunities in schools: Strategies that work. Preventing School Failure 51, 26-30. Irwin, L. K., Tobin, T. J. Sprague J. R., Sugai G. & Vincent C. G. (2004) Validity of office discipline referral measures as indices of school-wide behavioral status and effects of school-wide behavioral interventions. Journal 0/ Positive B e havior Interv e ntions, 6(3) 131-147. Kaufman, P., Bradbury D., & Owings J. (1992). Characteristics 0/ at-risk stud e nts in NELS:88. Nathional e duationlongitudinal study 0/1988. Contractor report. Berkely CA. Kaufinan A., & Kaufman F. (2004). Kaufman Test of Educational Achievement (2nd ed.) Minneapolis MN: AGS Publishing. Kaufinan, A., & Kaufman, F (2005). Kaufinan Test of Educational Achievement Brief Edition. Minneapolis MN: AGS Publishing. Landrum T. J., & Kauffman, J. M. (2006). Behavioral approaches to classroom management. In C. M. Evertson & C. S. Weinstein (Eds.), Handbook 0/ classroom management: Research practice and c ontemporOlY issues. Mahwah, NJ: Erlbaum. Lee Y., Sugai G. & Horner R. H. (1999). Using an instructional intervention to reduce problem and offtask behaviors. Journal 0.( Positive B e havior Inten1 entions, 1 (4) 195-204. Lee J. & Ware B. (2002). Op e n source d e velopment with LAMP: U s ing linux apache, m y SQL p e rl and PHP. Boston MA: Addison-Wesley Professional Leech, N. L., Barrett, K., & Morgan, G. A. (2011). SPSS/or interm e diate statistics: Use and interpr e tation (4th ed.). New York NY: Taylor and Francis 78

PAGE 88

Running Head: APPENDIX C Lehr C. A. & Lange C. M. (2003). Alternative schools and the s tu-dents they serve: Perceptions of state directors of special edu-cation. Policy Research Brief (University of Minnesota: Minne-apolis Institute on Community Integration) 14(1). Lewi s -Palmer, T ., Sugai G., & Larson S. (1999). Using data to guide decision s about program implementation and effectiveness. Effective School Practices 17(4) 4753. Luiselli J. Putnam, R., & Handler, M. (2001). Improving discipline practices in public schools: Description of a whole school and district wide model of behavior analysis consultation. The Behavior Analyst Today, 2(1) 18-26. Luiselli, J. K., Putnam R. F. & Sunderland, M. (2002). Longitudinal evaluation of behavior support intervention in a public middle school. Journal of Positive Behavior Interventions 6(3) 182-188. Luiselli J. K., Putnam R. F., Handler, M. W. & Feinberg A. B. (2005). Whole-school positive behavior support: Effects on student discipline problems and academic performance. Educational Psychology 25(2-3) 183-198. 81 Lewis-Palmer, T., Horner R. H., Sugai G., Eber, L. & Phillips D. (2002). Illinois Positiv e Behavior Interventions and Support Project: 2001-2002 Progress Report. University of Oregon: OSEP Center on Positive Behavior Support. McClure C. T. (2009). Behavior needs school wide effort. National staff development council: Teachers teaching teachers. Research Brief. May 2009, pp. 8-9 Retrieved from www.nsdc.org Mcintosh K. (2005 March). Use of DIBELS ORF traj e ctories to predict office disciplin e referrals. Paper presented at DIBELS Summit 2005 Ratin N. M. McKinney, J. D. (1989). Longitudinal research on the behavioral characteristics of children with learning disabilities Journal of L e arning Disabilities 22(3) 141-150,165. McLeod S. A. (2015). Skinner Operant Conditioning. Retrieved from www.simplypsychology.orgloperant-conditioning.html Muscott, H. (2006). Implementing PBIS with fidelity in PBIS NH schools. [PowerPoint Slides]. Retrieved from www.nhcebis.seresc.net Nelson J R., Benner G. J., Lane K. & Smith, B. W. (2004). Academic achievement ofK-12 students with emotional and behavioral disorders Exceptional Children 71(1),597 3 Nelson, C. M., Sprague J. R., Jolivette, K. Smith, C. R. & Tobin T. J. (2009) Positive behavior support in alternative education, community-based mental health and juvenile justice settings. In G. Sugai, R. Horner, G. Dunlap, and W. Sailor (Eds ) Hand-book of positive behavior support (pp. 465-496). New York NY: Springer. 79

PAGE 89

Running head: APPENDIX C Nel son, J. (1996). Designing sc hools to meet the need s of students who exhibit dis ruptive behavior. Journal of Emotional and Behavioral Disorders 4 147-161. No C hild Left Behind (NCLB) Act of 200 I Pub. L. No. 107-110 115, Stat. 1425 (2002). National School Climate Center (NSCC). (2013). How do we define school climate ? Retrieved from http ://www.schoolclimate.org/climate / Oliver, R. M. Wehby J. H., Reschly D. J. (2011). Teacher classroom management practices Effects on disruptive or aggressive student behavior. Campbell Systematic Reviews 2011.4. doi I 0.4073/esr.20 11.4 Putnam R. F Handler M., & O'Leary-Zonarich, C. (2003). Improvin g academic achievement using school-wide behavioral support interventions. Paper presented at the Annual Conference of the Association of Behavior Analysis. San Francisco, CA. Putnam R. F., Handler M. Rey 1., & O'Leary-Zonarich, C. (2002). C lass w id e behavior support interventions: Using functional assessment practice s to design effective inten1entions in general classroom setti ngs. Paper presented at the Annual Conference of the Association of Behavior Analysis. Toronto C anada. Putnam R. F., Homer, R. H., Algozzine, R. (2006) Academic achievement and the implementation of school-wide behavior support. Positive Behavioral Supports Newsletter, 3(1),1-2. Reynolds, R., & Kamphaus, C. (2003). R ey nolds In tellectual Assessment S c ales and Re y nold s Int ellectual Screenin g Test Profess ional Manual. Psychological Assessment Roid G. (2003) Stanford Binet Intelligenc e Scale s, Fifth Editi on. Roeser, R W. Eccles, J. S. & Sameroff, A. J. (2000). School as a context of early adolescents academic and social-emotional development: a summary of research findings. The Elementary School Journal 100(5) 443-471 Simonsen, 8., Britton, L., & Young, D. (2010) School-wide posi-tive behavior support in a non-public school setting: A case study Journal of Positive Behavior Interventions, 12, 180-191. doi:l0.117711098300708330495 Simonsen 8., & Sugai G. (2013). PBIS in alternative education settings: Positive s upport for youth and high-risk behavior. Education and Treatment of C hildren, 36(3), 3-14. doi: 10. I 353/etc.2013 0030 Skinner, B.F (1950). Are theorie s oflearning nece ssa ry ? Psychological Review 57(4) 193-216 Skinner, B .F. (1954). The sc ience oflearning and the art of teaching. Harvard Educational Review 24(2) 86-97. 80

PAGE 90

Running Head: APPENDIX C Skinner B.F. (1968). The Technology of Teaching. New York: Appleton-Century-Crofts. Sprague, J., Walker, H., Golly, A., White, K., Myers, D., & Shannon T. (2001). Translating research into effective practice: The effects of universal staff and student intervention on indicators of discipline and school safety. Education &Treatment of Children, 24, 495 511. Stevens J (2007) Interm e diate statisitcs: A mod e rn approach (3'd ed.). New York NY: Taylor & Francis Staddon J. (1995) On responsibility and punishment. The Atlantic Monthly, Feb., 88-94 Sugai, G., Lewis Palmer, T., Todd A. & Homer, R (200 I). School-wide eva luation 1001. University of Oregon Sugai, G., & Homer, R. H (1999). Discipline and behavioral support: Preferred processes and practices. Effective School Practices 17(4) 10-22. Sugai G., Homer R. H., Dunlap, G., Hieneman, M., Lewis, T. J., Nelson, C., et al. (2000). App l ying positive behavior support and functional behavioral assessment in schools. Journal of Positive Behavior Interventions 2(3) 131143. Sugai, G., Homer, R. H., Dunlap, G. (2003). Effective behavioral support (EBS) survey version 2.0. Educational and Community Supports: University of Oregon. Tobin T. & Sugai G. (1999). Predicting violence at school, chronic discipline problems, and high school outcomes from sixth graders' school records. Journal of Emotional Disorders. 7,40-53. Todd, A Haugen L., Anderson, K., & Spriggs M. (2002) Teaching recess : Low cost efforts producing effective results. Journal of Positive Behavior Interventions 2, 233-245. Wi l kinson, G. S., & Robertson, G. J. (2006). Wide rang e achiev e m e nt testfourth edition. Lut z, FL: Psychological Assessment Resources Wech s ler D. (2005). Wech s ler Indi v idual Achie v emen t Test 2nd Edition (WIAT II). London: The Psychological Corp. Wech s l er, D. (2004). The Wechsler intelligence scale for children fourth ed ition. London: Pearson Assessment. Wechsler, D. (2008). The W ec hsl e r abbrevaited sc ale of intelligen ce. London: Pear son Assessment. 81

PAGE 91

Running head : APPENDIX C Woodcock R. & Johnson M. ( 1 977). Woodcock-Johnson te s t s of acha i v m ent. Riverside Publi s hing. Yi M S. (2012) Evaluation of therapeutic progress in at-risk youth in a behavioral day treatment sc hool program Oran ge, CA. Chapman University. 82