Citation
Effects of participation in a guided reflective writing program on middle school students academic self-efficacy and self-regulated learning strategy use

Material Information

Title:
Effects of participation in a guided reflective writing program on middle school students academic self-efficacy and self-regulated learning strategy use
Creator:
Anderson, Katheryn L
Publication Date:
Language:
English
Physical Description:
xv, 166 leaves : ; 28 cm

Subjects

Subjects / Keywords:
English language -- Composition and exercises -- Study and teaching (Middle school) ( lcsh )
Self-efficacy ( lcsh )
Learning strategies ( lcsh )
Sex differences in education ( lcsh )
English language -- Composition and exercises -- Study and teaching (Middle school) ( fast )
Learning strategies ( fast )
Self-efficacy ( fast )
Sex differences in education ( fast )
Genre:
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )

Notes

Bibliography:
Includes bibliographical references (leaves 160-166).
General Note:
School of Education and Human Development
Statement of Responsibility:
by Katheryn L. Anderson.

Record Information

Source Institution:
|University of Colorado Denver
Holding Location:
Auraria Library
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
57495051 ( OCLC )
ocm57495051
Classification:
LD1190.E3 2004d A52 ( lcc )

Full Text
EFFECTS OF PARTICIPATION IN A GUIDED REFLECTIVE WRITING
PROGRAM ON MIDDLE SCHOOL STUDENTS ACADEMIC SELF-
EFFICACY AND SELF-REGULATED LEARNING
STRATEGY USE
by
Katheryn L. Anderson
B.A., University of Maryland, 1973
M.A., Adams State College, 1977
M.A., University of Colorado, Colorado Springs, 1987
A thesis submitted to the
University of Colorado at Denver
in partial fulfillment
of the requirements for the degree of
Doctor of Philosophy
Educational Leadership and Innovation
2004


2004 by Katheryn L. Anderson
All rights reserved


This thesis for the Doctor of Philosophy
degree by
Katheryn L. Anderson
has been approved
by
VsatH&tn

Date


Anderson, Katheryn L. (Ph.D., Educational Leadership and Innovation)
Effects of Participation in a Guided Reflective Writing Program on Middle
School Students Academic Self-Efficacy and Self-Regulated Learning
Strategy Use
Thesis directed by Associate Professor Lyn Taylor
ABSTRACT
This six-week, quasi-experimental study examined the effects of
instruction in a guided reflective writing program and gender on middle school
students development of academic self-efficacy beliefs and self-regulated
learning strategies use. The data collected included: (a) a pretest-posttest
self-efficacy inventory (b) responses to writing prompts, (c) audio-taped
interviews, and (d) observations. The intentions of this study were: (a) to
determine if instruction in guided reflective writing would help students
activate self-referent processes to self-regulate learning, (b) to determine how
reflective writing impacts students use of self-regulation of learning
strategies, and (c) to determine if there are gender differences in self-referent
processes and self-regulated learning strategy use.
The data from the pre-test and post-test self-efficacy inventory were
analyzed using a repeated measures factorial ANOVA. There was statistical
significance for the dependent variable attributions for success and failure,
There were no statistically significant results for the dependent variables goal
orientation, intrinsic motivation, or persistence and effort. Statistical
significance for the dependent variable ability attributions was inconclusive.
vt


Data collected from the structured survey interviews, responses to the
reflective writing prompts (RWP), observations, and students grades yielded
more positive results. The chi square association tests on responses to the
interview questions and n responses indicated that the students in the
experimental group more readily identified goal orientation of learning and
skill mastery and the strategies of goal setting and time management and
planning than those in the control group. Girls in the treatment group were
more likely to attribute success to persistence and effort and to state interest
in learning more often than the boys in the treatment group.
Chi square tests indicated that the students in the treatment group
were more likely to improve their grades than the students in the control
group. Chi square tests also indicated that the girls in the treatment group
were more likely to have improved grades than the boys in the same group.
Students from the treatment group made gains in terms of time on task, task
completion, and persistence in the face of difficulties. These gains were not
observed in the control group.
This abstract accurately represents the content of the candidates thesis. I
recommend its publication.
VII


DEDICATION
I dedicate this thesis to the memory of my sister, Dorothy Anderson, for
teaching me what is important, to the memory of my father, Boyd Anderson,
who was always there, to Laura Dulcie for her unwavering support, and to all
my students who helped me envision possibility.


ACKNOWLEDGEMENT
My thanks to my advisor, Lyn Taylor, mentor and friend; to my committee co-
chair Brent Wilson, Alan Davis, and Jeff Farmer for their support and
guidance. Thank you to the faculty and teammates at my school for support
through this process, especially, Rick Price, Fred Biebesheimer, Rebecca
Micheletti, Amy Jenkins, and Malcolm Davis..


CONTENTS
Figures...................................................xiii
Tables....................................................xiv
CHAPTERS
1. INTRODUCTION.............................................1
The Problem...........................................1
Purpose...............................................2
Research Questions....................................3
Conceptual Framework..................................4
Operational Definitions..............................10
Methodology..........................................11
Overview of Dissertation.............................14
2. REVIEW OF THE LITEATURE.................................16
Introduction.........................................16
Reflection and Reflective Writing....................16
20
21
Self-Efficacy.............................
Self-Efficacy and Self-Regulated Learning
Learned Helplessness......................
29


Girls' Self-Efficacy..................................33
Summary...............................................34
3. METHODOLOGY...............................................37
Introduction..........................................37
Review of Study Purpose...............................37
Research Questions....................................39
Study Design..........................................42
Setting...............................................43
Sampling Procedures and Subjects......................45
Variables.............................................49
Independent Variables...........................49
Dependent Variables.............................50
Instrumentation.......................................51
Childrens Academic Intrinsic Motivation
Inventory.......................................51
Morgan-Jinks Student Self-Efficacy
Scale...........................................53
Study Measure...................................54
Data Collection Procedures............................61
ix


Data Analysis Procedures
64
T Tests............................................64
Repeated Measures Factorial ANOVA..................65
Interviews and Reflective Writing Prompts..........65
Reliability and Validity.................................67
Summary..................................................70
4. RESULTS......................................................72
Introduction.............................................72
Measurement Strategies and Preliminary
Procedures...............................................74
Repeated Measures Factorial ANOVA Results................82
External Attributions For Success or Failure.......83
Goal Orientation...................................85
Intrinsic Motivation...............................86
Persistence and Effort.............................87
Ability Attributions...............................88
Post Hoc Tests...........................................89
Summary..................................................93
x


5. INTERVIEW, REFLECTIVE WRITING PROMPTS,
AND OBSERVATION RESULTS.................................96
Introduction.........................................96
Student Responses....................................96
Chi Square Tests of Association for Self-Referent
Responses............................................99
Chi Square Test of Association for Interview
Responses.....................................100
Chi Square test of Association for Reflective
Writing Prompt Responses......................101
Chi Square Tests of Association for Self-Regulated
Learning Strategy Use...............................103
Chi Square Test of Association for Self-Regulated
Learning Strategy Us Interview Responses......103
Chi Square Test of Association for Self-Regulated
Learning Strategy Use Reflective Writing Prompt
Responses.....................................107
The Students........................................109
Summary.............................................115
xi


6. FINDINGS AND DISCUSSION
117
Introduction............................................117
Findings................................................117
Self-Referent Processes.................................119
Self-Regulated Learning.................................122
Gender..................................................123
2001-2002.............................................. 125
Limitations.............................................127
Future Studies..........................................129
Concluding Thoughts and Pedagogical Implications........131
APPENDIXES.........................................................135
A. Parent Consent and Student Assent Letters................135
B. Writing Prompts...................................................140
C. Interview Questions..............................................143
D. Self-Efficacy Inventory..........................................144
E. Written Reflection, Interview, and Observation Rubric............150
F. Codes for Self-referent Characteristics and
Self-Regulated learning Strategies...............................151
G. Interview Responses..............................................152
REFERENCES.........................................................160
xii


FIGURES
Figure
1.1 Conceptual model of learning, self-efficacy, and
self-regulated learning.................................................6
3.1 Questions and measure for quantitative data...........................41
3.2 Questions and sources of structured interview survey data.............42
3.3 Ethnicity of treatment and control groups and school..................47
3.4 Gender by group.......................................................48
3.5 Survey items by self-referent component...............................55
3.6 Prompts and interview questions.......................................59
4.1 Interaction of external attributions between groups...................85
4.2 Grade change by group.................................................90
4.3 Grade change by team and gender.......................................92
5.1 Treatment group profiles for interviewed students....................110
5.2 Control group profiles for interviewed students......................111
xiii


TABLES
Table
4.1 Levene Test of Homogeneity on Pre-Test by Group..................75
4.2 T-Tests on Pre-Tests by Group...................................76
4.3 Pre- and Post-Test Means and Standard Deviations for Males.......77
4.4 Pre- and Post-Test Means and Standard Deviations for Females.....78
4.5 Pre- and Post-Test Means and Standard Deviations by
Group.............................................................79
4.6 Levenes Test of Homogeneity of Variance.........................82
4.7 Repeated Measures Factorial ANOVA for External Attributions
for Success or Failure...........................................84
4.8 Repeated Measures Factorial ANOVA for Goal Orientation...........86
4.9 Repeated Measures Factorial ANOVA for Intrinsic Motivation.......87
4.10 Repeated Measures Factorial ANOVA for Effort and Persistence....88
4.11 Repeated Measures Factorial ANOVA for Ability Attributions......89
4.12 Chi Square Test of Associations Grade Change by Gender..........91
4.13 Chi Square Test of Association Gender by Grade Change...........93
5.1 Chi Square for Interview responses by Group......................100
XIV


5.2 Chi Square for Interview Responses by Gender....................101
5.3 Chi Square for Self-Referent Processes RWP Responses
by Gender........................................................102
5.4 Chi Square for Self-Regulated Learning Strategies
Interview Responses by Group.....................................104
5.5 Chi Square for Self-Regulated Learning Strategies
Interview Responses by Gender....................................105
5.6 Chi Square for Self-Regulated Learning Strategies
RWP Responses by Gender..........................................108
xv


CHAPTER 1
INTRODUCTION
As a veteran teacher, I have observed that many learners at Cedar
Middle School (pseudonym) exhibit characteristics associated with low self-
efficacy beliefs. These characteristics include limited persistence at difficult
tasks, reliance on extrinsic rewards, attributions of success or failure to
external conditions such as luck, being liked by a teacher, or going to a good
school rather than abilities, and the unwillingness to take risks in their own
educational processes. Because of these factors, there has been little change
in these students' learning behaviors. These behaviors have led to limited
student success, perceptions of low academic self-efficacy, and limited self-
regulated learning strategy use.
The Problem
Middle school students may have difficulty completing learning tasks
because of low academic self-efficacy beliefs or difficulty in activating learning
strategies that would assist them in regulating their own learning. Difficulty in
activating these skills may, in part, involve an inability for middle school
1


students to reflect upon their own learning and learning strategies. Without
the means to develop and activate necessary self-regulated learning
strategies, middle-school learners may be poorly equipped for learning
(Borkowski and Thorpe, 1994). Further, patterns of self-regulation resist
change unless interventions are used (Zimmerman, 1998). Success or failure
may be linked to learners abilities to organize what they already know and
their personal learning strategies in order to complete new learning tasks.
Purpose
The purpose of this study was to determine the relative effects of
participation in a guided reflective writing program on middle school students'
self-efficacy constructs and their use of self-regulated learning strategies. The
intentions of this study were: (a) to determine if participation in guided
reflective writing would help middle school students activate self-referent
processes to self-regulate learning, (b) to determine how reflective writing
impacted students use of self-regulation of learning strategies, and (c) to
determine if there were gender differences in self-referent processes and self-
regulated learning strategy use in the treatment group and/or the control
group.
The guiding assumptions for this study were:
2


1) Reflection is closely linked to learning.
2) Reflection is integral to evaluating self-efficacy beliefs, using self-
regulated strategies, and developing concepts and skills.
3) Middle school learners may be unable to assess their own learning
capabilities.
4) Many learners are aware of an array of learning strategies, but
middle school learners may be unable to choose a strategy that will
work for the given learning task.
Research Questions
This study investigated the effects of participation in a guided reflective
writing program on middle school students' self-efficacy constructs and their
use of self-regulated learning strategies. The guiding question for this quasi-
experimental study was: What are the relative effects of participation in a
guided reflective writing program (RWP) and gender on sixth grade middle-
school students self-efficacy and use of self-regulated learning strategies.
The sub-questions measured by the pre- and post-test questionnaire were:
1) What are the relative effects of participation in a guided RWP on
academic self-efficacy beliefs including: goal orientation,
3


attributions for success or failure, intrinsic motivation and the self-
regulated learning strategy persistence and effort?
2) Are there differences between male and female students academic
self-efficacy beliefs including the components goal orientation,
attributions for success or failure, intrinsic motivation and the self-
regulated learning strategy persistence and effort?
3) Is there a significant interaction between participation in a guided
RWP between participants in the RWP and non-participants and/or
gender?
The sub-questions that the structured survey interview questions and
observations were to answer included:
4) What are the relative effects of participation in a guided RWP on
self-regulated learning strategy use?
5) Are there differences between male and female students self-
regulated learning strategy use?
Conceptual Framework
Setting the stage for task completion and learning through reflection
coupled with self-regulated learning strategies (Bandura, 1977; Schunk &
Zimmerman, 1994,1998) assists learners in focusing on learning and
4


becoming more involved in their own learning (Brookhart, 1996; Lewison,
1996; Vojnovich, 1997; Wolfe, 1996). These beliefs and strategies are
embedded in the personal experiences of the learners and imbued by culture
and tacit understandings (Bruner, 1990; Honebein, P. C., Duffy, T. M., &
Fishman, B. J., 1993; Grabinger, 1996). Success or failure of learners is
linked to their ability to organize what they already know and personal
learning strategies in order to complete new learning tasks (Bandura, 1993;
Dweck, 1975).
The relationship among self-referent processes of academic self-
efficacy, including the components of goal orientation, attributions, and
intrinsic motivation and self-regulated learning strategy use (Bandura, 1993;
Shunk, 1987,1990a, 1990b; Shunk & Zimmerman, 1994; Zimmerman, 1998)
can be modeled by a tetrahedron in which the task is introduced to the learner
(Figure 1.1). Before learners engage in a learning task, they reflect to
appraise their academic self-efficacy for completing that task as well as
evaluating their goal orientation, attributions, and motivational constructs
through reflective processes. Once learners evaluate their capabilities, goal
orientation, and intrinsic motivation, they may engage a variety of self-
regulatory processes such as goal setting, planning, and time management,
5


and adjusting effort and persistence (Shunk, 1990a, 1990b; Zimmerman,
1998).
Self-monitoring is a reflective process that continues throughout task
engagement. A supposition of this study is that reflection is integral for: self-
efficacy evaluation, self-regulated learning strategy use, and concept and skill
formation. Self-monitoring provides an ongoing evaluative process to assist
the learner in adjusting goals, managing time, and managing effort and
persistence. The self-regulated learner continues self-monitoring, adjusting,
and managing until the task is complete. Once the task is complete, the
learner may, once again, appraise his or her self-efficacy and its components
of goal orientation, attributions, and intrinsic motivation (Bandura, 1982,1993,
1997; Shunk, 1990a, 1994,1996).
Figure 1.1. Conceptual model of learning, self-efficacy, and self-regulated
learning.
6


In this model, the processes of reflective processes for self-efficacy
appraisal, self-regulatory strategy use, the learning task, and learning are the
vertices of the tetrahedron. Reflective processes necessary for cognition
(Bruner, 1990; Dewey, 1938) self-referent appraisals and activation of self-
regulatory strategies are at each vertex. If any of the vertices are removed,
the structure becomes less stable, so the learners approach to a learning
task may be less effective if (s)he is unable to appraise his/her academic self-
efficacy or activate self-regulatory strategies (Shunk 1990a, 1990b,; Shunk &
Zimmerman, 1998).
Learning may be viewed as a process in which learners engage in
reflection, self-efficacy appraisal, and self-regulated learning strategy use
(Zimmerman, 1998). Reflective processes are necessary for self-efficacy
evaluation and the components goal orientation, attributions for success or
failure, and intrinsic motivation. Performance is monitored through self-
reflection on self-regulated learning strategies including goal setting, time
management and planning, and persistence and effort management
(Zimmerman, 1998; Zimmerman, Bonner, & Kovach, 1996).
Once the learner engages in the learning task, reflection on the
learners capabilities and use of self-regulated learning strategies become
recursive processes until the task has been completed. Without ongoing self-
7


evaluations, the learners are unable to determine their progress in learning
and to make connections to prior learning (Dewey, 1938; Shunk, 1999;
Zimmerman, 1998).
Reflection on previously used self-regulated learning strategies before
actual engagement in learning and/or completing a task may assist the
learner in completing new learning tasks. Goal setting and goal orientation
are closely related. Goal orientation focuses on learning outcomes or
competition (Zimmerman, Bonner, & Kovach, 1996). This orientation
determines the specific outcomes and focus for learning. The selection of
processes necessary for task completion aids the learner and learning in
strategic planning.
Time management is an important factor in self-regulation. Because
teachers often determine completion, the learner has to establish schedules
for the stages for task completion within the designated deadlines
(Zimmerman, Bonner, & Kovach, 1996). Closely linked to time management
is persistence. When a learner persists at a task, (s)he continues to focus and
work on the task to maintain the schedule even when encountering problems
(Dweck, 1975, 1986).
The individual learners self-reflective processes include self-
evaluation, attributions, and self-reactions. Self-efficacy appraisal is one of
8


the initial processes. Self-evaluative processes also include comparing self-
monitored information and feedback with an established goal. Attributions are
the causal meanings of the results of success or failure of a learning activity
to luck, ability, or effort. Attributions may lead to self-reactions that assist the
learner in identifying the sources of errors (Dweck, 1975; Dweck & Chui,
1986, Shunk 1987). As the learner self-refiects, his or her understandings of
the learning task may be enhanced so that he or she can modify the
strategies being used in the learning task (Bandura, 1993; Bandura &
Cervone, 1983).
Self-efficacy is the appraisal on ones capabilities for the completion of
a learning task. Self-referent constructs related to academic self-efficacy are
goal orientation, attributions for success or failure, and intrinsic motivation.
Goal orientation is the learners reason for engagement in a task, i.e., whether
the goal is mastery, a product, or a reward. Attributions are the learners view
of why a task was successfully or unsuccessfully completed. Intrinsic
motivation is the learners natural interest in a learning task (Baundura, 1986;
Shunk, 1987, 1990a, 1990b).
The self-regulatory strategies examined in this study were goal setting,
planning and time management, and persistence and effort. Goal setting is
the process of establishing short-term and long-term goals for the learning
9


task. Planning and time management are the processes for setting timetables
for task completion. Persistence and effort are the processes by which the
learner continues engagement even though difficulties are encountered
(Garcia & Pintrich, 1991,1994).
Operational Definitions
The operational definitions for this study were derived from the above
concepts. These definitions represent the dependent variables that were
measured in the pre- and post-test questionnaire:
Academic self-efficacy is the appraisal on ones capabilities for the
completion of a learning task.
Goal orientation is the learners purpose for engagement in a task, i.e.,
whether the goal is mastery, a product, or a reward.
AWibutions for success or failure are the learners view of why a task is
successfully or unsuccessfully completed, task difficulty, ability, effort,
or persistence.
Intrinsic motivation is the learners natural interest in a learning task.
Goal setting is establishing short-term and long-term goals for
completion of a learning task.
10


Planning and time management are processes for setting targets for
task completion in time periods or stages.
Persistence and effort are the processes by which the learner
continues engagement even though difficulties are encountered
(Bandura, 1993; Csikszentmihalya, 1990; Dweck, 1975; Garcia & Pintrich,
1993; Shunk, 1987).
Two additional operational definitions were developed to inform the
sixth grade teachers during the presentations to the test-site staff
members:
Reflection is purposeful, intentional thinking through which learners
make cognitive connections between previously learned skills and
knowledge and new learning situations in order to make behavioral
and motivational decisions about their own learning (Dewey, 1938).
Reflective writing was defined as the product resulting from writing
about purposeful, evaluative thinking the self-referents and learning
strategies posed in the writing prompts (Hullfish & Smith, 1961).
Methodology
In this study, a group of middle school students who participated in the
six-week guided reflective writing program (RWP) was compared with
ll


students who did not participate in this program on their self-perceptions of
academic efficacy. The components of self-efficacy that were examined in
this study were: (a) goal orientation, (b) attributions for success or failure, and
(c) intrinsic motivation. The study also compared the experimental and control
groups for use of self-regulated learning strategies, specifically, goal setting,
planning and time management, and persistence and effort. Finally, both
academic self-efficacy and self-regulated learning were examined for gender
differences.
This study used a quasi-experimental 2x2x2 design in which pre- and
post-test data from a Likert-type, self-report inventory of academic self-
efficacy was completed by sixth-grade middle-school students. This inventory
yielded a total self-efficacy score and sub-scores for goal orientation,
attributions, persistence and effort, and intrinsic motivation. These constructs
were measured in both the experimental and control groups at the beginning
of the participation in the RWP and at the end of the RWP. The inventory was
drawn from existing inventories of academic self-efficacy and intrinsic
motivation. Structured survey interview research (LeCompte & Priessle, 1993)
data from student responses to writing prompts and interviews and
observations of students were collected to determine student awareness of
self-regulated learning strategies use.
12


Quantitative data gathered from the Likert-type self-reporting inventory
were analyzed to determine differences in student perceptions of academic
self-efficacy and the component constructs goal orientation, attributions, effort
and persistence, and intrinsic motivation. The quantitative data were
examined to determine relationships between these self-perceptions and self-
regulated learning strategy use. For example, were there differences in boys
and girls ability to identify the self-regulated learning strategies they used; or
were students who scored high on the self-efficacy inventory more able to
identify the self-regulated learning strategies they used than those students
who scored low on the academic self-efficacy inventory.
The responses to writing prompts and observational data from the
treatment group were collected for 30 school days. Observational data were
collected on the sixth-grade control group for 30 days. Twelve students in
both the experimental and the control group were interviewed at the end of
the study. Data gathered from the observations and interviews were used to
determine which self-regulated learning strategies were used by the students
in each group and to identify any differences in strategy use between groups
and genders. The students who were interviewed were selected on the basis
of their scores on the inventory. There were an equal number of girls and
boys who, by their responses to the inventory, indicated a high degree of
13


academic self-efficacy and its component self-referents, and an equal number
of girls and boys who indicated low academic self-efficacy and its component
self-referents for this in-class project.
Overview of Dissertation
Chapter 1 introduces the problem and purpose for this study. It includes a
description of the conceptual framework, the research questions, and an
overview of the methodology.
Chapter 2 is a review of the literature on academic self-efficacy, self-regulated
learning strategies, and gender differences in academic self-efficacy and self-
regulated learning.
Chapter 3 presents descriptions of the setting and sample population, the
independent and dependent variables, the instrumentation, the data collection
methods, data analysis methods, and estimates of reliability.
Chapter 4 includes the results of quantitative analyses of the pre- and post-
test inventories including the results of the repeated measures factorial
ANOVA and reliability analyses.
Chapter 5 presents the results of the structured interview survey research that
includes students interview responses, frequency of responses, and
classroom observations.
14


Chapter 6 summarizes the study, discusses the findings, and presents
conclusions and recommendations for future studies.
15


CHAPTER 2
REVIEW OF THE LITERATURE
Introduction
This chapter presents a review of the literature on reflection and
reflective writing, self-efficacy and self-regulated learning, the relationship
between them self-efficacy to self-regulated learning strategy use, and
literature on girls and self-referent processes.
Reflection and Reflective Writing
Reflection may be central to heightened academic self-efficacy beliefs,
activation of self-regulated learning strategy use, and learning. According to
Dewey (1938) the difference between remembering and reflecting is the
assessment and evaluation of the assumptions or tacit knowledge that
underpin our remembrances. Language is a tool of inner communication
necessary for reflection (Vygotsky, 1934). Active reflection may help learners
construct new schemas that ground judgments, attitudes, behavior, and
learning in differing contexts.
16


Reflection differs from the looser kinds of thinking primarily because it
is directed or controlled by a purpose (Hullfish & Smith, 1961), in this case
exploration of self-referent constructs. Reflection is one way to augment the
ability to make the absent present and to reform the elements of reality into
alternatives. When a learner selects from among alternatives and acts on one
or more of them, (s)he is able to change his or her reality. As learners reflect,
they discover ways in which a hypothetical idea may be tested against what is
known. The testing of hypotheses against previously acquired knowledge or
skills leads to the development of new knowledge and skills (Hullfish, 1961)
Reflection facilitates the construction of new knowledge and skills and
becomes an essential factor of learning. Learning engenders change,
transition, and an understanding of context. The learner acknowledges and
values the sense of personal identity and efficacy that emerge from reflective
interactions. Deep reflection is more than remembering; it is the assessment
and evaluation of prior experience and knowledge in light of new contexts
(Dewey, 1909,1938; Lambert, 1995; Mezirow, 1991).
Reflection on self-referent processes is central to the construction of
knowledge (Lambert, 1995). The changing characteristics of our reflections
embody a capacity for reinterpretation and change. Reflection provides a
framework for thinking, imagining, and making cognitive choices (Bruner,
17


1990). Reflective thinking helps students generate links to relate new ideas to
each other and to prior knowledge; judge understanding, whether or not links
have been generated; and activate a context for interpreting new information
(Dewey, 1909, 1938).
Reflective thinking is central to proficiency in written language, and
writing is one way in which learners may crystallize their thoughts and save
them for future examination and change. Writing assists in the development
of critical thinking (Qian, 1990). The act of writing shapes meaning making
and is essential for discovering new ideas and relationships. Writing provides
possibilities for engaging students in activities that validate their knowledge,
so they may become active doers and planners (Hillocks, 1995).
According to Albert Bandura (1993) there were strong connections
between writing and enhanced self-efficacy beliefs. He wrote that cognitive
development depended heavily on writing literacy and that the development
of proficient writing skills informed the discernment of the way processes of
self-efficacy operate.
The act of writing shapes meaning making and is essential for
discovery of new ideas and relationships. Writing provides possibilities for
engaging students in activities where what they know is valued and
respected, so they become active doers and planners (Hillocks, 1995).
18


Enhancement of perceived writing efficacy by instruction raises, through
different paths of influence, perceived self-efficacy for academic activities,
personal standards for the quality of writing, and academic goals and
attainments. Whereas verbal aptitude affects academic attainments only
indirectly by raising personal standards of writing, the increased sense of
academic efficacy promotes academic attainments both directly and by
heightening aspirations (Bandura, 1993, p. 137).
The curriculum model used for this project was presented by Hillocks
(1995) in Teaching Writing as Reflective Practice. I chose this model
because it has been used for reflective writing and provided a foundation for
instruction in reflective writing. This model assisted this researcher in
clarifying the reflective writing process for the treatment group students
teacher. Her instruction provided support for the students in writing their
responses. It was useful for organization in the early stages of the
instructional design. This framework for developing a reflective writing
curriculum is a six-step process:
1) analyzing current student progress in relation to goals,
2) positing some change or range of possible changes sought in the
traits and skills,
3) selecting or devising a teaching strategy or set of strategies to
19


implement the desired change,
4) devising a plan for implementing the strategies,
5) assessing the impact of the teaching strategy in order to discover
consequences and implication of the chosen strategy, and
6) confirming or changing of the strategy used.
Self-Efficacv
Perceptions of academic self-efficacy are the personal judgments of
capabilities to accomplish learning tasks, and these perceptions deal with
activities across different academic domains (Bandura, 1986,1997). In order
to appraise and make decisions about ones capabilities, learners may reflect
upon their academic abilities. A learner's self-efficacy beliefs and the nature
of instruction across academic areas are central to motivational orientation.
Learners differentiate self-perceptions of ability within the differing domains
and learning tasks. Perceptions that schoolwork is worthwhile and interesting
as well as endorsement of learning goals varied by learning situations
(Young, Arbeton, & Midgely, 1992).
According to researchers (Csikszentmihalyi, 1990; Pendvaris, 1990;
Schunk, 1985,1989,1990), there are general characteristics of learners who
20


have high academic self-efficacy beliefs across domains. Characteristics of
students with enhanced self-efficacy beliefs include:
1) the ability to focus on a goal
2) a positive expectation of a goal
3) a willingness to take a risk
4) engagement in a task
5) greater persistence when encountering difficulties
6) a loyalty of teamwork toward a goal
7) self-control
8) self-confidence and an understanding of potential.
The general characteristics of self-efficacious learners indicate connections to
self-regulated learning strategy use.
Self-Efficacv and Self-Reaulated Learning
The self-efficacy and the reflection on individual capabilities may
enable control over thoughts, feelings, and actions. Self-referent behavior and
self-regulation involve the ability to self-reflect and plan alternative strategies.
Motivational factors for these constructs include goal orientation, attributions
for success or failure, and intrinsic motivation (Bandura, 1993; Dweck, 1975,
1986; Shunk, 1989). According to Bandura (1986,1993), self-directed
21


learning requires motivation as well as cognitive and metacognitive strategies.
The motivational facet of self-directed learning encompasses a variety of
inter-linked self-referent processes including self-monitoring, self-efficacy
appraisal, personal goal setting, outcome expectations, and affective self-
motivation.
Academic self-efficacy involves the capability in which reflection on
component cognitive, social, and behavioral skills engender action for varied
purposes. Academic proficiency requires integration and continuous
management of multiple skills in changing learning environments. Task
engagement and regulation of strategies are partially managed by self-
judgments of capabilities (Bandura, 1982,1997).
People initially rely on past performance for evaluating their efficacy
and setting achievement goals (Bandura, 1982). They form schemas
concerning their efficacy. As they gain experience, they strengthen their self-
efficacy beliefs through performance attainments, and their personal efficacy
beliefs become more complex. Accurate appraisals of personal efficacy
determine how well one can invoke courses of action required for dealing with
future situations. Efficacy analysis requires reflective assessment of the level,
strength, and breadth of capabilities. Analyses are made for a variety of
22


activities in varied situations in advance of behavior, rather than immediately
prior to each task performance (Bandura, 1993,1997).
Students who believe in their efficacy tend to regulate their own
learning. The level of mastery of academic activities determine students
goals, motivation, and academic accomplishments (Schunk, 1994). The
influences of self-beliefs affect the selection and construction of learning
conditions. The impact of environmental influences affect human motivation
and mediate action through self-referent processes. This mediation gives
meaning and merit to external events and operates as an important proximal
factor. Efficacy beliefs influence how people feel, think, motivate themselves,
and behave (Bandura, 1993,1997; Schunk, 1994).
Garcia and Pintrich (1991) found that higher student motivation and
self-regulated learning were related to the integration of motivational and
cognitive components of student learning. Higher levels of motivation were
associated with higher levels of cognitive engagement. When engaging in
self-regulated learning as a set of learning strategies, for example,
monitoring, elaboration, and effort management, students approached a
learning task more effectively and flexibly.
Academic self-efficacy has consistently been found to be positively
related to effective use of strategies as well as academic success, so
23


reflection on individual capabilities may enhance self-regulated learning
strategy use. When engaging in self-regulated learning, intrinsic motivation is
optimized, and intrinsic motivation, in turn, impacts future use of learning
strategies (Csikszentmihalyi, 1990). As a result of their analyses of the
relationships among self-efficacy, intrinsic motivation, and self-regulated
learning, Garcia and Pintrich (1991) proposed that intrinsic motivation was the
trigger to the mediation among middle-school students perceptions of value,
importance, and interest; self-efficacy beliefs; and the use of learning
strategies. They concluded that intrinsic motivation and self-efficacy had
strong positive effects on self-regulated learning strategy use. They found that
intrinsic motivation had strong positive influences on academic self-efficacy,
and students' adoption of general learning and mastery orientation played a
significant role in their perceptions of self-competence and success in their
courses. The adoption of learning goals led to positive efficacy beliefs and
attributions, but the impact of self-regulated learning on self-efficacy was less
clear.
Children who focus on learning for mastery of a new skill or concept
appear to have their performance facilitated by increased challenge because
they view intelligence as incremental. These learners progress and master
skills through persistence and effort. They may be more willing to take risks in
24


order to attain skills and knowledge. Obstacles may be cues for increased
effort, so challenging tasks may be best for using and increasing their
repertoire of skills. Mastery learners develop autonomy in learning and
attribute success to effort and perseverance (Dweck, 1975,1986; Dweck &
Chiu, 1993). Dweck (1986) found that learners who have mastery goals
attained significantly higher scores on the transfer test, produced 50% more
work, and produced more rule-generated answers.
Young, Arbreton, and Midgely (1992) and Young and Urdan (1993)
found that students who were focused on learning were more intrinsically
interested in completing tasks for the purpose of understanding and
enhancing their knowledge and skills. They tended to work harder, persist
longer, and engage in more challenging tasks than students who were
focused on external factors such as grades or praise. These students also
used deeper processing strategies such as self-monitoring, making new
connections with prior knowledge, and discriminating between important and
unimportant information, all of which are reflective thought processes.
Dweck (1975,1986) and Dweck and Chiu (1993) found that if ability
was viewed as an innate quality in which performance was indicative of
intellectual capacities, the learner was more likely to choose tasks that
minimized errors and revealed their proficiencies. Learners who were
25


performance-oriented often were characterized by helplessness patterns.
These patterns were manifested in challenge avoidance and low persistence
that resulted in negative affect and self-cognitions and anxiety (Bandura,
1997). Learners who were ability focused or viewed their own individual ability
by comparing and competing with others were more concerned with grades
than learning. Young, Arbreton, and Midgely (1992) and Young and Urdan
(1993) also found that learners who were ability focused were more likely to
engage in irrelevant and diversionary verbalizations and used ineffective
strategies than learners who were mastery focused.
Tasks that are perceived as high evaluative threats and related to
deficient performances may lead to the individuals belief in a lack of basic
intelligence. Low-challenge learners prefer tasks that are less challenging at
the expense of extending knowledge and competencies. For the entity
learner, learners who view ability as unchanging, high effort is threatening
because it may reveal a lack of ability. The successes of others may diminish
their own perceived ability (Dweck, 1975,1986; Dweck & Chiu, 1993). The
entity learners focus is aimed at protecting a positive evaluation of personal
competence. Seeing ones self exceeded by others undermines self-efficacy,
increases erratic analytic thinking, and progressively impairs performance
(Bandura, 1993; Schunk, 1989,1990a, 1990b, 1996).
26


Dweck (1975,1986) suggested that when performance was the
primary goal, acquisition and display of cognitive skills were seriously
restricted. Children who viewed intelligence as fixed, focused on ability level.
This view worked against challenge by demanding childrens perceptions of
ability be high and remain high before choosing a challenging task. Entity"
learners chose personally easy tasks in which success was ensured, or
excessively difficult ones in which failure was unrelated to low ability. These
learners were more likely to interpret errors or failures in terms of ability, and
high effort was negatively related to satisfaction (Bandura, 1997; Dweck &
Chiu, 1993).
Perception of ability as an acquired or fixed attribute is related to
perceived controllability. The extent to which people perceive controllability
determines how the learner reacts to the aspects of control: (a) the level and
strength of personal efficacy to produce changes by persistent effort, (b) the
creative use of capabilities and resources, and (c) the variablity of the
learning environment (Dweck, 1975,1986; Dweck & Chiu, 1993). These
aspects represent the limitations and opportunities for self-appraisals. Self-
doubters anticipate the futility of effort to change their life situations. Those
with a strong belief in personal efficacy figure out ways to control and change
27


the environment and have resilient academic self-efficacy beliefs (Bandura,
1993).
Self-efficacy and reflective processes play a strong role in the
motivation processes. Reflections on cognitions may translate into incentives
and action (Bandura, 1993; 1997). Cognitive motivators are mediated by
causal attributions. Learners with high efficacy beliefs are more likely to
attribute success to increased effort and persistence and failure to lack of
effort. Learners with low self-efficacy attribute success or failure to perceived
ability (Dweck, 1975,1986; Dweck and Chiu, 1993). Learners outcome
expectancies and expectancy-value are related to the attributions of mastery
learners and performance learners. Both types of learners have the
expectation their behaviors would produce certain outcomes, and they place
value on those outcomes (Schunk, 1989,1990a, 1990b, 1994,1996).
The selection processes are related to personal reflection and are
essential to academic self-efficacy. The choice of activities and environments
lead to different competencies, interests, and social networks. Social
influences operate in selected environments and promote competencies,
values, and interests long after efficacy has influenced the choice.
Collaborative groups provide the learner with a social arena for learning
(Bandura, 1982). Perceived self-efficacy fosters engagement of social support
28


and mediated its effects on psychological wellbeing and functioning (Bandura,
1993). Influences on choice include verbal modeling of cognitive strategies,
proximal goal setting, ability, appropriate attributional feedback, positive
incentives, and analysis of task strategies (Bandura, 1993; Bandura &
Cervone, 1983).
High academic self-efficacy beliefs and challenging personal goals are
influenced by acquisition of skills and academic achievement. Children who
are able to reflect upon their own capabilities and have a high sense of
academic efficacy and self-regulatory strategies in place are more socially
active, are more popular, and experience less rejection by peers. It is difficult
for children to be socially active and maintain emotional wellbeing when they
encounter repeated academic failures and alienation of peers. Peer
relationships promote developmental courses that depend on the values,
standards of conduct, and lifestyles that are modeled by peers (Bandura,
1993).
Learned Helplessness
Motivational patterns of learned helplessness are not solely evident in
task performances of the low-achieving, failure prone child. There are strong
sex differences in task preferences. Many girls prefer tasks they know they
29


are good at, while many boys prefer those in which they would have to work
harder to master. These girls are more likely to ascribe to a fixed theory of
intelligence than boys. Being a high achiever and knowing one has done well
in the past does not appear to translate directly into high academic self-
efficacy beliefs in current difficulties or later challenges. Self-efficacy does not
clearly predict the maintenance of ability to perform or learn under these
conditions (Bandura, 1997; Dweck, 1986).
The girls learned helplessness patterns become more apparent in
later school achievement when they avoid challenging courses of study, drop
out of courses that pose a threat of failure, or show reduction of performance
under real difficulty. The differences often are demonstrated when differing
motivational patterns contribute to achievement discrepancies. For example,
sex differences in mathematics achievement are greatest among the brightest
students, and sex differences of motivational patterns appear among the
brightest students, especially bright girls. Many bright girls exhibit less
confidence in expectancies, lower preference for new or challenging tasks,
and more frequent debilitation when confronted with failure or confusion.
These girls frequently attribute failure to lack of ability. Bright girls often are
not advancing as rapidly as bright boys. Greater novelty and difficulty of the
30


future courses compared to previous courses result in declines in confidence
for bright girls but not for bright boys (AAUW, 1999; Dweck, 1986).
With increasing age, students make increasingly crucial decisions.
Helplessness patterns may limit achievement and future choices. Girls may
fail to enhance general intellectual growth. Teacher expectancies and
impressions about student ability affect student performance (AAUW, 1999).
Performance begins to align with these expectancies (Dweck, 1986). Females
frequently have more problems developing self-esteem because of their
reliance on others, not self-evaluations. They appraise themselves as having
high self-worth rather than high self-efficacy (Diedrick, 1998). This may result
from perceived opportunities related to academic self-efficacy and power over
one's environment (Dweck, 1975,1986; Dweck & Chiu, 1993). although low
self-efficacy beliefs about new learning tasks often lead to less self-regulated
learning strategy use, Zimmerman and Martinez-Pons (1990) found that girls
used strategies for self-regulation of learning to a greater extent than boys
even though girls felt less self-efficacious. Many women and girls were
conditioned to expect to fail, and they are more likely to attribute success to
luck and failure to lack of ability (Olson, 1988).
Langer (1979) identified situational factors that accompanied poor
performance. Among these factors were presence of a highly confident
31


individual and attending to what was strange rather that what was familiar and
within range of ability. A critical factor was assigning people to subordinate
roles or inferior labels. This implies limited competence and may lead to
negative effects on choice, effort, persistence, and self-disabling thought
(Bandura, 1982).
The negative effects of experiencing subordinate roles are reinforced
through inequities in response opportunities and the type and precision of
feedback. Inequities in interactions were studied over a six-year period and
included more than 100 classrooms from urban and rural elementary to
college level. There were no significant differences related to sex and/or
ethnicity of the teachers (Sadker & Sadker, 1986,1994). However, the
findings showed significant inequities in teacher-student communication
related to the gender of the student.
Sadker and Sadker (1992) found that many boys were being trained to
be assertive while many girls were being trained to be passive spectators and
listeners. This communication gender gap affected self-esteem, educational
attainment, career choice, and future income. The effects of girls invisibility in
teacher-student interactions were demonstrated in achievement test scores.
In early elementary years, girls scores were equal to or greater than boys.
32


By the end of high school, boys score were higher (Shakeshaft, 1986;
Grayson & Martin, 1988).
Girls Self-Efficacv
High self-efficacy beliefs may be derived from girls and womens
abilities and roles. Their standing in the community results from their
achievements. Knowledge, transformational operations, and component skills
are necessary for achievement, but they may be insufficient. A sense of
personal efficacy is essential for people to be productive and to regulate their
lives. The difference between possessing knowledge and skills and being
able to use them well directly affects achievement (Bandura, 1977,1993,
1997). Self-referent reflection is necessary to mediate knowledge and action.
How people judge their capabilities and how, through their self-efficacy
perceptions, they affect their motivation and behavior is especially important
(Ancis& Phillips, 1996; Bandura, 1982).
Girls with enhanced self-efficacy have a sense of their potential, their
capabilities, and their intrinsic value as individuals, but there are many girls
who do not have high self-efficacy beliefs. Many studies have show that
women and girls are less likely to have high self-efficacy beliefs than men and
boys. Too often women and girls belittle their own abilities, and small failures
33


only confirm low self-efficacy beliefs and feelings of inadequacy (AAUW,
1999; Brown, 1998; Gilligan, 1982; Brown & Gilligan, 1992; Mann, 1996;
Orenstein, 1994; Taylor, Gilligan, & Sullivan, 1995).
Although inequities have been addressed in educational settings,
subtle bias and stereotyping still exist. The differences in sex role
socialization between boys and girls are especially marked in the relationship
of self-efficacy beliefs that affect academic achievement, especially in
mathematics, science, and technology (AAUW, 1999). Early adolescents
experienced conflicting messages regarding girls. Girls are to be feminine,
submissive, and caring, and successful in school at a time when success is
dependent on competition and assertiveness. Brown (1998) proposed that
older adolescent girls were subtly told that academic achievement and career
preparation should be secondary to family and motherhood. Girls often
experienced social pressures to conceal their intelligence and were told that
girls who were good in mathematics and science were not.
Summary
Self-regulated learning strategies such as goal setting and effort-
management have been positively and significantly correlated with greater
enhancement of self-efficacy, mastery and performance goal orientation,
34


attributions of effort and persistence for success or failure, and intrinsic
motivation.
Self-regulated learning requires motivation and cognitive and
metacognitive strategies. It encompasses a variety of interrelated self-analytic
processes, including self-efficacy appraisal, personal goal orientation,
outcome attributions, affective self-incentives (Garcia & Pintrich 1991,1994;
Schunk, 1985,1990a, 1990b, 1996). The academically self-efficacious
learner applies self-regulatory strategies persistently when confronted with
stressors, difficulties, and competing attractions (Bandura, 1993). When the
ultimate goal of education is for students to use self-regulated learning
strategies and to enhance intrinsic motivation for learning, to develop the
foundation for becoming life-long learners, it becomes crucial for learners to
develop the self-referent and reflective processes for activating self-regulatory
learning strategies (Bandura, 1997; Csikszentmihalyi, 1990; Pajares, 1995).
The constructs of self-efficacy, goal orientation, attributions, and
intrinsic motivation have been critical factors for girls academic performance.
Middle-school girls have been particularly vulnerable to these self-referent
constructs, and this vulnerability has affected their performance in learning
tasks (Dweck, 1975,1986). In several studies (AAUW, 1999; Brown, 1998;
Taylor, Gilligan, & Sullivan, 1995) students mastery and performance goal
35


orientation and their intrinsic value have predicted young adolescents
cognitive strategy use.
36


CHAPTER 3
METHODOLOGY
Introduction
This chapter reviews the purpose of this study and presents the
research questions, hypotheses and data sources. The study design is
depicted and followed by descriptions of the setting, sampling procedures,
and subjects. The variables for the study are followed by data analysis
procedures. The data analysis procedures include preliminary analyses and
the repeated measures factorial ANOVA for the data collected on the pre- and
post-test inventory. The methodology for the structured survey interview
research (SSI) is presented. The sampling procedures and rationale for the
SSI research is followed by the data collection procedures. A presentation of
the writing prompts and interview questions are organized by dependent
variables.
Review of Study Purpose
The purpose of this study is to determine the relative effects of
participation in a guided reflective writing program (RWP) and gender on
37


middle school students academic self-efficacy and their use of self-regulated
learning strategies. This study compared two groups of middle school
learners perceptions of academic self-efficacy and self-efficacy components.
The data collected on the pre-test and post-test surveys were analyzed for
the following components of self-efficacy: (a) goal orientation, (b) attributions
for success or failure, (c) intrinsic motivation, and one self-regulated learning
strategy, persistence and effort. A repeated measures factorial ANOVA was
calculated for these components and a total score for academic self-efficacy.
This study compared students use of the following self-regulating
strategies: goal setting, planning and time management, and persistence and
effort. Subjects in the treatment and control groups were observed, and 12
students from each group were interviewed. The interviews took place at the
end of the six-week guided RWP. Students in both groups were observed
throughout the six-week study. A rubric (Appendix E) that identified student
strategy use guided the researcher in identifying strategy use for frequency
and duration. This rubric was used in the examination of writing prompts and
interviews.
The data collected in the interviews and observations of students in the
treatment and control groups were compared for self-regulated learning
strategy use. The data collected through the interviews and observations
38


were examined to determine whether gender differences in self-regulated
strategy use existed and to determine if there were differences between
students who reported high and low scores on the survey. These self-
regulated learning strategies were goal setting, planning and time
management, and persistence and effort.
Research Questions
The over-arching question for this quasi-experimental study was: What
are the relative effects of participation in a guided reflective writing program
(RWP) and gender on sixth grade middle-school students self-efficacy and
use of self-regulated learning strategies? The sub-questions measured by the
pre- and post-test questionnaire were:
1. What are the relative effects of participation in a guided RWP on
academic self-efficacy beliefs including the components: (a) goal
orientation, (b) attributions for success or failure, (c) intrinsic
motivation, and (d) the self-regulated learning strategy persistence and
effort?
2. Are there differences between male and female students academic
self-efficacy beliefs including the components: (a) goal orientation,
39


(b) attributions for success or failure, (c) intrinsic motivation and (d) the
self-regulated learning strategy persistence and effort?
3. Is there a significant interaction between participation in a guided
RWP and gender?
The sub-questions that the structured survey interview questions and
observations were to answer included:
4. What are the relative effects of participation in a guided RWP on
self-regulated learning strategy use versus non-participation in a
guided RWP?
5. Are there differences between male and female students self-
regulated learning strategy use?
The following figure (Figure 3.1) organizes the research questions,
and measures for the analysis of quantitative data. The dependent variables
include external attributions for success such as going to a good school and
being liked by the teacher, persistence and effort, ability attributions, intrinsic
motivation, and a total score for academic self-efficacy.
40


Research Question Measure
What are the relative effects of participation in a guided RWP academic on self-efficacy beliefs including the components: goal orientation, attributions for success or failure, intrinsic motivation and the self- regulated learning strategy persistence and effort? Academic Self-Efficacy Inventory
Are there differences in the effects of participation in a RWP between male and female students' academic self-efficacy beliefs including the components: goal orientation, attributions for success or failure, intrinsic motivation and the self- regulated learning strategy persistence and effort? Academic Self-Efficacy Inventory
Is there an interaction between treatment and control groups? Academic Self-Efficacy Inventory
Is there an interaction between males and females by treatment and control groups? Academic Self-Efficacy Inventory
Figure 3.1. Questions and measure for quantitative data.
The following figure (Figure 3.2) presents the organization of research
questions and sources of structured survey research data.
41


Research Question Structured Interview Survey Data Sources
What are the relative effects of participation in a guided RWP on academic self-efficacy beliefs including the components: goal orientation, attributions for success or failure, intrinsic motivation and the self- regulated learning strategy persistence and effort? Responses to prompts and interviews and observations field notes
What are the relative effects of participation in a guided RWP on self-regulated learning strategy use, specifically goal setting, time management and planning, and persistence and effort? Responses to prompts and interviews and observations field notes
Are there differences between male and female students academic self-efficacy and the components: goal orientation, attributions for success or failure, intrinsic motivation and the self-regulated learning strategy persistence and effort? Responses to prompts and interviews and observations field notes
Are there differences between male and female students self-regulated learning strategy use specifically, goal setting, time management and planning, and persistence and effort? Responses to prompts and interviews and observations field notes
Figure 3.2. Questions and sources of structured interview survey data.
Study Design
This quasi-experimental study used a 2x2x2 (treatment x gender x
time) repeated measures factorial design in which the relative effects of
participation in a guided RWP were examined for the following learner self-
referent characteristics: (a) goal orientation, (b) attributions, (c) intrinsic
42


motivation, (d) persistence and effort, and (e) a total score for academic self-
efficacy. These self-referent characteristics were the dependent variables
calculated in the repeated measures factorial ANOVA from the data collected
in the student self-report.
Structured survey interviews of 12 students in the treatment group and
12 students from the control group were conducted at the end of the
treatment. Observations of the students who were interviewed were
conducted during the six-week treatment. Field notes of the observations
were examined to compare treatment-group students use of self-regulated
learning strategies at the beginning and at the end of the treatment. The
interviewed students from the control group were observed throughout the
six-week study period. The field notes were examined to determine
differences in self-regulated learning strategy use between the treatment
group and the control group. Finally, the field notes were examined to
compare self-regulated learning strategy use of boys and girls in the
treatment and control groups.
Setting
Cedar Middle School (pseudonym) is in an urban setting in the second-
largest school district in a mid-sized city. The neighborhood is a combination
43


of single-family dwellings, many of which are subsidized by the county as low-
income housing and low-rent apartment complexes. The school is culturally
diverse, approximately 68% minority (Figure 3.2), and 78% of the students
participate in the free and reduced breakfast and lunch program.
The school has consistently scored in the D" or F categories in the
Colorado Student Assessment Program (CSAP). Because of this the school
has adopted a variety of programs to address student achievement:
Accelerated Reading, Destinations Mathematics, a before- and after-school
daily mathematics program, an after-school academic extended day program,
and Saturday school. In addition to these programs, all teachers were trained
in reading and writing across the content areas and Six-Traits Writing
Assessment.
Programs for students with special needs include English Language
Acquisition program, a self-contained program for students with significant
emotional and/or behavioral problems, a pull-out program for students with
severe learning disabilities, and an in-class assistance program with students
who have mild to moderate academic, social, and/or emotional needs.
44


Sampling Procedures and Subjects
Each grade level in Cedar Middle School has two teams and each
team has a teacher for each of the following disciplines: language arts,
mathematics, science, social studies, and reading. The students were placed
on the teams so that the number of students was balanced by pupil-teacher
ratio between the two grade-level teams. The classes were heterogeneous in
terms of ethnicity, academic achievement, and students with special needs.
The total number of sixth-grade subjects was approximately 230. There were
115 students on each team. Sixth-grade students were chosen because
research indicates that many students experience a reduction in positive
academic self-efficacy beliefs in the transition between elementary and
middle school (AAUW, 1999; Borkowski & Thorpe, 1994; Brown & Gilligan,
1992; Finders, 1997; Taylor, Gilligan, & Sullivan, 1995).
The teachers from both teams received information about the study
purpose and procedures. Academic self-efficacy, self-regulated learning
strategy, reflection and reflective writing, and the operational definitions were
presented and discussed. The social studies teacher from one sixth-grade
team volunteered to instruct the students in the guided RWP, so this team
became the experimental group. Because the other sixth-grade team did not
45


participate in the reflective writing activities, this team served as the control
group.
The volunteer teacher and this researcher worked together on how to
present the guided RWP. The students were to respond in a journal to the
prompts. Explanations of the prompts were given to individual students when
they asked for assistance. Observations of the student participation and
teacher explanations of prompts occurred throughout the study. Each day for
30 days the prompt was presented at the beginning of the class period as a
warm-up activity. The journals were collected weekly and at the end of the
six-week RWP for transcription and analysis.
The treatment and control groups generally reflect the ethnic
composition of the school (Figure 3.3). The students identified their own
ethnicities on the demographic portion of the self-efficacy inventory. The
categories were: (1) American Indian, (2) African American, (3) Hispanic,
(4) Asian, Pacific Islander, (5) White, and (6) Mixed. Many students had
expressed that traditional ethnic categories did not recognize their mixed
heritage, particularly students whose parents were Asian and African
American or Hispanic and African American. The mixed" category was added
to the demographic information to accommodate these students. The
differences in the ethnic demographics among the treatment and control
46


groups and the school may be a result of the addition of the mixed" category
and because some of the Hispanic students identified themselves as Native
American.
Ethnicity Treatment Group N % Control Group N % School N %
Native American 8 11.4 5 5 23 4
African American 22 31 15 21 183 32
Hispanic 16 24 27 32 161 28
Asian or Pacific Islander 1 1 1 1 6 1
European 4 8 21 27 103 18
American
Mixed 19 27 10 13 97 17
Figure 3.3. Ethnicity of treatment and control groups and school.
After preliminary procedures, data were examined for 70 students in
the treatment group, 30 males and 40 females, and 79 students in the control
group, 38 males and 41 females (Figure 3.4).
47


Gender Treatment Group N % Control Group N % N Total %
Male 30 44 38 56 68 46
Female 40 50 41 50 81 54
Total 70 47 79 53 149 100
Figure 3.4. Gender by group.
This study had a structured interview survey research measurement
component to deepen understanding of middle school learners characteristic
self-referent processes: self-efficacy, goal orientation, attributions for success
or failure, and intrinsic motivation; and self-regulated learning strategy use:
goal setting, planning and time management, and persistence and effort.
Twelve students from the reflective writing group and twelve from the control
group were selected for observation and interviews. Six boys and six girls
were chosen from the experimental groups and the control group. These
students were chosen by their results of pre-test inventory (Appendix D):
three girls and three boys who scored high on the inventory and three girls
and three boys who scored low on the measure in each group. These
students participated in interviews with the researcher and were observed for
48


strategy use. Both groups were observed throughout the study. The selected
students were interviewed after the post-study inventory was administered.
Variables
Independent Variables
The independent variables for this 2x2x2 study are method, gender,
and time. Each of these has two levels. Method had two levels (1)
participation in the guided RWP and (2) no participation in the guided RWP.
The treatment group responded to the reflective writing prompts daily for six
weeks. Each day the writing prompt (Appendix B) was placed on the
overhead projector. The students responded to the question during the first 5
minutes of class, the class time normally used for sponge" or warm-up
activities. Monday through Thursday the prompts were related to academic
self-efficacy and self-regulated learning strategy use. On Fridays, the prompt
asked students to evaluate their answers for the week and write about what
they learned about themselves from their answers for the week.
Gender was an attribute independent variable. The two levels of
gender were male and female students. The third independent variable was
time also had 2 levels. Level 1 of time was the pre-test inventory scores.
The pre-test inventory was administered to the treatment group and the
49


control group before the treatment group began participation in the guided
RWP. The second level of time was the post-test inventory scores. The post-
test inventory was administered to the treatment and control groups at the
end of the six-week RWP treatment.
Dependent Variables
The self-referent characteristics for which the data were analyzed
were: external attributions for success or failure, goal orientation, ability
attributions, and intrinsic motivation. Goal orientation is the learners view of
the goal of learning task as task or learning oriented. External attributions are
the learners perceptions of the success or failure, i.e., learners may attribute
success or failure to a variety of causes, such as persistence and effort, luck,
or ability (Dweck, 1975; Schunk, 1994; Schunk & Zimmerman, 1994). Intrinsic
motivation is derived from the task and self-efficacy beliefs about completing
that particular task (Bandura, 1997; Csikszentmihalyi, 1990). Ability
attributions are the skills and knowledge to which the learner attributes
success or failure (Dweck, 1975; Bandura, 1997).
The self-regulating strategies identified as dependent variables were
goal setting, planning and time management, and persistence and effort. Goal
setting targets both short- and long-term goals. Closely related to goal setting
50


were planning and time management, processes in which learners establish
timetables for task completion. Finally, persistence and effort are the learners
continuous engagement in a learning activity (Schunk & Zimmerman, 1994,
1998).
Instrumentation
An 5-point Likert-type inventory of self-referent constructs i.e.
academic self-efficacy, goal orientation, intrinsic motivation, attributions for
success or failure, and effort management (Appendix D) was adapted from
two existing measures, the Morgan-Jinks Student Efficacy Scale
(MJSES) (Jinks & Morgan, 1996) and the Childrens Academic Intrinsic
Motivation Inventory (CAIMI) (Gottfried, 1982). The measure used a five-point
Likert continuum ranging from strongly disagree (1) to strongly agree (5).
Children's Academic Intrinsic Motivation Inventory
(CAIMI1
The CAIMI (Gottfried, 1982) was developed to measure enjoyment of
learning, interest in novelty, curiosity, persistence in difficult tasks, and
feelings of competence, mastery, and challenge. Two validation studies were
conducted on the instrument (Gotfried, 1982). Internal consistency and test-
51


retest reliability were calculated for both studies. Gotfried (1982) found that
coefficient alpha ranged from .70 to .83, depending on the academic content
area, in the first study; and with the addition of several items addressing
content areas for the second study, coefficient alpha ranged from .74 to .96
when compared for reading, math, social studies, science, and general
learning. These coefficients were consistent across grades, genders, and
ethnicities. Her findings indicated high internal consistency within the scales.
Gottfrieds (1982) findings indicated moderately high stability over the
two-month test-retest period. Test-retest reliability for the CAIMl was
estimated through Pearson product-moment correlations between the initial
and second studies conducted in the development of the CAIMl. Gottfried
found that the correlations in the initial study one ranged from .66 to .76 for
qc.01 for reading, mathematics, social studies, science, and general
academic learning. The correlations for the second study ranged from .68 to
.80 for pc.01 for the same academic programs. These correlations were
consistent across grade levels, between genders, and among ethnicities.
Gotfried (1982) presented findings for criterion-related and construct
validity. She found that the CAIMl correlated with a wide variety of
achievement measures and school-related variables. The construct validity
was demonstrated through correlations with anxiety and related measures.
52


Morqan-Jinks Student Efficacy Scale (MJSES)
The Morgan-Jinks Student Efficacy Scale (MJSES) (Morgan & Jinks,
1996) was designed to help educators gain insight into grade 4-8 learners
self-perceptions of their academic performance. Morgan and Jinks (1996)
administered the MJSES at three field test sites. The student population in the
urban site was 100% African-American, and 100% of the students
participated in the free and reduced lunch program. The second site was a
suburban school in which 12% of the students were identified as belonging to
ethnic minorities and 19% participated in the free and reduced lunch program.
The third site was in a rural setting, all the students in this setting were
Caucasian and 12% were from low-income families.
The factor analysis conducted by Jinks and Morgan revealed three
major factors: talent, context, and effort. An item analysis was conducted, all
items with an item correlation below .30 were eliminated from the inventory.
This resulted in a 30-item scale with an internal consistency coefficient of .82.
The sub-scale coefficients were talent, .78; context, .70; and effort, .66. The
students scores correlated with grade achievement. No test-retest findings
were presented.
The authors of the MJSES used three panels to estimate content
validity. The first panel consisted of five university teacher educators. The
53


second panel was four middle school teachers, and the third panel consisted
of 15 grades 4-8 teachers. Two groups of children, one from grades 4 and 5
and one from grades 6-8, were asked to determine if the items on the test
were readable, clear, and relevant to their school experience. Items that were
categorized low were rewritten or eliminated. The resulting Likert-type scale
consisted of 53 of the original items and 4 items requesting grade
performance.
Study Measure
The measure for this study (Appendix D) was adapted from questions
from the MJSES and the CAIMI. Demographic data for gender, ethnicity, and
age were collected. Students were asked their favorite subject and the
reason for choosing that subject: Was it their favorite because it was
challenging, interesting, or easy? The students were asked to describe their
grades as mostly As, Bs, Cs, Ds, or Fs.
The students were asked to complete the 59-question Likert-type
inventory. The survey items were categorized according to the self-referent
construct to be measured. The statements that measured intrinsic motivation
were adapted from the CAIMI. The items that measured attributions, goal
orientation, persistence and effort, and ability attributions were adapted from
54


the MJSES. Morgan and Jinks (1996) identified the items categorized as
persistence and effort and ability. A panel of 19 teachers categorized the
remaining items as attributions and goal orientation.
Self-Referent (Dependent Variable) Survey Items
External Attributions I would get better grades if my teacher liked me better.* I go to a good school. My teacher thinks I am smart. No one cares if I do well in school.*
Goal Orientation I will graduate from high school. Adults who have good jobs probably were good students when they were kids. When I am old enough, I will go to college. It is important to go to high school. What I learn in school is not important.* I keep working on a problem until I understand it because I want to please my parents.
* polarity reversed for coding and analysis
Figure 3.5. Survey items by self-referent component.
55


Self-Referent (Dependent Variable) Survey Items
Intrinsic Motivation 1 enjoy doing new work in school because 1 want to get good grades.* 1 enjoy doing new work in school because 1 like a challenge. 1 enjoy doing new work in school because 1 want to please my teacher.* 1 enjoy learning new things because 1 want to please my parents.* 1 enjoy learning new things because 1 want to please my parents.* 1 like classes that are challenging. 1 enjoy new things because 1 am curious. 1 enjoy doing new work in school because 1 want to please my parents.* 1 enjoy doing new work in school because 1 like a challenge. 1 do not feel good inside when 1 know 1 have learned something new.* 1 keep working on a problem until 1 understand it because solving problems is interesting to me. 1 enjoy learning new things because 1 want to get good grades. 1 enjoy learning new things in Math Science Social Studies Language Arts Reading 1 enjoy learning new things. 1 like to learn. 1 dont like to work on new problems.*
* polarity reversed for coding and analysis
Figure 3.5. cont.
56


Self-Referent (Dependent Variable) Survey Items
Intrinsic Motivation When 1 get bored, 1 do not look for new things to do.* When 1 dont have new things to do in school, 1 get bored. New ideas are not interesting to me.* When 1 know 1 have learned something new, 1 feel good inside.
Persistence and Effort 1 work hard in school. 1 could get my best grades if 1 tried enough. Most of my classmates work harder on their homework than 1 do.* 1 always get good grades when 1 try hard. 1 keep working on a problem until 1 understand it so that 1 can get a good grade. 1 like to review work 1 already know. *1 keep working on a problem until 1 solve it. 1 try to learn more about something that 1 dont understand right away so that 1 will understand it. 1 get bored when 1 dont understand new things to do in school.* When 1 dont understand a problem, 1 give up right away.* When 1 cant understand something right away, 1 try to learn more about it so 1 can understand it.
Ability Attributions 1 am a good science student. Sometimes 1 think an assignment is easy when my classmates think it is hard. 1 am a good social studies student. 1 am one of the best students in my class. 1 am a good math student. My classmates get better grade that 1 do*
* polarity reversed for coding and analysis
Figure 3.5. Cont.
57


Self-Referent (Dependent Variable) Survey Items
Ability Attributions I usually understand my homework assignments. I am a good language arts student. I usually do not get good grades because school is too hard.* I am a good reader.
* polarity reversed for coding and analysis
Figure 3.5. Cont.
Each of the writing prompts and interview questions were developed to
address the self-referent processes and self-regulatory strategies (Figure
3.6). The panel of 19 teachers coded the writing prompts and interview
questions for the self-referent constructs, academic self-efficacy (SE), goal
orientation (GO), attributions for success or failure (SA), and intrinsic
motivation (IM). The students responses were also coded for self-regulated
learning strategies, goal setting (GS), time management and planning (TM),
and persistence and effort (PE). This panel then coded student responses for
the self-referent processes and self-regulated learning strategies stated by
the students. Figure 3.5 presents the dependent variables addressed by
each writing prompt and interview question.
58


Self-Referent and Self- Regulated Learning Strategy (Dependent Variable) RWP Prompts and Interview Questions
Academic Self-Efficacy Writing Prompts: Things 1 can do as a writer: What is something you can do now that you couldnt do last year? Look at your answers for this week and think about the writing assignments you worked on this week. What do your answers tell you about yourself? Describe what your answers and the ways you worked on the activities tell you about yourself. What could you teach others to do? How would you teach them? Describe yourself as a student. 1 am an important part of my class because... One thing 1 really like about my writing is... What did you have trouble with on your writing project? Interview Questions: What is one thing that you feel you need to work on for your next project?
Goal Orientation Interview Questions: Thinking about the writing project you just completed, what did you learn?
Figure 3.6. Prompts and interview questions.
59


Self-Referent and Self- Regulated Learning Strategy (Dependent Variable) RWP Prompts and Interview Questions Writing Prompts: When my writing is going well, 1 feel... Writing Prompts:
Attributions for Success or Failure 1 am proud about my accomplishment in because... 1 am aettina better at because... When 1 am successful at a project, it is because... 1 could be a better writer if 1
Intrinsic Motivation Writing Prompt: Name 1 thing you enjoy learning about and explain why. What has been a great learning experience in your life? Why? I like to write about...
Goal Setting Writing Prompt: What are your goals for writing? Interview Question: How can you improve your next writing project?
Planning and Time Management Writing Prompts: What do you do when you are learning something new that you are good at? Explain your answer. How will you achieve your goal? Writing Prompts: What steps do you take in completing homework assignments? What do you do when you are learning something new?
Figure 3.6. cont.
60


Self-Referent and Self- Regulated Learning Strategy (Dependent Variable) RWP Prompts and Interview Questions
Planning and Time Management What do you do to organize for in class assignments? What do you do to begin a writing assignment? What can you improve in your study skills? What do you do when you are trying to learn something difficult? Explain your answer. Interview Questions: What did you do to get ready to complete this writing project? What things would you do differently on your next writing project? What could you do to make your next writing project easier for you to do?
Figure 3.6. cont.
Data Collection Procedures
Students in both sixth-grade groups completed a self-report, Likert-
type inventory for self-referent constructs (Appendix D). The inventory
measured the dependent variables: goal orientation, attributions for success
or failure, intrinsic motivation, persistence and effort, ability attributions, and a
total score for academic self-efficacy. The inventory was administered to both
61


groups of sixth-grade middle school students at the beginning and at the end
of this six-week study.
The sixth-grade students from the treatment and control groups were
asked to complete the survey during class by their regular classroom
teachers. The inventories were administered to all the students on the same
day. The classroom teachers read the instructions to the students. The
inventory was administered the week before the onset of the treatment. The
treatment group began the RWP on the Monday following the administration
of the pre-test inventory.
This study had a structured interview survey research measurement
component to provide additional data to deepen understanding of middle
school learners characteristic self-referent processes: self-efficacy, goal
orientation, attributions for success or failure, and intrinsic motivation; and
self-regulated learning strategy use: goal setting, planning and time
management, and persistence and effort. Twelve students from the reflective
writing group and twelve from the control group were selected for observation
and interviews. Six boys and six girls were chosen from the experimental
groups and the control group. These students were chosen by their results of
pre-test inventory (Appendix D): three girls and three boys who scored high
on the inventory and three girls and three boys who scored low on the
62


measure in each group. These students participated in interviews with the
researcher and were observed for strategy use by the researcher. Both
groups were observed throughout the study. The selected students were
interviewed after the post-study inventory was administered.
The three boys and three girls with the highest scores on the pre-test
measure and three boys and three girls who scored the lowest on the pre-test
were selected from the treatment group for participation in the interviews. The
control group had 12 students chosen in the same configuration as the
experimental group. The interviews were conducted after the post-test survey
was administered. The 24 students from both the RWP and the control
groups were observed throughout the study. This researcher interviewed
each of the students individually in a quiet, private office. The interviews were
audio-taped for the purpose of later transcription. The students were
informed that the interview would be taped, and they were assured that the
tapes would be heard by the interviewer only. Each student was asked each
of the 7 interview questions (Appendix C).
Observations were conducted throughout the six-week RWP
treatment. Each of the students identified for interviews were observed 4
times. The observations were conducted in 5-minute sweeps. The behaviors
63


exhibited during the each of the sweeps was recorded on the observation
rubric (Appendix E).
The writing prompts (Appendix B), interview questions (Appendix C),
and observation rubric (Appendix E) addressed the learners academic self-
efficacy beliefs, goal orientation, and intrinsic motivation; as well as the self-
regulated learning strategies, goal setting, planning and time management,
and persistence and effort.
Data Analysis Procedures
All data were examined to determine if there were gender differences
in self-perceptions of academic self-efficacy and the self-efficacy
components: goal orientation, attributions for success or failure, and intrinsic
motivation. These data were compared for boys and girls in each of the
treatment and control groups and between the experimental and control
groups.
T Tests
T tests were conducted on the mean scores of the treatment and
control groups. The t tests compared the mean scores for the pretests for
total test and the sub-scales for external attributions, goal orientation, intrinsic
64


motivation, persistence and effort, and ability attributions. The hypothesis that
was tested was that two population means for the treatment and control
groups are equal.
Repeated Measures Factorial ANOVA
The repeated measures factorial ANOVA was employed to increase
the detection of treatment effects. The repeated measures design requires
independent groups of participants of the treatment but treats a combination
of variables which, in turn, allows for a smaller number of participants. This
analysis of variance treats each variable as a within-subjects variable and
eliminates from estimates of error variance due to individual differences
across the levels of the variables. Because of this power is gained. In
addition, a repeated measures ANOVA tests for differences in means among
measures of main effects in the treatment by measure (Gabriel & Hopkins,
1974; Kesselman, Huberty, Lix, Olejnik, Cribbie, Donahue, Kowalchuk,
Lowman, Petrosky, Kesselman, & Levine, 1998).
Interviews and Reflective Writing Prompts
The experimental group students responses to reflective writing
prompts, interviews, and observations were examined for student perceptions
65


of self-efficacy, goal orientation, attributions for success or failure, and
intrinsic motivation. The interview responses and observations were
examined for self-regulated learning strategies: goal setting, planning and
time management, and persistence and effort.
The SSI data yielded information about how the students viewed
themselves as learners and how they proceeded when engaging in learning
tasks at school. The SSI data examined for the treatment group included
responses to the RWP prompts, observational data, and responses to
interview questions. Seventy-five students in the treatment group responded
to the RWP prompts. Twelve students from the treatment group were
interviewed. These 12 students responses to the RWP prompts were
examined. Their responses to the writing prompts were compared with their
interview responses to assess similarities in their views of themselves as
learners. Responses to the prompts and interview questions were also
evaluated to determine if the student could identify which self-regulated
learning strategies they used to complete a learning task. Observational data
gathered from the all students who were interviewed were analyzed.
Structured interview survey data were collected through students
responses to the writing prompts and interviews. Responses for the
interviews and self-assessments were coded for the characteristics of self-
66


referent constructs and strategies of self-regulation (Appendix F). Analyses of
the writing prompts, observations, and student self-assessments were
conducted for the purpose of determining the effects of reflective writing on
self-efficacy, goal orientation, attributions for success or failure, and intrinsic
motivation. An analysis of the interviews conducted at the end of the study
was completed using the same codes.
Cross tabs were calculated on the frequency of responses related to
the frequency of responses related to academic self-efficacy and the self-
efficacy constructs and self-regulated learning strategies examined in this
study. Chi square of association tests were conducted on the interview
responses from the treatment and control groups and on the treatment
groups responses to the reflective writing prompts. These procedures were
used to determine associations between responses and group and/or gender.
Reliability and Validity
To address concerns of internal validity for the inventory a panel of 19
educators, school administrators and counselors, teacher educators, and
educational consultants were trained in the operational definitions used for
this study. These professionals were asked to categorize the statements on
the academic self-efficacy inventory and the interview questions. The
67


categories that evolved from the inventory were used for the dependent
variables for the repeated measures factorial ANOVA. The calculation for the
inter-rater agreement for the inventory was J!ctuala$Teeme'}t?..,. where the actual
possibleagreements
agreements were 975 and the possible agreements were 1121. The inter-
rater agreement quotient was 87%.
A reliability analysis was calculated to determine the extent to which
the items in the self-efficacy inventory were related to each other.
Chronbachs alpha was calculated to determine internal consistency. This
reliability analysis is based on the average inter-item correlation. Intra-class
correlation coefficients can be used to compute interrater reliability estimates
(Crocker & Algina, 1986). A Chronbachs alpha reliability analysis was
calculated for each sub-scale. The results for the external attributions, .30,
intrinsic motivation, .51, and effort and persistence, .41, indicated a moderate
degree of internal consistency. The calculated Chronbachs alpha on the
sub-scales for goal orientation and ability attributions yielded .78 and .77
respectively. These statistics indicated a high degree of internal consistency.
The reflective writing prompts and interview questions targeted the
students perceptions of their own learner characteristics and strategy use.
Responses were coded using the following codes: SE, self-efficacy: GO, goal
orientation, SA, attributions, IM, intrinsic motivation, GS, goal setting, TM,
68


time management and planning, and PE, persistence and effort. Interviews
and observations of the 24 students responses chosen for structured
interview survey research analysis were audio-taped and coded using the
same codes. Strict adherence to the operational definitions and consistency
in coding student behaviors and responses aided in increasing inter-rater
reliability. Frequency counts of the self-referent constructs academic self-
efficacy, goal orientation, attributions for success or failure, and intrinsic
motivation were drawn from the interviews and RWP responses.
To estimate inter-rater reliability in the analysis of interview questions,
four teachers were trained for coding the self-referent processes and self-
regulated strategy use. The training consisted of discussions of the
operational definitions and coding example responses. Each teacher
volunteer and the researcher coded the students responses to writing
prompts and interview questions. The inter-rater agreement was calculated
using the formula, actuala8reeTrtents. The actual agreements were 572 and the
possibleagreements
possible agreements were 680. This yielded an 84% agreement among the
coders. For both the categorizing of inventory items and the responses to the
interview, there was a high degree of inter-rater agreement.
69


Summary
This study compared the self-referent constructs of academic self-
efficacy and the components of goal orientation, attributions for success or
failure, persistence and effort, and intrinsic motivation and the use of self-
regulated learning strategies between a quasi-experimental group and a
control group. The subjects were sixth-grade middle school students from a
school in a suburban setting in Colorado. The students in the experimental
group participated in a reflective writing program. Although the students in the
control group maintained writing journals, their prompts were related to self-
efficacy, its components, or self-regulated learning strategies.
Preliminary analyses included t-tests on the pretest inventory and sub-
scale means and violations of assumptions for the repeated measures
factorial ANOVA. An examination of data for outliers was conducted.
Repeated measures factorial ANOVA procedures were conducted on pre-test
and post-test data gathered from the academic self-efficacy inventory.
Students from both groups were observed to identify self-regulating
strategy use. Twelve students from each group participated in interviews at
the end of the treatment period. The experimental groups student reflective
writing, and the interviews and observations of students from both groups,
were coded and examined to determine relationships between the treatment
70


and control groups awareness of themselves as learners and their use self-
regulated learning strategies.
Reliability coefficients indicated a high degree of interrater reliability,
87% for the self-efficacy inventory and 84% for the interview questions. The
obtained coefficient from Chronbachs alpha .79 on the pre-test and .84 on
the post-test indicate a high degree of internal consistency on the self-efficacy
inventory. Chronbachs alpha on the sub-scales of the inventory indicated
71


CHAPTER 4
RESULTS
This chapter presents the results of the data gathered in this study.
The introduction reviews the derivation of the self-efficacy inventory used for
the study and the research questions. The first section presents
measurement strategies, the results of t- tests conducted on the pretest data,
the preliminary procedures, and the obtained descriptive statistics. The last
section presents the results of the repeated measures factorial ANOVA.
Introduction
For this study data from a self-reporting self-efficacy survey and
demographic information were analyzed. The self-efficacy inventory was
derived from two existing surveys, the Morgan-Jinks Student Efficacy Scale
(MJSES) (Jinks & Morgan, 1996) and the Childrens Academic Intrinsic
Motivation Inventory (CAIMI) (Gottfried, 1982). The survey was administered
as a pre-test and a post-test to determine the effects of: a) participation in a
guided reflective writing program (RWP) and 2) gender on academic self-
efficacy beliefs. The factors under consideration were student pre-test and
72


post-test self-efficacy beliefs, gender, and group (treatment or control). The
dependent variables are the pre-test and post-test scores for student self-
efficacy. The independent variables are gender (male and female), group
(treatment and control), and time (pre-test and post-test). The analyses were
conducted to determine if there were main effects on student academic self-
efficacy components, attributions for success or failure, goal orientation, and
intrinsic motivation; the self-regulated learning strategy effort and persistence
by gender, group, and/or time.
These factors lead to the following research questions:
What are the relative effects of participation in a guided RWP on
student academic self-efficacy beliefs?
What are the relative effects of participation in a guided RWP on self-
regulated learning strategy use?
Are there significant differences between male and female students in
terms of academic self-efficacy?
Are there significant differences between male and female students in
terms of self-regulated learning strategy use?
Are there significant interactions between: participation in a guided
RWP and gender in academic self-efficacy beliefs, participation in a
73


guided RWP and group, and participation in a guided RWP, gender,
and group?
Measurement Strategies and Preliminary Procedures
Prior to conducting the repeated measures factorial ANOVA,
quantitative data were reviewed to determine which students had moved to
other schools during the six weeks between the pre-test and post-tests. The
data were also examined for outliers (Leik, 1997; Lindsey, 1999; Hair,
Anderson, Tatham & Black, 1998). The box plots on the pre-test data by
gender showed that the scores of 2 male students and 2 female students had
values that were outliers. Box plots for the post-test data indicated that the 2
male students who had outlier scores were members of the treatment group,
and that the female student who had an outlier score was from the control
group. These outlier scores were deleted from the data for subsequent
analyses as recommended by Leik (1997) and Lindsey (1999).
The Levene test for homogeneity (Table 4.1) was conducted with the t-
tests on the total pretest score and the sub-scales for (a) external attributions,
(b) goal orientation, (c) intrinsic motivation, (d) persistence and effort, and (e)
ability attributions. The null hypothesis: H0: the variance of the mean scores
of the treatment group is equal to the variance of the mean scores of the
74


control group on the total pre-test scores and each of the sub-scales was
retained because there were no significant differences indicated: total pretest
score, F(1,147)=,26, p=.62; external attributions, F(1,147)=.30, p=.59; goal
orientation, F(1,147)=.25, p=.62; intrinsic motivation, F(1,147)=.86, p=.35;
persistence and effort, F(1,147)=. 16, p=.70; and ability attributions,
F(1,147)=.001, p=.97.
Source dfl df2 F Si&
ATT Pre-test 1 147 .295 .59
GO Pre-test 1 147 .253 .62
IM Pre-test 1 147 .863 .35
PE Pre-test 1 147 .155 .70
AAT Pre-test 1 147 .001 .97
ATT=External Attributions, GO=Goal Orientation, IM=lntrinsic Motivation,
PE=Persistence and Effort, AAT=Ability Attributions
Table 4.1 Levene Test of Homogeneity on Pre-Test bv Group
The t tests (Table 4.2) were conducted to test the null hypothesis that
the means for the pretests were equal for the treatment and control groups.
The t tests indicate that there was no significant difference in the mean score
75


between the treatment group and the control group: total pretest score,
t(147)=.77, p=.44; external attributions, t(147)=-1.4, p=.16; goal orientation,
t(147)=.36, p=.72; intrinsic motivation, t(147)=.03, p=.98; persistence and
effort, t(147)=.30, p=.77; and ability attributions, t(147)=1.8, p=.07. These
results suggest that there were no statistically significant differences of the
means between the treatment and control groups at the onset of the study.
Measure M RWP Darticipation SD M No RWP participation SD t (147)
ATT 7.4 2.2 7.9 2.1 -1.4
GO 8.9 2.3 7.7 2.0 .36
IM 75 12.3 74.2 11.7 .03
PE 19.4 3.1 19.8 2.6 .30
AAT 30.5 5.1 30.0 5.4 1.8
ATT=External Attributions, GO=Goal Orientation, IM=lntrinsic Motivation,
PE=Persistence and Effort, AAT=Ability Attributions
Table 4.2 T-Tests on Pre-Tests bv Group
The descriptive statistics were calculated on the pre-test and post-test
scores for gender and group (Table 4.3, Table 4.4, and Table 4.5).
76


Participation in RWP No RWP Participation
Pre-test Post-test Pre-test Post-test
Construct
by gender M SD M SD M SD M SD
Male
ATT 7.3 2.2 9.4 2.2 7.9 2.1 7.8 2.0
GO 72.5 10.4 73.7 13.8 73.0 10.2 72.1 13.2
IM 62.1 7.3 63.2 5.7 61.4 7.2 62.7 7.5
PE 19.5 2.8 20.0 3.5 19.3 2.7 19.5 3.0
AAT 32.5 4.7 29.2 4.6 31.5 5.5 33.2 4.8
ATT=External Attributions, GO=Goal Orientation, IM=lntrinsic Motivation,
PE=Persistence and Effort, AAT=Ability Attributions
Table 4.3 Pre- and Post-Test and Standard Deviations for Males
77


Participation in RWP No participation in RWP
Pre-test Post-test Pre-test Post-test
Construct
by gender M SD M SD M SD M SD
Female
ATT 7.5 2.2 8.4 1.9 7.1 1.9 7.3 2.0
GO 74.9 9.2 76.1 11.1 73.7 8.0 76.1 10.0
IM 62.0 4.9 61.6 4.8 62.6 6.4 62.9 6.4
PE 19.9 2.4 19.0 2.8 19.3 2.2 20.0 2.2
AAT 35.1 5.2 31.4 5.3 33.2 4.8 30.8 4.3
ASE 211.3 17.1 207.2 19.3 209.0 15.0 207.4 18.1
ATT=Attributions, GO=Goal Orientation, IM=lntrinsic Motivation,
PE=Persistence and Effort, AAT=Ability Attributions, ASE=Total Academic
Self-Efficacy Score
Table 4.4 Pre- and Post-Test Means and Standard Deviations for Females
78


Participation in RWP Non-Participation in RWP
Pretest Post-test Pretest Post-test
ATT 7.5 9.0 8.0 7.9
GO 74.4 73.9 74.5 74.5
IM 62.2 61.8 61.3 63.1
PE 19.8 19.5 19.5 19.8
AAT 34.1 30.2 32.1 29.5
ATT=Attributions, GO=Goal Orientation, IM=lntrinsic Motivation,
PE=Persistence and Effort, AAT=Ability Attributions
Table 4.5 Pre-Test and Post-Test Means and Standard Deviations bv Group
Box's test of equality of covariance was conducted to determine
significance of homogeneity of covariance. This test is used to test the
hypothesis that the covariance of the dependent variables are equal across
groups. The statistic was calculated on the covariance for gender, team, and
time. The resulting statistic, F(234, 40139.14)=1.126, p=.092 indicated that
there was no significance among the scores for the covariances in the
dependent variables.
79


The Levene test for homogeneity (Table 4.6) was conducted on the
pre-test and post-test data because of the unequal numbers of male and
female students in both the treatment and control groups and the differences
in numbers of participants in the treatment and control groups. The null
hypotheses for the Levene test of homogeneity of variance are: H0: the
variance of the mean scores of the treatment group is equal to the variance of
the mean scores of the control group on the pre-test; H0: the variance of the
mean scores of the treatment group is equal to the variance of the mean
scores of the control group on the post-test; H0: the variance of the mean
scores of the male participants is equal to the variance of the mean scores of
the female participants on the pre-test; and H0: the variance of the mean
scores of the male participants is equal to the variance of the mean scores of
the female participants on the post-test.
The Levene test of homogeneity indicated that there were no violations
of homogeneity of variance for the pre-tests on the dependent variables of:
attributions for success or failure, F (3,145)=. 12, p=.95; goal orientation, F(3,
145)=1.13, p=.34; intrinsic motivation, F(3, 145)=1.87, p=.14; persistence and
effort, F(3,145)=1.48, p=.22; and ability attributions, F(3, 145)= .59, p=.62at
a=.05. Similarly, the post-test scores for the total academic self-efficacy
score, F(3,145)=1.55, p=.20; attributions for success or failure, F (3,
80


145)=2.25, p=..09; goal orientation, F(3,145)=2.43; intrinsic motivation,
p=.07; F(3,145)=1.87, p=.14; persistence and effort, F(3,145)=1.19, p=.32
indicated that the null hypotheses for these dependent variables should be
retained.
However, The Levene statistic for post-test intrinsic motivation
variances F(3,145)=3.13, p=.03 and ability attributions, F(3,145)=3.21, p=.03
indicated a violation of the assumption of homogeneity of variance for these
constructs. Because the null hypotheses for significant differences among
gender, method, and time, for the repeated measures factorial ANOVA for
intrinsic motivation was retained, this violation of the homogeneity of variance
did not pose a problem for further analyses. The Levene statistic for the post-
test ability attributions warranted concern because the repeated measures
indicated significance for time and gender.
81


Source dfi df2 F Sio.
ATT Pre-test 3 145 .117 .950
ATT Post-test 3 145 2.250 .085
GO Pre-test 3 145 1.132 .338
GO Post-test 3 145 2.433 .067
IM Pre-test 3 145 1.867 .138
IM Post-test 3 145 3.125 .028*
PE Pre-test 3 145 1.483 .222
PE Post-test 3 145 1.193 .315
AAT Pre-test 3 145 .590 .623
AAT Post-test 3 145 3.210 .025*
* significant at p<.05
ATT=Attributions, GO=Goal Orientation, M=lntrinsic Motivation,
PE=Persistence and Effort, AAT=Ability Attributions
Table 4.6 Levene's Test of Homogeneity of Variance
Repeated Measures Factorial ANOVA Results
The repeated measures factorial ANOVA was conducted for Time,
Group, Gender, Time x Group, Group x Gender, Time x Gender, and Time x
Group x Gender on the dependent variables attributions for success or failure,
82


goal orientation, intrinsic motivation, persistence and effort, and ability
attributions (Table 4.7-Table 4.11).
External Attributions for Success or Failure
The repeated measures factorial ANOVA yielded significant results for
attributions for success or failure (Table 4.7). Statistical significance was
found for Time, F(1,145)=9.53, p=.002, Timex Group, F(1,145)=15.616,
p<.001, and Gender x Time, F(1, 145)=4.59, p=.04. Therefore, the alternate
hypotheses, Time, H0: pPretes^ipoto*- Time x Group, H0: !*,****!**>** and
Mreatment^M-Cntroli nd Time by Gender, H0: M-Pretest^M-postest^1"1^ Female Were
accepted. The results for Group F(1,145)=1.643, p=.20; Gender F(1,
145)=2.431, p=.12; Gender x Group F(1,145)=.002, p=.97; and Time x Group
x Gender F(1,145)=.311, p-.578 indicated no statistical significance, so the
null hypotheses for attributions for success or failure for Group, H0:
M'TreaBnent=M'Controi> Gender, H0: {iMaie=FFemaie. Gender x Group ; Group x Gender,
H0: Mtreatmant:~Mcontroi and H0: and Time x Group x Gender, H0:
fAPretest=lAPostest and H0: MTreatment=M'Cortr0i and H0. M-Maie=M'Femaie are retained because
there are no statistically significant differences for these hypotheses.
83


Source df S£ MS F Stq,
PT 1 33.543 33.543 9.531 .002**
T 1 9.116 9.116 1.643 .20
G 1 13.491 13.491 2.431 .12
T X PT 1 54.96 54.96 15.616 <.001**
G X PT 1 16.169 16.169 4.594 .04*
G XT 1 1082E-02 1082E-02 .002 .97
PT XT X G 1 1.096 1.096 .311 .578
Error 145 804.58 5.549
*pc.05, ** fx.01
PT=Time (pretest/post test), T=Group, G=Gender
Table 4.7 Repeated Measures Factorial ANOVA for External Attributions for
Success or Failure
Plots of the pre- and post-test means for external attributions showed
significant interactions for the treatment and control groups (Figure 4.1).
Figure 4.1 indicates that means of the treatment group increased their scores
for the sub-scale for external attributions to a greater extent than the means
for the control group.
84


Estimated Marginal Means of ATT
FACTOR 1
Factor 1,1=pre-test, 2=post-test
Figure 4.1. interaction of external attributions between groups
Goal Orientation
The ANOVA for the dependent variable goal orientation yielded no
statistically significant results (Table 4.8). The null hypotheses for goal
Orientation, H0. M-pretest-Fpostest> ^0- ^Treatment-Fcontroti ^0- FMale=FFemalei 31" retained
because there were no statistically significant differences in the means. The
null hypotheses for Time x Group H0: ^^=11****, and nTreatment=Fcniro1; Time x
85