Citation
The role of self-efficacy in paraeducators' career decisions

Material Information

Title:
The role of self-efficacy in paraeducators' career decisions
Creator:
Sandoval-Lucero, Elena M
Publication Date:
Language:
English
Physical Description:
245 leaves : ; 28 cm

Subjects

Subjects / Keywords:
Teachers' assistants ( lcsh )
Self-efficacy ( lcsh )
Self-efficacy ( fast )
Teachers' assistants ( fast )
Genre:
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )

Notes

Bibliography:
Includes bibliographical references (leaves 230-245).
General Note:
School of Education and Human Development
Statement of Responsibility:
by Elena M. Sandoval-Lucero.

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:
57707803 ( OCLC )
ocm57707803
Classification:
LD1190.E3 2004d S36 ( lcc )

Full Text
THE ROLE OF SELF-EFFICACY IN PARAEDUCATORS CAREER
DECISIONS
by
Elena M. Sandoval-Lucero
B.A., University of Northern Colorado, 1985
M.A., University of Colorado at Denver, 1995
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 Elena M. Sandoval-Lucero
All rights reserved


This thesis for the Doctor of Philosophy
degree by
Elena M. Sandoval-Lucero
has been approved
by
I
4 5
Date


Sandoval-Lucero, Elena M. (Ph.D., Educational Leadership and Innovation)
The Role of Self-efficacy in Paraeducators Career Decisions
Thesis directed by Research Professor Nancy K. French
ABSTRACT
This mixed methods research study examined the career paths, work
environments, educational experiences, family background, career plans, and
demographic characteristics of paraeducators who became teachers and current
paraeducators. The study also measured their teacher, academic, and general self-
efficacy. The purpose of the study was to explore the possibility that self-efficacy
plays a role in paraeducators career decisions. Data was collected through three
sources: a survey, career goals statements, and interviews. Twenty-two
participants were included in the survey phase of the study, and fourteen
participants were included in the interview phase.
There were some qualitative differences and significant quantitative
differences between the groups on general self-efficacy and teacher efficacy.
However, the groups did not have differences in their academic self-efficacy. All
paraeducators received support from career ladder programs or school district
sponsored programs to attend college. They identified support from their cohorts
and their families as the biggest factors in their college success.
Paraeducators in this study who became teachers, and paraeducators who
planned to become teachers described work environments that provided mastery
teaching experiences, teachers who were social role models, and school
professionals who encouraged them to advance their careers. Mastery experiences
occurred when teachers worked collaboratively with their paraeducators to plan
and deliver instruction to students in their classrooms. Teachers also included
IV


paraeducators in other activities such as EEP work and parent teacher conferences.
These teachers were identified by paraeducators as social role models because
they provided good supervision for the paraeducator role and encouraged the
paraeducators to become teachers.
Paraeducators who planned to remain in the paraeducator role spent more
of their work day performing general clerical duties. They worked less
collaboratively with teachers to deliver instruction. They also received less
encouragement to become teachers.
The study has implications for two areas of practice. First, the quality of
supervision of paraeducators is important for their career development. Second,
the study also highlighted the need to clearly define paraeducators roles and
responsibilities in ways that utilize their skills, abilities, and interests, and
promote their career development.
This abstract accurately represents the content of the candidates tfy! ;. I
recommend publication.
Signed
Nancy
v


DEDICATION
This dissertation is dedicated in loving memory,
To my grandmother, Amada P. Sandoval
She was not able to finish her education.
Yet, she dreamed of a better life for her ten children.
She gave them the ability to realize her dream by encouraging and supporting
their education and by giving them the gift of literacy.
To my father, Carmel Enrique Sandoval
A teacher who not only fulfilled his mothers dreams for him, but who made his
children and his students believe in the possibilities that existed for them.
Dad, because of you, I never thought I was tilting at windmills.
And also,
To my husband, Frank J. Lucero, Ph.D.
An educator who believes all students can achieve their dreams.
You are my best friend, my greatest champion, and my source of inspiration.
Your love, support, and encouragement helped make this dream a reality for me.


ACKNOWLEDGEMENTS
This dissertation would not have been possible without the help of so
many important people in my professional and personal life. I would like to take
this opportunity to thank them for their support, encouragement, and contribution
to this undertaking.
First of all, I gratefully acknowledge my advisor and mentor Nancy
French. Who would have thought that when we met ten years ago to discuss
teacher recruitment that our paths in life would converge in this unique way?
What a lucky day that was in my life. Your support and guidance throughout this
process has been invaluable to me. Thank you a thousand times for always being
there when I needed to talk something through, share an exciting outcome, or just
have a brief moment of panic. Most of all, I want to thank you for introducing me
to a very special group of people who not only stimulated my intellectual
curiosity, but who captured my heart, paraeducators. Your passion for and
commitment to this field is inspiring.
I also wish to thank the rest of my committee members: Lynn Rhodes,
Andy Helwig, Sally Nathenson-Mejia, and Bob Leonetti. Each of you made your
own unique contribution to this dissertation. Without your support, advice, and
constructive feedback I could not have completed this project. It was a pleasure to
work with such a dedicated group of educators.
My family members have always created a strong support system for me.
Since they have been hearing me talk about this process for four long years, they
deserve a very special thank you. Frank, my husband, my friend, you knew I
needed to do this even before I did. Thank you for letting me realize it on my
own. This experience has been even better because I was able to share it with you.
Thank you for always believing in me. To my mother, Neva, your love is the
definition of unconditional positive regard. My whole life you have always been
there for me. Thank you for encouraging me, for telling me that I could be
whatever I wanted to be, for respecting my choices, and for always being there
when I needed you. You are a truly amazing and special person. Last but certainly
not least, I want to thank my sisters. Loretta, your passionate commitment to
lifelong learning has been an inspiration to me. Lynda, you have always had the
courage to follow your dreams. You may be my tiny baby sister, but you will
always be my hero.
There are also several colleagues who provided support, encouragement,
and resources to help me complete this study. A special thank you goes to
Lorenso Aragon, Helen Berg de Balderas, Christine Tanguay, and Debbie


Tschida. You helped me identify study participants and smoothed the way for me
to contact them. Without you, this study would not have been possible. Your
assistance is greatly appreciated. To my friend and colleague Ritu Chopra, words
cannot express the depth of my gratitude to you for your mentorship, advice,
encouragement, and friendship throughout each step of this process. I cannot
begin to tell you how much your support meant to me. I am sure we were sisters
in another life. My appreciation goes to Jamie Finn, who along with Ritu, served
as my raters for interrater reliability. Thank you for taking the time out of your
busy schedules to contribute to my study. I would also like to give a big thank you
to Willie Hepworth for his technical assistance with my statistical analyses. My
sincere appreciation also goes to my editor, Jamie Dawkins, for her excellent
work in helping me put this study into its final format.
I also wish to thank several of my fellow students. To the Paraeducators
in Education doc lab members, your support and encouragement throughout this
process is greatly appreciated. To my friends Karen Johnson and Maria Uribe, we
started this program together, and we finished it together. My journey through the
process was better because I was able to share it with the two of you. I wish you
luck with all your future endeavors.
Finally, I wish to thank the teachers and paraeducators who participated in
this study. You were the key to the success of this project. Your willingness to
give of your precious time and share your experiences was incredible. Your
passion for the work you do was evident in your words. The children of Colorado
are lucky to have such dedicated people working with them.


CONTENTS
Figures.......................................................xv
Tables........................................................xvi
CHAPTER
1. INTRODUCTION.................................................1
Statement of the Problem.................................1
Background and Significance of the Problem...............4
Conceptual Framework.....................................8
The Research Questions..................................14
Operational Definitions.................................14
Overview of Methodology.................................17
Overview of Remaining Chapters..........................22
2. REVIEW OF LITERATURE........................................25
Introduction............................................25
Introduction to Self-efficacy...........................25
Academic Self-efficacy..................................29
Peer Influence on Academic Self-efficacy.........29
Academic Motivation and Achievement..............31
Gender and Academic Self-efficacy................31
IX


Cultural Influences on Academic-Self-efficacy
32
Career Self-efficacy......................................33
Gender, Ethnicity, and Career-Self-efficacy........34
Career Decision-making Self-efficacy...............35
Teacher Efficacy..........................................36
General Self-efficacy.....................................38
Summary...................................................39
Paraeducators in Education................................39
Introduction and Definition........................39
History of Paraeducators...........................40
Past and Current Roles.............................42
Training and Education.............................44
Career Ladder Programs.............................46
Summary and Unanswered Questions..........................48
3. METHODOLOGY..................................................53
Introduction..............................................53
Overall Approach and Rationale............................53
Types of Data......................................54
Mixed Method Strategy..............................57
Research Questions........................................58
Sample Selection..........................................60
x


The Participants.....................................60
The Sampling Method..................................64
Data Collection Techniques................................67
Phase I-The Survey and Career Goals Statements.......67
Phase II-The Interviews..............................69
Data Management and Protection Procedures...................72
Data Analysis Procedures....................................73
Demographic Analysis.................................73
Quantitative Analysis................................75
Qualitative Analysis.................................75
Trustworthiness of the Research...........................77
Tri angulation.......................................79
Negative Case Analysis...............................79
Rich and Thick Description...........................80
Purposive Sampling...................................80
Detailed Description of Participants.................81
Interrater Reliability...............................82
Instrumentation......................................85
xi


The Researchers Bias and Role
87
Limitations..................................................89
Summary......................................................91
4. RESULTS.........................................................92
Introduction.................................................92
Section One: Background, Characteristics, and Experiences of
Participants.................................................94
Description of Participants..........................94
The Three Subgroups..................................99
Career Path.........................................102
Paraeducator Role-Likes and Dislikes................105
Educational Experiences.............................107
Factors that Contributed to College Success.........Ill
Career and Educational Goals........................113
Section Two: Self-efficacy Beliefs of Participants........118
Mastery Experiences.................................118
Social Role Models..................................125
Verbal Persuasion...................................129
Emotional Control...................................133
Self-efficacy Scales................................137
Section Three: The Decision to Teach...................142
Xll


The Paraeducators Who Will Remain in that Role.......143
The Paraeducators Who Planned to Become Teachers... 146
The Paraeducators Who Became Teachers...............147
Views of Career Opportunities........................149
Summary.....................................................151
5. DISCUSSION AND IMPLICATIONS....................................155
Introduction................................................155
Research Questions..........................................156
Discussion of Results.......................................157
Demographic Characteristics..........................157
Career Path..........................................162
Classroom Experiences................................163
Educational Experiences..............................166
Educational Goals....................................168
T eacher Efficacy....................................169
Academic Self-efficacy...............................172
General Self-efficacy................................173
Views of Career Opportunities........................174
The Decision to Become a Teacher.....................175
Conclusions.................................................178
Limitations of the Study....................................182
xiii


Implications for Future Practice...........184
Implications for Future Research...........185
APPENDIX
A. LETTERS FROM THE PROJECT COORDINATORS........188
B. SURVEY COVER LETTER..........................191
C. INFORMED CONSENT.............................193
D. SURVEYS......................................205
E. REMINDER POSTCARD............................219
F. INTERVIEW PROTOCOLS..........................220
G. HUMAN SUBJECTS APPROVAL MEMO.................222
H. NVIVO CODING NODES FOR QUALITATIVE ANALYSIS..224
I. INTERRATER RELIABILITY SURVEY................226
REFERENCES............................................230
xiv


FIGURES
Figure
1.1 Conceptual Framework...............................................8
1.2 Dissertation Structure............................................23
3.1 Conceptual Framework..............................................58
4.1 Conceptual Framework..............................................93
xv


TABLES
Table
3.1 Number of participants in each phase of the study........................71
3.2 Quantitative and qualitative criteria for evaluating research
quality..................................................................78
4.1 Paraeducators who became teachers subgroup profile and non-
exemplars........................................................................95
4.2 Current paraeducator subgroup profile and non-exemplars..................97
4.3 Work experience in education of study participants prior to entering
college..................................................................98
4.4 Demographic information of participants in the three subgroups..........101
4.5 Evidence of mastery experiences found in participants statements.......119
4.6 Evidence of social role models found in participants statements........126
4.7 Evidence of verbal persuasion found in participants statements.........130
4.8 Evidence of emotional control found in participants statements.........134
4.9 One-way ANOVA results for the three self-efficacy scales................138
4.10 Means and standard deviations for the three self-efficacy scales........139
xvi


CHAPTER 1
INTRODUCTION
Statement of the Problem
Paraeducators have been employed in schools to provide various support
functions since the 1950s (Nittoli & Giloth, 1997; Pickett, 2003). Since that time,
their numbers have been steadily growing. Paraeducators filled 1.3 million full
and part-time positions in 2000, and their employment is expected to grow (U. S.
Department of Labor, 2002). Their roles have evolved from primarily clerical
duties to a wide variety of functions including instructional delivery (Boomer,
1994; Doyle, 1997; French, 1999; Rogan & Held, 1999). With the
implementation of the federal New Careers program in the 1960s, the idea of
generating more paraprofessional jobs for low-income populations and
subsequently providing additional training and education to create career ladders
into the professions became popular (Nittoli & Giloth, 1997).
Recruiting paraeducators into teaching has many benefits for schools.
Paraeducators who become teachers meet the needs of schools in many different
ways. They are more diverse than the current teaching workforce and tend to be
indigenous to the communities in which they work (Nittoli & Giloth, 1997;
Smith, 2000). As a result, they can add needed linguistic and cultural diversity to
1


the teaching workforce and an understanding of the community served by the
school (French & Pickett, 1997; Genzuk & Baca, 1998; Nittoli & Giloth, 1997;
Miramontes, 1990; Rubin & Long, 1994; Rueda & DeNeve 1999). In addition,
their work experience is mostly in fields where there are teacher shortages such
as bilingual and special education (Haselkom & Fideler, 1996). Finally, once
they enter the teaching profession, they tend to remain in the field, reducing
teacher attrition rates (Clewell & Villegas, 2001; Haselkom & Fideler, 1996).
There are also benefits for the paraeducator who enters a professional career that
offers personal growth and career development opportunities as well as financial
security.
Yet paraeducators face many obstacles before they can enter teaching
(Haselkom & Fideler, 1996; Nittoli & Giloth, 1997). Many do not have
education beyond high school, lack basic skills, and are unsure about their
academic abilities (Aragon, 2003; Bernal & Aragon, In press; Haselkom &
Fideler, 1996). The largest barrier for paraeducators entering a teacher education
program is financial. Paraeducators cannot afford to give up a salary and benefits
to attend school and need help paying for tuition and books (Gordon, 1995;
Haselkom & Fideler, 1996). Many colleges do not offer flexible course
scheduling or day care services that allow paraeducators to coordinate work,
family, and school responsibilities (Gordon, 1995; Haselkom & Fideler, 1996).
2


Career ladder programs are designed specifically to help paraeducators
complete teacher education programs and enter the professional ranks. Research
has identified the key features that help paraeducators successfully complete
career ladder programs and enter teaching (Aragon, 2003; Bernal & Aragon, In
press; White, In press). Successful career ladder programs are collaborations
between K-12 schools and higher education. They employ recruitment and
selection processes specifically tailored to paraeducators, offer financial support,
provide personal and academic counseling, and have flexible course scheduling
to accommodate paraeducators work schedules (Dandy, 1998; Genzuk & Baca,
1998, Safarik, 2001; Villegas & Clewell, 1998).
In 1997, the National Education Association completed a survey of their
6500 members who were education support personnel, also called paraeducators.
Nearly 4000 responded to the survey and 50% indicated they were interested in
becoming teachers. That same year, Nittoli and Giloth (1997) identified 150
career ladder programs nationwide that enrolled a total of 9000 paraeducators.
There is certainly a need to continue developing career ladder programs for
paraeducators. However, most programs specifically tailored to the needs of
paraeducators are funded by grants, and this type of support is dependent upon
political priorities and teacher shortages (White, In press). Often once the
funding ends, the programs do not continue (Haselkom & Fideler, 1996).
3


Some paraeducators overcome financial, academic, and social obstacles
to complete college and enter teaching. Yet, there are still unanswered questions.
How are paraeducators who become teachers different and similar from their
peers who do not follow this career path? What influences their decision to
become teachers? This study identified the factors that influence paraeducators
transition to teaching by examining the differences and similarities between
paraeducators who become teachers and paraeducators who do not want to
become teachers.
Background and Significance of the Problem
For the past decade, literature in the field of teacher recruitment and
training has been foretelling a teacher shortage for the beginning of the 21st
century (Darling-Hammond, 2000). A demand for 2 million new teachers has
been cited due to increasing enrollments, immigration, increased retirements, and
attrition (Clewell & Villegas, 2001; Darling-Hammond, 2000). Urban districts
are described as the most in need of new teachers (Recruiting New Teachers,
Inc., 2000) because they tend to have lower salaries, overcrowded classrooms,
older facilities, and poor reputations, which make it difficult to find and retain
teachers. (Schuerman, 2000).
This projected shortage has spawned a wide range of new teacher
recruitment strategies. Urban districts are recruiting adults from other career
4


fields who are changing careers or immigrants with college degrees and
providing on the job training (Kantrowitz & Wingert, 2000). This training is
referred to as alternative teacher licensure. Many states have approved alternative
routes to teacher licensure (Rosenberg & Sindelar, 2001). They are usually in the
form of programs that prepare people without college level teacher preparation
for licensure, while they teach in classrooms (Roth & Lutz, 1986). According to
Rosenberg and Sindelar (2001), three major factors have contributed to the
marked increase of alternative licensure programs. One is the perception of a
persistent shortage of qualified teachers. The second is an extreme need to
diversify the teaching workforce, and the third is a distrust of the value of
traditional teacher education programs delivered by colleges and universities.
Conceptually, the argument that alternative certification programs are an
excellent way to attract more teachers, especially men and minorities, to the
teaching profession is promising (Chesley, Wood, & Zepeda, 1997; Hawk,
Burke, & Brent, 1999; Shepard, 1999). Building a teaching workforce that
mirrors the demographics of the growingly diverse American population is an
important goal. Alternative programs are attractive because they offer the
prospect of immediate employment. Similar to paraeducators, many adults
considering a career change, especially people of color, cannot give up
employment and benefits to attend school fulltime (Shepard, 1999).
5


Critics of alternative programs are concerned with the quality of these
programs. Alternatively licensed teachers have extremely high attrition rates
(Banks & Necco, 1987). In fact, only 34 percent of the teachers certified
through alternative programs remain in the field after three years, compared to 84
percent of those who have completed five years of trainingnamely a bachelors
degree in a subject field and a masters degree in education (Darling-Hammond,
2000, as cited in Schuerman, p. 32). This creates a revolving door in many
schools especially in those that can least afford such instability and inexperience.
Poor quality alternative licensure programs that promise quick entry into the
teaching profession often fail to produce highly qualified teachers and this
shortchanges students (Berry, 2001).
Studies have confirmed that one of the main factors determining
students educational success is the quality of their teachers. Yet
data show that the least qualified teachers are usually assigned to
schools with the fewest fiscal resources and the neediest students.
(Presidents Advisory Commission on Educational Excellence for
Hispanic Americans, 2000, p. 23)
More recently, the teacher recruitment literature has shifted its focus from
the shortage of new teachers to the retention of teachers. The National
Commission on Teaching and Americas Future (2003) noted that the teacher
shortage is only a symptom of the real problem, which is teacher retention. Their
report No Dream Denied suggested that hiring unqualified and poorly prepared
teachers who leave the profession quickly only exacerbates these high turnover
I
6


rates. The shift to retention creates a stronger argument for focusing recruitment
and training efforts on paraeducators. In an evaluation of the national Pathways
to Teaching Careers program, Clewell and Villegas (2001) examined the
retention rates of three groups of teacher recruits. Those groups were
paraeducators, returning Peace Corps volunteers, and unlicensed teachers. Their
evaluation revealed some important findings. First, of the three groups in the
Pathways program, paraeducators were the most likely to have chosen to work in
urban areas and more likely to remain in teaching for more than three years.
Second, returning Peace Corps volunteers were the least likely to be teaching in
urban schools. Third, minority graduates from all three groups were more likely
to continue in teaching after three years than white graduates of the Pathways
program. Fourth, paraeducators were rated more highly by principals than any
other group of novice teachers (Clewell & Villegas, 2001).
These findings, coupled with a renewed focus on retention of teachers to
solve shortages in urban areas, build a strong argument for investing in
paraeducators as future teachers. Examining the similarities and differences
between paraeducators who become teachers and those who do not will add to
our understanding of how to encourage more paraeducators to enter the
professional teaching ranks.
7


Conceptual Framework
The conceptual framework for this study was derived from research in the
areas of self-efficacy and paraeducator employment and training. The framework
can be depicted using the graphic below.
Figure 1.1 Conceptual Framework
The center of the graphic represents the major focus of this study, which
was paraeducators decision to become teachers. What goes into the decision
making process for paraeducators? How do they come to believe that they can
complete the necessary education and become teachers? In order to discover the
8


answers to these questions, this study examined the experiences of paraeducators
who became teachers and those who had not become teachers.
The middle layer of the graphic contains the ideas included in the broad
conceptual focus of the study, which is based on Banduras (1977, 1997) social
learning theory. According to the theory, human behavior, especially behavioral
change, is influenced by self-reflective thought through which people evaluate
their interactions with the environment and develop efficacy expectations about
themselves (Bandura, 2001). Bandura (1994) defines perceived self-efficacy as
peoples beliefs about their capabilities to produce designated levels of
performance that exercise influence over the events that affect their lives (p.
71).
Experiences that build efficacy come from four sources. They are mastery
experiences, social role models, verbal persuasion and emotional control
(Bandura, 1994). Bandura asserts that mastery experiences have the most
powerful impact on the development of self-efficacy because they give the
person direct evidence of their ability and probability of success. For example,
successfully designing and delivering a lesson to a group of students would be a
mastery experience for a student teacher.
After direct experience, the most powerful way to gain information about
capabilities is through what Bandura (1977, 1994) calls vicarious experience.
People watch and evaluate the experiences of social role models to gauge their
9


own abilities (Schunk, 1987). Social role models for paraeducators in a school
building could be teachers, principals, and other paraeducators who are attending
college.
The third source, verbal persuasion, is not powerful enough to build and
maintain a strong sense of efficacy alone, but it can support the development of
self-efficacy in conjunction with other sources (Bandura, 1995, 1997). Verbal
persuasion is more powerful once a sense of self-efficacy is developed as
encouragement from others can help people persevere in the face of difficulties
(Bandura, 1995, 1997). If a paraeducator was encouraged to earn a college
degree and become a teacher by her supervising teacher, this would be an
example of verbal persuasion.
People also rely on their own emotional and physiological signals to build
efficacy (Bandura, 1977, 1997). That is, people can interpret the emotional and
physical signs of stress as a lack of ability (Bandura, 1994). However, by
changing ones personal reactions to stress and mood, a person can increase a
sense of self-efficacy (Bandura, 1997).
Self-efficacy is context specific (Zimmerman, 1995). It has been applied
to many areas of human functioning. The three areas that are of most interest to
this study are teacher, academic, and general self-efficacy. In order to choose a
particular career, a person has to believe that they can accomplish both the
educational requirements for the job and the specific tasks associated with it
10


(Betz & Hackett, 1981). Many paraeducators have completed little or no college
(Loschert, 2003). Paraeducators who choose to enter teaching must believe that
they can complete both the college requirements and the job tasks necessary for
teaching thus, they need to have high academic and teacher efficacy. They must
also be able to persevere in the face of obstacles they encounter as they complete
their education. To do this they need to have high general self-efficacy, which is
a broad and stable sense of personal competence to deal effectively with a
variety of stressful situations (Schwarzer & Scholz, 2000, p. 1). The middle ring
of the graphic contains the types of self-efficacy that are the focus of this study.
Finally, the outer circle of the graphic represents how self-efficacy can be
applied to paraeducator career choices. The ideas contained in this portion of the
graphic were developed based on both research in the field of paraeducator
employment and training and experience working with paraeducators who are
becoming teachers. It represents my hypotheses about the factors that may
impact the development of teacher, general, and academic self-efficacy in
paraeducators that influence their decision to enter teaching and their ability to
achieve this goal. These factors include work environment, family background,
school experiences, and personal characteristics.
First, it is important to consider the impact of the work environment on
paraeducators teacher efficacy. In numerous recent studies, paraeducators report
inadequate paraeducator training, confusion about roles and responsibilities, low
11


salaries, limited advancement, lack of support, and under appreciation of their
contribution (Chopra, Sandoval-Lucero, Aragon, Berg de Balderas, Bernal, &
Carroll, In press; French & Cabell, 1993; French & Chopra, 1999; Hadadian &
Yssel, 1998; Morehouse & Albright, 1991; Passaro, Pickett, Latham, & HongBo,
1994). Is it possible that poor working conditions and lack of respect have a
negative impact on self-efficacy? If paraeducators are not trained to do their jobs,
does this lead to confusion and frustration that would limit their ability to have
mastery experiences, one of the sources of self-efficacy? Role confusion could
also raise their anxiety levels, contributing to negative emotions that could lower
their career self-efficacy. If paraeducators are not respected or supported by
teachers and administrators, does this limit the social role models in their
schools, people who they might, in other circumstances, look up to and emulate?
On the other hand, paraeducators who are well trained report satisfaction
with learning and are able to apply new skills on the job (Hall, McClannahan, &
Krantz, 1995; Reinoehl & Halle, 1994; Storey, Smith, & Strain, 1993).
Paraeducators who describe being valued members of the team are better able to
fulfill important and needed roles such as helping to create strong connections
between the school and the community (Chopra, et. al., In press). When
paraeducators receive adequate training and are integrated into the school
community, do their work experiences build efficacy and contribute to the
likelihood that they might become teachers?
12


Paraeducators personal and family background could also influence their
general, teacher, and academic self-efficacy. Many are first generation college
students and face numerous obstacles to college completion, including lack of
financial resources, child care difficulties, limited English proficiency, family
opposition, lack of basic skills, fear of technology, and low self-esteem
(Haselkom & Fideler, 1996). All of these issues could have a negative impact on
self-efficacy. First generation college students may have a lack of role models for
college completion. Family opposition could be considered negative verbal
persuasion and could impact self-efficacy. Bandura (1994) states that verbal
persuasion alone cannot increase self-efficacy and negative verbal persuasion is
more likely to contribute to low-self efficacy. Lack of basic and study skills, fear
of technology, limited English proficiency can all create negative emotions.
People create self-efficacy expectations based on their own emotional signals and
reactions to stress (Bandura, 1977, 1997).
The question to be answered is: how do work environment, family
background, school experiences, and personal characteristics contribute to the
development of high general, teacher, and academic self-efficacy for
paraeducators who enter the professional teaching ranks? How are they different
from their peers who do not? From where do they derive and maintain the self-
efficacy beliefs required to enter and complete a teacher education program? This
study sought to answers this question.
13


The Research Questions
To investigate the role of self-efficacy in paraeducators decisions to
become a teacher, the following research questions were answered in this study:
1) In what ways are paraeducators who become teachers similar and
different from paraeducators who do not become teachers in terms of:
1 a) Demographic characteristics
lb) Career path
lc) Classroom experiences
ld) Educational experiences
1 e) Educational goals
1 f) General self-efficacy
1 g) Academic self-efficacy
lh) Teacher efficacy
li) Views of career opportunities
Operational Definitions
Academic self-efficacy is defined as, students perceptions of their competence
to do their class work (Midgley, Maher, Hruda, Anderman, Anderman,
Freeman, Gheen, Kaplan, Kumar, Middleton, Nelson, Roeser, & Urdan, 2000,
14


pg. 27) and was measured in this study by the academic self-efficacy scale from
the Patterns of Adaptive Learning Scales survey (Midgley, et. al., 2000).
Career opportunities are defined as choices or activities presented to
paraeducators that might enhance their ability and potential to move into a
professional role and were documented by paraeducators descriptions of these
choices.
Career path is defined as the path the participant took to become a paraeducator
and was documented by participants descriptions of how they became
paraeducators.
Classroom experiences are defined as instructional and other roles assumed by
paraeducators while working in the classroom that either increase or decrease
their desire and potential to move into a professional role and were documented
by paraeducators descriptions of their roles.
Demographic characteristics include gender, age, ethnicity, level of education,
marital status, number of children and other dependents, personal and family
income, grade point average in high school, native language, and number of
years working as a paraeducator, and were reported by paraeducators on the
demographic portion of the survey instrument.
Educational goals are defined as descriptions of plans for continuing education
for career advancement and future success and were documented by
paraeducators descriptions of these plans.
15


General self-efficacy is defined as a broad and stable sense of personal
competence to deal effectively with a variety of stressful situations (Schwarzer
& Scholz, 2000, p. 1) and was measured by the General self-efficacy scale
(Schwarzer & Jerusalem, 1995).
Paraeducators are school employees 1) whose positions are either instructional
in nature or who provide other direct services to children and youth and/or their
families; and 2) who work under the supervision of teachers or other professional
practitioners who are responsible for a) the design, implementation, and
assessment of learner progress, and b) the evaluation of the effectiveness of
learning programs and related services for children and youth and/or their
families (Pickett, 1989, p. 2).
Other school experiences are defined as paraeducators own educational
experiences as well as experiences they have had while working in a school
setting that either increase or decrease their desire and potential to move into a
professional role and were documented by paraeducators descriptions of these
experiences.
Teacher efficacy is defined as, teachers beliefs that they are contributing
significantly to the academic progress of their students, and can effectively teach
all students (Midgley, et. al., 2000, pg. 27) and was measured by the Patterns of
Adaptive Learning Scales (Midgley, et. al., 2000).
16


Overview of Methodology
The type of research questions derived from a review of the literature and
the development of the conceptual framework naturally led to the use of mixed
research methods because the study was primarily concerned with finding a
solution to a problem (Patton, 1990). For this study, the problem was
encouraging more paraeducators to enter teaching. The factors that influence
paraeducators transition to teaching need to be known and understood in order
to facilitate more paraeducators becoming teachers. The focus was on the
problem of recruiting more paraeducators into teaching and finding potential
solutions. This focus made it necessary to use multiple approaches to learn about
the problem (Tashakkori & Teddlie, 1998). According to Creswell (2003), there
are several forms of mixed methods research where the qualitative and
quantitative data can be collected concurrently or sequentially. This study was
designed using the concurrent nested strategy of mixed methods research where
both types of data were collected concurrently and one method was embedded
in the other (Creswell, 2003).
In nested research designs one method is predominant and the second
method is embedded in the first (Creswell, Plano Clark, Gutmann, & Hanson,
2003). The study of career choice and educational experiences of paraeducators
is necessary for constructing explanations as to why they do or do not choose
teaching as a profession. There are few answers in the professional literature
17


about the reasons paraeducators choose certain career paths. These answers may
be based on assumptions gained from only superficial information about
paraeducators. Therefore, we should be looking more deeply into the influences
on decision making in order to find explanations. The explanations represented in
the study emerged from learning about the experiences, perceptions, and
thoughts that shape the reality of the paraeducators being studied. Their view of
reality can be gathered though qualitative research on this topic. Therefore,
qualitative data was the predominant method in this study. The quantitative data
collected was used to enrich the description of the participants (Creswell, et. al,
2003, p. 230). This deeper description was useful in generalizing the results to
other paraeducators who share similar characteristics with the participants in this
study.
The sample came from three groups of paraeducators. One was a group
who had recently completed a grant-funded paraeducator career ladder program
and entered the professional teaching ranks. These participants represented the
paraeducators who became teachers. The other participants came from three
cohorts of paraeducators who were in programs designed to help them earn an
Associates degree in Elementary Education. One of the cohorts was funded by a
grant. The other two were funded by the school districts in which the
paraeducators were employed. The participants from these three cohorts
represented the paraeducators who plan to remain in that role. The coordinators
18


of each cohort agreed to assist with access to the participants by first contacting
all members of each cohort to receive their permission to release their names and
addresses for the study. They then provided names and current contact
information of all students who agreed to participate in the study.
Three types of data were collected from each group:
1) A survey designed to measure general, academic, and teacher efficacy;
2) A statement regarding their career goals; and
3) In-depth interviews with six to eight select participants who were the
representatives of each group.
All participants in each group were asked to complete a brief survey. The
survey for the paraeducators who were teachers contained two parts. Part I
contained a measure of general self-efficacy (Schwarzer & Jerusalem, 1995),
Part II contained two scales from the Patterns of Adaptive Learning Scales
(PALS) survey (Midgley, et. al., 2000), one measuring academic self-efficacy
and the other measuring teacher efficacy. In addition to the two sections
described above, the survey for the paraeducators contained one open ended-
question asking them to describe their career goals and aspirations. Similar
information was available from and collected for the paraeducators who became
teachers. It came from their application to the career ladder project. At the time
of admission, five to six years ago, they were asked to write a statement about
why they wanted to become a teacher. Reviewing those statements better
19


reflected their thinking at the time they made the decision to become teachers.
Finally, selected individuals participated in in-depth interviews about their
career, classroom experiences, educational goals, and views of career
opportunities. Six participants from the teacher group were interviewed. Eight
participants from the paraeducator group were interviewed. This was due to the
fact that some participants in the paraeducator group were planning to become
teachers. Four who were interviewed were planning to continue as paraeducators.
The remaining four were planning to continue their education beyond the
Associates degree and become teachers. Subjects were selected for interviews
based on their answers to the survey, their career goals statement, and their
willingness to participate in an in-depth interview.
The sampling method for each phase of the study was purposive. The
intent of purposive sampling is to find groups of participants in settings where
the phenomenon under study is most likely to occur (Denzin & Lincoln, 2000).
Those groups were identified through four funded cohorts designed specifically
for paraeducators. For each stage of the study there was a selection process.
Before any data was analyzed, the survey responses and demographic
information were reviewed to ensure the respondents met the criteria to be
included in each subgroup.
Criteria for Paraeducators Who Became Teachers
1) Worked as a paraeducator before entering a teacher education program.
20


2) Worked as a paraeducator while completing the teacher education
program.
3) Had two years or less teaching experience at the time of the study.
Criteria for Paraeducators Who Will Remain in the Paraeducator Role
1) Currently working as a paraeducator.
2) Planned to remain in the paraeducator role.
3) Not currently enrolled in and have no plans to enter a teacher licensure
program.
A third set of criteria was developed after the initial data analysis phase as some
of the paraeducators in the second group had decided to become teachers. Those
criteria are outlined below.
Criteria for Paraeducators Who Are Planning to Become Teachers
1) Currently working as a paraeducator.
2) Interested in becoming a teacher.
3) Planned to enroll in a teacher education program upon completion of
Associates degree.
The selection process for the interview portion of the study included
further selection criteria beginning with willingness to be interviewed as
indicated by the participants on the informed consent. Using typical case
selection (LeCompte & Preissle, 1993, Miles & Huberman, 1994), interview
participants were evaluated to determine if they best represented the
21


characteristics of their subgroup based on the themes that emerged in the analysis
of the survey and career goals statement data. Finally, they continued to meet the
three criteria to be included in their respective subgroup. Some participants chose
not to participate in the interview phase of the study.
Interviews were recorded, transcribed, and entered into QSR*NVivo, a
computerized data analysis program. Career goal statements were also entered
and analyzed. The four sources of self-efficacy: mastery experiences, social role
models, verbal persuasion, and emotional control (Bandura, 1994) were used as
an initial lens to view the responses. Additional themes based on the interview
protocol questions themselves were also included in the analysis. From this
analysis, a picture of the personal experiences of these paraeducators was
created. Answers to the survey questions were entered into SPSS 12.0 and means
on each scale for both groups were compared using one-way analysis of variance
(ANOVA) to test for differences.
Overview of Remaining Chapters
The remaining chapters contain additional information about the study.
The structure of this study resembled the shape of an hourglass. Figure 1.2 is a
graphic representation of the structure.
22


Figure 1.2 Dissertation Structure
Chapter 1 provided an overview of the study, including portions of the
supporting literature. Chapter 2 begins at a wider point of the top of the hourglass
by summarizing the literature from which the research questions were derived,
beginning with self-efficacy, the broad conceptual focus of the study. It then
narrows the topic of self-efficacy down to three specific areas: academic, career,
and teacher efficacy. Next, the literature on paraeducators is reviewed beginning
with a history of paraeducators in education. It also examines the evolution of
their roles, their ongoing training and education, and the development of career
ladder programs to transition them into teaching. The chapter ends with a
summary and unanswered questions that pinpoint the focus of the study, which
was the role of self-efficacy in paraeducators career decisions. At this point, the
research questions are at the narrowest point of the hourglass.
23


Chapter 3, Methodology, outlines in detail the methods used to answer the
research questions. Chapter 4 summarizes the findings of the study. These
chapters are still focusing on the narrowest point of the hourglass. Chapter 5,
Discussion and Implications, begins to expand the hourglass back out to the
wider base by relating the findings of the study back to the broad literature base.
24


CHAPTER 2
REVIEW OF LITERATURE
Introduction
This study examined the similarities and differences between
paraeducators who became teachers and those who have not become teachers,
with the specific goal of understanding how self-efficacy plays a role in the
decision to become a teacher. This chapter reviews the literature that informed
the study. Two broad areas of literature are included. The first area is the
literature on self-efficacy. The second area is the body of literature on
paraeducators in education.
Introduction to Self-efficacy
Bandura (1977) first introduced the concept of self-efficacy as part of his
social learning theory. Self-efficacy is a persons set of beliefs about his/her
capabilities to perform or accomplish certain tasks and how these beliefs directly
affect a persons ability to persist and succeed (Bandura, 1994). Self-efficacy
differs in specific ways from other concepts contained in theories of the self.
Self-concept is a global view of the self developed from experiences and
25


evaluations from others and self-esteem is an assessment of self worth (Bandura,
1997). Self-efficacy is highly predictive of future behavior where self-concept is
not (Pajares & Kranzler, 1995; Pajares & Miller, 1995). Additionally, high self-
esteem does not necessarily produce achievement (Bandura, 1997). Self-efficacy
is context specific (Zimmerman, 1995). In other words, a person may have high
self-efficacy in one context, but have low self-efficacy in another context. For
example, a person may have high academic self-efficacy, but have low self-
efficacy for smoking cessation. Efficacy beliefs can also change over time; they
are not fixed (Bandura, 1997). Self-efficacy is resilient over time if a persons
experiences have helped him/her learn the skills to persevere in the face of
obstacles (Bandura, 1995).
According to Bandura (1977, 1994) self-efficacy beliefs are developed
from four main sources: mastery experiences, vicarious experience through
social role models, social persuasion, and affective or emotional states. Mastery
experiences are the most powerful because they provide the most direct evidence
that a person has the skills for success (Bandura, 1994, 1995). For example,
completing a first marathon will give a person the best and most direct evidence
that he/she is capable of doing so. The idiom success breeds success would
certainly apply here. Success builds a strong sense of self-efficacy; however,
failure is damaging to ones efficacy, especially if it happens before a strong
sense of efficacy is developed (Bandura, 1994, 1995). Learning how to ride a
26


bike or how to navigate the internet are examples of mastery experiences because
the person is actually participating in the activity.
The second source from which a person can draw self-efficacy beliefs is
vicarious experience through social role models (Bandura, 1977, 1994).
Observing people succeed, even in the face of obstacles, can increase a persons
belief in his/her own efficacy (Schunk, 1987). Similarly, seeing people fail can
lower self-efficacy beliefs (Brown & Inouye, 1978). The impact of social
modeling on self-efficacy is stronger if the person perceives the other as being
more like oneself (Bandura, 1994, 1995). The greater the perceived similarity,
the more impact there is on personal efficacy (Bandura, 1995). For example,
observing the first person in the family graduate from college will increase a
persons efficacy beliefs for college graduation more than seeing others who are
less like him/her graduate.
The third influence on the development of self-efficacy is verbal
persuasion (Bandura, 1994, 1995). People can be persuaded that they possess the
skills to accomplish certain goals or tasks, and verbal encouragement can make
them more likely to increase their efforts than to give up when they encounter
obstacles (Litt, 1988; Schunk, 1989). Verbal persuasion alone cannot create high
efficacy beliefs; in fact, negative verbal persuasion can more easily undermine
ones beliefs (Bandura, 1994). Verbal persuasion in conjunction with other
sources of self-efficacy can help build stronger efficacy beliefs (Bandura, 1994).
27


The final source from which people can develop efficacy beliefs is from
their own emotional and physiological states (Bandura, 1994). Stress, tension,
and mood can all affect a persons assessment of personal efficacy (Bandura,
1995; Ewart, 1992; Kavanaugh & Bower, 1985). How the person interprets
his/her own physical and emotional states impacts his/her self-efficacy (Bandura,
1995). For example, a negative experience in school can lower academic self-
efficacy unless a person can reduce stress and change his/her emotional
interpretations of the situation.
Because self-efficacy is specific to different domains, it has been studied
in many contexts. The research that is important to examine for this study is
contained in those areas dealing with academic self-efficacy, career self-efficacy,
specifically, teacher efficacy, and general self-efficacy. In order to become a
teacher, a paraeducator must have high academic self-efficacy to believe in
his/her ability to complete the educational requirements to enter the field. The
paraeducator must also have high teacher efficacy and believe he/she can
complete the tasks and requirements of the job. Paraeducators also need to have
high general self-efficacy in order to cope with the stresses of completing college
while balancing work, family, and school. The following sections examine the
research in the areas of academic, career, teacher, and general self-efficacy in
more depth.
28


Academic Self-efficacy
The development of academic self-efficacy begins in childhood and the
first experiences that contribute to its development are at home (Schunk &
Pajares, 2001). Parents are the first role models, and interaction with them affects
how children can leam problem solving, persistence, and coping strategies
(Banrura, 1997; Meece, 1997). Children can begin to develop academic self-
efficacy by observing their parents overcoming obstacles and solving daily
problems (Bandura, Barbaranelli, Capara, & Pastorelli, 2001). If parents also
provide an environment rich with mastery experiences, they greatly contribute to
the development of their childs academic self-efficacy (Bandura, 1997; Meece,
1991). Additionally, parents, through verbal persuasion, can help children
develop higher academic self-efficacy by encouraging them to participate in
varied activities that present new and different challenges that the child can
attempt and eventually master (Schunk & Pajares, 2001).
Peer Influence on Academic Self-efficacy
Once children enter school, they also encounter peers who can offer
additional experiences and feedback that help or hinder the development of
academic self-efficacy (Shunk, 1987). Peers serve as powerful social role models
and a great deal of social learning takes place among peers (Bandura, 1997).
Through peers, children assess the probability of their academic success.
29


Children tend to associate with others of similar age, interests, and
abilities, and they judge the attainments of others who are more similar to
themselves as predictive of their own abilities (Bandura, 1997). Their friends
accomplishments increase their belief in their capability to perform (Schunk,
Hanson, & Cox, 1987). Similarly, by vicariously experiencing failures of those
who are perceived to be similar to them, they lower their judgments about their
own abilities (Brown & Inouye, 1978). The successes and failures of the peers
who are deemed most similar to the child have the greatest impact on the
development of his/her academic self-efficacy (Bandura, 1997).
Once in high school, peers become an even stronger influence on
academic self-efficacy. Students who choose to associate with a group that has
high academic self-efficacy will have higher academic success than those who
choose a group that is less academically motivated (Steinburg, Brown, &
Dombusch, 1996). In high school, classes can be larger and teachers may spend
less time with students than in elementary. When there is less time for
individualized instruction and personal attention, academic self-efficacy can
sometimes decrease due to the fact that the teacher has less time for verbal
persuasion and cannot provide as many mastery experiences (Pintrich & Schunk,
1996).
30


Academic Motivation and Achievement
A second area of investigation related to academic self-efficacy has
studied the relationships of academic-self-efficacy, academic motivation, and
achievement (Pajares, 1996). Studies in this area have shown that students
beliefs that they are capable of performing certain academic tasks affect both
their thinking strategies and academic persistence (Pintrich & Garcia, 1991).
Several investigations have demonstrated that academic self-efficacy impacts
basic cognitive skills, performance in academic coursework, and performance on
standardized achievement tests (Zimmerman, 1995, Moulton, Lent, & Brown,
1991). The relationship between perceived efficacy and academic achievement
was stronger for high school and post-secondary education students than it was
for elementary school students (Multon, et al., 1991). The researchers
hypothesized that this is because older students are better able to judge their
academic abilities since they have had more school experiences.
Gender and Academic Self-efficacy
Numerous studies have investigated the relationship of academic self-
efficacy to gender. Gender differences did not emerge for academic tasks and
measures except in the areas of math and science (Hackett, 1985; Hackett, Betz,
OHalloran, & Romac, 1990; Pajares & Miller, 1994; Post-Kammer & Smith,
1986). However, additional research has determined that these differences can be
31


influenced by numerous factors, including previous achievement and age. In fact,
differences in male and female academic self-efficacy are significantly reduced
when previous achievement is controlled (Pajares, 1996). Age also influences
academic self-efficacy. There is little difference found in the academic self-
efficacy of elementary school aged boys and girls (Multon, et. al., 1991). Gender
differences do not emerge until adolescence when academic self-efficacy for
girls declines (Wigfield, Eccles, & Pintrich, 1996).
Cultural Influences on Academic-Self-efficacy
Ethnic and cultural influences on academic-self efficacy have also been
examined (Bores-Rangel, Church, Szendre, & Reeves, 1990; Church, Teresa,
Rosebrook, Szendre, 1992; Gloria & Hird, 1999; Lauver & Jones, 1991;
Oettingen, 1995; Post, Stewart, & Smith, 1991). Self-efficacy is a cross-cultural
phenomenon, but culture may influence its development differently in two ways.
Culture may impact the type of information given from the four sources of
efficacy beliefs and the way it is evaluated and integrated into a persons
assessment of personal efficacy (Oettingen, 1995). A variety of studies have
examined the self-efficacy of minority students as compared to non-minority
students. Generally, minority students have lower academic self-efficacy than
non-minority students; however, those differences disappear when researchers
control for socioeconomic status (Graham, 1994). Additionally, strong
32


relationships have been found between positive ethnic/racial identity and
academic self-efficacy (Okech & Harrington, 2002). Hackett, Betz, Casas, &
Rocha-Singh (1992) investigated self-efficacy among diverse engineering college
students. They found that non-minority college students did have higher levels of
self-efficacy for engineering careers than Hispanic students, largely due to
differences in academic preparation for college. Therefore, this strongly suggests
that ethnicity alone is not related to academic self-efficacy but is related to access
to quality education. Lack of educational access can inhibit the development of
high academic self-efficacy (Hackett, 1995).
Career Self-efficacy
Career and academic self-efficacy are related, and much of the research
on career-self efficacy has also focused on academic persistence. This is because
the perceived ability to both complete the educational requirements and the job
tasks for a given career impacts the career options a person will consider
(Bandura, 1997). Betz & Hackett (1981) first applied the concept of self-efficacy
to career development. They found that self-efficacy did play a part in the career
choices of college men and women, with womens self-efficacy expectations
leading them to limit their career options to traditionally female careers even
when there were no differences in their scores of verbal and math ability on
standardized tests. Subsequent studies found that women had lower career self-
33


efficacy expectations for careers requiring math and science ability (Campbell &
Hackett, 1986; Hackett, 1985; Hackett, Betz, OHalloran, & Romac, 1990; Post-
Kammer & Smith, 1986).
Gender, Ethnicity, and Career-Self-efficacy
Gender differences are more likely to arise when women are exposed
only to stereotypical tasks and careers or when they feel pressure to conform to
gender specific roles (Betz & Hackett, 1983; Hackett, et. al., 1990; Wheeler,
1983). For example, Stickel and Bonnett (1991) found gender differences in
perceived ability to combine a family and a career. Women may have chosen not
to pursue nontraditional careers because of their perceived ability to perform in
that career and also because they are unsure of their ability to combine the
careers requirements with family responsibilities. Men did not change their
career goals based on whether they wanted to have a family or not. These gender
differences in career self-efficacy were considered to be present largely due to
female socialization, which provides less access to the sources of self-efficacy
that build strong expectations of efficacy for certain careers (Bandura, 1997;
Betz, 1992; Juntunen, 1996, Scheye & Gilroy, 1994).
Subsequent studies found that gender differences in career self-efficacy
were not as pronounced as in the initial studies (Kelly, 1993; Lent, Brown, &
Larkin, 1984, 1986). This could be due to the changing roles of women in society
34


over the past several decades. Gender differences in career self-efficacy are also
not found among homogenous groups, such as high achieving students in
graduate or professional programs (Clement, 1987; Lent, et. al., 1984, 1986). As
with race and ethnicity, gender alone is not related to career self-efficacy.
Womens career expectations can be shaped by the type of self-efficacy
information to which they are exposed (Banruda, 1997). Scheye and Gilroy
(1994) found that women enrolled in single sex educational institutions in high
school and college, who had male teachers who served as role models had higher
career self-efficacy for nontraditional careers.
Other studies have confirmed that the self-efficacy expectations of
diverse groups of high school and college students do predict the range of
occupations they will consider as well as the individual careers and college
majors they may choose (Betz & Hackett, 1981, 1983; Bores-Rangel, et. al.,
1990; Lauver & Jones, 1991; Lent, et. al., 1986). Most of these studies have
focused on careers in the math and science fields due to the continued under
representation of women and minorities in these fields.
Career Decision-making Self-efficacy
Self-efficacy for making career decisions has also been studied
extensively. Career decision-making self-efficacy involves developing skills for
making career decisions and developing confidence in decision-making abilities
35


(Hackett, 1995). Aspects of effective career decision making include goal setting,
career exploration, problem solving, planning, and self-appraisal (Crites, 1981;
Zunker, 1998). Career decision-making self-efficacy and efficacy for specific
occupations are only slightly related because the career decision-making process
is not dependent upon which career is being considered (Hackett, 1995; Taylor &
Pompa, 1990). The extent to which a person will engage in career exploratory
behavior is influenced by career decision-making self-efficacy (Blustein, 1989).
Differences in career decision-making self-efficacy have been identified
by race and major in college students (Gloria & Hird, 1999). For minority
students, having a strong other orientation, meaning being able to identify with
an ethnic group other than their own, increased their career decision-making self-
efficacy (Gloria & Hird). The researchers hypothesized that this was probably
due to the fact that identifying with another group expands the students number
of social role models. Career decision-making self-efficacy is also a factor in the
retention of nontraditional college students, with those who have been able to
make a career choice more likely to persist in college (Sandler, 2000).
Teacher Efficacy
The career field of particular interest to this study is teaching and the
research that has been conducted in the area of teacher efficacy. Tschannen-
Moran, Woolfolk Hoy, & Hoy (1998) defined teacher efficacy as the teachers
36


belief in his or her capability to organize and execute courses of action required
to successfully accomplish a specific teaching task in a particular context (p.
233). Having a strong sense of teacher efficacy is beneficial for both teachers and
students. Teachers with high efficacy put forth more effort, were more
innovative, had better planning and organization, were more persistent, resilient,
enthusiastic, and were more committed to their careers and to helping students
with difficulties and challenges (Tschannen-Moran, et. al., 1998). For students,
high teacher efficacy impacts student achievement, motivation, and academic
self-efficacy (Ashton & Webb, 1986; Guskey & Passaro, 1994). The sources of
teacher efficacy have been examined from various perspectives. The type of
teacher education program completed, the student teaching experience, and
mentoring are related to higher teacher efficacy in novice teachers (Darling-
Hammond, Chung, & Frelow, 2002; Knobloch & Whittington, 2002; Yost,
2002). For veteran teachers, supervision is important for teacher efficacy.
Ebmeier (2003) found it was important for teachers to feel that their
administrators were responsive and understanding of the teachers job in the
classroom. Interestingly, in this study, school climate and the teachers peers
were not closely related to high teacher efficacy.
37


General Self-efficacy
General self-efficacy is a concept that is associated with a persons
coping skills. If a person has high general self-efficacy, he/she has positive and
optimistic beliefs about his/her ability to cope with stressful situations and
challenges (Schwarzer, 1998). The other areas of self-efficacy discussed in this
review were context specific. General self-efficacy refers to an overall sense of
confidence to cope with a variety of stressful situations (Schwarzer, 1998). The
events themselves are not interpreted as stressful; rather it is a persons
perception that the life events are beyond ones ability to cope that causes the
stress (Bandura, 1997). If people believe they can handle environmental
demands, they do not become overwhelmed and stressed by them (Bandura,
1997).
General self-efficacy has been studied in relation to adaptation in the face
of stressful life transitions (Jerusalem & Mittag, 1995). During stressful life
transitions, a persons general self-efficacy beliefs may provide him/her with
coping resources if the beliefs are strong or may make him/her more vulnerable
to stress if the beliefs are weak (Bandura, 1986; Jerusalem, 1993; Schwarzer,
1992). High general perceived self-efficacy allows people to trust their
capabilities, face environmental challenges confidently, and interpret problems as
challenges, not threats (Bandura, 1997; Jerusalem & Mittag, 1995). Jerusalem
and Mittag (1995) found that low general self-efficacy beliefs can cause
38


depression, anxiety, and low self-esteem. They determined that these beliefs can
develop from repeated past failures and lack of feedback or negative feedback
from teachers, parents, and peers. For paraeducators, returning to school as an
adult with multiple life demands can be a stressful life transition. Their general
perceived self-efficacy beliefs could impact their ability to cope with the
demands of work, family, and school while they are attending college.
Summary
Research has also shown that self-efficacy is a strong predictor of future
behavior (Bandura, 1997; Pajares & Kranzler, 1995; Pajares & Miller, 1995).
Empirical evidence strongly suggests that perceived efficacy contributes to career
choice and development (Hackett, 1995). The focus of this study was to examine
the role of teacher, academic, and general self-efficacy on the career paths of
paraeducators from the perspective of paraeducators who became teachers and
those who chose to remain in the paraeducator role.
Paraeducators in Education
Introduction and Definition
To be able to understand how self-efficacy plays a role in paraeducators
decision to become teachers, we must first review their history, roles, training
39


and education beginning with a definition. Pickett (1989) defines paraeducators
as:
School employees 1) whose positions are either instructional in
nature or who provide other direct services to children and youth
and/or their families; and 2) who work under the supervision of
teachers or other professional practitioners who are responsible
for a) the design, implementation, and assessment of learner
progress, and b) the evaluation of the effectiveness of learning
programs and related services for children and youth and/or their
families (p. 2).
History of Paraeducators
The history of the use of paraeducators in instructional support roles in
schools begins in the 1950s. During World War II some teachers changed
careers to manufacturing fields that would help the countrys war effort (Smith,
2000). This created a post-war teacher shortage (Pickett, 2003). Paraeducators,
then called teacher aides, were initially non-minority, middle class, college
educated women without teacher licensure who were recruited to perform routine
clerical duties and classroom monitoring in order to allow teachers to spend more
time in direct instruction (Frith, 1982; Lindsey, 1983; Morehouse & Albright,
1991; Pickett, 2003; Smith, 2000). Although their numbers continued to grow
over the next decade, critics were concerned that they would be used to cheaply
replace qualified teachers rather than assist them (Pickett, 2003).
During the 1960s and 1970s, rapid social change brought about the
implementation of many new educational and human service programs that
40


increased both the number and the diversity of the paraeducator workforce
(Pickett, 2003; Smith, 2000). The civil rights movement, the shift of services for
people with disabilities from institutional to community based, the
implementation of Head Start, the reauthorization of the Elementary and
Secondary Education Act to include programs for children who were
educationally and economically disadvantaged, and the War on Poverty all
contributed to the increased use of paraeducators (Gartner & Reissman, 1974;
Pickett, 2003; Smith, 2000). While the educational reform efforts of the 1980s
called for improving student achievement and preparing a highly qualified and
competent teaching workforce, the growing reliance on paraeducators was
largely ignored (Pickett, 2003).
From the 1990s to the present, the number of paraeducators employed in
schools continued to increase due to a growing number of English Language
learners in classrooms, an increase in at-risk students, a continued and persistent
shortage of special education teachers, and the need to have school personnel that
reflected the diversity of the school age population (Smith, 2000). In the late 20th
and early 21st centuries, increased reliance on paraeducators and continuing
reform efforts resulted in legislation being enacted that specifically addressed the
qualifications of paraeducators. This legislation includes the reauthorization of
the Individuals with Disabilities Act in 1997 and the reauthorization of the
Elementary and Secondary Education Act, also called No Child Left Behind, in
41


2002 (Pickett, 2003). Both set higher standards for the employment and training
of paraeducators.
Past and Current Roles
The growth of the paraeducator workforce has also resulted in an
evolution of their roles in the classroom. Initially, their roles were mostly clerical
in nature (Pickett, 2003; Smith, 2000). Today, some paraeducators still provide
services similar to their 1950s predecessors, such as copying, lunchroom and
playground monitoring, and creating materials (Pickett, 2003). However, their
use has increased in specific educational programs such as special education,
Title I programs, speech therapy, and preschool. With teacher supervision, they
have become more involved in the instructional process, including small group
instruction, assessment, and behavior management (Blalock, 1991; Downing,
Ryndak, & Clark, 2000; Marks, Schrader, & Levine, 1999; Miramontes, 1990;
Passaro, et., al., 1994; Rogan & Held, 1999).
In addition, the school population has also become more diverse since
paraeducators were first employed in education. Because they tend to live in the
communities in which they work and they are demographically more like their
students, paraeducator roles have grown to include cultural liaison between
teacher and students (French & Pickett, 1997; Genzuk & Baca, 1998; Nittoli &
Giloth, 1997; Miramontes, 1990; Pickett, 1989; Rubin & Long, 1994).
42


Paraeducators often serve as connectors between home and school, providing
linguistic and cultural assistance to parents, teachers, and students (Chopra, et.
al., In press; Rueda & DeNeve, 1999).
The rapid expansion of the use of paraeducators in education and their
widely diverse roles among various schools and districts has caused continued
confusion on the part of paraeducators and teachers about appropriate roles,
including which duties are appropriate for teachers and which for paraeducators
(Chopra, et. ah, In press; French & Pickett, 1997; Riggs & Mueller, 2001; Rogan
& Held, 1999). Lack of clarity about paraeducator roles causes tension between
paraeducators and professionals (Chopra, et. ah, In press; Nittoli & Giloth, 1997;
Riggs & Mueller, 2001). The tension in part can be attributed to teachers fears
of being replaced by non-certified staff, a problem that dates back to the very
beginning of the history of paraeducators in classrooms (Nittoli & Giloth, 1997;
Pickett, 2003). As a result, paraeducators often are not properly or fully
integrated into a team of educators working to serve the needs of students (Nittoli
& Giloth, 1997; Riggs & Mueller, 2001). Role confusion and its associated
problems result in a loss for paraeducators. If their skills are not used
appropriately, or they do not work in a team environment, they miss out on
important career development opportunities through working with students and
using teachers as their models for success.
43


Training and Education
Along with an increase in responsibility comes an increased need for
training. (French & Chopra, 1999; French & Pickett, 1997; Giangreco, Edelman,
Broer & Doyle, 2001; Nittoli & Giloth, 1997; LeTendre, 1998; Pickett, Likins, &
Wallace, 2003; Pickett, Vasa, & Steckelberg, 1993; Riggs & Mueller, 2001;
Rubin & Long, 1994). Training for paraeducators has not kept up with their
increases in numbers or their changing roles in education, and many
paraeducators continue to fulfill roles in schools for which they are not properly
trained (Blalock, 1991; Fletcher-Campbell, 1992; French & Pickett, 1997,
Haselkom & Fideler, 1996). Training does lead to improved job performance.
Research on training indicates that paraeducators who are well trained can apply
new skills on the job and this results in increased student achievement and
increased job satisfaction on the part of the paraeducators (Hall, McClannahan, &
Krantz, 1995; Reinoehl & Halle, 1994; Storey, Smith, & Strain, 1993).
From the 1950s through the end of the 20th century, training for
paraeducators has been fragmented and unregulated (Nittoli & Giloth, 1997).
Few states have legislation that mandates training and education for
paraeducators (Pickett, 2003). However, with the reauthorization of the
Elementary and Secondary Education Act-No Child Left Behind, a legislative
step was taken to specify the training and education necessary for paraeducators
44


(P. L. 107-110). As of January 2002, newly hired paraeducators, and as of
January 2006, current paraeducators working in Title I programs must meet the
following specific educational requirements:
1) Completed 2 years of study at an institute of higher education
(demonstrating knowledge of and the ability to assist, i.e., instructing in
reading, writing and math). OR
2) Obtained an associate's (or higher) degree (demonstrating knowledge of
and the ability to assist, i.e., instructing in reading, writing and math). OR
3) Demonstrated knowledge of reading, writing and mathematics, as well as
the ability to assist in reading, writing and mathematics through a formal
academic assessment. (Colorado Department of Education, 2003)
While this legislation currently only applies to paraeducators working in Title I
schools and programs, it is expected that similar guidelines will be included in
the future reauthorization of educational programs at the federal level. Meeting
these requirements could be challenging for paraeducators. Many have a long
way to go to meet the minimums. Of the paraeducators who are members of the
National Education Association, 66% have less than a two-year degree (Loschert,
2003). On the other hand, completing the educational requirements could be a
first step for paraeducators who want to go on to become teachers.
45


Career Ladder Programs
Career ladder programs are training programs for paraeducators that are
designed to assist them with completing college and earning teacher licensure.
They originally grew out of the War on Poverty initiatives of the 1960s (Pickett,
2003). Pearl and Reissman (1965) first recommended career ladders as
educational programs designed to help low-income and minority workers obtain
training to work in human services positions in their own neighborhoods. The
goal was not only to lift low income neighborhoods out of poverty by providing
jobs to residents who came from the populations that they served, but to also
provide additional education and training that would create career ladders into
the professional ranks (Nittoli & Giloth, 1997). While they existed, the new
careers programs created jobs, provided training, and spawned paraprofessional
training programs at community colleges (Reissman, 1984). However, they did
not create the expected systemic reform of human services. There was very little
movement of the low income workers into professional ranks, and many of them
were eventually replaced by college educated professionals (Nittoli & Giloth,
1997; Reissman, 1984).
The Career Opportunities Program (COPS) was the first career ladder
program for paraeducators that grew out of the War on Poverty initiatives in the
1970s (Pickett, 2003). With the passage of the federal Education Professions
46


Development Act in 1967, the funds for training paraeducators to become
teachers became available (Haselkom & Fideler, 1996). Over the course of seven
years, the COPS program trained about 15,000+ paraeducators across the nation
in programs that were a collaborative effort between school districts and colleges
(Kaplan, 1977). This collaboration was an innovation on past programs. The goal
of the COPS program was to train paraeducators who were community residents
in low income neighborhoods to advance into the professional ranks and become
teachers (Pickett, 2003). COPS teachers were well trained and highly sought after
teachers, but unfortunately, many programs did not continue after the funding
ended (Kaplan, 1977; Pickett, 2003).
Today, career ladder programs for paraeducators still exist, but only for
paraeducators who work in areas where there are critical shortages of teachers,
such as bilingual and special education (Nittoli & Giloth, 1997). Funding for
career ladder programs comes and goes according to current teacher recruitment
trends and political agendas (White, In press). The most current data on funded
career ladder programs indicates that they enroll about 9000 paraeducators in
about 150 programs nationwide (Nittoli & Giloth, 1997).
The career ladder programs of today have learned much from their early
predecessors and from research on their successes. Many of the successful
components of the early programs are still important today. To be successful,
model paraeducator to teacher programs must include in their design strong
47


collaboration between K-12 schools and universities, recruitment and selection
processes specifically tailored to paraeducators, financial support for participants,
personal and academic advising, and flexible course scheduling to accommodate
work schedules (Dandy, 1998; Genzuk & Baca, 1998; Genzuk, Lavadenz, &
Krashen, 1994; Safarik, 2001; Villegas & Clewell, 1998).
Paraeducators face many obstacles to program completion, including lack
of financial resources, family resistance to education, lack of basic academic and
study skills, and balancing multiple roles (Haselkom & Fideler, 1996). The
supports built into career ladder programs are important to overcoming these
obstacles. Paraeducator-to-teacher programs need to provide academic, social,
and financial supports for students to help them be successful (Genzuk, et. al.,
1994; White, In press). Those who succeed become valued teachers who fulfill
many needs of todays schools. They reflect the diversity of their students, they
are willing to work in low-income and urban districts, and they tend to remain in
the profession (Clewell & Villegas, 2001; Genzuk, et. al. 1994; Haselkom &
Fideler, 1996; White, In press).
Summary and Unanswered Questions
The research on self-efficacy and on paraeducators could contribute to
our understanding of why some paraeducators choose to become teachers and
others do not. Self-efficacy has been studied in many educational settings with a
48


variety of student groups, including college men and women (Betz & Hackett,
1981, 1983; Clement, 1987; Foss & Slaney, 1986; Matsui, Ikeda, & Ohnishi,
1989, Sheye & Gilroy, 1994), high school equivalency diploma students (Bores-
Rangel, et. al, 1990; Church, et. al., 1992), minority college students (Gloria &
Hird, 1999; Lauver & Jones, 1991; Post, et. ah, 1991), math and science students
(Hackett, 1985; Hackett, et. ah, 1990; Post-Kammer & Smith, 1984, 1986), and
community college students (Rotberg, Brown & Ware, 1987). Paraeducators
could fall into all of these groups, and what has been learned from these studies
could help in understanding paraeducator career and academic development.
However, paraeducators also have characteristics that make them unique and the
development of their career and academic self-efficacy bears further scrutiny.
We know that the paraeducators in todays schools tend to be
overwhelmingly female and come from diverse backgrounds (French & Pickett,
1997; Genzuk, & Baca, 1998, Genzuk, et. ah 1994; Nittoli & Giloth, 1997;
Miramontes, 1990; Pickett, 1989; Rubin & Long, 1994; Smith, 2000; White, In
press). From the research on self-efficacy, we have learned that women and
minorities have lower career and academic self-efficacy, not due to these factors
alone, but due to the fact that because of their gender and ethnicity they are often
exposed to fewer sources and models of self-efficacy through socialization and
poor academic preparation (Bandura, 1997; Hackett, 1985, 1995; Hackett et. ah,
1992). As a result of the findings of these studies, we might expect paraeducators
49


to have low career and academic self-efficacy due to similar factors. However,
the question that remains is how and where do the paraeducators who become
teachers develop the academic and teacher efficacy to complete a teacher
education program and enter teaching?
The workplace environment could also have an influence on the
development of paraeducators career self-efficacy. The rapidly changing
paraeducator roles described in the literature could impact their development of
self-efficacy in several ways. On the negative side, inadequate training, role
confusion, and tensions between paraeducators and teachers are problems
paraeducators face at work that have been identified in the literature (Blalock,
1991; Chopra, et. al., In press; Fletcher-Campbell, 1992; French & Pickett, 1997;
Haselkom & Fideler, 1996; Nittoli & Giloth, 1997; Riggs & Mueller, 2001;
Rogan & Held, 1999). Confusion and tensions at work could lead to stress and
negative emotional reactions. Stress, tension, and negative emotions can lower
self-efficacy (Bandura 1977,1986, 1995). Tensions and distrust between
paraeducators and teachers could limit the paraeducators access to social role
models at work who could increase their efficacy for entering teaching.
Conversely, over time paraeducators have moved into less clerical and more
instructional roles (Blalock, 1991; Downing, et. al., 2000; Marks, et. al., 1999;
Miramontes, 1990; Passaro, et. al.; Rogan & Held, 1999). This role expansion
has possibly given them job responsibilities that provide more mastery teaching
50


experiences that could contribute to the development of their teaching efficacy,
provided that they are adequately trained for these roles.
The research on career ladders for paraeducators is also important in
understanding paraeducator career and academic development. The personal,
academic, and financial obstacles that paraeducators face when they enter school
are well documented (Haselkom & Fideler, 1996; White, In press) as are the
supports built into career ladder programs that are important to overcoming these
obstacles (Genzuk, et. al. 1994). It would be interesting to determine if the social,
academic, and financial supports offered by career ladder programs are
successful because they are available or because they build the academic self-
efficacy of students by helping them learn to overcome the inevitable problems
experienced by college students. This learning could take place through mastery
experiences, such as successfully completing a math class after receiving
tutoring. Program advisors could also provide multiple sources of self-efficacy.
They could serve as social role models, provide verbal persuasion for remaining
in school, and help paraeducators cope so that these stresses do not create
negative emotions. Peers in the career ladder cohort could also fulfill the same
roles for each other.
No research has specifically studied the teacher and academic self-
efficacy of paraeducators. Until that happens, all of the ideas about how
paraeducators develop the teacher and academic self-efficacy necessary to
51


become teachers are speculative. If self-efficacy does play a role in a
paraeducators career decisions, then there should be differences in the teacher
and academic self-efficacy of paraeducators who become teachers and those who
do not want to become teachers. For those who decide to become teachers, what
are the sources of their teacher and academic-self efficacy? This study examined
these questions by learning about the teacher and academic self-efficacy beliefs
of paraeducators who became teachers and those who did not become teachers. It
also examined the contexts in which they lived and worked to determine the
sources of their efficacy beliefs.
The research questions for the study were:
1) In what ways are paraeducators who become teachers similar and
different from paraeducators who do not become teachers in terms of:
1 a) Demographic characteristics
lb) Career path
lc) Classroom experiences
1 d) Educational experiences
1 e) Educational goals
1 f) General self-efficacy
1 g) Academic self-efficacy
lh) Teacher efficacy
li) Views of career opportunities
52


CHAPTER 3
METHODOLOGY
Introduction
This chapter describes the methodology of the research. The study
employed a mixed methods approach. Included are sections on overall approach
and rationale, research questions, sample selection, data collection techniques,
data management and protection procedures, data analysis procedures,
trustworthiness of the research, limitations, and a summary.
Overall Approach and Rationale
Because they are derived from a review of the literature and the
development of the conceptual framework, the research questions naturally led to
the use of mixed research methods. Morse (2003) states that a qualitative
method used simultaneously with quantitative method with an inductive
theoretical thrust is used when some portion of the phenomenon being examined
can be measured, and this measurement enhances the qualitative description or
interpretation (p. 202). The theoretical focus of this study was inductive. I
wanted to learn from the paraeducators themselves if self-efficacy played a role
in their career decisions, specifically the decision to enter teaching. This required
53


the use of qualitative methods. However, self-efficacy is a phenomenon that can
be measured, and doing so enhanced the qualitative description in this study.
Types of Data
Quantitative.
Through twenty five years of research, a variety of scales have been
developed that measure self-efficacy in different contexts. To determine if
teacher and academic self-efficacy play a role in a paraeducators decisions to
enter teaching, this study measured their efficacy in these areas. Bandura (1997)
recommends that self-efficacy scales should measure peoples beliefs in their
abilities to fulfill different levels of task demands within the psychological
domain selected for study (p. 44). Hackett (1995) asserts that self-efficacy
assessments designed to measure specific domains are better at predicting future
behavior than broad generic measures (p. 247).
Based on these recommendations, this study was designed to measure
academic self-efficacy and teacher efficacy using scales developed for these two
domains from the PALS survey (Midgley, et. al., 2000) and general self-efficacy
using the General Self-Efficacy scale (Schwarzer & Jerusalem, 1995). I chose
teacher efficacy because that is the specific career field on which this study
focused and academic self-efficacy because paraeducators who want to become
teachers will have to complete significant educational requirements to enter the
54


profession. General self-efficacy was included to assess their ability to cope with
more general stressful situations because, as nontraditional students, they face
inevitable difficulties and obstacles that arise as they try to juggle work, family,
and school responsibilities to complete their education. General self-efficacy is a
measure of how people cope in difficult situations. High self-efficacy
expectations mean a person is more likely to persevere rather than give up when
faced with negative environments or obstacles (Betz, 1992). The scale is a
general measure that does not identify behavior change in specific areas;
therefore, it was designed to be used with other survey items that cover the
specific behavior of interest (Schwarzer & Fuchs, 1996). In this case it was used
along with scales that measure academic and teacher efficacy.
Qualitative.
The qualitative portion of the study drew information from two data
sources. Participants provided career goals statements asking them to describe
their past work experiences in education, including teaching or other experiences
related to working with children, their reasons for wanting to attend college, and
their educational goals. Selected participants who best represented each group
also participated in interviews to leam more about their overall career self-
efficacy, their work experiences, educational goals, and views of career
opportunities.
55


The need for mixed methods arose to enhance the description that
emerged from the qualitative data, to overcome the limitations of survey
research, and to explore this topic in depth. It was important to measure teacher,
academic, and general self-efficacy and to collect demographic information in
order to provide a description of the differences and similarities between the two
groups in the study. However, the survey results did not provide rich detail,
identify subtle cultural differences, or provide in-depth information (Hernandez,
2000; Marshall & Rossman, 1999) about the complexities of the career decision-
making process for paraeducators. Little is known about the role of self-efficacy
in the transition of paraeducators into teaching. This creates a need for
exploratory studies. To gain in-depth knowledge about career self-efficacy, work
environment, educational goals, views of career opportunities, and sources of
career self-efficacy for paraeducators, qualitative methods were used.
The qualitative approach enabled the study to be more flexible (Marshall
& Rossman, 1999). In the event that the survey data did not produce significant
results between the two groups, the qualitative data could shed light on the
reality. In fact, there were some unexpected outcomes in phase I of data
collection and the use of qualitative methods allowed me to pursue these
outcomes to gather more detailed information. Krathwohl (1998) stated that both
quantitative and qualitative methods could be used in exploratory studies. This
study was essentially an exploratory study using both quantitative and qualitative
56


data to better understand the role that self- efficacy plays in paraeducators
decisions to enter teaching.
Triangulation.
The two types of data also gave me the opportunity to triangulate the
data. Creswell (2003) noted that this type of triangulation uses both quantitative
and qualitative methods to offset the weaknesses inherent in one method with
the strengths of the other method (p. 217). The results of the quantitative portion
assisted with triangulation by providing quantitative details to the qualitative
portion (Morse, 2003).
Mixed Method Strategy
The mixed method strategy used in this study was the concurrent nested
model of research. In this strategy, one method takes priority and the other
method is nested in the primary method (Creswell, 2003). Since this study is
exploratory, the qualitative method was the predominant method. The
quantitative method was nested. In this type of research, the nested method can
be used to provide a detailed description of the participants (Creswell, 2003,
Creswell et. al., 2003). The differences or similarities in the two groups as
measured on the self-efficacy survey, along with the demographic information,
helped to describe the participants in great detail. The combination of both sets of
57


data provided an in-depth picture rich with detail about the career paths of
paraeducators and the factors that influence their decision to become teachers.
Research Questions
Using information gained from the development of the conceptual
framework and the review of literature, I developed the research questions. To
explain how the research questions were devised I will refer to figure 3.1.
Figure 3.1 Conceptual Framework
The figure is the graphic representation of the conceptual framework that was
discussed in Chapter 1. The factors in the outer circle are variables and contexts
58


from or in which the participants in the study may have an opportunity to
develop self-efficacy beliefs. These were examined qualitatively.
To examine school experiences, participants were questioned about their
classroom experiences while they worked as paraeducators and about their own
school experiences. Their work environment was studied by asking questions
about their career path in education, their duties in the classroom, and work
influences on their career decisions. Information about their family background
and personal characteristics was gathered by asking demographic questions.
The variables in the middle circle are the types of self-efficacy beliefs
pertinent to this study. These were examined quantitatively using self-efficacy
scales that measured each type of efficacy.
The central issue of the study, the decision to become a teacher, lies in the
center circle of the graphic. The research questions were derived from this
framework. I examined the variables in the outer circles to gain information
about the inner circle using the following research questions.
1) In what ways are paraeducators who become teachers similar and
different from paraeducators who do not become teachers in terms of:
la) Demographic characteristics
lb) Career path
lc) Classroom experiences
ld) Educational experiences
59


le) Educational goals
1 f) General self-efficacy
lg) Academic self-efficacy
lh) Teacher efficacy
li) Views of career opportunities
Sample Selection
The Participants
The sample of teachers and paraeducators came from four cohorts of
paraeducators who have been identified because of their participation in several
different grant-funded or school district sponsored paraeducator training projects.
The paraeducators who became teachers.
The sample of paraeducators who became teachers came from one of the
cohorts that completed a federally funded career ladder project through the
BUENO Center for Multicultural Education at the University of Colorado,
Boulder. The grant was a Title VII grant designed to train paraeducators to
become bilingual or ESL teachers in the northeast region of the state. Participants
came from school districts in Brighton, Ft. Lupton, and Greeley, Colorado.
The twenty students in the BUENO cohort were all paraeducators who
began their education at Aims Community College in Ft. Lupton and completed
their degrees at the University of Northern Colorado in Greeley, Colorado.
60


Twenty three students were provisionally admitted to the BUENO cohort in
1998. After completion of one quarter of coursework, twenty were formally
admitted to the project and went on to complete their degrees, giving the cohort
an attrition rate of 13% (L. Aragon, personal communication, October 8, 2003).
Nineteen are now in their first or second year of teaching. The last member of the
cohort is still completing her undergraduate education.
While the completion rate of this cohort was very high, it is not unusual
among paraeducator career ladder projects to have relatively low attrition rates
(Haselkom & Fideler, 1996). The attrition rate among paraeducators enrolled in
the Collaborative Bilingual Education Teacher Training projects housed in The
PAR2A Center at the University of Colorado at Denver is 12%. The two largest
paraeducator career ladder programs in the nation had similarly low attrition
rates. Paraeducators enrolled in the Pathways to Teaching Careers projects
nationwide had an 18% attrition rate (Clewell & Villegas, 2001). The Latino
Teacher Project at the University of California, Los Angeles had a 12% attrition
rate for paraeducators in elementary certification and 18% for paraeducators in
secondary certification (R. Baca, personal communication, October 12, 2003).
Findings for both of these large scale projects are based on over ten years of data.
Therefore, the completion rate of paraeducators in the BUENO projects, while
exceptional, is similar to paraeducators in career ladder projects nationwide.
These low attrition rates are most likely due to the comprehensive array of
61


support services, both formal and informal, provided for participants in most
career ladder programs (Haselkom & Fideler, 1996).
The current paraeducators.
The second group of participants came from three cohorts of
paraeducators. One cohort has completed a fifteen credit hour paraeducator
certificate program through The PAR2A Center at the University of Colorado at
Denver. The certificate consists of fifteen one credit hour paraeducator training
courses delivered by trained educators at school district sites. With the
implementation of the federally funded CO-TOP*ELA grant they now have the
opportunity to continue their education and complete their Associates degree at
the Community College of Denver. The grant is a Title III grant designed to help
paraeducators meet the requirements of the NCLB Act of 2001 (P. L. 107-110).
The remaining paraeducators in this subgroup came from two district
sponsored cohorts of paraeducators who are also earning their paraeducator
certificates and Associates degrees at the Community College of Denver. These
cohorts are also attending college in order to meet the requirements of the NCLB
Act of 2001 (P. L. 107-110). The districts are both located in the northeastern
Denver metro area.
There were forty two potential participants from these cohorts. All were,
at the time of the study, in various stages of completing their degrees at the
Community College of Denver and were following the same curriculum, which
62


will result in a paraeducator certificate in either special education or bilingual
education and an Associates degree in elementary education.
By choosing these particular paraeducators for the study, I ran the risk
that some of them might decide, during the course of the study, that they wanted
to become teachers. Bandura (1997), reviews research that indicates that changes
in perceived efficacy can produce behavioral changes after experiencing personal
mastery of academic tasks. In other words, once these paraeducators experienced
some success in college, their perceived academic self-efficacy may have
changed, influencing them to expand their career and educational goals.
However, I needed to use paraeducators who were taking courses at the time of
the study because the questions on the academic self-efficacy scale refer to
participants beliefs about their abilities to compete the work in their classes. It
would have been difficult for paraeducators not enrolled in classes to answer
those questions. The risk that some in the paraeducator group might want to
become teachers was acceptable to me. By examining their answers to the survey
and interview questions in comparison to the teachers and the paraeducators who
do not want to become teachers, I was able to see a continuum of the career
decision making process.
Paraeducators not included in the study.
An interesting possibility would have been to also include a group of
paraeducators who started a teacher education program but dropped out before
63


completion. However, given the extremely low attrition rates among
paraeducator career ladder programs and the number of projects to which I had
access, it would have been difficult to find participants for that subgroup in
numbers that equaled the size of the other groups.
The Sampling Method
The sampling method of identifying groups of paraeducators who met the
criteria of the study was purposive. The goal of the study was to identify
similarities and differences between two very specific groups of paraeducators.
The intent of purposive sampling is to find groups of participants in settings
where the phenomenon under study was most likely to occur (Denzin & Lincoln,
2000). Therefore, I specifically chose a group of paraeducators who have
recently become teachers and a group that was still working as paraeducators,
both of whom were currently enrolled in classes and working in schools. This
ensured that they would be able to answer the survey questions dealing with
courses (i.e. academic self-efficacy) and working with children (i.e. teacher
efficacy).
The goal of this study, as with many qualitative studies, was not to make
broad generalizations, but to gain an in depth understanding of the phenomenon
being studied (Kemper, Stringfield, & Teddlie, 2003; LeCompte & Preissle,
1993). What was needed were small samples of people that could meet the
64


specific criteria to be included in each group and could provide in-depth
information on the phenomena being studied (Kemper, et. al., 2003; Marshall &
Rossman, 1999; Patton, 1990).
The particular sampling technique used for this study was stratified
purposive sampling. This technique involves dividing the target population, in
this case paraeducators, into categories with the goal of identifying similarities
and differences across subgroups (Kemper, et. al., 2003; Marshall & Rossman,
1999). The cohorts selected were the first step of the process as they were
designed to attract a) paraeducators who want to become teachers and b)
paraeducators who primarily want to meet the educational requirements of the
NCLB act (P. L. 107-110) and do not necessarily want to become teachers,
respectively. Further criteria for being included in each subgroup were
developed.
Inclusion criteria.
Before any data was analyzed, the survey responses and demographic
information were used to screen the respondents and determine whether they met
the following criteria. To be included in the subgroup of paraeducators who
became teachers, participants had:
1) Worked as a paraeducator before entering a teacher education
program.
65


2) Worked as a paraeducator while completing their teacher education
program.
3) Had less than two years teaching experience.
To be included in the subgroup of paraeducators who do not want to become
teachers, participants were:
1) Currently working as a paraeducator.
2) Planning to remain in the paraeducator role.
3) Not currently enrolled in or no plans to enter a teacher licensure
program.
A third set of criteria was developed after the initial data analysis phase because,
as I suspected, some of the paraeducators in the second group began planning to
become teachers. Those criteria are outlined below. To be included in the
subgroup of paraeducators planning to become teachers, participants had to be:
1) Currently working as a paraeducator.
2) Interested in becoming a teacher.
3) Planning to enroll or considering enrollment, in a teacher education
program upon completion of Associates degree.
The selection process for the interview portion of the study included further
selection criteria. The first criterion was willingness to be interviewed as
indicated by the participants on the informed consent. The second criterion was
that the person chosen best represented the characteristics of the subgroup of
66


either teachers or paraeducators based on the themes that emerged in the analysis
of the survey and career goals statement data. This type of selection is called
typical case selection (LeCompte & Preissle, 1993, Miles & Huberman, 1994).
From the data, the researcher develops a profile of the average or typical case and
then chooses participants based on those criteria (LeCompte & Preissle, 1993).
The third set of criteria was the same as for the inclusion for each subgroup
outlined above. When participants were contacted they were asked the three
questions for each subgroup to ensure that they continued to meet the criteria.
Data Collection Techniques
Phase I-The Survey and Career Goals Statements
As a first step the Project Coordinator of each grant agreed to contact
their group. Lorenso Aragon contacted the BUENO Center students, Helen Berg
de Balderas contacted the CO-TOP*ELA students, and Debbie Tschida contacted
the students in the two district sponsored cohorts. The letter from each
coordinator identified me as the researcher, explained the purpose of the
research, and asked each person for their permission to release their name and
address for the purpose of being contacted for the study (See Appendix A for
examples of the letters from the project coordinators.) Only those who gave
permission to provide their name and address were contacted for the study.
67


Sixteen participants in each subgroup gave permission to be included in
the study and were asked to complete the survey and career goals statement.
Each survey was mailed with a cover letter explaining the purpose of the study, a
form to complete if they were interested in participating in an interview, the
informed consent, preaddressed return envelopes to return the survey and
informed consent separately, and $5.00 in cash. (See Appendix B for copies of
the survey cover letter and Appendix C for copies of the informed consent). The
survey contained one page of demographic information, the three self-efficacy
scales, and a section for the career goals statement, with instructions for
completion of each section (See Appendix D for copies of each survey
document). I numbered each survey so that I could identify who had returned
them. The teacher group surveys were numbered A1 through A16. The
paraeducator surveys were number B1 through B16.
The participants were given a deadline to return the survey that was three
weeks from the date of mailing. One week after the surveys were mailed, I called
each participant to notify them that the surveys had been mailed, remind them to
return the informed consent and survey, and asked them to call me if they had
any questions. The week after the return deadline, I mailed reminder postcards to
all participants who had not yet returned their survey. (See Appendix E for a
copy of the reminder postcard).
68


The BUENO Center group completed a career goals statement at the time
of application to the career ladder program five years ago. This statement was
used for the study, as it best represented their thinking at the time they were still
paraeducators making the decision to become teachers. As the surveys were
returned, I contacted the BUENO Center coordinator, and he emailed me
electronic versions of their statements.
Of the twenty participants in the BUENO Center grant project, sixteen
agreed to be contacted for the study, and ten completed the survey and gave
permission for me to review their career goals statement. Of the forty two
participants in the CO-TOP*ELA project and the district sponsored cohorts,
sixteen agreed to be contacted for the study, and twelve completed the survey
and career coals statement.
Phase II-The Interviews
The goal of the interview phase of the study was to interview at least five
participants from each subgroup. Because the Office of English Language
Acquisition (OELA) is the funding source for the two grant projects selected for
the study, and the recruiting for participants is targeted to paraeducators working
in Bilingual/ESL programs, the groups were fairly similar in demographic
characteristics. Not all, but the majority of participants were female, Latina,
nontraditional age students, who were parents and were working as paraeducators
69


when recruited. The demographics of the district sponsored groups were similar
with the exception that the more were white, not of Hispanic origin. Patten
(1997) states that if the population for a study is homogenous, then sample size
can be small. Based upon the number of surveys returned, ten people represented
approximately one third of the entire sample. However, in qualitative research
sample size needs to remain flexible (Marshall & Rossman 1999). With this in
mind, initially six participants from the teacher group were interviewed and five
participants from the paraeducator group were interviewed.
After an initial review of the data, I determined that saturation had been
reached for the teacher group. Saturation is a term used in qualitative research
that means that data continues to be collected until relatively little new
information is emerging (Krueger, 1994). The six interviews conducted for the
teacher group represented sixty percent of the total respondents and that
subgroup category was well saturated.
Of the five paraeducators who were interviewed for the second subgroup,
four wanted to remain as paraeducators and one was considering becoming a
teacher. The subgroup of paraeducators who wanted to remain in the
paraeducator role was saturated at this point. However, the emerging subgroup of
those planning to become teachers provided new information that needed to be
pursued. Consequently, new criteria were developed for the subgroup of
paraeducators who were planning to become teachers, and three additional
70


paraeducators were selected and interviewed. (See Appendix F for the interview
protocols) Table 3.1 presents a summary of the number of study participants.
Table 3.1
Number ofparticipants in each phase of the study
Phase Potential Actual Number by
Phase I Survey Number Number Subgroup Total
BUENO Center 16 10 10a 10
CO-TOP*ELA 10 7 4b 3C
District A cohort 4 4 lb 3C
District B cohort Phase II Interview 2 1 lc 12
BUENO Center 9 6 6a 6
CO-TOP*ELA 6 5 3b 2C
District A cohort 2 2 lb lc
District B cohort 1 1 lc 8
Subgroup 1 = paraeducators who became teachers. bSubgroup 2 = paraeducators who planned to
become teachers. cSubgroup 3 = paraeducators who planned to remain in the paraeducator role.
71


Compensation to participants.
Participants were offered a gift for participation in each phase of the data
collection process. Participants in phase I received five dollars for completing the
survey. Participants in phase II received the choice of a gift certificate to either
Borders Bookstore or Safeway in the amount of $20.00 to thank them for
volunteering for the interview. The human subjects review committee suggested
referring to each as gifts rather than payments in all study documents, including
the informed consent. (See appendix G for the Human Subjects approval memo.)
Data Management and Protection Procedures
Data from the survey was recorded on the survey instrument and entered
into an SPSS 12.0 database. Interviews were tape recorded, transcribed and
entered into QSR*NVivo. I received the career goals statements from the
BUENO Center group electronically. This meant that I only had to save them as
rich text files and then import them into QSR*NVivo. Career goals statements
from the CO-TOP*ELA and school district groups were hand written in the
survey instrument itself. These were typed into a word processing program,
saved as rich text files, and then entered into QSR*NVivo for analysis. The
demographic data was entered into a word processing program and put into table
format.
72


I decided to do my own transcription of the interviews for several
reasons. First, my preferred transcriber was not available. Second, several of the
people I interviewed spoke English with an accent. This is not a problem for me,
but I have experienced a loss of data when my alternate transcriber processed
interview documents for another study in which some of the participants spoke
with accents. I wanted to avoid this problem. Finally, following Wolcotts advice
(1990a) I wanted to immerse myself into the details before I broke it into parts
(p.128). Spending time transcribing the interviews verbatim facilitated this
immersion. After the interviews were transcribed, they were entered into
QSR*NVivo and coded.
Data will be maintained for a minimum of three years. Audiotapes and
paper surveys will be kept in a locked cabinet at the home of the researcher.
Coded data will be stored on Zip disks, and back up copies will be made to be
kept in two alternate locations.
Data Analysis Procedures
Demographic Analysis
Answers to the demographic questions on the survey were entered into a
word processing program and put into a table. A detailed demographic
description of the participants appears in Chapter 4. Demographic information
included gender, age, ethnicity, level of education, high school grades, number of
73


years worked as a paraeducator, personal income, family income, number of
children and other dependents, and native language.
Also included was a question about which careers the participants might
consider in the future. The teachers were asked what careers they saw themselves
in ten years from now. Their choices included teacher, administrator, counselor,
and special service provider (i.e. speech pathologist, audiologist, school
psychologist, etc.). I chose these careers because they are all options in K-12
education for teachers who want to move into different roles. However, some
may require education beyond the Bachelors degree. Paraeducators were asked
the same question and given the options of paraeducator, teacher, counselor, or
day care director. I chose these options because the current degree program they
are completing can facilitate their entry into each of these careers. Again, some
may require education beyond the Associates degree. I also summarized the
results of the answers to this question in a table using a word processing
program.
For the demographic analysis I used two subgroups, the paraeducators
who became teachers and the current paraeducators. During this analysis I
discovered that half of the current paraeducators wanted to become teachers.
Based on this new information, the remaining analyses used three subgroups; the
paraeducators who became teachers, the paraeducators who are planning to
become teachers, and the paraeducators who wish to remain in that role.
74


Quantitative Analysis
The data from the surveys was entered into SPSS 12.0. For this study, a
was set at .05. Mean scores on the general self-efficacy scale, the academic self-
efficacy scale, and the teacher efficacy scale were computed for each participant.
An overall group mean was also computed for each group on each of the three
self-efficacy scales. Group means were analyzed using one-way ANOVA to test
for differences between the subgroups on the three measures of self-efficacy.
Qualitative Analysis
My process for analyzing the qualitative data involved several steps.
First, I determined that a quote would constitute one complete thought or idea. If
that thought or idea was part of a story, and needed to be contextualized, I
included the story as part of the quote. If a participant made a statement that was
very long I tried to look for different themes in the quote to break it into smaller
ideas.
I was already extremely familiar with the data because I had done all the
transcription. In the first round of coding I went through each document and
coded based on my first instinct or reaction to the quote. The four sources of self
efficacy mastery experiences, social role models, verbal persuasion, and
emotional control (Bandura, 1977, 1997) were used as an initial lens to view the
75


career goals statements and the responses to interview questions. I created nodes
for each of the four sources of self-efficacy. In addition, I created nodes based on
the specific questions asked in the interview. As themes outside of this initial
framework emerged, I added to the node tree in QSR*NVivo. From this a picture
of the personal experiences of these paraeducators and teachers was created. (See
Appendix H for a listing of nodes for both the career goals statement analysis and
the interview analysis.) Next, I went back and reviewed certain quotes that I had
questions about and either recoded them, added them to multiple nodes, or left
them coded as they were.
Next, I put the documents into sets. For the analysis of the data I put the
documents into three groups. One set was the paraeducators who became
teachers. There were ten participants in this group. The second set was current
paraeducators who wanted to become teachers. There were six participants in this
group. Finally, I created a set for current paraeducators who wanted to remain in
the paraeducator role. There were also six participants in this group.
After creating sets, I ran reports for each set. I printed a report for each of
the nodes I created for the analysis. I combed the reports to determine if each
quote truly belonged in that node. If there were quotes that I identified that
needed to be moved or included in more than one category, I went back into the
project in QSR*NVivo, made the changes, and then printed new node reports.
76


Finally, I reviewed each node report again to determine major themes,
similarities, and differences between the two groups.
Trustworthiness of the Research
As a researcher, it is important to establish that I have conducted high
quality research (Johnson & Turner, 2003). Qualitative researchers use the terms
trustworthiness, credibility, transferability, and confirmability to evaluate
research (Denzin & Lincoln, 2000; Johnson & Christensen, 2000; Lincoln &
Guba, 1985). Quantitative researchers focus on reliability and validity (Campbell
& Stanley, 1963). The predominant method in this mixed methods study was
qualitative, so I chose to rely on the trustworthiness criteria in evaluating the
study, although reliability and validity data were provided for the survey scales in
the instrumentation section that follows. Creswell and Miller (2000) developed a
list of eight procedures that can be used by qualitative researchers to document
credibility, transferability, dependability, and confirmability. They are; a)
prolonged engagement in the field; b) triangulation; c) peer review or debriefing;
d) negative case analysis; e) clarifying researcher bias; f) member checks, g)
thick description; and h) external audits. Other procedures identified in the
literature include purposive sampling, code-recode strategy, and detailed
description of participants (Anfara, Brown, & Mangione, 2002, LeCompte &
Goetz, 1982). Creswell (1998) recommends that a study should use at least two
77


validation procedures. In this study, I used eight procedures to ensure
trustworthiness of the research. The following table 3.2, adapted from Anfara, et.
al., depicts both types of criteria and the methods I used in the study to document
each one.
Table 3.2
Quantitative and qualitative criteria for evaluating research quality
Quantitative Methods Qualitative Methods Strategy Used
Internal Validity Credibility Tri angulation Negative case analysis
External Validity Transferability Rich, thick description Purposive Sampling Participant description
Reliability Dependability Interrater reliability Tri angulation Reliability data on scales
Objectivity Confirmability Tri angulation Researchers bias & role Negative case analysis
78


Triangulation
Triangulation is a method of cross checking the data (Lincoln & Guba,
1985). One purpose of the mixed methods in this study was to triangulate using
the three types of data to confirm and crosscheck the data (Creswell, 2003;
Erzberger & Kelle, 2003). To gather evidence from multiple sources using
different methods for the purpose of triangulation, I used the surveys, the career
goals statements, and the in-depth interviews.
Negative Case Analysis
During data analysis, the researchers must examine and analyze not only
the positive occurrences of the phenomenon being studied, but also the negative
occurrences (Lincoln & Guba, 1985). Initially, I discussed with my committee
interviewing one person from each of the subgroups who did not fit the profile of
his/her group as a form of negative case analysis. However, as I examined the
results of each phase of data collection, I realized that there was not one
individual in each group who did not fit the profile. Most of the participants in
each subgroup were very similar to each other. Certain individuals in each group
may have had one or two characteristics that were not in common with the
others, but none deviated from the norm overall.
Alternatively, as I analyzed the demographic information, career goals
statements, and the interviews, I identified characteristics, themes, and statements
79


that did and did not support my working hypotheses and findings. This ensured
that the findings emerged entirely from the participants, independent from my
biases and beliefs (Denzin & Lincoln, 2000; Lincoln & Guba, 1985; Miles &
Huberman, 1994). The negative evidence that emerged from the data analysis is
included in Chapter 4 and is referred to as non-exemplars.
Rich and Thick Description
When describing the findings, the researcher should use rich and thick
description to increase trustworthiness by giving the reader authentic detail about
the phenomenon being studied. (Creswell, 2003; Denzin & Lincoln, 2000). To
report the qualitative findings of this study I used detailed description and direct
quotes from the participants to tell the story in their own words and from their
perspective. The quantitative data was used to supplement and support their
descriptions.
Purposive Sampling
Purposive sampling allows the research to use specific criteria to select
the participants or cases that will best answer the questions or test the hypothesis
by giving in depth information (Kemper, et. al, 2003; Maxwell & Loomis, 2003).
I carefully selected participants for this study first by identifying training projects
that attracted paraeducators with different career goals and second by using clear
80


criteria for each person to be included in the two subgroups. My goal was to
obtain in depth information on the role of self-efficacy in the career decision-
making process for paraeducators who become teachers and those who do not.
Additional criteria were developed when a third subgroup of paraeducators who
were planning to become teachers emerged during the data analysis. Along with
the criteria that were developed for participants to be included in the study the
subgroup criteria was used to screen participants for the interview phase of the
study. Participants were checked at the beginning of each phase of the research
study to ensure that they continued to meet the criteria for inclusion.
Detailed Description of Participants
LeCompte and Goetz (1982) state that a detailed description of the
participants helps to eliminate possible informant bias by identifying the groups
from which the participants come. I asked the participants to complete
demographic information as part of the survey process so that I could construct a
detailed description of the groups. That description can be found in chapter 4.
Included in chapter 3 is information on the attrition rates of career ladder
programs for paraeducators both in Colorado and the nation to demonstrate the
similarities the BUENO Center cohort has to participants in other career ladder
projects. This information will also be useful for replicating this study.
81


Interrater Reliability
Peer examination or debriefing is another method of establishing
trustworthiness of the research by testing the quality of the researchers
inferences (Creswell, 2003; LeCompte & Goetz, 1982; Lincoln & Guba, 1985;
Miller, 2003). In this study, I used interrater reliability to check inferences of the
coded data. There are several methods of calculating interrater reliability
beginning with a simple percentage of agreement. Goodwin (2001) states that it
is preferable to use multiple sources of data to build evidence of interrater
reliability, but there are only two methods that can be used with nominal data. I
chose to use them both. I calculated percentage of agreement and a Cohens
kappa (1960) coefficient. Unlike simple percentage of agreement, the Cohens
kappa method takes into consideration the amount of agreement that can be
attributed to chance (Goodwin, 2001).
Cohens kappa calculation.
The formula for calculating Cohens kappa is as follows:
K = (Fo-Fc)/ (N Fc)
In the equation:
N = the total number of codes assigned by each rater.
Fo the number of codes on which the raters agree, and
Fc = the number of codes that can be expected to agree by chance
82


Cohens method of determining the Fc was to create a contingency table to
examine the judges ratings across categories. Then the row and column totals
are multiplied to determine marginal probabilities (Goodwin, 2001). Cohens
kappa coefficients can be interpreted like Pearson coefficients. Generally a kappa
of .70 or higher is considered an appropriate level of interrater agreement (Harris,
Pryor, & Adams, n.d.).
Two colleagues, both of whom have conducted research in the
paraeducator field, served as my raters for this process. I developed a survey to
be given to each rater. The survey explained the purpose of interrater reliability
and included definitions of each category of coding node. I provided definitions
for the raters to ensure that the raters were using the same criteria for classifying
the quotes that I did. There were two sections to the survey, one for the career
goals statements and one for the interviews. Each contained selected passages
from a portion of the coded data. To determine which quotes to include in the
survey, I systematically selected every third quote from reports printed for each
node in the analysis. (See Appendix I for a copy of the interrater reliability
survey).
A total of seventy quotes were used for this analysis. This number
represented about 15% of my data. A general rule of thumb in determining
interrater reliability is to use 10% of the data (L. Goodwin, personal
communication, April 14, 2003). I searched the literature to find a source for this
83


recommendation, but was unsuccessful. Interrater reliability is discussed in many
research methods texts and articles. However, a step-by-step process for carrying
out your interrater reliability analysis is not usually included. As an alternative, I
chose to use a more conservative amount or 15% of my data.
My colleagues ratings were compared to mine. Evidence of the
reliability of the coding was calculated using the two methods chosen for this
study. The results of this analysis for rater number one produced a simple
percentage of agreement between my coding and rater number ones coding of
86%. The Cohens kappa calculation resulted in a coefficient of .83. For rater
number two, the simple percentage of agreement between my coding and rater
number twos coding was 89%. The Cohens kappa calculation resulted in a
coefficient of .85.
Most notable after this analysis were the ratings in the emotional control
category. Both the raters were very consistent with their ratings in this category,
and they consistently disagreed with me. I rated eleven quotes in the interrater
survey as mastery experiences. They rated only six of those as mastery
experiences. The remaining five they coded as emotional control. When I saw
this consistency in their ratings, I discussed it with them and reviewed my
coding. The raters considered quotes in which participants spoke of their feelings
about their jobs as emotional control. If their experiences helped them develop
positive feelings about teaching, they could have integrated this emotional and
84