Relationships of attitude toward a behavior, subjective norm, and perceived behavioral control as antecedents to compute...

MISSING IMAGE

Material Information

Title:
Relationships of attitude toward a behavior, subjective norm, and perceived behavioral control as antecedents to computer use by elementary teachers in a public school setting
Physical Description:
xi, 160 leaves : ill. ; 29 cm.
Language:
English
Creator:
Geddes, James Michael
Publication Date:

Subjects

Subjects / Keywords:
Curriculum and Instruction (ISC) thesis, Ph. D   ( lcsh )
Dissertations, Academic -- Teaching and Learning -- UF   ( lcsh )
Genre:
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )

Notes

Thesis:
Thesis (Ph. D.)--University of Florida, 2004.
Bibliography:
Includes bibliographical references.
Statement of Responsibility:
by James Michael Geddes.
General Note:
Printout.
General Note:
Vita.

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 003163331
System ID:
AA00011856:00001


This item is only available as the following downloads:


Full Text










RELATIONSHIPS OF ATTITUDE TOWARD A BEHAVIOR, SUBJECTIVE NORM,
AND PERCEIVED BEHAVIORAL CONTROL AS ANTECEDENTS TO COMPUTER
USE BY ELEMENTARY TEACHERS IN A PUBLIC SCHOOL SETTING














By

JAMES MICHAEL GEDDES


A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA


2004
































Copyright 2004

by

James Michael Geddes














ACKNOWLEDGMENTS

Many individuals and influences have made this achievement possible at this

particular point in my life. I first thank God for the many blessings and abundant grace

He has bestowed on me. I firmly believe that every good gift comes from above (James

1:17) and that the greatest knowledge (Gk. gnosis) that we can have is knowledge of

Him. My greatest earthly appreciation goes to my wife, Phyllis. I would not be what I am

or where I am today without her unselfish love, encouragement, and support. I doubt she

bargained for a life-long student when we got married in 1977, but she has never

complained and has always been proud of my achievements, which I consider our

achievements. I thank my mother, Frances, for her encouragement to persist early in my

academic career, and for her example as an educator and an adult student. My thanks go

to my doctoral committee chair, Dr. Kara Dawson, who was a great source of

information, direction, inspiration, and motivation when needed, despite her own

challenges of a busy schedule and a young family. My committee overall was extremely

supportive and helpful and in this regard I thank Dr. Jeff Hurt for his leadership in my UF

work, and the contribution to my life and academic work of Dr. Lee Mullally, Dr.

Colleen Swain, and Dr. David Honeyman. Finally, I express my appreciation and thanks

to my employer, the Citrus County School District. I work with a great group of

colleagues and professionals who value life-long learning and they have been extremely

supportive in my pursuit of an advanced degree. We work in a wonderful, real-life

laboratory and I learn from them on a daily basis.













TABLE OF CONTENTS

Page

ACKNOWLEDGMENTS ................................... .......................................... iii

LIST OF TABLES ................................................ vii

LIST O F FIG U R E S ........................................ ......................... .......................... ix

A B S T R A C T ..................................................................................... ................. ........... x

CHAPTER

1 IN T R O D U C T IO N .................................................................... .............................

Statem ent of the Problem ................................... ........... ..........................
N eed for the Study................................................................... ...................... 3
Significance of the Study........................................................... ......................5
D definition of Term s ........................................ .................................................5
L im itations................................. ........................................... ......................... 7
A ssum options .............................................................................. .................... 7
R research Q questions ...................................................................... ...................8
State ent of H ypotheses ........................................................... ...................... 9

2 REVIEW OF LITERATURE................................................. ......................10

Computer Use by Teachers .............................................................
Teacher Use of Computers as an Innovation..................... ....................11
The Context of Teacher Computer Use..................... ......... .............. 12
The Teacher and Computer Use................................ ......................13
Stages of Technology Use............................... ..... ..... .................... 16
T teacher B eliefs .................................................................................................... 19
Theory of Planned Behavior................................................. ........................21
Attitude Toward the Behavior........................... ... .......................22
Subjective Norm...............................................................22
Perceived Behavioral Control .................................. .......................24

3 RESEARCH DESIGN AND METHODOLOGY.................. ...........................27

Research D esign .................................................................................................. 27








Population and Sample .............................................. ......................28
Instrumentation Development .................................................... ...................29
Behavior of Interest......................... ...... .... ....................30
Behavioral Intention ................................... .... .. ....................30
Theory of Planned Behavior Variables.......................... ........................31
Elicitation Study......................... ................... ............. ............32
Beliefs and Strength Evaluations ..................... ...........................33
Questionnaire Development........................................................ ...........34
Levels of Computer Use.................... .................................35
Procedures for Piloting the Instrument......................... ... .....................37
C ontent V alidity ................................................................ .....................38
R eliab ility ........................................................................ .............................3 9
D ata C collection ................................................................... .................................45
D ata A nalysis.................................................................... ...................................46

4 R E S U L T S ................................................................................. ........................... 4 8

Research Questions ............................................. ..........................................48
Survey Item Responses...................................... ............................................49
Demographic Characteristics of Respondents...................... ..................... 54
Criterion Variable...................................................................... ......................56
Predictor V ariables .................................................................... .................... 57
A ttitu d e ...................................................................... ...................................58
Subjective Norm..........................................................................................59
Perceived Behavioral Control ........................................... .....................59
Statistical A analysis ..................................................... .................................. 62
S um m ary ......................................................................... .....................................66

5 DISCUSSION, CONCLUSIONS, AND RECOMMENDATIONS .......................68

Overview of the Study.............................................................. .....................68
Research Questions .................................................................... ...................70
K ey F ind in g s ............................................................................ ...........................70
D iscussion....................................... ...... ... ............................................... 72
Influence of Teacher Beliefs on Practice ..................... ......................73
Contribution of TPB Variables ........................................... .......................78
Levels of Computer Use................... ...................................79
Implications .......................................... ....................................80
Recommendations for Further Research .......................................................82
S um m ary ......................................... ....................................................................83

APPENDIX

A DISTRICT PERMISSION LETTER ................................... .................84

B UNIVERSITY INSTITUTIONAL REVIEW BOARD FORM ................................ 87








C ELICITATION STUDY QUESTIONNAIRE ..........................................89

D ELICITATION STUDY COVER LETTER ................... ..........................91

E ELICITATION STUDY RESULTS ................................... .....................92

F D O M A IN EX PERTS .............................................. ...................................... 99

G DOMAIN EXPERT COVER LETTER............................................................ 100

H PILOT SURVEY FEEDBACK FORM .............................. .......................104

I SAMPLE TPB QUESTIONNAIRE...................... .... ...................107

J PILOT SURVEY QUESTIONNAIRE........................... ....... ............. 116

K REVISED PILOT SURVEY QUESTIONNAIRE.............................................127

L FINAL SURVEY QUESTIONNAIRE.................................. ......................137

M FINAL SURVEY COVER LETTER............................ ....................147

N FINAL SURVEY REM IN DER ........................................... .............................149

O CORRELATIONS BETWEEN TPB FACTORS AND SURVEY ITEMS............150

P STATISTICAL RELATIONSHIPS OF FINAL SURVEY..................................153

REFERENCES ............. ........................................154

BIOGRAPHICAL SKETCH ...................................................... .....................160














LIST OF TABLES


Table Page

3-1. Theory of Planned Behavior Constructs and Related Final Survey Items ...............35

3-2. Pilot Survey Instrument Measures, Item Numbers, and Alpha Coefficients..............40

3-3. Revised Pilot Survey Instrument Measures, Item Numbers, and
A lpha C oeffi clients ...................... ....................................... ....................... .... 4 1

3-4. Final Survey Instrument Measures, Item Numbers, and Alpha Coefficients.............42

3-5. Alpha Coefficients for Pilot Survey and Final Survey.......................................43

4-1. Final Survey Items, Response Options, Scaling, Means, and Standard Deviations...49

4-2. Demographic Characteristics of Respondents...................... .....................55

4-3. Levels of U se Reported ............................................................... ...................56

4-4. Descriptive Statistics for Belief-based Attitude Measure.........................................59

4-5. Descriptive Statistics for Belief-based Subjective Norm Measure ..........................59

4-6. Descriptive Statistics for Belief-based Control Measure .........................................60

4-7. Correlation of Belief-based and Direct Measures .......................................61

4-8. Correlations among Attitude, Subjective Norm, Perceived Behavioral Control and
B behavioral Intention............................................................ .......................63

4-9. Bivariate and Partial Correlations of the Predictors with Behavioral Intention.........64

4-10. Descriptive Statistics for Intention and Levels of Use Measures...........................66

4-11. Correlation of Behavioral Intention and Levels of Use..........................................66

E-1. Advantages of Using the Computer to Present a Lesson During Instruction...........92

E-2. Disadvantages of Using the Computer to Present a Lesson During Instruction......94









E-3. Approve of My Using a Computer to Present a Lesson During Instruction ............95

E-4. Disapprove of My Using a Computer to Present a Lesson During Instruction........96

E-5. Factors or Circumstances that Enable You to Use a Computer to Present a Lesson
D during Instruction ............................................................. ...................... 97

E-6. Factors or Circumstances that make it Difficult or Impossible for You to Use a
Computer to Present a Lesson During Instruction ......................................... 98

0-1. Correlations Between TPB Factors and Attitude Survey Items...........................150

0-2. Correlations Between TPB Factors and Subjective Norm Survey Items...............151

0-3. Correlations Between TPB Factors and Perceived Behavioral Control Survey
Item s ................................................................................... ............. ........... 152














LIST OF FIGURES


Figure Pae

2-1. Theory of Planned Behavior Model .......................................... ............. 25

3-1. Scree Plot of TPB Variable Factor Analysis..................... ...................44

4-1. Belief-based M measure of Attitude.................................................58

4-2. Belief-based Measure of Subjective Norm.........................................59

4-3. Belief-based Measure of Perceived Behavioral Control ....................................... 60

4-5. Relationship of Behavioral Intention to Actual Behavior ......................................... 66

P-I. Statistical Relationships of Final Survey................... ........................153














Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy

RELATIONSHIPS OF ATTITUDE TOWARD A BEHAVIOR, SUBJECTIVE NORM,
AND PERCEIVED BEHAVIORAL CONTROL AS ANTECEDENTS TO COMPUTER
USE BY ELEMENTARY TEACHERS IN A PUBLIC SCHOOL SETTING

By

James Michael Geddes

December 2004

Chair: Kara M. Dawson
Major Department: School of Teaching and Learning

Despite significant financial investments made in recent years to equip public

schools with computers, some observers believe we are not seeing an equivalent return in

terms of the percentage of teachers who use computers for instruction and related

improvements in teaching and learning. A number of factors, primarily external to

teachers, have been suggested as reasons for lower levels of computer use. The purpose

of this study was to investigate internal variables related to elementary teacher use of

computers for instruction. The predictor variables for this study were derived from

Ajzen's Theory of Planned Behavior, and they included attitude toward the behavior

(ATT), subjective norm (SN), perceived behavioral control (PBC) and behavioral

intention (INT). The criterion variable was levels of computer use as measured by

Marcinkiewicz' Levels of Use instrument. The sample consisted of classroom-based

elementary teachers (n = 203) in a central Florida school district.







A 73-item survey instrument was developed and validated to capture data for this

study. The survey captured indirect (belief-based) and direct measures of the variables of

interest as well as pertinent demographic data. Construct validity for the instrument was

investigated through item analysis and confirmatory factor analysis. Internal consistency

estimates for item groups were calculated through the use of Cronbach's alpha (a).

Descriptive statistics were also calculated for each item. Relationships among the

variables were analyzed using Pearson's correlation coefficient (r), and multiple linear

regressions (R).

The findings revealed a positive correlation between attitude toward the behavior

(ATT) and intention, between subjective norm (SN) and intention, and between perceived

behavioral control (PBC) and intention. The three variables of ATT, SN, and PBC

combined accounted for 65% of the effect size on behavioral intention (INT). The

attitude variable had the largest influence on intention. A significant correlation was also

identified between behavioral intention and levels of computer use reported by teachers.

Results of this study seem to validate the Theory of Planned Behavior as a

predictive model in this context. The findings show the importance of internal,

personological variables on the successful integration of computer technology in a K-12

setting and they suggest that more research is needed in this area.














CHAPTER 1
INTRODUCTION

Computers have a ubiquitous presence in school classrooms in 2004, yet many

teachers have not embraced this technology as an instructional tool. My study analyzed

elementary teacher use of instructional technology, specifically computers, from a

behavioral perspective. The Theory of Planned Behavior (Ajzen, 1988, 1991) was used as

the theoretical framework. The constructs of attitude toward the behavior, subjective

norm, and perceived behavioral control were investigated as antecedent factors related to

intention and demonstrated behavior of teacher computer use in the classroom.

Statement of the Problem

Investment in instructional technology by K-12 public education has been

enormous in the last 15 years (Office of Technology Assessment, 1995). Total

technology spending among U. S. public school districts for the 2002-2003 school year

was $5.74 billion, with projections of $5.80 billion for the 2003-2004 school year

(Quality Education Data, Inc., 2003). This investment has been made primarily in

computers and in associated costs such as networking and Internet access. The claim of

promoters has been that instructional technology will improve teaching and learning and

that it will make the process of education more efficient and effective. Instructional

technology in the form of computers is not the first technology promoted by such claims.

Similar results were anticipated with earlier technologies such as instructional radio,

instructional television, and various "teaching machines" (Ohles, 1985; Cuban, 1986).








Like those earlier technologies, computers have been embraced in classrooms by only a

small percentage of teachers as effective users (Marcinkiewicz, 1994; Zhao & Frank,

2003). A quick walk-through of schools reveals large numbers of computers in

classrooms. In 2002, student-to-computer ratios in Florida's K-12 public school systems

averaged 3.7 : 1 (Council for Education Policy, Research, and Improvement, 2002).

Despite their ubiquity, a relatively small percentage of teachers are using computers on a

regular basis; and when they are used, computers are often used to support traditional

curriculum and teaching methodologies and not in transformational ways (Hadley &

Sheingold, 1993; Cuban, Kirkpatrick, & Peck, 2001; Zhao & Cziko, 2001).

Although the use of computers has become more widespread in the workplace,

elementary and secondary teachers are less likely to use computers than are persons

employed in other managerial or professional fields (USDOE, 2002). Significant

investments in computer hardware and teacher training have been made in K-12 schools,

yet a relatively small percentage of teachers are using computers effectively in

instruction. In their discussion of policy in relation to technology, Means et al. state

"technology's potential for profound influences on instruction is yet to be realized"

(Means, Roschelle, Penuel, Sabelli, & Haertel, 2003). This perceived low return on

investment has become a source of concern for some, which has developed into criticism

that financial resources could be better used in other ways to improve teaching and

learning (Oppenheimer, 1997; Cuban, 2001). Implementers have been successful in

getting computers in classrooms but they have failed to adequately consider factors that

might predict or influence effective use by teachers.








Need for the Study

There is a need to better understand why teachers are not using computers in

greater numbers so we can develop processes and strategies to help teachers become

more effective in the use of these instructional tools to increase student achievement. An

examination of instructional technology in an instructional setting reveals three related,

interactive domains: the teacher, the innovation, and the context (Zhao, Pugh, Sheldon, &

Byers, 2002). These three domains provide a useful framework for considering effective

instructional technology use in a K-12 educational setting, and they facilitate analysis

consistent with the approach taken by Ertmer (1999) in terms of internal and external

factors. Of these three domains, the two most examined in relation to effective use of

instructional technology in the classroom setting are innovation and context, because

these domains are the easiest to visualize, examine, and quantify. There has been much

less analysis of the teacher domain in relation to instructional technology use due to the

complexities of human nature and the number of variables involved.

The use of computers in the classroom has been considered from the perspective of

the computer as an educational innovation (Cuban, 1986; Marcinkiewicz, 1994). Rogers

(1995) identified characteristics of innovations that facilitate higher levels of adoption,

and these characteristics have merit in relation to teacher use of computers in the

classroom. From a consideration of context, there is evidence suggesting that the

traditional time, space and curriculum structures of K-12 public education are limiting

factors in terms of teachers adopting computers in classrooms (Bossert, 1996; Cuban,

Kirkpatrick, & Peck, 2001). Ely (1999) identified a number of conditions that facilitate

change, and most of these conditions are external, or contextual to the user. Many

contextual factors related to instructional technology in the classroom have been








addressed (such as training for sufficient knowledge and skills, accessibility, time to

learn, and support). Despite the availability of computers in classrooms, and higher levels

of teacher training, adoption and use levels of computer technology are still quite low.

Becker (1994) observed that only a very small percentage of classroom teachers could be

classified as exemplary computer users in 1993. Ten years later, the percentage of

teachers using computers for instruction in transformational ways had not increased

significantly (Florida Department of Education, 2003). To date, a number of studies have

examined the phenomenon of low teacher use of computers in the classroom, but these

studies have largely looked at first-order (external) barriers (Ertmer, 1999) such as

limited resources and support. Second-order barriers to technology integration are

intrinsic to teachers and include "beliefs about teaching, beliefs about computers,

established classroom practices, and unwillingness to change" (Ertmer, 1999, p.48).

There has been much less research into second-order (internal) barriers to technology use

such as belief systems, attitudes, values, and perceived control. Although change and the

diffusion of an innovation is a social process (Rogers, 1995), it ultimately involves an

individual going through an evaluation process and making a decision to change. In the

change process, what people do and do not do is the crucial variable (Fullan &

Stiegelbauer, 1991). Teachers are critical components of the change process in education

(Fullan & Stiegelbauer, 1991), and their perspective is often overlooked when a decision

is made to deploy a technology in a classroom setting (Cuban, 1986). Because internal

and personological variables related to levels of computer use by teachers have often

been overlooked (Marcinkiewicz, 1994), this study is needed to enable us to better

understand these variables in the teacher domain to facilitate higher levels of use and







more effective use of computers by teachers. A more complete understanding of internal

factors related to computer use by teachers will enable us to better predict how teachers

will use computers in the instructional process, and it will facilitate the development of

more effective implementation strategies for this technology in classrooms.

Significance of the Study

This study is important because it examines internal variables related to elementary

teacher use of computers for instruction. Specifically, it looks at the constructs of attitude

toward the behavior, subjective norm, perceived behavioral control, and behavioral

intention as delineated in the Theory of Planned Behavior (Ajzen, 1988). Olher :rudlJc

have measured individual internal factors related to computer use such as attitude or self-

efficacy, but a search of the literature found no study that measures attitude toward the

behavior, subjective norm, and perceived behavioral control in relation to teacher

computer use. Attitude toward the behavior, subjective norm, and perceived behavioral

control, collectively, are the antecedents to behavioral intention which is highly

correlated to actual behavior (Ajzen, 1988, 1991). The instrument developed from this

study will be a useful tool for researchers and implementers interested in more effective

computer use by teachers; and it will facilitate a better understanding of the-related

internal and personological variables. This study also extends previous research that used

the Theory of Planned Behavior in other disciplines, and it tests the validity of this

theoretical framework in relation to teacher use of computers in the classroom.

Definition of Terms

To provide a clear understanding of the concepts and terms related to this study, the

following definitions are provided:







Theory of Planned Behavior (TPB) is a framework for the study of human behavior

(Ajzen, 1988, 1991). This theory proposes that human behavior is guided by three belief

areas: behavioral, normative, and control. TPB looks at the constructs of attitude,

subjective norm, and perceived behavioral control as antecedents to behavior. TPB has

been validated in a number of studies across various disciplines.

Attitude toward the behavior is the degree to which a person has a favorable or

unfavorable evaluation or appraisal of the behavior in question (Ajzen, 1988, 1991).

Subjective norm is the perceived social pressure to perform or not perform a

particular behavior (Ajzen, 1988, 1991).

Perceived behavioral control is a person's perception of the ease or difficulty of

performing the behavior of interest (Ajzen, 1988, 1991). Perceived behavioral control is

assumed to reflect a person's previous experiences, as well as anticipated challenges and

obstacles.

Behavioral intention is an individual's intention to perform a given behavior

(Ajzen, 1988, 1991). In this definition, intention is influenced by the constructs of

attitude toward the behavior, subjective norm, and perceived behavior control.

Teacher computer use in my study refers to the use of a computer for instructional

purposes by a classroom teacher. Teacher instructional use of computers involves having

students use software programs as part of their instructional experience, as well as for

demonstration or presentation of a lesson (Becker & Riel, 2000). It is operationally

defined in this study as non-use, utilization, and integration as determined by the Levels

of Use (LU) instrument (Marcinkiewicz, 1991).








Teacher beliefs are the ideas teachers believe in and are committed to. They are

sometimes called core values (Lumpe & Chambers, 2001). Beliefs "shape goals, drive

decisions, and create discomfort when violated"(Loucks-Horsley et al., 1998, p.18).

Belief implies the mental acceptance of something as true. There is a critical relationship

between the beliefs of teachers and the instructional decisions they make (Haney,

Czemiak, & Lumpe, 1996).

Limitations

1. The study focused on elementary teachers in a public school district, and the

results may not generalize to secondary teachers due to differences in curriculum needs,

organizational climate, and school culture.

2. Though the overall response rate was good, the study was limited by the

willingness of participants to complete the research instrument.

3. The generalizability of the results is limited to the sample of respondents from

one school district that participated in the study.

4. The study was limited by the degree to which participants understood the

research instrument. For example there could be different interpretations by teachers of

what it means to use a computer for instruction.

5. The study was limited by the degree to which participants could objectively

analyze themselves in relation to the behavior of interest using a self-report (survey)

based instrument.

Assumptions

This study is based on the following assumptions:








1. The instruments used measured attitude toward the behavior, subjective norm,

and perceived behavioral control toward teacher use of computers in the classroom for

instruction.

2. The participants adequately understood the research instrument.

3. The participants answered the research instrument honestly.

4. Behavioral intention is related to and predicted by attitude toward the behavior,

subjective norm, and perceived behavioral control of a person toward the behavior.

5. Attitude as a construct in this study is a function of the strength of a person's

belief about the consequences of the behavior and evaluation of the consequences.

6. The participants are able to adequately rate themselves using a self-report

instrument.

Research Questions

1. What is the relationship between attitude toward the behavior and an elementary

teacher's intention to use computers in classroom instruction?

2. What is the relationship between subjective norm and an elementary teacher's

intention to use computers in classroom instruction?

3. What is the relationship between perceived behavioral control and an elementary

teacher's intention to use computers in classroom instruction?

4. Do the constructs attitude toward the behavior, subjective norm, and perceived

behavioral control have equal influence on an elementary teacher's intent to use

computers in classroom instruction?

5. Is there a correlation between an elementary teacher's intent to use a computer

for instruction and the teacher's actual use of a computer for instruction?








Statement of Hypotheses

This study tested the following null hypotheses:

1. There is no correlation between attitude toward the behavior and behavioral

intention to use computers by elementary teachers.

2. There is no correlation between subjective norm and behavioral intention to use

computers by elementary teachers.

3. There is no correlation between perceived behavioral control and behavioral

intention to use computers by elementary teachers.

4. There is no difference in the influence of the constructs attitude toward the

behavior, subjective norm, and perceived behavioral control on an elementary teacher's

intention to use computers for classroom instruction.

5. There is no correlation between behavioral intention and actual computer use by

elementary teachers.













CHAPTER 2
REVIEW OF LITERATURE

Beginning in the early 1980s, relatively powerful personal computers became

available for home use. The succeeding years have shown that this technology has had a

dramatic and pervasive impact on our society, and on how we live, work, play,

communicate, and function in general (International Society for Technology in

Education, 2000). The development of and maturation of the Internet is a prime example

of the ubiquity of computer technology in our society. K-12 public education is, in many

ways, a mirror of society, and the same technological advances that have impacted

society have had a pervasive and powerful impact on education as well. The introduction

of computers into public education has held high hopes as a tool to make the instructional

process more efficient and effective (Ohles, 1985; Cuban, 1986; Cuban, Kirkpatrick &

Peck, 2001). There has been a considerable effort to provide computers to schools in high

numbers, to network schools, and to connect schools to the "information superhighway."

The assumption has always been "if you build it they will come" and that if we provided

the appropriate technology tools in the appropriate amounts to provide sufficient access,

they would be used. In fact, this has not been the case. Despite significant investments in

technology infrastructure and large numbers of computers in classrooms, teachers are

using this technology for instruction in rather limited numbers and ways (Hadley &

Sheingold, 1993; Becker, 1994).








Computer Use by Teachers

Three related, interactive domains (the innovation, the context, and the teacher)

(Zhao, Pugh, Sheldon, and Byers, 2002) provide a useful framework for the consideration

of technology implementation in a K-12 educational setting.

Teacher Use of Computers as an Innovation

In his research on the diffusion of innovations through social systems, Rogers

(1995) identified characteristics of innovations that facilitate faster and higher levels of

adoption: relative advantage, compatibility, complexity, trialability, and observability.

* Relative advantage refers to the degree to which an innovation is perceived as
better than the idea it supersedes.
* Compatibility is the degree to which an innovation is perceived as being consistent
with the existing values, past experiences, and needs of potential adopters.
* Complexity refers to the degree to which an innovation is perceived as difficult to
understand and use.
* Trialability is the degree to which an innovation may be experimented with on a
limited basis.
* Observability is the degree to which the results of an innovation are visible to
others.

A large body of research over 3 decades has validated these factors as characteristics of

innovations more likely to be adopted by potential adopters (Rogers, 1995). These

characteristics are also valid and important when considering teacher use of computers in

the classroom. Teachers must see a benefit and advantage in using a computer for

instruction over more traditional methods; the computer and related software systems

must be compatible with existing classroom practice and educational goals; the computer

system must not be too complex for teachers to use; teachers need to be able to "try out"

the computer during the process of integration into classroom practice; and, teachers need

to have opportunities to observe the effective use of computers in the classroom by peers

and in settings similar to their own. Certainly some teachers have not adopted the use of








computers in their classroom because the computer or related software systems failed to

meet these criteria (Cuban, 1986; Cuban, Kirkpatrick, & Peck, 2001). There are,

however, instances when computers and software systems in the classroom do meet these

criteria, and they still are not adopted by teachers for use in the instructional process. This

scenario implies that it is necessary to consider more than just the computer as innovation

when attempting to understand what is involved in teachers' effective computer use in the

classroom.

The Context of Teacher Computer Use

Another domain from which to consider effective use of instructional technology is

the context of that technology use (Zhao, Pugh, Sheldon, and Byers, 2002). There is

evidence suggesting that the traditional time, space, and curriculum structures of K-12

public education are limiting factors in terms of teachers adopting computers in

classrooms (Bossert, 1996; Cuban, Kirkpatrick, & Peck, 2001). Ely (1999) identified a

number of conditions that facilitate change. Many of these conditions are external, or

contextual to the user:

1. There must be "dissatisfaction with the status quo."

2. The individuals who will ultimately implement any innovation must possess
sufficient knowledge and skills to do the job.

3. The things that are needed to make the innovation work should be easily accessible.

4. Implementers must have time to learn, adapt, integrate, and reflect on what they are
doing.

5. Rewards or incentives must exist for the participants.

6. Participation in the change process must be expected and encouraged.

7. An unqualified go-ahead and vocal support for the innovation by key players and
other stakeholders is necessary.








8. Leadership must be evident. (Ely, 1999).

Other than dissatisfaction with the status quo and sufficient knowledge and skills,

these facilitating conditions are factors external to the individual in relation to the change

process. Many contextual factors related to instructional technology in the classroom

have been addressed (such as training for sufficient knowledge and skills, accessibility,

time to learn, and support). Despite the availability of computers in classrooms, and high

levels of teacher training, adoption and use levels of computer technology are still quite

low. Becker (1994) suggested that only 3-5% of classroom teachers could be classified as

exemplary computer users. More recent studies reveal similar evidence that the number

of classroom teachers using computers for instruction in exemplary or transformational

ways is minimal (Florida Department of Education, 2003). Thus contextual factors alone

cannot provide the explanation for this phenomenon; and other domains must still be

considered.

The Teacher and Computer Use

Teachers are critical components of the change process in education (Fullan &

Stiegelbauer, 1991), and their perspective is often overlooked when a decision is made to

deploy a technology in a classroom setting (Cuban, 1986; Tobin & Dawson, 1992). An

evaluation of the innovation characteristics and external considerations described above

reveal that some of them have internal influence that involves individuals (and in the case

of computers in the classroom, teachers). Relative advantage has motivational

implications by addressing what advantages an innovation may have for an end user

(teacher) and her goals, thus making a learning curve investment worthwhile. This is also

reflective of expectancy theory, which states that an individual will act in a certain way

based on the expectation that the act will be followed by a given outcome and on the








attractiveness of that outcome to the individual (Vroom, 1964). Compatibility has internal

implications because it relates to the values, beliefs, previous experiences, and needs of a

user. Dissatisfaction with the status quo is also a powerful internal factor that can

influence an individual to change and accept a new technology or innovation. The

internal dissonance created by internal dissatisfaction can be a powerful motivating force

for change (Schunk, 2000). On the other hand, the lack of this dissonance inhibits

motivation to change. A number of psychological and social factors may influence

teacher adoption and use of instructional technology in the classroom; and many have

been studied in this regard, including self-efficacy (Jorde-Bloom, 1988; Pajares, 1996),

computer anxiety, locus of control, level of innovativeness, self competence, perceived

relevance, attitude toward computer technology, and levels of experience related to

computer technology (Delcourt & Kinzie, 1993; Rogers, 1995; Marcinkiewicz, 1996).

Although change and the diffusion of an innovation is a social process (Rogers,

1995), it ultimately involves an individual going through an evaluation process and

making a decision to change. A great deal of literature addresses organizational, or

institutional, change. Fundamentally, organizations are made up of individuals. In the

change process, what people do and do not do is the crucial variable (Fullan &

Stiegelbauer, 1991). Hall and Hord (1987) address elements related to the individual in

their Concerns Based Adoption Model (CBAM). In this model, intended adopters may

go through seven stages:

* Awareness: Individual may know the innovation exists, but have little concern or
involvement with it.

* Informational: Individual decides they would like to know more about the
innovation.








* Personal: Prospective adopter's uncertainty about the demands of the innovation,
their ability to meet them, and their role in the innovation.

* Management: The administrative and logistical challenges of innovation use.

* Consequence: Individual begins to ask how the innovation use is affecting students.

* Collaboration: How the individual adopter can coordinate and cooperate with
others in the use of the innovation.

* Refocusing: When the adopter begins to have ideas about replacing or improving
on the innovation.

A valuable perspective offered by the CBAM is that individuals are the critical

consideration in the change process. In implementing change or an innovation, it is

important to understand the point of view of the participants in the change process. The

CBAM also makes the assumption that "to change something, someone has to change

first" (Hall & Hord, 1987, p.10).

For analysis related to instructional technology integration, Ertmer (1999) describes

first- and second-order barriers to integration. First-order barriers to technology

integration are described as external to the teacher. For example, not having enough

computers or not having enough training in how to use computers. Second-order barriers

to technology integration are intrinsic to teachers and include "beliefs about teaching,

beliefs about computers, established classroom practices, and unwillingness to change"

(Ertmer, 1999, p.48). For many years it was assumed that providing technology

resources in adequate quantities would automatically lead to effective integration in

classrooms. Time and experience has shown this to not be the case. Technology is

prevalent in schools and large amounts of training have been provided, and yet effective

use, which is transformational and integrated, is minimal. The concept of second-order

barriers to technology integration begins to recognize the social and psychological








complexity related to change and teacher practice. Personal change is complex because

people are complex by nature. In this light, we need to focus on internal (motivational)

considerations to see higher levels of implementation and more effective use of

instructional technology among teachers. Internal variables related to levels of computer

use by teachers have often been overlooked (Marcinkiewicz, 1994).

Stages of Technology Use

A number of instruments have been developed that measure stages of technology

use among teachers. Examples of these models include research from Concerns Based

Adoption Model (Hall & Hord, 1987); Apple Classrooms of Tomorrow (ACOT) (Dwyer,

1995); Texas Center for Educational Technology (Christensen, 1997) based on Russell

(1995); Technology Maturity Model (Kimball & Sibley, 1997); and Technology

Adoption Model (Hooper & Rieber, 1995). The models are similar in that they reflect a

progression from non-use or awareness of technology in the classroom to a high level of

dependence on technology in the classroom as a tool to improve teaching and learning.

They typically provide a description of each stage to facilitate application and

classification for specific individuals and settings. Stages from these various models are

listed below for comparison.

Concerns Based Adoption Model-Levels of Use (Hall & Hord, 1987)

* Non-use
* Orientation
* Preparation
* Mechanical
* Routine
* Refinement
* Integration
* Renewal








Apple Classroom of Technology (ACOT) (Dwyer, 1995)

* Entry
* Adoption
* Adaptation
* Appropriation
* Invention

Texas Center for Educational Technology (Self-evaluation instrument) (Christensen,
1997)

S Stage 1: Awareness
Stage 2: Learning the process
Stage 3: Understanding and application of the process
Stage 4: Familiarity and confidence
Stage 5: Adaptation to other contexts
Stage 6: Creative application to new contexts

Technology Maturity Model (Kimball & Sibley, 1997)

Emergent Systems Stage
Islands of Technology Stage Level 1
Integrated Systems Stage Level 2
Intelligent Systems Stage Level 3

Technology Adoption Model (Hooper & Rieber, 1995)

Familiarization
Utilization
Integration
Reorientation
Evolution

A 2002 study by the Council for Educational Policy, Research, and Improvement

delineates 3 levels of computer use by teachers (Council for Educational Policy,

Research, and Improvement, 2002). The levels are based on categories from the Milken

Family Foundation's Teaching in American Schools: Seven Dimensions for Gauging

Progress. Dimension 3 of this series is termed "Professional Competency Continuum

(PCC), Professional Skills for the Digital Age Classroom". These levels were captured in








the Florida Department of Education 2001-2002 Technology Resources Survey (Florida

Department of Education, 2003). The levels, description and statewide percentages

reported are

* Stage I Entry. Operate computers at a basic level, with instruction mostly teacher-
centered and tasks are structured as exercises. (Statewide-36.5% teachers were in
this category in 2001-2002).

* Stage II Adaptation. Technology is integrated into the classroom in support of
existing practices. Educators use a variety of applications. (Statewide-48.2%
teachers were in this category in 2001-2002).

* Stage III Transformation. Adept at transferring skills from current technology
tools to new ones and often learn independently. (Statewide-15.3% teachers were
in this category in 2001-2002; Council for Educational Policy, Research, and
Improvement, 2002).

A more recent measure of teacher computer usage was measured by the State of

Florida Department of Education's 2003 Florida School Technology and Readiness

(StaR) Survey (Florida Department of Education, 2004). This measure consisted of a

rubric to evaluate levels of computer use. The four levels of the rubric and their

descriptions are

Level I Entry

Teachers use e-mail and word processing programs.
Technology not used to review student assessment information.

Level 2 Intermediate

Streamlined administrative tasks (grades, attendance, lesson planning, etc.).
Technology used infrequently to review student assessment information.

Level 3 Advanced

Technology used for research; creating templates for students; multimedia and
graphical presentations and simulations; and correspondence with experts,
peers, and parents.
Technology frequently used to review student assessment information.








Level 4 Target

Teachers explore and evaluate new technologies and their educational impact;
technology used for inquiry, analysis, collaboration, creativity, content
production, and communication.
Technology regularly used to review student assessment information which
results in needed changes in instruction. (Florida Department of Education,
2004).

The 2003 state average for elementary teachers on this rubric of computer use was

2.2, orjust beyond the Level 2 use reflected in the above description.

The models listed above are helpful in classifying stages of technology use on an

individual or organizational level. These models have not, however, been validated as

measurement instruments that might be useful in a more rigorous analysis of teacher

levels of computer use in the classroom. One instrument that has been validated as useful

to measure teacher use of computers is Levels of Use (LU) (Marcinkiewicz, 1991). The

LU assessment tool consists of four pairs of two cross-matched items each. Results from

this instrument can be used to categorize teacher use of technology as nonuse, utilization,

and integration. The LU will be described in greater detail in Chapter 3.

Teacher Beliefs

Teachers are critical components in the educational change process (Fullen &

Stiegelbauer, 1991), and this is especially true in efforts to implement the use of

technology in the instructional process (Ertmer et al., 1999). Real and lasting change in

classrooms must be driven by changes in teachers' beliefs about the purpose and nature

of instruction, and teacher belief systems are very resistant to change (Ringstaff,

Sandholtz, and Dwyer, 1991). There are strong relationships between teachers' beliefs

and their planning, instructional decisions, and classroom practices (Pajares, 1992).

Rokeach (1992), referenced in Albion & Ertmer (2002), describes belief systems as








comprising five types of beliefs ordered along a dimension ranging from central to

peripheral. At the central end are primitive beliefs (consensual), which are core beliefs

"that are formed through personal experiences, reinforced through social consensus, and

thus, most resistant to change" (Albion & Ertmer, 2002, p.35). Moving toward the

peripheral end of the scale from these core beliefs are primitive beliefs (private), beliefs

about authorities, beliefs derived from authorities, and inconsequential beliefs. Belief

types toward the peripheral end of the scale are much more malleable and less resistant to

change. Beliefs about the nature of teaching and classroom practice are formed over

many years, and reside at the central end of the scale. These are core values and beliefs

for teachers, and they are quite resistant to change. One interesting observation by Pajares

(1992) is that "change in beliefs follows, rather than precedes, change in behavior"

(p.321). The implications for this in terms of teacher use of computers is that teacher use

of a computer for instruction may lead to changes in beliefs about teaching. The theory of

planned behavior (Ajzen, 1988, 1991) is a useful theoretical framework for understanding

behavior because it considers the underlying belief structures that motivate a person to

perform a particular behavior. According to Ajzen (1991),

"... the theory postulates that behavior is a function of salient information, or
beliefs, relevant to the behavior. People can hold a great many beliefs about any
given behavior, but they can attend to only a relatively small number at any given
moment (see Miller, 1956). It is these salient beliefs that are considered to be the
prevailing determinants of a person's intentions and actions." (p.189).

The TPB constructs of attitude toward the behavior, subjective norm, and perceived

behavioral control are determined by behavioral beliefs, normative beliefs, and control

beliefs (Ajzen, 1988, 1991). The underlying belief structures of the Theory of Planned

Behavior and their relation to TPB constructs are described next.








Theory of Planned Behavior

The Theory of Planned Behavior (TPB) (Ajzen, 1988, 1991) is a theoretical

framework designed to understand and predict human behavior. It is based on the premise

that humans are rational beings who make systematic use of the information available to

them. According to the theory, humans behave in predictable ways, and consider the

implications before choosing to engage or not engage in a particular behavior or activity.

A first step in understanding behavior is to identify and measure the behavior of interest.

According to TPB, once the behavior has been identified, it is possible to investigate

what determines the behavior. The theory suggests that a person's intention to perform

(or not perform) a behavior is the immediate determinant of that behavior. Merely

acknowledging, however, that intent predicts behavior does not address the reasons for

the behavior. Aprimary goal of TPB is to understand the "why" of human behavior. To

this end it is then important to identify the determinants of intentions. TPB is concerned

with the causal antecedents of volitional behavior (Ajzen, 1988). The theory postulates

that a person's intention is a function of three basic determinants: attitude toward the

behavior, suhjectivernorm, and perceived behavioral control. The nature of these three

determinants in an individual is a product of the individual's belief system. People

develop a multitude of beliefs toward various objects, activities, and events during the

course of their lives, but they can only attend to a relatively small number at a time

(Miller, 1956). Ajzen (1988) refers to these as salient beliefstTo understand why a

person holds a certain attitude toward an object, it is necessary to assess their salient

beliefs about that object. Beliefs related to attitude toward a behavior are behavioral

beliefs, beliefs related to subjective norm are normative beliefs, and beliefs related to

perceived behavioral control are control beliefs.







Attitude Toward the Behavior

Attitude toward the behavior is a person's general feeling of favorableness or

unfavorableness toward that concept: a person's judgment that performing the behavior is

good or bad, or that they are in favor of or against performing the behavior. The more

favorable a person's attitude is toward a behavior, the more he should intend to perform

that behavior. The more unfavorable his attitude is, the more he should intend to not

perform the behavior. A frequent technique used to assess a person's attitude toward a

behavior is the semantic differential (Osgood, Soci, & Tannenbaum, 1957). It is critical

that these scales be evaluative in nature (Ajzen & Fishbein, 1980).

The measures of attitude result in a single score, which represents a person's

general evaluation or overall feeling of favorableness or unfavorableness toward the

behavior in question. An estimate of the attitude toward the behavior can be obtained by

multiplying belief strength and outcome evaluation, and summing the resulting products

(Ajzen, 1988). This process is shown in the following equation:

AB oc Cbe

A8 = Attitude toward behavior B
b, = Belief that performing behavior B will lead to outcome i
e, = Evaluation of outcome i
S= Sum is over n salient beliefs.

Subjective Norm

Subjective norm is a person's perception that other people he considers important

desire the performance or nonperformance of a specific behavior. This perception may or

may not reflect what the important others actually think they should do. The more a

person perceives that others who are important to him think he should perform a








behavior, the more he will intend to do so. Subjective norm as a determinant is social in

nature. In TPB, people are viewed as intending to perform those behaviors they believe

that important others think they should perform. If they believe important others think

they should not perform a behavior, they will usually intend not to do so. The measure of

subject norm is shown in the following equation:

SN c bmj

SN = Subjective norm
b = Normative belief concerning referentj
m, = Person's motivation to comply with referentj
n = Number of salient normative beliefs.

TPB is an extension of a theory base called the Theory of Reasoned Action (TRA)

(Ajzen & Fishbein, 1980). TRA was developed to deal with behaviors that are under

volitional control (Ajzen, 1988) and included the constructs of attitude toward the

behavior and subjective norm. A consideration of attitudes and subjective norms leads to

the beginning of understanding why people behave the way they do (Ajzen, 1988). It

must be acknowledged, however, that not all behavior is under complete volitional

control. Many behaviors require the availability of necessary opportunities and resources.

People should succeed in performing a behavior to the extent they intend to perform the

behavior and have the required opportunities and resources (Ajzen, 1988). With the

understanding that factors beyond a person's volitional control can disrupt the intention-

behavior relationship, Ajzen (1988) added another antecedent of intention called

perceived behavioral control to his theory. The extended theoretical framework

containing perceived behavioral control is the theory of planned behavior (TBP).







Perceived Behavioral Control

Perceived Behavioral Control as a construct is used to account for factors beyond a

person's volitional control that can disrupt the intention-behavior relationship (Ajzen,

1988). According to Ajzen, this element "refers to the perceived ease or difficulty of

performing the behavior and it is assumed to reflect past experience as well as anticipated

impediments and obstacles" (Ajzen, 1988, p.132). The more resources and opportunities

individuals believe they possess, and the fewer obstacles or challenges they expect, the

greater will be their perceived control over the behavior. TPB does not address the actual

control an individual has in regard to a particular behavior, but their perceived behavioral

control on achievement of behavior goals. Beliefs about resources and opportunities

underlie perceived behavioral control (Ajzen, 1991). Perceived behavioral control can be

symbolized in the following equation:


PBC a CiP

PBC= Perceived behavioral control
C= Control belief
P = Perceived power of the particular control factor to facilitate or inhibit
performance of the behavior
n = Salient control beliefs.

In TPB, behavior is predicted and determined by intention, which mediates the

effects of attitude toward the behavior, subjective norm, and perceived behavioral

control. The stronger the intention to perform a behavior or achieve a goal, the more

likely that behavior is expected. Elements and relationships of the theory of planned

behavior are shown in Figure 2-1.










Behavioral \ Attitude toward
Beliefs \ the behavior





Normative Subjective Intention Behavic
Beliefs,




Control Perceived
Beliefs behavioral
control


Figure 2-1. Theory of Planned Behavior Model.

Ajzen (Ajzen & Fishbein, 1980; Ajzen, 1988) provides methodologies to capture

and quantify a person's attitude, subjective norm, and level of perceived behavioral

control towards a behavior. Using these methodologies, TRA and TPB have been

validated in their ability to predict behavior through analysis of intention (Ajzen &

Fishbein, 1980; Ajzen, 1988). A number of studies, most using multiple linear regression

procedures, have supported the link between attitude and subjective norm as predictors of

intent and behavior. According to Ajzen (1988), consideration of attitudes and subjective

norm permit "highly accurate prediction", with correlations in representative studies

ranging from 0.73 to 0.89. In nil-.ij..rl. L ,.f 87 studies, Sheppard, Hartwick, and

Warshaw (1988) found the average correlation between attitude/subjective norm and

intention was 0.66. Armitage & Conner (2001)-concluded from their meta-analysis of 185

independent studies that the theory of planned behavior is efficacious as a predictor of

intentions and behavior. They calculated an average multiple correlation of attitude,








subjective norm and perceived behavioral control with intention at R = .63 (Armitage &

Connor, 2001).

The theory of reasoned action and the theory of planned behavior have been used

extensively to analyze and predict a broad range of behaviors (Sheppard, Hartwick, and

Warshaw, 1988; Armitage & Connor, 2001). A May 2003 search of an extensive online

database, UMI ProQest Digital dissertations, returned 433 citations on the theory of

reasoned action and 277 citations on the theory of planned behavior. A wide variety of

behaviors were examined in the dissertation literature, including abstinence from vaginal

intercourse, participation in cardiac rehabilitation, shopping on the Intemet, intention to

use condoms, physical activity among older women, and fast food consumption. Studies

relating to K-12 education included teachers' beliefs regarding the implementation of

constructivism in their classroom, teacher perceptions of play, the use of the Internet for

information retrieval, and the use of computers in mathematics instruction. A search on

TRA and TPB in relation to technology use revealed a number of studies. Representative

studies related to technology include Internet user attitudes, faculty use of web-based

instruction, user acceptance of information technology, and the use of educational

technology in hospitals. There were very few studies addressing teacher use of computer

technology in the classroom. A representative study in this area related to the use of TRA

for the development of an instrument to measure teachers' attitudes toward their use of

computers in mathematics instruction (Long, 1994). The premises of the theory of

planned behavior make it an attractive theoretical framework with which to analyze the

use of computers by elementary teachers in public education.













CHAPTER 3
RESEARCH DESIGN AND METHODOLOGY

This research was conducted to examine the relationship of attitude toward a

behavior, subjective norm, perceived behavioral control, and behavioral intent to

elementary school teacher use of computer technology for instruction. This chapter

describes the research design and methodology used in this study. The topics covered are

research design, population and sample, instrument development, procedures for piloting

the survey instrument, data collection, and data analysis.

Research Design

This was a non-experimental, predictive study that examined variables related to

the instructional use of computers by elementary school teachers. The theoretical base for

my study was Ajzen's Theory of Planned Behavior (Ajzen, 1980, 1988). Predictive

research designs enable the investigator to predict the value of a dependent variable based

on the values obtained from other independent variables (Bums & Grove, 1997).

Prediction attempts to explore causal relationships between different variables. Non-

experimental research allows us "(I) to increase our understanding of relationships

among variables (to fulfill the explanatory purpose), (2) to predict a criterion outcome

based on predictor information (to fulfill the predictive purpose), and (3) to test the

contribution made by variables in a theoretical model that is being proposed to explain a

phenomenon or behavior (to perform model crn,.i I (Keppel & Zedeck, 1989, p.389) A

paper-based survey instrument was developed and distributed to capture data relevant to

the variables of interest for the study. The variables of interest for this study are derived








from the Theory of Planned Behavior (Ajzen & Fishbein, 1980; Ajzen, 1988). The

predictor variables are attitude toward the behavior (ATT), subjective norm (SN),

perceived behavioral control (PBC), and behavioral intention (INT). The criterion

variable for this study is actual computer use as a self-report by elementary teachers and

measured using the Levels of Use (LU) instrument (Marcinkiewicz, 1991). Descriptive

statistics, correlations, and multiple linear regressions were used for statistical analysis in

my study. For statistical significance, a confidence level of 95% (a = .05) was

determined in advance.

Population and Sample

The population of interest for this study was elementary teachers in a medium-sized

school district in central Florida. Permission was requested, and granted, from the

district's Superintendent of Schools (Appendix A) to perform the study in the school

district. The elementary teachers targeted for the study were classroom-based, and they

represented 7 grade levels; Pre-kindergarten through 5th grade. These teachers were from

the school district's 10 elementary schools, and collectively, they taught 6,794 Pre-K

through grade 5 elementary students, which represents 43.6% of the district's overall

student population. Of the elementary teacher population used for this study, 94.1% were

female (n = 304) and 5.9% were male (n = 18). An electronic list of all classroom-based

elementary teachers was obtained from the school district's Information Services

Department. A paper-based survey instrument was developed and distributed to all

elementary teachers in the district (n = 322). Because this study involved variables related

to computer use, a paper-based survey methodology was chosen over alternative survey

methodologies, such as computer-mediated web-based surveys, to avoid any bias that







computer users or non-users may have. The sample consisted of those teachers who

completed and returned the survey instrument by the designated deadline. Elementary

teachers were chosen because they teach a variety of subjects, and they are less likely to

be influenced in their degree of computer use because of a particular subject area or more

complex instructional needs (Marcinkiewicz, 1991).

Instrumentation Development

A questionnaire was developed to measure the constructs of attitude, subjective

norm, perceived behavioral control, and behavioral intention using the methodology

outlined by Ajzen (Ajzen & Fishbein, 1980; Ajzen, 1985, 1988, 1992, 2002) for Theory

of Planned Behavior (TPB) research. This methodology has been used in hundreds of

studies across a variety of subject areas and it has demonstrated utility as a predictive

model. Two large-scale meta-analyses conducted by Sheppard, Hartwick, and Warshaw

(1988) provide strong support for the predictive utility of the Ajzen and Fishbein model.

They calculated a frequency-weighted average correlation for the intent behavior

relationship at 0.53, based on 87 separate studies with a total sample of 11,566 and a

significance level of 0.01. Sheppard, Hartwick, and Warshaw calculated a frequency-

weighted average correlation for the (attitude + subjective norm) intent relationship at

.66, based on 87 separate studies with a total sample of 12,624 and significance at the

.001 level (Sheppard, Hartwick, & Warshaw, 1988).

Detailed procedures for instrument development are provided in Ajzen (Ajzen &

Fishbein, 1980; Ajzen, 1988). The general steps are listed below and they are elaborated

on in the following sections. In general, the steps involved for instrument development

include:








1. Define the behavior of interest in terms of its action, target, context and time
elements.

2. Define the corresponding behavioral intention.

3. Define the corresponding TPB variables of attitude, subjective norm, and
perceived behavioral control.

4. Elicit salient outcomes and referents.

5. Define beliefs and belief strength evaluations including behavioral beliefs,
outcome evaluations, normative beliefs, motivation to comply, control beliefs,
and perceived power of control factors.

6. Construct a questionnaire based on the set of beliefs identified.

Behavior of Interest

The first step of the instrument development process involved developing a

definition of the behavior of interest that addressed the elements of action, target, context

and time. The behavior of interest was defined as a teacher using a computer to present a

lesson in your classroom during instruction. Action = Using a computer; Target = present

a lesson; Context = in your classroom; and Time = during instruction. To facilitate an

understanding of this definition for teachers taking the survey, the following explanation

was provided in the survey instructions: "The phrase 'present a lesson in your classroom'

refers to teacher use of a computer for the presentation of a lesson or a teacher involving

students in using a computer as part of a lesson. It does not refer to management uses of a

computer like keeping attendance records or electronic grade books".

Behavioral Intention

The second step in instrument development was the development of a definition for

the corresponding behavioral intention related to the behavior of interest. The common

meaning of "intention" is the act or fact of intending, or a determination to do a specified

thing or act in a specified manner (Guralnik, 1980). In the Theory of Planned Behavior,








behavioral intention is an individual's intention to perform a given behavior (Ajzen,

1988, 1991). In the TPB definition, intention is influenced by the constructs of attitude,

subjective norm, and perceived behavioral control. Behavioral intention was defined as

an elementary teacher's intention to use a computer to present a lesson in their classroom.

Theory of Planned Behavior Variables

The third step in the instrument development process involved defining the

variables of interest in the context of the behavioral intention: using a computer to present

a lesson in your classroom. These variables were defined as

Attitude toward using a computer to present a lesson in your classroom.

Subjective Norm toward using a computer to present a lesson in your classroom.

Perceived Behavioral Control toward using a computer to present a lesson in your
classroom.

These steps seem rather simplistic, but they are a formal declaration of the behavior

of interest and of the four variables of interest: attitude toward the behavior, subjective

norm, perceived behavioral control and behavioral intent, in relation to that behavior.

According to Ajzen (Ajzen & Fishbein, 1980) the above steps permit prediction and

explanation of behavior at a general level. The variables of interest are measured in a

direct manner using the survey method of semantic differential. Any standard attitude

scaling procedure such as Likert or Thurstone could be used, but the semantic differential

is commonly employed because of its ease of construction (Ajzen, 2002). Semantic

differential scales use a series of adjectives to describe the topic or object of interest

(Alreck & Settle, 1995). Adjectives must define a single dimension, and each pair must

be bipolar opposites labeling the extremes. Whereas the steps described above permit

prediction and explanation of behavior at a general level, the following steps in the TPB








instrument development process provide information about the cognitive foundations

underlying the behavior of interest (Ajzen & Fishbein, 1980).

Elicitation Study

During the fourth step of the instrument development process, an elicitation study

was performed to elicit salient outcomes, referents, and control factors related to the

behavior of interest. The elicitation study in this process is similar to a pilot study in the

development of an evaluation instrument. Ajzen and Fishbein (1980) provide four

guidelines to assist in the development of the elicitation study:

* The elicitation study population and the main study population should be similar
with respect to the participants' demographic characteristics (ex. type of
population, age, sex, race/ethnicity, and socioeconomic status).

* Open-ended statements are recommended because they allow the participants to
record multiple behavioral, normative, and control beliefs about a behavior (ex.
"List the advantages of using a computer for classroom instruction").

* A content analysis (frequency count) is used to rank-order the participants' beliefs.
They can be rank-ordered into larger concept themes (e.g., improved student
engagement) and raw data themes (e.g., use of electronic grade book, use of word
processing for student work).

* The five to ten most common behavioral, normative, and control beliefs that
emerge from the participants' responses are used to develop the beliefs instrument
for the main theory of planned behavior study.

An open-ended questionnaire was developed to capture data for this phase of the

instrument development process (Appendix C). The phrasing of the open-ended questions

was patterned after questions provided by Ajzen (Ajzen & Fishbein, 1980) for this kind

of study.

A random sample of teachers (n = 60) was sent this questionnaire along with a

cover letter (Appendix D), an approved University of Florida Institutional Review Board

Form (Appendix B), and a return envelope. The elicitation study questionnaire was







distributed through the school district's courier-based delivery system. A total of 34

completed surveys were returned that were usable for an effective response rate of 57%.

The data captured from the elicitation study was aggregated and responses for each item

were then analyzed to look for patterns that could be consolidated and synthesized as a

basis for what Ajzen (Ajzen & Fishbein, 1980) calls "modal accessible beliefs" from the

population of interest. Ajzen describes this process,

"Once the respondents have listed their beliefs, we have to make decisions
concerning the number and kind of beliefs to be included in the model set. The first
step is analogous to a content analysis of the various beliefs emitted by different
individuals. It involves organizing the responses by grouping together beliefs that
refer to similar outcomes and counting the frequency with which each outcome in a
group was elicited." (Ajzen & Fishbein, 1980, p.68).

He goes on to state "To make the decision you have to use your common sense, since no

clear rules can be provided" (Ajzen & Fishbein, 1980, p.69). Frequency lists developed

from the elicitation study data are in Appendix E.

Beliefs and Strength Evaluations

In the fifth step of the instrument development process, the synthesized data from

the elicitation study was used to define beliefs and the related strength evaluations of

those beliefs. Beliefs related to the attitude measure are behavioral beliefs, and the related

strength evaluations are outcome evaluations. Beliefs related to the subjective norm

measure are normative beliefs, and the related strength evaluations are motivation to

comply. Beliefs related to the perceived behavioral control measure are control beliefs,

and the related strength evaluations are perceived power of control factors. A pilot

questionnaire (Appendix J) was then developed based on the set of beliefs identified. The

statements were phrased such that the beliefs corresponded to the constructs related to the

behavior of interest. The pilot survey was administered to a random sample of elementary







teachers (n = 100) from the population of interest. In the survey instrument, the

respondents were asked to evaluate each of the outcomes and to indicate the subjective

probabilities that their performing the behavior would lead to each of the outcomes. A

bipolar evaluative scale was used for this assessment. It is essential to use a bipolar scale

in order to assess modal beliefs about the behavior of interest and the strength of these

beliefs (Ajzen & Fishbein, 1980). An example of an item to measure a belief would be

"My principal thinks that I should use a computer to present a lesson in my classroom

during instruction (likely unlikely)." An item to measure the strength of that belief

would be "Generally speaking, how much do you want to do what your principal thinks

you should do (not at all very much)." By measuring belief strength and evaluations

with respect to the modal salient beliefs, a given individual's attitude toward a behavior,

subjective norm, and perceived behavioral control can be predicted. Information about

the determinants of these constructs can also be obtained. Use of the modal set of beliefs

enables comparisons across individuals.

Questionnaire Development

The sixth step of instrument development involved the construction of a

questionnaire based on the beliefs identified for the behavior of interest. A sample Theory

of Planned Behavior Questionnaire provided by Aizen' (2004) was used as the model for

survey development in this study (Appendix I). A TPB instrument, as specified by Ajzen

(2002), consists of indirect measures and direct measures of the variables of interest. The

indirect (belief-based) measures of Attitude Toward the Behavior (ATT) are captured in

items 1-10 and 33-41 on the final survey. The indirect (belief-based) measures of

i In recent years, Dr. Ajzen changed the spelling of his surname to Aizen. I reference both names
depending on publication attribution.








Subjective Norm (SN) are captured in items 28-32 and 59-63. The indirect (belief-based)

measures of Perceived Behavioral Control (PBC) are captured in items 43-58. The direct

measures of Attitude Toward the Behavior (ATT), Subjective Norm (SN), Perceived

Behavioral Control (PBC), and Behavioral Intention (INT) are captured in items 12

through 27 in the final survey instrument (Table 3-1). Based on the guidelines provided

by Ajzen (2002), the items for the different variables were intermingled in this section.

Table 3-1. TPB Constructs and Related Final Survey Items
TPB Construct Survey Items
Outcome Evaluation (OE) 1-10
Behavioral Beliefs (BB) 33-41
Attitude- Direct 14,17,21,25,27
Normative Beliefs (NB) 59-63
Motivation to Comply (MC) 28-32
Subjective Norm -Direct 13,20,22,26
Control Beliefs (CB) 43-50
Power of Control Beliefs (PC) 51-58
Level of Control Direct 12,16,19,24
Levels of Use (LU) 64-67
Intention (INT) 15,18,23

Levels of Computer Use

This study evaluated the theoretical constructs of attitude toward a behavior,

subjective norm, perceived behavioral control, and behavioral intention in relation to

levels of computer use by elementary teachers. The instrument development process

described above is designed to measure and quantify attitude, subjective norm, perceived

behavioral control, and behavioral intent. It was still necessary to have an instrument to

measure teacher levels of computer use in some objective, quantifiable way for

comparison of predicted intent to actual computer use. The Levels of Use (LU)

assessment tool (Marcinkiewicz, 1991) was used to measure teacher levels of computer

use based on a self-evaluation by the teacher. Permission was granted by the author to use








the Levels of Use instrument in this study. Marcinkiewicz (1993) provides background

on the development of this instrument:

"The Levels of Use (LU) assessment. Computer use was the criterion of interest; it
was operationally defined as the integration of computers in the classroom. The LU
assessment (Marcinkiewicz, 1991) was used to categorize teacher's level of use.
The LU used the paired comparisons method because the levels of computer use
that had been defined were mutually exclusive and exhaustive. Nunnally (1959)
considered this method to be the "most exact psychophysical tool" (pp. 20-21)
useful for precise information concerning judgments or preferences. The levels of
utilization and integration were each represented by two items. One item from the
utilization level was alternately paired with an item from the integration level. The
response procedure used a forced-choice method. The design of the measure
allowed the anticipation of the two patterns of response. Responses following either
of the anticipated patterns indicated consistency; inconsistent responses were
analyzed. Because subjects responded to each item twice, any inconsistent
responses were readily evident. An item was included on the questionnaire as a
measure with which criterion-related validity could be calculated for the LU.
Respondents selected one statement from three that best described their use of
computers in teaching. This item echoed the critical attribute of expendability of
use as in the LU. A measure of association, Cohen's kappa, was computed between
the subjects' responses to this item and their levels of use (K=.72) to estimate the
consistency of classification of the measures (Crocker & Algina, 1986; Suen
1990)" (Marcinkiewicz, 1993, pp. 225-227).

According to Marcinkiewicz (1993), this instrument is validated and has high

reliability as measured with the Coefficient of Reproducibility (CR = .96). A CR of .90 is

considered the criterion for demonstrating that the items of an assessment form an

ordered scale of allowable response patterns (Torgerson, 1958; Bailey, 1987).

The LU assessment tool consists of four pairs with two cross-matched items each:

1. a. In my instruction, the use of the microcomputer is supplemental.
b. The microcomputer is critical to the functioning of my instruction.

2. a. The use of the microcomputer is not essential in my instruction.
b. For my teaching, the use of the microcomputer is indispensable.

3. a. The microcomputer is critical to the functioning of my instruction.
b. The use of the microcomputer is not essential in my instruction.








4. a. For my teaching, the use of the microcomputer is indispensable.
b. In my instruction, the use of the microcomputer is supplemental.

Teachers who do not use computers at all in their teaching were instructed to enter letter

"c" on their response sheets for these items. This established them at the nonuse level.

From Marcinkiewicz (1991, pp. 36-37), "The levels of utilization and integration are

represented by two items each. One item from the utilization level is paired with an item

from the integration level. The response procedure is forced choice; therefore, the subject

is directed to select the statement that he or she most strongly feels is true to him or her.

Because the subject was asked to respond to each of the items twice, any inconsistency in

responding was readily evident". Using these criteria, a respondent's score can be either 4

or 8. Scores of 6 or 7 are possible, but they would indicate an inconsistency. A score of 4

would indicate the utilization category and 8 would indicate the integration category. The

results from the LU instrument can be used to categorize teacher use of technology as

nonuse, utiiI,:ji..ni and integration. The utilization level is achieved when a teacher

begins to use computers in their teaching. The integration level is realized when the

teacher's computer use becomes critical to his or her teaching. Membership in these latter

two categories is determined by how expendable computers are to a teacher's teaching

(Marcinkiewicz, 1993).

Procedures for Piloting the Instrument

Using the results from the elicitation study, a pilot survey instrument was

developed. As a first step, the pilot survey instrument was evaluated by domrainexperts

for content validity. Based on their feedback (detailed in Content Validity section below),

the pilot *ur'.. %., hcn revised to improve focus, ,.. ari.. clarity, and readability. The

pilot survey content was then organized and published using a computer desktop








publishing application. Once published, an optical character recognition program (OCR)

software application was used to create an electronic template of the survey for later

automated capture of the survey data. The pilot survey was next administered to a

random sub-sample (n = 100) of the population of interest, and it was then evaluated for

reliability and validity using a computer statistical software application. Based on

statistical analysis, the pilot survey was revised and became the final survey instrument.

Content Validity

Content validity addresses the extent to which the instrument actually relates to the

content of the area or issue under investigation (Gable, 1986). The best way to ensure

content validity is to subject the instrument to judgmental validation by experts in the

area (Wiersma, 1991; Nardi, 2003). The pilot survey instrument was submitted to six

domain experts, prior to administration, for item analysis, review, and feedback relative

to domain coverage. The domain experts were drawn from university and school district

level instructional technology programs. Of the six domain experts invited to participate,

four responded with feedback related to content validity. See Appendix F for a list of the

domain experts who participated in the study. A cover letter (Appendix G) and pilot

survey was sent via email to the domain experts. The domain experts provided helpful

fc dbjk and iu.'Pt -in'r f.r ;fnpri. ;.i li.e pilot version of the survey instrument on a

form that was provided to them for this purpose (Appendix H). Their suggestions

included the revision of some items, deletion of some items that were somewhat

redundant, rewording of some items for consistency,. I.,riin, jii,.,n and ic;n..rn,. -,ei of a two

scale descriptors, and recommendations for page layout of the survey. Based on this

feedback revisions were made to the pilot survey instrument prior to distribution for the

pilot study. The sample (n = 100) for the pilot study was randomly selected from the







population of interest. The pilot survey (Appendix J) was sent to the sample recipients,

along with a cover letter, approved University of Florida Institutional Review Board form

(Appendix B), and a return envelope, using the school district's courier-based mail

system. The surveys were individually coded for tracking purposes. Approximately one

week after the pilot survey was distributed, an email reminder and a paper reminder was

to sent to each recipient that had not returned their survey. Sixty-three (n = 63) of the 100

pilot surveys were returned by the deadline for an effective return rate of 63%. This

higher than average return rate was attributed to the reminder sent out to pilot survey

recipients, the convenience of the district courier-based mail system, and the fact that the

survey was possibly viewed as an in-district study due to the role of the researcher in the

school district as Coordinator of Instructional Technology.

Raw data from the pilot survey was captured using a scanner and the data file was

exported into SPSS Version 12.0 statistical software for analysis of descriptive

statistics, reliability and validity.

Reliability

Internal consistency estimates of reliability and scale stability of the instrument

were assessed using the coefficient alpha (Green, Salkind, & Akey, 2000; Netemeyer,

Bearden, & Sharma, 2003) as calculated in the SPSS statistical software program. The

coefficient alpha (Cronbach, 1951) is an index of reliability associated with the variation

accounted for by the true score of the "underlying construct." Alpha coefficients range in

value from 0 to 1 and may be used to describe the reliability of factors extracted from

dichotomous and/or multi-point formatted questionnaires or scales (i.e., rating scale: 1 =

poor, 5 = excellent). The higher the score, the more reliable the generated scale is.

Nunnaly (1978) has indicated 0.7 to be an acceptable reliability coefficient but lower







thresholds are sometimes used. Variables related to the Theory of Planned Behavior

(TPB) can be evaluated directly through survey questions and indirectly through belief-

based responses related to the constructs. Table 3-2 shows the constructs of the TPB,

related pilot survey items, and their corresponding alpha coefficient.

Table 3-2. Pilot Survey Measures, Item Numbers, and Alpha Coefficients (n = 63)
TPB Construct Survey Items Pilot Survey
Outcome Evaluation (OE) 1-11 .678
Behavioral Beliefs (BB) 35-45 .823
Attitude- Direct 15,19,23,27,29 .932
Normative Beliefs (NB) 64-68 .944
Motivation to Comply (MC) 30-34 .796
Subjective Norm Direct 14,18,22,24,28 .687
Control Beliefs (CB) 46-54 .553
Power of Control Beliefs (PC) 55-63 .581
Level of Control Direct 13,17,21,26 .550
Levels of Use (LU) 69-72 .932
Intention (INT) 16,20,25 .910

As shown in Table 3-2, with the exception of the Perceived Behavioral Control

variables (Control Beliefs, Power of Control Beliefs, and Level of Control Direct),

reliability was relatively high (Nunnaly, 1978) for the constructs measured with the pilot

survey instrument. Ajzen (2002) states that the direct measures in the TPB instrument

should demonstrate reliability, but that the indirect measures may not, since they are

belief-based, and thus more subjective. An item analysis of the pilot instrument was done

using the SPSS@ software and revisions were made to improve reliability. Revisions to

the pilot survey instrument (Appendix J) included the elimination of items 11, 45, 53, and

63 to increase reliability for their corresponding construct measure. Instructions and an

example were added for item 12 to enable respondents to better understand how to

complete the grid for that item, which was in a different format than the other response

types. Item 18 was eliminated because it did not contribute to the reliability of the








Perceived Behavioral Control (PBC) attribute it was intended to measure. Items 13, 17,

21, and 26 were revised in an attempt to strengthen reliability for the PBC direct measure.

Items 55 through 62 were revised to make them more consistent and understandable.

And, finally, a "K" for kindergarten level was added to item 78 (grade level) in the

demographic section. The elimination of items also required a renumbering of the revised

pilot instrument. After changes were made, the revised version of the pilot survey

instrument (Appendix K) was administered to a small random sample (n = 12) from the

population for evaluation. Reliability for the revised pilot survey instrument is shown in

Table 3-3.

Table 3-3. Revised Pilot Survey Measures, Item Numbers, and Alpha Coefficients
(n = 12)
TPB Construct Survey Items Alpha
Outcome Evaluation (OE) 1-10 .863
Behavioral Beliefs (BB) 33-41 .724
Attitude- Direct 14,17,21,25,27 .854
Normative Beliefs (NB) 59-63 .897
Motivation to Comply (MC) 28-32 .875
Subjective Norm Direct 13,20,22,26 .652
Control Beliefs (CB) 43-50 .490
Power of Control Beliefs (PC) 51-58 .513
Level of Control Direct 12,16,19,24 .145
Levels of Use (LU) 64-67 .985
Intention (INT) 15,18,23 .918


The statistical results from the revised pilot survey were somewhat inconclusive in

terms of overall improved reliability, probably due to the small sample size (n = 12).

Increases in alpha occurred in Outcome Evaluation, Motivation to Comply, Levels of

Use, and Intention. Behavioral Beliefs, Attitude Direct, Normative Beliefs, Subjective

Norm -Direct, Control Beliefs, Power of Control Beliefs, and Level of Control Direct

all decreased in alpha. What was clear from the statistical analysis was that the Perceived








Behavioral Control measure continued to reflect a low alpha coefficient, which indicated

that the items still needed some improvement to adequately measure the construct. To

this end, items 12, 16, 19 and 24 were completely reworded for the final version of the

survey as follows:

* Item 12. Using a computer in my classroom to present a lesson during instruction is
within my control. True False

Item 16. I have control over using a computer in my classroom to present a lesson
during instruction. Agree Disagree

Item 19. Factors beyond my control determine whether or not I can use a computer
in my classroom to present a lesson during instruction. Disagree Agree

* Item 24. The use of a computer in my classroom to present a lesson during
instruction depends on factors beyond my control. True False

The changes made to these items significantly increased the reliability for the

control measure and specifics will be discussed below in relation to the final survey

instrument. The revision of these four items were the only changes made to the second

pilot survey instrument, which then became the final survey instrument (Appendix L).

Alpha coefficients for the final survey instrument are listed in Table 3-4.

Table 3-4. Final Survey Measures, Item Numbers, and Alpha Coefficients (n = 203)
TPB Construct Survey Items Alpha
Outcome Evaluation (OE) 1-10 .745
Behavioral Beliefs (BB) 33-41 .859
Attitude- Direct 14,17,21,25,27 .928
Normative Beliefs (NB) 59-63 .894
Motivation to Comply (MC) 28-32 .763
Subjective Norm Direct 13,20,22,26 .756
Control Beliefs (CB) 43-50 .689
Power of Control Beliefs (PC) 51-58 .824
Level of Control Direct 12,16,19,24 .789
Levels of Use (LU) 64-67 .935
Intention (INT) 15,18,23 .905








Table 3-5 provides a comparison of reliability measures for the pilot survey and the

final survey instrument.

Table 3-5. Alpha Coefficients for Pilot Survey and Final Survey
TPB Construct Pilot Survey (n = 63) Final Survey (n = 203)
Outcome Evaluation (OE) .678 .745
Behavioral Beliefs (BB) .823 .859
Attitude Direct .932 .928
Normative Beliefs (NB) .944 .894
Motivation to Comply (MC) .796 .763
Subjective Norm Direct .687 .756
Control Beliefs (CB) .553 .689
Power of Control Beliefs (PC) .581 .824
Level of Control Direct .550 .789
Levels of Use (LU) .932 .935
Intention (INT) .910 .905

Construct validity for the survey instrument was investigated by doing an item

analysis to compute item discrimination, and by performing a factor analysis to identify

groups of items that have variance in common. The dimensionality of the 13 items related

to attitude toward the behavior, subjective norm, and perceived behavioral control was

analyzed using confirmatory factor analysis. Three criteria were used to determine the

number of factors to rotate: the a priori hypothesis that the measures were

multidimensional, the scree test, and the interpretability of the factor solution. In factor

analysis, variance of the factors is measured in Eigenvalues (Norusis, 2003). Eigenvalues

are useful in determining how many factors should be used in the analysis (Green,

Salkind & Akey, 2000). One criterion used to determine how many factors should be

retained is to keep factors that have Eigenvalues greater than 1. A scree test is a graphic

plot of Eigenvalues for a combination of variables. It frequently yields more accurate

results than the Eigenvalue > 1 criteria by allowing the researcher to examine the plot of

the Eigenvalues and retain all factors with values in the hard descent part of the plot








before the Eigenvalues start to level off (Green, Salkind & Akey, 2000; Norusis, 2003).

The scree plot for the TPB variables of ATT, SN, and PBC (Figure 3-1) descended

sharply before leveling off at Factor Number 4, indicating that the initial hypothesis of

multidimensionality was correct and suggesting three factors. Consequently, three factors

were rotated using a Varimax rotation procedure. The rotated solution (Appendix O)

yielded three interpretable factors; attitude, perceived behavioral control, and subjective

norm. The attitude factor and the perceived behavioral control factor were clearly

identifiable in the correlational analysis. For the subjective norm construct, two of the

four items stood alone, and two of the items appeared to align with the attitude factor.







4-











1 2 4 4. 7 B 41 10 11 12 1I
Factor Number


Figure 3-1. Scree Plot of TPB Variable Factor Analysis.

The finalsuvey instrument consisted of 73 items. Six of the items captured

demographic information including gender, age, number of years teaching, number of

years using a computer, location of computer use, and grade level taught. The final







survey instrument also contained a coded field used to track which survey forms were

returned.

Data Collection

After finalization of the survey instrument, it was distributed to the entire

population of classroom-based elementary teachers (n = 322) in the school district. The

Salant and Dillman (1994) survey methodology was used in an effort to maximize

response rate. All mailing and correspondence related to the study was distributed

through the school district's courier-based mail system. All elementary teachers received

a personalized advance notice letter explaining the study and requesting their

participation. Approximately one week later, the elementary teachers were sent a cover

letter (Appendix M), the survey questionnaire (Appendix L), the approved University of

Florida Institutional Review Board form (Appendix B), and a return envelope. Eight days

after this mailing, a follow-up letter was sent to the teachers. This letter thanked those

who had responded and requested a response from those who had not yet responded.

Timing of the survey distribution was carefully considered so as not to conflict with other

important events for elementary teachers such as statewide high-stakes testing. Of the

total number of surveys distributed (n = 322), 203 completed surveys were returned by

the deadline for an effective return rate of 63%. It is interesting to note that this was the

same return rate as the initial pilot study during instrument development. Like the pilot

study, this higher than average return rate was attributed to the reminder sent out to pilot

survey recipients, the convenience of the district courier-based mail system, and the fact

that the survey was possibly viewed as an "in-district" study due to the role of the

researcher in the school district as Coordinator of Instructional Technology. Raw data








from the final survey was captured using a scanner and the data file was then imported

into SPSS Version 12.0 statistical software for analysis.

Data Analysis

The variables of interest in this study were analyzed using descriptive statistics,

correlations, and multiple linear regressions. Multiple regression and correlation

statistical techniques are preferred over ANOVA techniques in non-experimental

research because the former are not restricted by the conditions of collecting the data, the

way in which subjects are assigned, or the nature or type of data collected (Keppel &

Zedeck, 1989). The Pearson correlation coefficient (r) was used because it facilitates an

understanding of the linear relationship between two variables (Keppel & Zedeck, 1989).

Using correlational analysis (r), the direction and strength of a relationship between two

variables can be evaluated. The Pearson correlation coefficient (r) was used to analyze

the variable relationships in this study between attititude (ATT) and intention (INT),

subjective norm (SN) and intention (INT), perceived behavioral control (PBC) and

intention (INT), and intention (INT) to levels of use (LU).

Multiple linear regression is a more advanced statistical technique that facilitates

the evaluation of multiple relationships among variables. The multiple correlation (R) is

an index that indicates the strength of the relationship between the predicted scores and

the observed scores for a sample (Green, Salkind & Akey, 2000). It enables the

prediction of values of a dependent variable as a linear combination of the values of

multiple independent (predictor) variables (Norusis, 2003). Because of this capability,

multiple linear regression was used to analyze the combined influence of attitude (ATT),

subjective norm (SN) and perceived behavioral control (PBC) on the TPB behavioral

intent (INT) variable.





47


Because this is a non-experimental study, the variables are referred to as predictor

and criterion versus independent and dependent (Green, Salkind & Akey, 2000). The

predictor variables of interest are attitude toward the behavior (attitude), subjective norm,

perceived behavioral control, and behavioral intent. Demographic information was also

captured including age, gender, number of years teaching, number of years using

computers, and grade level. The criterion variable is computer use as established by the

Levels of Use (LU) assessment tool. Three levels of use were identified: nonuse,

utilization, and integration. Interaction of the predictor and criterion variables was

analyzed using a regression model. Results from the data analysis are detailed in

Chapter 4.













CHAPTER 4
RESULTS

The Theory of Planned Behavior (TPB) (Ajzen, 1988, 1991) can be used to predict

behavior based on the antecedent variables of attitude, subjective norm, and perceived

behavioral control. The purpose of my study was to test this predictive theory on the

instructional use of computers by elementary school teachers. A survey instrument was

developed and validated for use in this study. It was necessary to create a survey instrument

for this research because a search of the literature revealed the unavailability of a validated

instrument to test the Theory of Planned Behavior in relation to elementary teacher use of

computers for instruction. Descriptive statistics, correlations, and multiple linear regression

statistical analysis were used to examine the relationship between the predictor variables of

attitude, subjective norm, perceived behavioral control, and behavioral intention to the

criterion variable of computer use. The purpose of this chapter is to share the results from

this study and it will include a review of the research questions, descriptive statistics for

individual survey items, demographic characteristics of the population, statistical analysis of

the predictor and criterion variables, statistical results of the hypotheses, and a summary.

Research Questions

The research questions for this study were

1. What is the relationship between attitude toward the behavior and an elementary
teacher's intention to use computers in classroom instruction?

2. What is the relationship between subjective norm and an elementary teacher's intention
to use computers in classroom instruction?








3. What is the relationship between perceived behavioral control and an elementary
teacher's intention to use computers in classroom instruction?

4. Do the constructs attitude toward the behavior, subjective norm, and perceived
behavioral control have equal influence on an elementary teacher's intent to use
computers in classroom instruction?

5. Is there a correlation between an elementary teacher's intent to use a computer for
instruction and his or her actual use of a computer for instruction?

Survey Item Responses

A survey instrument was developed for this study using procedures recommended by

Ajzen (Ajzen & Fishbein, 1980; Ajzen, 1988) for Theory of Planned Behavior (TPB)

research. The survey development process, described in Chapter 3, addressed reliability

during the development process, content validity, and construct validity. Reliability for the

final survey instrument was determined by using the alpha coefficient. Overall reliability of

the final survey instrument was high, at a = .928, ranging from .689 to .935 for individual

TPB constructs. Nunnally (1978) has indicated a = .70 to be an acceptable reliability

coefficient. Table 4-1 reflects the final survey items, response options, scaling, means, and

standard deviations of the final survey instrument.

Table 4-1. Final Survey Items, Response Options, Scaling, Means, and Standard Deviations
Item Variable Items, Response
# Name Options and Scaling M SD
I oel Holding students' attention and keeping them interested is... 6.52 0.72
(7) good bad (1)
2 oe2 Reducing student discipline problems in the classroom is... 6.19 1.11
(7) good bad (1)
3 oe3 Enhancing the curriculum with the vast amount of information 6.12 0.74
available from the Internet is ... (7) good bad (1)
4 oe4 Demonstrating to students how computers can be used for 6.28 0.79
learning is... (7) good- bad (1)
5 oe5 Presenting a lesson to the whole class at once is... 5.86 1.08
(7) good- bad (1)
6 oe6 Helping students develop needed computer literacy skills because 6.44 0.75
computers are an integral part of our society is ...
(7) good bad (1)








Table 4-1. Continued
Item Variable Items, Response
# Name Options and Scaling M SD
7 oe7 Encouraging student participation in the lesson is ... 6.46 0.80
(7) good bad (1)
8 oe8 Accommodating different student learning styles is... 6.44 0.84
(7) good -bad (1)
9 oe9 Having technical problems with computers during my lesson is .. 2.27 1.31
(7) good -bad (1)
10 oelO Enhancing the learning experience of my students is ... 6.55 0.65
(7) good- bad (1)
11 pastbeh During the current school year, how many times have you used a 9.19 18.28
computer to present a lesson?
12 pbcl Using a computer in my classroom to present a lesson during 5.54 1.80
instruction is within my control. (7) true -false (1)
13 snl Most people who are important to me think that I 4.88 1.32
use a computer to present a lesson in my classroom during
instruction. (7) should should not (1)
14 attl For me to use a computer to present a lesson in my classroom 5.48 1.14
during instruction is ... (1) harmful beneficial (7)
15 intl I intend to use a computer to present a lesson in my classroom 4.84 1.80
during instruction. (1) unlikely likely (7)
16 pbc2 1 have control over using a computer in my classroom to present a 5.44 1.89
lesson during instruction. (7) agree disagree (1)
17 att2 For me to use a computer to present a lesson in my classroom 5.05 1.55
during instruction is ... (7) pleasant unpleasant (1)
18 int2 I will try to use a computer to present a lesson in my classroom 5.46 1.65
during instruction. (7) true -false (1)
19 pbc3 Factors beyond my control determine whether or not I can use a 3.73 2.06
computer in my classroom to present a lesson during instruction.
(7) disagree agree (1)
20 sn2 The people in my life whose opinions I value would 6.19 1.12
of my using a computer to present a lesson in my
classroom during instruction. (7) approve disapprove (1)
21 att3 For me to use a computer to present a lesson in my classroom 5.76 1.33
during instruction is ... (7) good bad (1)
22 sn3 Most people who are important to me use computers personally or 6.32 1.18
in their work. (7) true -false (1)
23 int3 I plan to use a computer to present a lesson in my classroom 5.20 1.72
during instruction. (1) disagree agree (7)
24 pbc4 The use of a computer in my classroom to present a lesson during 4.01 2.01
instruction depends on factors beyond my control.
(1) true -false (7)








Table 4-1. Continued


tem Variable Items, Response
# Name Options and Scaling
25 att4 For me to use a computer to present a lesson in my classroom
during instruction is (1) worthless valuable (7)


M SD
5.61 1.30


26 sn4 The people in my life whose opinions I value 6.32 1.09
computers personally or in their work. (7) use- do not use (1)
27 att5 For me to use a computer to present a lesson in my classroom 5.12 1.61
during instruction is ... (7) enjoyable not enjoyable (1)
28 mel Generally speaking, how much do you want to do what your 6.15 1.03
principal thinks you should do? (1) not at all- very much (7)
29 mc2 Generally speaking, how much do you want to do what parents of 5.27 1.16
your students think you should do? (I) not at all very much (7)
30 mc3 Generally speaking, how much do you want to do what other 4.69 1.31
teachers think you should do? (1) not at all very much (7)
31 mc4 Generally speaking, how much do you want to do what your 4.81 1.32
students think you should do? (1) not at all- very much (7)
32 mc5 Generally speaking, how much do you want to do what your 5.08 1.24
school technology and/or media specialist thinks you should do?
(1) not at all very much (7)
33 bbl My use of a computer to present a lesson in my classroom during 5.83 1.13
instruction will hold students' attention and keep them interested.
(7) likely unlikely (1)

34 bb2 My use of a computer to present a lesson in my classroom during 4.94 1.19
instruction will reduce student discipline problems.
(7) likely- unlikely (1)
35 bb3 My use of a computer to present a lesson in my classroom during 5.66 1.13
instruction will enhance the curriculum with the vast amount of
information available from the Internet. (7) likely unlikely (1)
36 bb4 My use of a computer to present a lesson in my classroom during 6.11 0.97
instruction will demonstrate to students how computers can be
used for learning. (7) likely unlikely (1)

37 bb5 My use of a computer to present a lesson in my classroom during 5.84 1.45
instruction will enable me to present a lesson to the whole class at
once. (7) likely unlikely (1)
38 bb6 My use of a computer to present a lesson in my classroom during 5.61 1.33
instruction will help students develop the computer literacy skills
they need in today's society. (7) .': .... .. ;.' (1)
39 bb7 My use of a computer to present a lesson in my classroom during 5.66 1.12
instruction will encourage student participation in the lesson.
(7) likely unlikely (1)


I








Table 4-1. Continued
Item Variable Items, Response
# Name Options and Scaling M SD
40 bb8 My use of a computer to present a lesson in my classroom during 5.46 1.22
instruction will accommodate different student learning styles.
(7) likely unlikely (1)

41 bb9 My use of a computer to present a lesson in my classroom during 3.37 1.43
instruction will increase the chances that my lesson will be
disrupted because of technical problems. (1) likely- unlikely (7)
42 bblO My use of a computer to present a lesson in my classroom during 5.85 0.92
instruction will enhance the learning experience of my students.
(7) likely unlikely (I)
43 cbl Having enough technology equipment in my classroom is 6.44 0.87
important. (7) agree- disagree (1)

44 cb2 Training on the use of technology in instruction is important. 6.64 0.62
(7) agree disagree (1)

45 cb3 Adequate time is needed to try new or different instructional 6.72 0.55
strategies in the classroom. (7) agree disagree (1)

46 cb4 Technical glitches and equipment problems are a hindrance 5.85 1.25
when using technology in the classroom. (7) agree disagree (1)

47 cb5 A teacher needs support to use technology in the classroom. 6.47 0.80
(7) agree disagree (1)

48 cb6 A teacher needs knowledge on how to use 6.76 0.50
technology in the classroom. (7) agree disagree (1)
49 cb7 Classroom technology should be easy to learn and use. 6.61 0.56
(7) agree disagree (1)
50 cb8 Adequate resources are needed to use technology in the 6.67 0.74
classroom. (7) agree disagree (1)
51 pcl Having enough technology equipment in my classroom would 6.13 1.12
make it to use a computer to present a lesson
during instruction. (1) more difficult easier (7)
52 pc2 Having more training on the use of technology in instruction 6.08 1.08
would make it to use a computer in my classroom
to present a lesson during instruction.
(1) more difficult easier (7)
53 pc3 Having more time would make it to use a computer 6.17 1.10
in my classroom to present a lesson during instruction.
(1) more difficult easier (7)
54 pc4 Having fewer technical glitches and equipment problems would 6.18 1.03
make it to use a computer in my classroom to
present a lesson during instruction. (1) more difficult easier (7)








Table 4-1. Continued


I


M SD
6.25 0.88


tem Variable Items, Response
# Name Options and Scaling
55 pc5 Having support would make it to use a computer in
my classroom to present a lesson during instruction.
(I) -.. *., .i, .. easier (7)
56 pc6 Lack of knowledge on how to use technology would make it
to use a computer in my classroom to present
a lesson during instruction. (1) more difficult easier (7)
57 pc7 Having classroom technology that is easier to learn and use would
make it to use a computer in my classroom to
present a lesson during instruction. (1) more difficult easier (7)
58 pc8 Having adequate resources would make it
to use a computer in my classroom to present a lesson during
instruction. (1) more difficult easier (7)
59 nbl My principal thinks that I should use a computer to present a
lesson in my classroom during instruction. (7) likely unlikely (1)
60 nb2 Parents of my students think that I should use a computer to
present a lesson in my classroom during instruction.
(7) likely- unlikely (1)
61 nb3 Other teachers think that 1 should use a computer to present a
lesson in my classroom during instruction. (7) likely unlikely (1)
62 nb4 My students think that I should use a computer to present a lesson
in my classroom during instruction. (7) likely- unlikely (1)
63 nb5 My school technology and/or media specialist thinks that I should
use a computer to present a lesson in my classroom during
instruction. (7) likely unlikely (1)
64 lul a. In my instruction, the use of the computer is supplemental.
(value= 1)
b. The computer is critical to the functioning of my instruction.
(value=2)
c. I do not use a computer in my teaching. (value=0)
65 lu2 a. The use of the computer is not essential in my instruction.
(value=l)
b. For my teaching, the use of the computer is indispensable.
(value=2)
c. I do not use a computer in my teaching. (value=0)
66 lu3 a. The computer is critical to the functioning of my instruction.
(value=2)
b. The use of the computer is not essential in my instruction.
(value=1)
c. I do not use a computer in my teaching. (value=0)


6.32 1.07


6.28 0.88


6.27 0.91


4.59 1.36

4.19 1.21


4.17 1.12

4.57 1.45

4.91 1.48


0.94 0.54




1.07 0.60




1.04 0.60







Table 4-1. Continued
Item Variable Items, Response
# Name Options and Scaling M SD
67 lu4 a. For my teaching, the use of the computer is indispensable. 0.97 0.55
(value=2)
b. In my instruction, the use of the computer is supplemental.
(value= 1)
c. I do not use a computer in my teaching. (value=0)


Demographic Characteristics of Respondents

The population of interest for this study included all classroom-based elementary

teachers (n = 322) in a medium-sized school district in central Florida. Two hundred three

(n = 203) useable surveys were returned for an effective return rate of 63%. Of the

respondents, 93.1% (n = 189) were female and 6.9% (n = 14) were male. This difference in

response rates by gender represents the reality that the vast majority of K-5 elementary

school teachers are female, and this result is representative of the population at large, which

was 94.1% female and 5.9% male. Respondents ranged in age from 22 years to 64 years old,

with the mean age being 43 years old. Years teaching ranged from 1 year to 40 years, and the

average years teaching from respondents was 14.2 years. Years using a computer ranged

from 0 to 25. The overall average of years using a computer from the sample was 12.4 years.

Of the total respondents, none (n = 0) reported using a computer at home only. Six percent of

teachers (n = 13) reported using a computer only at school, and 79% of teachers (n = 160)

reported using a computer at home and at school. These results suggest that approximately

15% of the sample (100% 85% reported above) do not use a computer at all, either at home

or at school. This assumption is somewhat consistent with the levels of non-computer use

(13.3%) that was reported on the Levels of Use measures of the survey, which will be

discussed below. Grade level taught by the teachers in the sample ranged from Pre-

Kindergarten through 5th Grade. Demographic results are shown in Table 4-2.








Table 4-2. Demographic Characteristics of Respondents
Characteristic N %
Gender
Female 189 93.1
Male 14 6.9
Age
20-29 29 14.3
30-39 47 23.2
40 49 62 30.5
50-59 55 27.1
60-69 10 4.9

Years Teaching
0- 5 43 21.2
6-10 44 21.7
11 15 39 19.2
16-20 29 14.3
21-25 22 10.8
26-30 12 5.9
31-35 10 4.0
36-40 4 2.0

Years Using Computer
0- 5 21 10.3
6-10 72 35.5
11-15 55 27.1
16-20 45 22.2
21 -25 8 3.9
26-30 2 1.0

Where Computer Used
Home Only 0 0.0
School Only 13 6.4
Home and School 190 93.6

Grade Level Taught
Pre-K 10 4.9
K 23 11.3
1 35 17.2
2 32 15.8
3 37 18.2
4 30 14.8
5 32 15.8
Other 4 2.0








Criterion Variable

The criterion variable for this study was levels of computer use. The Levels of Use

(LU) assessment tool (Marcinkiewicz, 1991) was used for this measure. The LU instrument

indicates three categories of computer use: non-use, utilization, and integration. A score of 0

indicates non-use, a score of 4 indicates the utilization category, and a score of 8 indicates the

integration category. While scores other than 0, 4, or 8 are possible, they indicate an

inconsistency that makes it difficult, if not impossible, to ascertain an accurate level of use.

In terms of computer use, the non-use level is axiomatic. The utilization level is achieved

when a teacher begins to use computers in their teaching, and the integration level is realized

when the teacher's computer use becomes critical to his or her teaching.

Levels of use results are reported in Table 4-3. Of the total respondents, 13.3% of the

teachers (n = 27) indicated the non-use category. Fifty-nine percent (59.1%) of the teachers

(n = 120) indicated the utilization level, and 8.4% of the teachers (n = 17) indicated the

integration level. Nineteen percent (19.2%) of the respondents (n = 39) indicated LU scores

other than 0, 4 or 8, which made it impossible to accurately categorize their level of computer

use. These responses were labeled "Uncertain" in Table 4-3. For statistical analysis the LU

score was treated as a continuous, versus a categorical, measure.

Table 4-3. Levels of Use Reported
Levels of Use N %
Non-User 27 13.3
Utilization Level 120 59.1
Integration Level 17 8.4
Uncertain 39 19.2







Predictor Variables

This section will address the relationship of indirect measures to direct measures in a

TPB study and it will include a description of how indirect measures were calculated. The

section will also include correlations between the indirect and direct measures for the

predictor variables of this study.)

The premise of the Theory of Planned Behavior (TPB) is based on an individual's

beliefs about a particular behavior, including behavioral beliefs, normative beliefs, and

control beliefs{Behavioral beliefs are the basis for an individual's attitude toward a behavior,

whether they have a positive or negative evaluation of it. Normative beliefs are the basis for

an individual's subjective norm toward a behavior, that is, what they perceive significant

others expect from them in regard to that behavior. And control beliefs are the basis for an

individual's perceived behavioral control toward a particular behavior; whether or not they

feel they have the resources or means to perform the behavior. It is difficult to directly

measure relevant, or salient (Ajzen, 1988) beliefs, in relation to a behavior because these

beliefs can change due to time, experience, and varying contexts. For this reason, indirect, or

belief-based, measures of the TPB constructs are measured in addition to direct measures of

the constructs. Even though either calculation, indirect or direct, can be used to predict

intention, direct measures are usually preferred because the intention construct is assessed

directly (Aizen, 2004). In the Theory of Planned Behavior (TPB), belief-based (indirect) and

direct measures of attitude, subjective norm, and perceived behavioral control are alternative

ways of measuring the same underlying constructs. These two types of measures should

correlate according to Ajzen (2002). A positive correlation between the belief-based

(indirect) measure and direct measure of a particular TPB construct provides evidence that

the direct measure is a reflection of an individual's salient beliefs toward that behavior A







correlational analysis was performed on the belief-based measures and direct measures of

attitude, subjective norm, and perceived behavioral control. The indirect measures for each

construct were calculated using the formulas provided by Ajzen (2002).

Attitude

Attitude toward the behavior is the degree to which a person has a favorable or

unfavorable evaluation or appraisal of the behavior in question (Ajzen, 1988, 1991). Using

Ajzen's (1988) methodology, two items comprise an indirect measure of attitude; a

behavioral belief (b) and the evaluation of an outcome related to that belief (e). The scores

for the behavioral belief and corresponding outcome evaluation are multiplied, and the total

of all indirect measures related to that variable are summed, resulting in an overall value for

the indirect measure of attitude. The belief-based measure of attitude is shown in Figure 4-1.

A, c b,e,

AB = Attitude toward behavior B.
b, = Belief that performing behavior B will lead to outcome i.
e, = Evaluation of outcome i.
S= Sum is over n salient beliefs.

Figure 4-1. Belief-based Measure of Attitude.

There is no a priori way to determine the optimal scaling of the items that make up

belief-based measures (Ajzen, 2002). Scaling analyses were performed for all three TPB

constructs using both unipolar (1 to 7) and bipolar (-3 to +3) scoring. Ajzen (2002)

recommends retaining the scores that produce the stronger correlation between belief-based

and direct attitude measures. Unipolar x unipolar scoring was used for the belief-based

attitude measure. Table 4-4 shows descriptive statistics for the belief-based attitude measure.








Table 4-4. Descriptive Statistics for Belief-based Attitude Measure
Variable Name M SD
AT UU TOT 332.22 63.53

Subjective Norm

Subjective norm is the perceived social pressure to perform or not perform a particular

behavior (Ajzen, 1988, 1991). Using Ajzen's (1988) methodology, two items comprise an

indirect measure of subjective norm; a normative belief related to a referent (b) and the

person's motivation to comply with that referent (m). The scores for the normative belief and

corresponding motivation to comply are multiplied, and the total of all indirect measures

related to that variable are summed, resulting in an overall value for the indirect measure of

subjective norm. The belief-based measure of subjective norm is show in Figure 4-2.

SNoc c bmn

SN = Subjective norm.
bi = Normative belief concerning referentj.
mi = Person's motivation to comply with referentj.
n = Number of salient normative beliefs.

Figure 4-2. Belief-based Measure of Subjective Norm.

Unipolar x unipolar scoring was used for the belief-based subjective norm measure. Table 4-

5 shows descriptive statistics for the belief-based subjective norm measure.

Table 4-5. Descriptive Statistics for Belief-based Subjective Norm Measure
Variable Name M SD
SN UU TOT 117.98 38.45

Perceived Behavioral Control

Perceived behavioral control is a person's perception of the ease or difficulty of

performing the behavior of interest (Ajzen, 1988, 1991). Using Ajzen's (1988) methodology,

two items comprise an indirect measure of perceived behavioral control; a control belief (C)

and the perceived power of the control factor (P). The scores for the control belief and







corresponding perceived power of the control factor are multiplied, and the total of all

indirect measures related to that variable are summed, resulting in an overall value for the

indirect measure of perceived behavioral control. The belief-based measure of perceived

behavioral control is show in Figure 4-3.


PBC c ,CP,

PBC = Perceived behavioral control
C= Control belief
P = Perceived power of the particular control factor to facilitate or inhibit
performance of the behavior.
n = Salient control beliefs

Figure 4-3. Belief-based Measure of Perceived Behavioral Control.

Unipolar x bipolar scoring was used for the belief-based control measure. Table 4-6 shows

descriptive statistics for the belief-based perceived behavioral control measure.

Table 4-6. Descriptive Statistics for Belief-based Control Measure
Variable Name M SD
PBC UB TOT 117.24 37.62

The correlational analysis showed mixed results for the belief-based and direct

measures of attitude, subjective norm, and perceived behavioral control (Table 4-7). For

attitude, the correlation for both measures was significant at r (201)= .62, p <.01. For

subjective norm, the correlation for both measures was significant at r (201) = .36, p <.01.

For perceived behavioral control, the correlation for both measures was not significant at

r (201) = .02, p = .753. t is unclear why there was a lack of correlation between the belief-

based (indirect) measures and the direct measures for the perceived behavioral control factor.

An example of a direct measure of this factor is item 12 of the final survey: "Using a

computer in my classroom to present a lesson during instruction is within my control, true -

false". An example of indirect items for PBC include item 45, "Adequate time is needed to







try new or different instructional strategies in the classroom, agree disagree" which is a

control belief statement, and the corresponding item 53, "Having more time would make it

more difficult easier to use a computer in my classroom to present a lesson during

instruction.", which is a power of control measure. It is possible that the phrasing of the

indirect items for this factor indicated a belief among respondents (based on information

captured in the elicitation study), but that they may not have perceived time as an element of

control related to this behavior. The same might be true for other PBC indirect measures that

addressed resources, training, technical problems, support, knowledge, and ease of use. The

lack of correlation between the PBC belief-based and direct items did not influence the

results of this study, because the direct measures appeared to more clearly address the control

construct, and, based on the recommendation of Aizen (2004), only the direct item measures

were used for hypothesis testing. The nature of the PBC indirect measures will be addressed

in the recommendations for further study section in Chapter 5.

Table 4-7. Correlation of Belief-based and Direct Measures
Indirect Direct
Measure Variable Variable r
Attitude AT UU TOT ATT DM TOT .62*
Subjective Norm SN UU TOT SN SM TOT .36*
Perceived Behavioral Control PBC UB TOT PBC DM TOT .02
*p<.01

Appendix P shows the overall construct reliability, correlations, and relationships between

the belief-based and direct measures on the final survey instrument.

Aizen (2004) suggests that either indirect or direct measures can be used to predict

intention, but because intentions are assessed directly, direct measures are usually preferred

for the sake of consistency. The direct measures of attitude, subjective norm, perceived

behavioral control, and behavioral intention were used for this statistical analysis. The data







were analyzed using Pearson product-moment correlation coefficients (r) and multiple linear

regressions. The relevant statistics are presented in summary tables below.

Statistical Analysis

This section will address the research questions and hypotheses of this study.

Correlational and multiple linear regression statistical procedures were used for hypothesis

testing.

Research Question 1. What is the relationship between attitude toward the behavior and

an elementary teacher's intention to use computers in classroom instruction?

Hypothesis 1. There is no correlation between attitude toward the behavior and

behavioral intention to use computers by elementary teachers.

A correlation coefficient was computed for the attitude measure and the intention

measure. The correlation between the attitude measure and the intention measure was

significant, r (201) = .80, 2 < .001 (Table 4-8). Therefore, this null hypothesis was rejected.

Research Question 2. What is the relationship between subjective norm and an

elementary teacher's intention to use computers in classroom instruction?

Hypothesis 2. There is no correlation between subjective norm and behavioral

intention to use computers by elementary teachers.

A correlation coefficient was computed for the subjective norm measure and the

intention measure. The correlation between the subjective norm measure and the intention

measure was significant, r (201)= .53, < .001 (Table 4-8). Therefore, this null hypothesis

was rejected.

Research Question 3. What is the relationship between perceived behavioral control

and an elementary teacher's intention to use computers in classroom instruction?








Hypothesis 3. There is no correlation between perceived behavioral control and

behavioral intention to use computers by elementary teachers.

A correlation coefficient was computed for the perceived behavioral control measure

and the intention measure. The correlation between the perceived behavioral control measure

and the intention measure was significant, r (201) = .32, p < .001 (Table 4-8). Therefore, this

null hypothesis was rejected.

Table 4-8. Correlations among Attitude, Subjective Norm, Perceived Behavioral Control and
Behavioral Intention
Attitude Subjective Norm Control


Attitude
Subjective Norm .61*
Control .32* .27*
Intention .80* .53* .32
*< .001

Research Question 4. Do the constructs attitude toward the behavior, subjective norm,

and perceived behavioral control have equal influence on an elementary teacher's intent to

use computers in classroom instruction?

Hypothesis 4. There is no difference in the influence of the constructs attitude toward

the behavior, subjective norm, and perceived behavioral control on an elementary teacher's

intention to use computers for classroom instruction.

A multiple regression analysis was conducted to evaluate how well the variables

attitude toward the behavior, subjective norm, and perceived behavioral control predicted

behavioral intention, which according to the TPB, are antecedent to actual behavior. The

linear combination of attitude, subjective norm, and control measures was significantly

related to behavioral intention, F (3, 199) = 124.61, p = 000. The sample multiple correlation

coefficient was .81, indicating that approximately 65% of the variance of the behavioral


Attitude Subiective Norm Control


*


r ..................








intention measure in the sample can be accounted for by the linear combination of the

attitude, subjective norm, and perceived behavioral control measures.

The general formula for a linear model for predicting the values of a dependent variable (Y)


from one or more independent variables (X) is Y = B0 +BL X, + B2X2 +...+B,X, (Norusis,

2003). The regression equation for my study is BI = -2.952 + (.575xATT_DMTOT) +

(.083xSN DM TOT) + (.051xPBCDMTOT). Table 4-9 presents indices to indicate the

relative strength of the individual predictors. All the bivariate correlations between the

predictor measures and the behavioral intention measure were positive, as expected, and all

three of the indices were statistically significant (p< <.001). Only the partial correlation

between the attitude measure and behavioral intention was significant. On the basis of these

correlational analyses, it is tempting to conclude that the only useful predictor of behavioral

intention is the attitude measure. It alone accounted 1...r (i4".. (.802= .4I ,: ih: variance of

behavioral intention, while the other variables contribute only an additional 1% (65% 64%

= 1%). Judgments about.the relative importance of these predictors, however, are difficult

because they are correlated. The correlations among the predictor variables ranged from .27

to .61 (Table 4-8 above).

Table 4-9. Bivariate and Partial Correlations of the Predictors with Behavioral Intention
Correlation Between each Predictor
Correlation Between each and BI Controlling for
Predictors Predictor and BI all Other Predictors
Attitude .80* .70*
Subjective Norm .53* .08

Perceived Control .32* .10

*p<.001

Figure 4-4 shows a graphic representation of the relationships of attitude, subjective norm,

and perceived behavioral control to behavioral intention.

























Figure 4-4. Correlation of ATT, SN, PBC to INT. *p < .001


Research Question 5. Is there a correlation between an elementary teacher's intent to

use a computer for instruction and their actual use of a computer for instruction?

Hypothesis 5. There is no correlation between behavioral intention and actual computer

use by elementary teachers.

A correlation coefficient was computed for the behavioral intention measure and the

levels of use measure. Three items on the final survey instrument, items 15, 18, and 23,

captured the direct measure of the respondent's intention to perform the behavior of using a

computer to present a lesson. Scaling on these items ranged from I (unlikely, false, disagree)

to 7 (likely, true, agree). The maximum potential value for the behavioral intention measure

was 21, and the actual mean was 15.5 (n = 203). For the Levels of Use (LU) measures (items

64-67 on the final survey instrument) the levels of utilization and integration were

represented by two items each. One item from the utilization level is paired with an item

from the integration level. Using this criteria, a respondent's score should be either 4 or 8.

Scores of 6 or 7 are possible, but they would indicate an inconsistency. A score of 4 would








indicate the utilization category and a score of 8 would indicate the integration category. For

statistical analysis the LU score was treated as a continuous measure, and the mean score was

3.88 (n = 203). Descriptive statistics for these measures are shown in Table 4-10. The

correlation between the behavioral intention measure and the levels of use measure was

significant, r (201)= .45,p < .01 (Table 4-11). Therefore, this null hypothesis was rejected.

Table 4-10. Descriptive Statistics for Intention and Levels of Use Measures
Variable Name M SD
INT DM TOT 15.50 4.74
LU DM TOT 3.88 2.06


Table 4-11. Correlation of Behavioral Intention and Levels of Use
Intention Levels of Use
Intention (INT DM_TOT) 1 .45*
Levels of Use (LU DM TOT) .45* 1
*p <.001

This relationship between behavioral intention and actual behavior (Levels of Use) is shown

graphically in Figure 4-5.



Behavioral .45* Actual
Intention Behavior


Figure 4-5. Relationship of Behavioral Intention to Actual Behavior. *p < .001

Summary

The purpose of this chapter was to examine the results of a survey designed to measure

TPB constructs in relation to elementary teacher use of computers for instruction. An

important part of this process involved the development of a survey instrument for this use.

The 73-item survey instrument was evaluated for content validity, construct validity, and

item reliability and there was evidence for strong validity and reliability. Statistical analysis

revealed significant correlations between the attitude (ATT) and intention (INT) variables,





67


the subjective norm (SN) and intention (INT) variables, and the perceived behavioral control

(PBC) and intention (INT) variables. The strongest correlation was between ATT and INT,

followed by SN-INT and PBC-INT, respectively. The attitude variable had a significantly

larger statistical effect on behavioral intention than did subjective norm or perceived

behavioral control, either individually or combined. The results also indicated a statistically

significant relationship between behavioral intention (INT) and actual levels of computer use

(LU) by elementary teachers. Statistical procedures were used to evaluate 5 null hypotheses,

and all 5 null hypotheses were rejected.













CHAPTER 5
DISCUSSION, CONCLUSIONS, AND RECOMMENDATIONS

The purpose of this chapter is to discuss the findings from this study as well as

conclusions and recommendations based on this research. The chapter will include an

overview of the study, key findings, discussion, implications, recommendations for

further research, and a summary.

Overview of the Study

A large amount of evidence suggests, that despite significant investments in

hardware and training, a relatively small percentage of teachers use computers in

significant ways for instruction. A number of ideas have been posited for this

phenomenon, and there is a need to better understand some of the reasons for the

relatively low level of adoption of computer technology by teachers. The purpose of this

study was to examine internal variables related to elementary teacher use of computers

for instruction and to evaluate the efficacy of the Theory of Planned Behavior (Ajzen &

Fishbein, 1980; Ajzen, 1988) in this context. The Theory of Planned Behavior suggests

that we can predict behavior based on our understanding of a person's intention to

perform that behavior. Behavioral intention is predicated on three variables that are

belief-based: attitude toward the behavior, subjective norm, and perceived behavioral

control. These variables can be measured indirectly through belief-based responses and

dJircil., .;. ith lijrL.icr responses. The indirect and direct measures of these constructs

should correlate. Though either measure can be used to analyze effect, Aizen (2004)

suggests using direct measures because they are more compatible with the direct measure








of behavioral intention specified in the research design. A 73-item survey instrument was

developed and validated for use in capturing data for this study. The population for this

study was all Pre-K through 5th grade teachers (n = 322) in a medium-sized school district

in central Florida. Of the total distributed, 63% (n = 203) completed surveys were

returned.

The data were analyzed using descriptive statistics, correlations, and multiple linear

regressions. The criterion variable for this study was actual computer use (LU) and the

predictor variables were attitude toward the behavior (ATT), subjective norm (SN),

perceived behavioral control (PBC), and behavioral intention (INT). Attitude toward the

behavior is the degree to which a person has a favorable or unfavorable evaluation or

appraisal of the behavior in question (Ajzen, 1988, 1991). Subjective norm is the

perceived social pressure to perform or not perform a particular behavior (Ajzen, 1988,

1991). Perceived behavioral control is a person's perception of the ease or difficulty of

performing the behavior of interest. It is assumed to reflect a person's previous

experiences, as well as anticipated challenges and obstacles (Ajzen, 1988, 1991).

Behavioral intention is an individual's intention to perform a given behavior. Intention is

influenced by the constructs of attitude, subjective norm, and perceived behavioral

control (Ajzen, 1988, 1991). Demographic data were also captured for this study

including gender, age, number of years teaching, number of years using a computer,

location of computer use, and grade level taught. The research questions for this study

involved examining the relationship between attitude and intention, subjective norm and

intention, perceived behavioral control and intention, and behavioral intention to actual

computer use.








Research Questions

1. What is the relationship between attitude toward the behavior and an elementary
teacher's intention to use computers in classroom instruction?

2. What is the relationship between subjective norm and an elementary teacher's
intention to use computers in classroom instruction?

3. What is the relationship between perceived behavioral control and an elementary
teacher's intention to use computers in classroom instruction?

4. Do the constructs attitude toward the behavior, subjective norm, and perceived
behavioral control have equal influence on an elementary teacher's intent to use
computers in classroom instruction?

5. Is there a correlation between an elementary teacher's intent to use a computer for
instruction and their actual use of a computer for instruction?

Key Findings

Research Question 1. What is the relationship between attitude toward the behavior

and an elementary teacher's intention to use computers in classroom instruction? The

findings indicated that there was a high correlation between the attitude toward the

behavior construct and behavioral intention (r = .80, < .001). This suggests that those

elementary teachers with a more favorable evaluation toward using a computer for

instruction have a higher intention to do so. This favorable assessment by elementary

teachers may include the practice of using the computer for instruction, as well as the

outcomes that result from the use of a computer in instruction. These findings indicate

that the first hypothesis, which states there is no correlation between attitude toward the

behavior and behavioral intention to use computers by elementary teachers, was rejected.

Research Question 2. What is the relationship between subjective norm and an

elementary teacher's intention to use computers in classroom instruction? The findings

indicated that there was a moderate correlation between the subjective norm construct

and behavioral intention (r = .53, E <.001). This suggests that those elementary teachers








who intend to use a computer in instruction have the perception that others who are

significant to them expect them to do so. These significant others could be their principal,

other teachers, parents of their students, their students, or their technology specialist or

media specialist. These findings indicate that the second hypothesis, which states there is

no correlation between subjective norm and behavioral intention to use computers by

elementary teachers, was rejected.

Research Question 3. What is the relationship between perceived behavioral control

and an elementary teacher's intention to use computers in classroom instruction? The

findings indicated that, though significant, there was a somewhat low correlation between

the perceived behavioral control construct and behavioral intention (r = .32, 1 < .001).

This suggests that an elementary teacher's perception of how easy or difficult it is to use

a computer for instruction has little to do with their intention to perform this behavior.

These findings indicate that the third hypothesis, which states there is no correlation

between perceived behavioral control and behavioral intention to use computers by

elementary teachers, was rejected.

Research Question 4. Do the constructs attitude toward the behavior, subjective

norm, and perceived behavioral control have equal influence on an elementary teacher's

intent to use computers in classroom instruction? The findings indicated that the linear

combination of attitude, subjective norm and perceived behavioral control account for the

majority of effect size (65%) of the behavioral intent construct. However, the

contribution of each variable appeared extremely disproportionate. Of the 65%, the

attitude construct contributed 64%, and subjective norm and perceived behavioral control

together contributed the other 1%. This suggests that those elementary teachers who







intend to use a computer for instruction are largely influenced by their evaluation of that

particular behavior, and to a much lesser e'.ienl .h j thlie, feel others expect of them in

regard to -hji bch. ;.r or by the level of control they believe they have in performing it.

The large effect size of attitude in comparison to subjective norm and perceived

behavioral control was a surprising result from this study. These findings indicate that the

fourth hypothesis, which states there is no difference in the influence of the constructs

attitude toward the behavior, subjective norm, and perceived behavioral control on an

elementary teacher's intention to use computers for classroom instruction, was rejected.

Research Question 5. Is there a correlation between an elementary teacher's intent

to use a computer for instruction and his or her actual use of a computer for instruction?

The findings indicated that there was a significant correlation between the behavioral

intention construct and actual use of a computer in instruction (r = .45, p < .001). This

suggests that those elementary teachers who intend to use a computer for instruction

actually do so, and the findings tend to validate the TPB as a predictive model. Intention

is predicated on their attitude toward that behavior, their perception of the expectation of

others in that regard, and their perceived level of control in using a computer for

instruction. These findings indicate that the fifth hypothesis, which states there is no

correlation between behavioral intention and actual computer use by elementary teachers,

was rejected.

Discussion

The findings from this study of elementary teachers and their use of computers for

instruction suggest the validity of the Theory of Planned Behavior as a predictive model.

The predictor variables of attitude, subjective norm, and perceived behavioral control all

had a valid and significant correlation to the behavioral intent construct, and behavioral







intent was significantly correlated to actual use of a computer. This section will elaborate

on some of the major findings from this study including the influence of teacher beliefs

on teacher practice, the respective contribution of the Theory of Planned Behavior

variables, and teacher levels of computer use.

Influence of Teacher Beliefs on Practice

Attitude, subjective norm and perceived behavioral control are mediated by

personal beliefs, and this study is significant in relation to elementary teacher use of

computers because so much of teacher practice is guided and motivated by personal and

professional beliefs. Teachers hold the key to effective technology integration in the

instructional process (Marcinkiewicz, 1993). Marcinkiewicz (1993) states, "Full

integration of computers into the educational system is a distant goal unless there is a

reconciliation between teachers and computers. To understand how to achieve

integration, we need to study teachers and what makes them use computers" (p.234). An

understanding of what makes teachers use a computer for instruction must take into

account the role of teacher beliefs. Beliefs play a significant role in determining teacher

practice (Albion & Ertmer, 2002). According to Cuban (1993) we have been less then

successful with technology integration in the classroom because we have failed to take

these beliefs, which dominate popular views of schooling, into account, including

"cultural beliefs about what teaching is, how learning occurs, what knowledge is proper

in schools, and the teacher-student (not student-machine) relationship" (p.186). The

tendency to disregard the culture of schooling and classrooms is a problem with most

curriculum reform efforts (Tobin & Dawson, 1992). Pajares (1992) noted the strong

relationships among teachers' beliefs and planning, their instructional decisions, and their

classroom practices. He observed, "beliefs are the best indicators of the decisions







individuals make throughout their lives" (p.307). Teacher beliefs are at the core of

teacher practice (Albion & Ertmer, 2002). The strong role of beliefs in teacher practice is

reflected in the large contribution of the attitude construct toward behavioral intention in

this study. An elementary teacher will not intend to use a computer for instruction until

she believes in the efficacy of the computer as a tool to enhance instruction.

One of the most interesting findings from this study was the extremely large effect

of an elementary teacher's attitude toward using a computer for instruction on their

intention to do so. The attitude-intention relationship in and of itself is not unusual and

seems rather intuitive. What was unusual to the researcher was the relatively small impact

that the subjective norm and perceived behavioral control components had on intention

for this behavior, in relation to the attitude component. Subjective norm addresses what a

teacher perceives as the expectations of others (i.e. social pressure), in regard to a

particular behavior. For an elementary teacher these significant others are their principal,

peer teachers, students, parents of students, and other school staff like their technology

specialist or their media specialist. The results suggest that, while elementary teachers

want to comply with their principal's expectations (M = 6.15), they do not feel there is an

extremely high expectation from principals that they use a computer for instruction (M =

4.69). In fact, the results indicate that these teachers believe their school technology

specialist has higher expectations for them in that regard than their principal (M = 4.91),

and yet the elementary teachers in this study have a lower motivation to comply with the

expectations of their technology specialist (M = 5.08) (Table 4-1, items 28-32, 59-63).

This perception by elementary teachers of mediocre expectations for computer use by

their principal is consistent with other research. Higgins & Russell (2003) state that the








elementary teachers (n = 1432) in their study of technology integration reported the

emphasis their principal placed on technology as: 27% "Heavy", 60.1% "Some", 11.3%

"Little" and 1.6% "None". In the same study, the elementary teacher population gave an

even lower self-report of the emphasis they place on technology: 18.1% "Heavy", 63.1%

"Some", 17% "Little" and 1.8% "None" (Higgins & Russell, 2003). This perception by

elementary teachers of low expectations for computer use from others may be valid for a

number of reasons. The message many elementary teachers are getting the most

reinforcement on in 2004 relates to accountability and preparing students for success on

high-stakes tests. It is very possible that principals and teachers do not feel they have the

luxury of time or opportunity to focus on technology integration in light of other, very

immediate, demands and priorities. This assessment is consistent with the observations

others make about the culture and organization of schooling being inconsistent with

technological innovation (Cuban, 1986, 1993).

Another possibility to consider regarding the low contribution of subjective norm to

behavioral intention relates to a teacher's perception of their role. In general, a teacher's

influence on classroom practice is significant and they perceive it as such (Martin &

Clemente, 1990). Teachers have a high locus of control (Rotter, 1966) in the instructional

process that may neutralize the influence of others regarding expectations for a particular

behavior. Teachers play a strategic role in what occurs in the classroom (Burkman, 1989).

Elementary teachers have strong beliefs about their significant role in student learning,

and any change in their practice, like adoption of computer use, will take these beliefs

into account. If a teacher believes a practice or innovation will enhance the teacher-

student relation and their instructional role, they are more likely to embrace a change








involving that practice or innovation. "The direct relationship between personal teaching

efficacy and change suggests that teachers are more likely to change their behavior in

directions that may improve their classroom effectiveness if they believe that they

themselves are instrumental to the learning of their students" (Smylie, 1988, p.22 as

quoted in Martin & Clemente, 1990). Regarding the practice of instructional systems

design as an innovation, Martin & Clemente (1990) note that teachers tend to prefer new

approaches that maintain their influence in a teacher-student relationship. Though not the

focus of their discussion, the same issues may influence teacher use of computers for

instruction. As the instructional leader in the classroom, an elementary teacher will not

take a secondary role to any other person, innovation, or practice. The role of the

classroom teacher must be considered and valued for the successful integration of any

instructional innovation, including computers. The innovation must be perceived to have

relative advantage and compatibility (Rogers, 1995) for the teacher for them to adopt its

use for instruction. This has significant implications for the way we approach technology

integration to enhance teaching and learning. The decision to integrate technology cannot

be a "top-down" or an "outside-in" decision. In many cases, innovations are attempted

without the careful examination of whether or not they address what is perceived as

priority needs by teachers (Fullan & Stiegelbauer, 1991). Teachers must have shared

decision-making in the decision to use computers for instruction, and they must have

flexibility to determine on their own the benefit this innovation will have for their

instructional practice. Ely (1999) calls this level of teacher involvement "participation" in

his list of necessary conditions that facilitate the implementation of educational

technology innovations. In many cases, a teacher's belief system about computer use and








any potential benefit will change only after they begin using the innovation. "Change in

beliefs follow, rather than precede, change in behavior" (Pajares, 1992, p.321). The

introduction of computers as a tool to enhance teaching and learning must be considered

in the context of teacher expertise, experience, existing practice, and what they are trying

to achieve in the classroom (Miller & Olson, 1994). Teachers must have a vision for how

an innovation like computers can improve their practice and their outcomes and they

must be allowed time for this vision to develop. Leadership is needed to facilitate the

development of this vision (Ely, 1999).

The small influence of a teacher's perception of control in using a computer for

instruction was also a surprising finding. There is evidence that teachers credit factors

beyond their control as reasons they do not use technology in the classroom (Cuban,

1993; Ely, 1999). Some of the factors mentioned include not having enough computers,

unreliability of technology, lack of time for planning and innovation, lack of training, and

lack of support. The findings from this study suggests that these barriers are of less

consequence in relation to the attitude factor, and this result seems congruent with the

observation of teachers who use technology in the classroom. Ertmer, et al. (1999)

discuss first-order and second-order barriers to technology use. First-order barriers

include factors external to the teacher such as lack of equipment and lack of time.

Second-order barriers relate to factors more internal to a teacher such as values and

beliefs about the role of a teacher and the role of teaching and learning. Ertmer points out,

however, that second-order barriers (i.e. values and beliefs), can mediate the impact and

influence of first-order barriers. Higher levels of computer use can be expected to take

place when perceived value is high and resources are low, than when perceived value is







low and resources high (Harrington, 1993; Ertmer, 1999). For example a teacher who

values the use of computers in the instructional process may scavenge to get their hands

on any hardware they are able to and make it work to serve their instructional purposes.

A teacher who does not value a computer in the instructional process will not go to that

effort and will tend to use the lack of equipment as an excuse to not use that particular

technology for instruction. The findings from this study indicate that a teacher's

perception of their control or lack of control has a small role in whether or not they intend

to, or actually use, a computer for instruction. Those teachers that value this behavior will

"find a way" to use it.

Contribution of TPB Variables

The contribution of the Theory of Planned Behavior (TPB) variables to behavioral

intention and actual use in this study is consistent with similar studies in other contexts. A

literature search was performed prior to the decision to perform this particular study to

see if there was any similar research done on computer use using the Theory of Planned

Behavior model. The study that was closest in nature involved college and university

park and recreation faculty intention to use instructional technology (Mak, 2000). The

population for the Mak study was higher education faculty members (n = 1,188). In the

Mak study, attitude, subjective norm, and perceived behavioral control accounted for

48% of the intention to use instructional technology, and the contribution of the

individual variables was somewhat consistent with my study. The strongest predictor of

intention was attitude (R = .43), followed by subjective norm (R = .29) and perceived

behavioral control (R = .12) (Mak, 2000). This comparison suggests there could be some

consistency in the factors that influence educators to use computers for instruction across

different contexts and levels. This would be an interesting topic for further research.








Similar to the Mak (2000) study, it was clear from the results of my study that the

three predictor variables of attitude toward the behavior, subjective norm, and perceived

behavioral control made significantly different contributions to the overall effect size of

the behavioral intention construct. This disparity may lead some to question the efficacy

of the Theory of Planned Behavior in this regard. Aizen (2004) addresses this

phenomenon and states,

"There is nothing in the theory to suggest that attitude, subjective norm, and
perceived behavioral control will each make a significant contribution to the
prediction of intention. The relative importance of these three factors is likely to
vary from one behavior to another and from one population to another. In some
cases, one or another of the three factors will be found to have no significant effect
on intention. Assuming that the factors were measured with equal reliability, lack
of predictive validity merely indicates that for this particular behavior and
population, the factor in question is not an important consideration in the formation
of intention." (Aizen, 2004, unnumbered).

Levels of Computer Use

The results of this study indicate that there was a positive and significant (r = .45)

relationship between teacher intent to use a computer for instruction and his or her

actual use of a computer for instruction. As suggested in the previous discussion

above, if an elementary teacher places value on a practice that they believe will

enhance the instructional process, they are likely to use that practice. Fifty-nine

percent (59.1%) of the elementary teachers in this study rated themselves at the

"utilization" level of computer use. This level indicates that a teacher has started to

use a computer in their classroom, but in non-essential or supplementary ways.

Slightly over eight percent (8.4%) of the elementary teachers in this study rated

themselves at the integration level of computer use, which indicates that the computer

is critical or indispensable to his or her teaching. There are no state of Florida

statistics using the Levels of Use instrument (Marcinkiewicz, 1993) for direct








comparison of levels of use statewide to teachers in this sample. The focus of this

study was not on overall computer use, but on elementary classroom teacher use of a

computer for instruction. With this difference in mind, the measures of levels of use

identified in this study appear consistent with other indicators used in the state of

Florida to measure teacher computer use (Florida Department of Education, 2003,

2004). The measures indicate that progress has been made in more teachers using

computers in the classroom, but that there is still a long way to go before higher

numbers of teachers are using computers in ways that transform teaching and

learning. Use of computers by elementary teachers to date has been more

evolutionary than revolutionary.

Implications

1. It would be helpful to evaluate a teacher's attitude during pre-employment

review to determine if it is compatible with technology use, change, and innovation. This

type of evaluation can be difficult to do in the hiring process. Experience with personnel

indicates that one can often influence the knowledge and skills of an individual if they are

teachable, but that a person's attitude is an internal element that is much more resistant to

external influence. In other words, it is quite difficult to influence a person's attitude

resulting in a change of behavior. A positive attitude toward change and innovation may

be more nature than nurture, thus it becomes important to evaluate this qualitative aspect

when making hiring and retention decisions.

2. For teachers, the use of computers for instruction will be internally motivated,

not externally mandated. In this light it is important to consider strategies that impact

teacher attitude toward computer use in positive ways. The research indicates that teacher

belief patterns do not precede, but follow, changes in behavior. It is necessary to help







teachers take "baby steps" in using a computer for instruction, through the use of mentors

and modeling, to give them opportunities to see how this technology can improve their

practice and their results in meaningful and significant ways. If the innovation cannot

stand up to the "teacher test" in term of relative advantage and compatibility with existing

practice, it will not be adopted.

3. The change agent needs to focus on the social, psychological, and personological

aspects of technology integration. Personal and professional values, attitudes, and beliefs

are powerful influences on teacher practice, and they cannot be ignored. There is an

abundance of evidence suggesting that characteristics of an innovation are important

factors that influence levels of adoption, as well as the environmental and contextual

elements related to a particular innovation. In the past, technology integrators and change

agents have focused on these two areas largely to the exclusion of personal and social

internal factors. One reason for this is that external elements are easier to identify and

measure. People and their related social contexts are complex entities that are difficult to

understand and quantify. The psycho-social aspect of teacher integration must be a

consideration in any strategies related to change in the area of higher levels of technology

adoption and instructional computer use by teachers.

4. Change facilitators (like principals and technology specialists) in elementary

schools need to take a leadership role to effect higher levels of teacher computer use for

instruction. Proactive leadership will include creating a climate for change,

experimentation, and innovation, and the expectation that teachers use computers for

instructional purposes. As mentioned above, significant teacher use of computers for








instruction will not be externally mandated, but a good leader can establish a climate and

expectation that will facilitate higher levels of computer use by teachers (Schiller, 2003).

Recommendations for Further Research

1. This study, related to the Theory of Planned Behavior (Ajzen & Fishbein, 1980;

Ajzen, 1988) and elementary teacher computer use, necessitated the development and

validation of a survey instrument to capture data relevant to the variables of interest.

Overall, the survey instrument demonstrated high levels of reliability and validity. One

area of question on the instrument was the lack of correlation between the indirect

(belief-based) and direct measures of perceived behavioral control. It would be beneficial

to evaluate and further test this survey instrument in this regard for additional

development and refinement.

2. The generalizability of the results from this study are limited to the sample of

respondents from one school district who participated in the study. It would be useful to

duplicate this study in other elementary teacher populations for comparison of results.

3. The focus of this study was on elementary teachers and their use of computers

for instruction. It would be informative to implement this study in other teacher

populations such as middle school and high school, to explore any differences that might

result in the context of different curriculum needs and school climates.

4. This study reflected the strong influence of teacher attitude on teacher use of a

computer for instruction. It would be informative to further explore the subjective norm

and control dimensions to determine the influence of factors such as strong principal

leadership on levels of teacher computer use for instruction.

5. This study reflected the strong influence of teacher attitude on their use of a

computer for instruction. Additional research is needed on the internal, personological,








and psychological aspects of teacher computer use for instruction. It would be useful to

explore ways that teacher attitude toward computer use might be mediated and

influenced.

Summary

This study examined internal variables identified in the Theory of Planned

Behavior (Ajzen & Fishbein, 1980; Ajzen, 1988) in relation to elementary teacher use of

computers for instruction. Significant correlations were identified between the variables

of attitude toward the behavior and behavioral intention, subjective norm and behavioral

intention, and perceived behavioral control and behavioral intention. Of the three, the

attitude variable had the largest influence on behavioral intention. There was also a

significant relationship between the predictor variable of behavioral intention and the

criterion variable of actual computer use by elementary teachers. The results from this

study seem to validate the Theory of Planned Behavior as a predictive model in the

context of elementary teacher use of computers for instruction. The results underscore the

importance of personal and social factors and their influence on teacher computer use,

and the study contributes to our overall understanding of factors that might facilitate

higher levels of instructional technology integration in the elementary classroom.













APPENDIX A
DISTRICT PERMISSION LETTER








James M. Geddes
1 N. Umber Point
Inverness, Florida 34450

November 3, 2003

Mr. David Hickey
Superintendent
Citrus County Schools
1007 W. Main Street
Inverness, Florida 34450

Dear Mr. Hickey:

I am currently conducting doctoral research under the supervision of Dr. Kara
Dawson through the School of Teaching and Learning, College of Education, University
of Florida in Gainesville. My research focus is on developing a better understanding of
the factors that influence elementary teachers to use computers in instruction. I am using
survey methodology to collect data for my study.

I am requesting permission from you to conduct this research in the Citrus County
School District, and to allow the elementary teachers in our school district to participate
in this study. I am also requesting that the completed surveys may be returned to me at
the Instructional Resource Center in self-addressed envelopes I will provide for that
purpose. Data collection for my study will involve three phases:

Phase One: An elicitation study involving approximately 50 elementary teachers. A brief
open-ended survey will be distributed and collected for this purpose.

Phase Two: A pilot instrument will be developed and distributed to approximately 50-100
elementary teachers. The completed survey will then be analyzed and revised to
maximize reliability and validity.

Phase Three: A final survey will be distributed to all elementary teachers in our school
district to capture data relevant to the variables of interest in my study.

Participation in this study will be completely voluntary on the part of elementary teachers
in the district, and they are allowed to not continue with the study at any point they
desire. There is no risk to elementary teachers participating in this study, nor will any
compensation be offered for participation. Individual identities will be protected to the
extent allowed by law. This study will only involve elementary teachers, and there will be
no student participation. I anticipate that the survey will take about 30 minutes for
teachers to complete.








If allowed to conduct this study in the Citrus County School District:

Elementary teachers will be invited to participate on a voluntary basis, and they
will be asked to provide their permission on an Informed Consent Form.
Surveys will not be distributed during critical instructional periods like FCAT
preparation or testing.
I will suggest that teachers complete this survey during planning time or other
non-instructional time so there will be no disruption of classroom instructional
time.
I will make contact with and seek permission from every building administrator
prior to distributing surveys or involving teachers from their school in this study.

The benefits of this study for our school district may include a better understanding of
what influences an elementary teacher to use computers in the classroom for instruction.
This understanding would help us develop better strategies to prepare elementary
teachers to use computers in instruction and it would help us better utilize existing
computer resources to improve teaching and learning.

I have attached a copy of the Informed Consent Form that will be used with teachers
participating in the study. Questions concerning the procedures for this research can be
directed to Dr. Kara Dawson, School of Teaching and Learning, Educational Technology,
2403 Norman Hall, University of Florida, Gainesville, Florida 32611. Phone (352) 392-
9191, ext. 261.

Thank you for your consideration of my request, and I anticipate your response.

Sincerely,








James M. Geddes


XC: Dr. Kara Dawson













APPENDIX B
UNIVERSITY INSTITUTIONAL REVIEW BOARD FORM










School of Teaching and Learning
2403 Norman Hall
University of Florida
Gainesville, Florida 32611

INFORMED CONSENT FORM

My name is Mike Geddes. I am currently conducting doctoral research under the
supervision of Dr. Kara Dawson in the School of Teaching and Learning at the
University of Florida. The focus of my research is aimed at better understanding what
influences an elementary teacher's intent to use computers in the classroom for
instruction. Based on the intended benefits to the Citrus County School District
instructional program, Superintendent David Hickey has approved the dissemination of
the survey.
If you agree to participate in the study, I will provide you a survey and ask you to
indicate your opinion and feelings about various aspects of using a computer in the
classroom for instruction. In addition, I will ask you to complete a brief demographic
questionnaire. You do not have to answer any questions you do not wish to answer. You
are also free to withdraw from the study at any time without consequence. A self-
addressed envelope will be provided for returning the survey and demographic
questionnaire to me via our county mail system. There are no perceived risks to you as a
participant in this survey, and there will be no compensation. The benefits may include a
better understanding of what influences an elementary teacher to use computers in the
classroom for instruction. The amount of time expected to complete the survey and
demographic information is approximately 30 minutes. Your identity will be kept
confidential to the extent provided by law. Results will be reported in the form of group
data.
Questions concerning the procedures for this research can be directed to Dr. Kara
Dawson, School of Teaching and Learning, Educational Technology, 2403 Norman Hall,
University of Florida, Gainesville, Florida 32611. Phone (352) 392-9191, ext. 261.
Questions or concerns regarding your right as a research participant may be directed to
the University of Florida Institutional Review Board, P. O. Box 112250, Gainesville, FL
32611-2250, Phone (352) 392-0433.

I have read the procedure described above. I voluntarily agree to participate in this study
and I have received a copy of the description.

Signature Date
(Participant)

I wish to receive a copy of the results of this study.

Signature Date
Mailing Address
Please return signed form to Mike Geddes at the Instructional Resource Center. Thank You.













APPENDIX C
ELICITATION STUDY QUESTIONNAIRE


COMPUTER USE RESEARCH QUESTIONNAIRE

Instructions:

Please answer the following "open-ended" questions in the space provided. You may use
additional space (ex. back of page) if needed.

There are no right or wrong answers to the questions. Please take your time and be
reflective. Answer candidly and to the best of your ability.

Note: The phrase "present a lesson in your classroom" refers to teacher use of a computer
for the presentation of a lesson or a teacher involving students in using a computer as
part of a lesson. It does not refer to management uses of a computer like keeping
attendance records or electronic grade books.

When you have completed the questionnaire, place it in the provided return envelope
along with the signed Informed Consent Form, and return it via county mail to Mike
Geddes at the Instructional Resource Center.


The deadline for returning the completed questionnaire is December 10, 2003.




Full Text
xml version 1.0 encoding UTF-8
REPORT xmlns http:www.fcla.edudlsmddaitss xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.fcla.edudlsmddaitssdaitssReport.xsd
INGEST IEID EH8OJPYBJ_FZAC1X INGEST_TIME 2012-09-24T13:33:30Z PACKAGE AA00011856_00001
AGREEMENT_INFO ACCOUNT UF PROJECT UFDC
FILES