Measuring the stages of concern in the development of computing expertise

MISSING IMAGE

Material Information

Title:
Measuring the stages of concern in the development of computing expertise
Uncontrolled:
Stages of concern in the development of computing expertise
Physical Description:
ix, 242 leaves : ill. ; 28 cm.
Language:
English
Creator:
Martin, Jean Buddington, 1943-
Publication Date:

Subjects

Subjects / Keywords:
Computers -- Study and teaching   ( lcsh )
Educational innovations   ( lcsh )
Educational Leaership thesis Ph. D
Dissertations, Academic -- Educational Leadership -- UF
Genre:
bibliography   ( marcgt )
non-fiction   ( marcgt )

Notes

Thesis:
Thesis (Ph. D.)--University of Florida, 1989.
Bibliography:
Includes bibliographical references (leaves 234-240)
Statement of Responsibility:
by Jean Buddington Martin.
General Note:
Typescript.
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 - 030477951
oclc - 21018743
System ID:
AA00014226:00001


This item is only available as the following downloads:


Full Text


















MEASURING THE STAGES OF CONCERN
IN THE DEVELOPMENT OF COMPUTING EXPERTISE









By

JEAN BUDDINGTON MARTIN









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

1989

































Copyright 1989

by

Jean Buddington Martin


















ACKNOWLEDGMENTS


The completion of a project such as this would be

impossible without the assistance of many individuals. I
would like each of them to know how much I appreciate the

help that was provided me. I was especially fortunate to

have Dr. James Heald as my committee chairman. He was always

available with sound advice and abundant technical knowledge.

The confidence that Dr. James Hensel expressed in the project

was gratifying. Dr. Lee Mullally asked questions that
required me to become more aware of my audience and thus to

write with greater clarity.

I would also like to express my gratitude to Dr. Gene

Hall, who generously shared his expertise and resources with

me. His enthusiasm and encouragement were deeply

appreciated. I also owe a great deal to Dr. Archie George

for sharing with me his wealth of statistical knowledge, good

humor, and encouragement. The excellent work of Sara Kolb

and Jerry Barucky provided the models for my work, and I am

very grateful to them.

Much of the work involved in this dissertation centered

on my ability to gain access to the subjects. My deepest












appreciation is extended to all of those who said "yes."

Thanks also go to all of my colleagues and friends who

listened to me, shared their ideas with me, and made me laugh

when I did not feel like doing so. Of course, special thanks

go to all of the participants in this study. In addition,

Leila Cantara, Kathy Carroll, and Evelyn Sparrow were

especially helpful to me in the completion of my doctoral

program.

Finally, my greatest thanks I have saved for my dearest

colleague, my best friend, my husband Ken. He provided

unwavering patience and encouragement, never complaining that

I almost always had work to do.



















TABLE OF CONTENTS

Page

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

ABSTRACT . . viii

CHAPTERS

I INTRODUCTION . . ... 1

Rationale for the Study ........... 1
Problem Statement .............. 6
Purposes of the Study . 6
Significance of the Study . 8
Study Questions ................. 9
Definitions of Terms ............. 10
Assumptions . . 11
Limitations of the Study . 11
Summary . . .. 11

II LITERATURE REVIEW . ... 13

Introduction . .. 13
Change Theory Models . .. 14
The Teacher Concerns Model . 18
Stages of Concern About the Innovation 22
Open-Ended Statements of Concern ... 27
The Stages of Concern About the
Innovation Questionnaire . 28
Extension and Diversity of the Stages of
Concern Model . .. 31
Concerns Among Computer Users .. 40
A Pilot Study . ... 43
Summary . . ... .. 45

III RESEARCH DESIGN . .. 47

Organization . . .. 47
Phase I--Identification of Computer Concerns 48
Phase II--Discussion of 78-Item Instrument 51
Questionnaire Construction and Modification
Process . . 51
Item Placement . 57












Page

Administration of the 78-Item Questionnaire .. 57
Analysis of the 78-Item Questionnaire 59
Phase III--Discussion of 32-Item Instrument 60
Data Analysis . ... 61
Study Population . ... 65
Summary . . ... .. 67

IV FINDINGS . . ... ... 69

Introduction ............. 69
Analysis of Responses to 78-Item Questionnaire 71
Questionnaire Respondent Demographics 71
Item Selection for Final Questionnaire 79
Analysis of Responses to Final 32-Item
Questionnaire . .. 95
Questionnaire Respondent Demographics 95
Results Pertaining to Question 1 ... 104
Results Pertaining to Question 2 ... .109
Results Pertaining to Question 3 ... .114
Results Pertaining to Question 4 ... .121
Results Pertaining to Question 5 ... 154
Results Pertaining to Question 6 ... 163

V CONCLUSIONS AND RECOMMENDATIONS ... .176

Introduction . . ... 176
Study Overview . ... .177
Conclusions . ....... 180
Appropriateness of the Concerns Model and
Reliability and Validity of the 32-Item
Questionnaire . .. 180
Relationship of Experience and Education to
the Sequence of Concern Stages and the
Intensities of Concern Stages 183
Relationship of Gender and Age to Stages of
Concern in the Development of Computing
Expertise . . 187
Relationships Between a Major System or
Software Change and a Computer User's
Concern Profile . .. 189
Discussion . . 191
Recommendations . .. 195
Summary . . 197
APPENDICES

A EXAMPLE CBAM PROJECT OPEN-ENDED STATEMENT OF
CONCERN . . 200












Page

B COMPUTER USE OPEN-ENDED STATEMENT OF CONCERN 203

C COMPUTER USE REQUEST FOR PARTICIPATION ... .207

D 78-ITEM COMPUTING CONCERNS QUESTIONNAIRE 209

E 32-ITEM COMPUTING CONCERNS QUESTIONNAIRE 225

REFERENCES . . 241














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

MEASURING THE STAGES OF CONCERN
IN THE DEVELOPMENT OF COMPUTING EXPERTISE

By

Jean Buddington Martin
May, 1989

Chairman: James E. Heald
Major Department: Educational Leadership

A three-phase study was undertaken to determine if the

Stages of Concern, a change theory model formulated in

earlier studies of teacher education and educational

innovation adoption, could be applied to the computing

experience, and to determine whether a valid and reliable

instrument for assessing Stages of Concern among computer

users could be developed within this framework. Additional

questions were formulated to determine whether the

intensities of the concern stages were dependent upon the

amount of experience a person had with a specific computer

application; to determine whether there was a sequential

pattern to the Stages of Concern which was associated with
computing experience or education; to determine if concerns

about using computers were associated with age and/or gender

of an individual; and to determine what effects a major


viii










system or software change had on a computer user's concerns

profile.

The identification of types of concerns held by a broad

range of individuals about computing was accomplished in

Phase I. An open-ended, statement-of-concern form was used
to collect the concerns from 222 respondents. From the data,

the researcher identified 156 usable statements, which were

Q-sorted into categories. In Phase II, the statements
receiving a high degree of agreement as to categorical fit

were included in a 78-item questionnaire which was

administered to 323 participants whose computing experience

ranged from none at all to extensive. Responses from those

subjects were factor analyzed to determine if factors

associated with the hypothesized stages emerged. Based on

the results from this preliminary questionnaire, a final,

32-item questionnaire was developed and administered in Phase

III to a separate sample of 388 participants with a broad

range of computing experience. Data from the responses again

were analyzed with factor analysis.

The construct validity of the instrument was measured
using factor analysis results and item/stage scores,

item/total scores, and interstate correlational analyses.

Reliability of the 32-item questionnaire was determined by

Cronbach's alpha correlation and by using test-retest

correlations. Other research questions were addressed by
using analyses of variance to determine differences among
mean stage score responses for each variable subgroup.

















CHAPTER I
INTRODUCTION


Rationale for the Study
Although we are still very much in the grip of the
"Second Wave," evidence of "Third Wave" changes envisioned by

Toffler (1980) abound. One needs only to browse through any
newspaper to find ample evidence of the decay that was
prophesied--the breakdown of the nuclear family, the
accelerated crime rate, and dangerous economic oscillations.

Within this environment, a new society reliant upon an
increased technological base is being formed. Powerful

computing systems are changing the workplace as prophesied in
Megatrends (Naisbitt, 1982). The tremendous growth and rapid

evolution of computer capabilities during the past few years
have created tremendous challenges for organizations and
their managers. The future promises to bring with it
accelerated technological transformation. The tumult,

created by technological innovation, requires a substantial
amount of "high touch" in order to successfully transcend
this turbulent period of history (Naisbitt, 1982).

Students, educators, professionals, managers, and
workers in almost every endeavor have been a part of or have












been affected by the computer revolution. The transformation

from a nonuser of the technology to one who has developed

whatever level of expertise is necessary or desirable is
often accompanied by feelings that range from slight

apprehension to almost paralyzing fear.

Attitudinal studies often have been focused on levels of
anxiety, aversion, or apprehension associated with the use of

computers. There is a growing body of knowledge to suggest

that concerns about computing do influence performance

(Meier, 1985; Nickell & Pinto, 1986), and there is strong

evidence to suggest that women, as a group, do not feel as

positive as men do about the technology. Further, a gender

gap seems to have developed regarding the use of computers

among children.

Girls are not taking advantage of opportunities to use
computers available to them in school. This has been

particularly conspicuous in middle and high school grades.

Students who have an aversion to the technology are likely to

become adults who are the most fearful and view the computer
as a threat rather than a tool. This condition is of utmost

importance because of the possibility of detrimental

occupational consequences (Sanders, 1986). To compound the

problem, additional researchers have supported the finding that

there is a gender difference among college students (Nickell

& Pinto, 1986; Sproull, Zubrow, & Kiesler, 1986), and other
researchers have provided some preliminary evidence of the












impact of computer aversion as it is manifested at that
educational level (Fritz, 1985; Meier, 1985; Wallace, 1986).

Newspaper columnist Lewis Grizzard (1987) provided a
passionate plea for understanding of people, like himself,
who suffer from high anxiety over the use of computers. One
does not have to be a columnist and author in order to be
expressive about his or her feelings regarding the use of
computers. One novice student reported his first experience

in the computer lab without his instructor:

I was afraid of destroying the disk because it
didn't look very sturdy and while I was typing,
I was worried about power fluctuations. And I
was worried about messing up somehow and making
an irreversible error and having to start all
over again. And helpless! Strained helpless!
I felt really helpless. And the poor girl that
worked in the lab, I was constantly needing her
help not out of necessity. I guess I could
have figured it out on my own, but I was so
afraid. I almost had to have her standing there
while I did this whole letter because it was the
first time on my own and I wanted her standing
there just in case I did something wrong. (Martin,
1987, p. 10)

Certainly not everyone who comes into contact with the
technology suffers from cyberphobia; however, observers have
noted numerous variations of fearful reactions to the use of

computers. For many people, male and female, young and old,
the initial computing experience can be traumatic.
Fortunately most people survive the experience, traumatizing
though it may be.

In a pilot study, Martin (1987) observed that computer
users developed different types of concerns as they learned












to use the computer, and these concerns seemed to follow a

sequential pattern. Initial concerns about their own
abilities to cope subsided, and other concerns surfaced to

replace the apprehensive ones. These new concerns focused

more on the task itself. More specifically, they centered on

factors such as time required to complete the task,

availability of help resources, and quality of the product

produced. Further, the types of concerns expressed by more

experienced computer users related to cooperative work with

peers or how to adapt what they had learned to a different

project. Finally, apprehension about one's own ability

frequently resurfaced when he or she was confronted with a

major procedural change in the use of hardware or software.

However, self-concerns were neither so overwhelming nor did

they take as long to subside when compared to the anxiety

experienced by someone new to the use of computers.

Insights into the change process have been acquired by
researchers over time. A number of different models have

been developed to aid in understanding what takes place when
change occurs (Havelock, 1971). Many of the behaviors and

perceptions documented among computer users are similar to

those identified as "concerns" in the teacher development

process noted by Fuller (1969) in her work with student

teachers. Subsequently, Fuller developed the Teacher

Concerns Model, which categorized various types of concerns
about teaching. These concerns are progressive and can be












grouped into three broad categories: self concerns, task
concerns, and impact concerns. This work was elaborated upon

and generalized to the process of the adoption of an
innovation by Hall, Wallace, and Dossett (1973). Hall,
George, and Rutherford (1979) described the adoption of an

innovation as a process through which individuals migrate.
The product of the endeavors of Hall et al. (1973) was the
Concerns Based Adoption Model (CBAM) of which one diagnostic
dimension recognizes seven distinct Stages of Concern (SoC)
about an innovation that are experienced during the change
process. Hall and others have demonstrated the diversity,
validity, and reliability of the model through a body of

research conducted both nationally and internationally (Hall
& Hord, 1987; Van den Berg & Vandenberghe, 1981).
Additionally, research related to the Stages of Concern
was extended to the nursing profession by Kolb in a 1983

study. Subsequently, Barucky (1984) added another dimension
to the concerns theory knowledge base in his study by

broadening the framework to encompass leadership development
in the Air Force professional military education programs.

Recognizing the diverse application of the concerns
model, there appears to be ample confirmation that the model

can be used to provide the theoretical framework to examine
the various concerns that are being observed among computer
users. One practical application would be the development of

appropriate interventions to mitigate the more harmful












effects of change that are related to the use of computers.
At the theoretical level, the introduction of computing is an

area of change that has not yet been used to test the

generalizability of the Stages of Concern framework.

Problem Statement

A priority in this investigation is to determine if the

Stages of Concern Model is an appropriate conceptual model to
assess concerns among computer users. Subsidiary tasks

closely related to this primary task are to determine if an

instrument for assessing Stages of Concern among computer

users can be developed within the framework; to determine if

there is a relationship between the level of computing

education and the sequence of concern stages; to determine if

there is a relationship between the intensities of various

concern stages and the computing application experience level
of an individual; to determine the associations between

age/gender and SoC scores; and to determine what effect a
major system or software change has upon a computer user's

concern stage and concern sequence.

Purposes of the Study
Technology can cause tremendous apprehension for some

people. It can severely handicap the choices available to

them as students, as evidenced through course selection, and

later in career choices. There may be some groups of people,
possibly women and older adults, who suffer

disproportionately from this handicap. Even when the












experience is not excessively traumatic, concerns can inhibit
the maximum benefits that could be obtained during the
learning process. A first step in mitigating the
debilitating consequences of this situation is to develop an

instrument that can be used to identify the current concerns
level of individuals using computers. Only after this is

accomplished can relevant intercessions be used to expedite
and ease the transition into the next stage of the process.
More research must be conducted in order to develop a valid
and reliable instrument to identify the variety of concerns
that are common to computer users who have varying levels of

expertise and experience with the technology. The findings
of this research can then be used to answer the question
about the feasibility of the Stages of Concern framework as a
theoretical structure for the development of an instrument to
identify the concerns of computer users and other related
questions.

Therefore, the purposes of this study are several.
The first purpose of this study is to corroborate and advance
the knowledge base regarding the concepts of the Concerns

Model by establishing a model that represents the concerns of
computer users. Second, this study is intended to develop an
instrument that can be used to determine the Stages of
Concerns about computing that exist in a wide range
of people using the technology. The instrument will be
subjected to appropriate tests to determine its validity and












reliability. Third, this study is intended to determine

if there is a relationship between the level of computing

education and the sequence of concern stages. Fourth, this

study is intended to determine if there is a relationship

between the intensities of various concern stages and the

computing application experience level of an individual.

Fifth, this study is intended to determine the associations

between age/gender and SoC scores. Finally, it is the intent

of this researcher to determine what effect a major system or

software change has upon a computer user's concern stage and

concern sequence.

Significance of the Study
The primary importance of this study is the extension of
concerns theory into yet another area of study. Extensive

research that has already been undertaken has shown that the

Stages of Concern model can be a powerful tool for

undertaking and advancing the educational process. Without a

doubt technology will continue to advance, and people must be

aided in the adaptation process. Change carries with it

major ramifications for the individual and society as the use

of the innovation becomes more prevalent. The ramifications

can take the form of a course of study not pursued, a career

not sought, a position of advancement not achieved, or

something more subtle such as unnecessary stress that may












manifest itself in negative ways. The individual might
suffer a lessening of self-confidence or a feeling of lack of
competence which can be so detrimental to the human
condition. None of these consequences is very desirable.

Certainly, additional research is necessary in order to
better understand the process of adapting to changes in the

computing environment. Only then can the maximum benefit be
achieved from the advancements that are being made in
computing technology. Only then can we develop the most

effective and relevant interventions that will allow users of
the technology to progress at an optimal pace.

Study Questions
1. Do identifiable "stages (factors) of concern" about
the development of computing expertise exist among

individuals with varying amounts of computing experience?

2. Can an instrument for assessing Stages of Concern
among computer users be developed within the framework which
is both reliable and valid?

3. What is the relationship of the level of computing

education and experience to the sequence of concern stages?

4. What are the relationships between the intensities
of various concern stages and the computing application

experience levels of individuals?
5. What are the associations between age and gender
with SoC scores?












6. What effect does a major system or software change

have upon a computer user's concern stage and concern

sequence?

Definition of Terms
In this study, the key terms will be defined as follows:

The definition of concern as used in this study was

developed by Hall et al. (1979). The composite

representation of the feelings, preoccupation, thought, and

consideration given to a particular issue or task is called

concern.

Stages of Concern are a set of sequential stages of

concern through which a person facing change is believed to

pass.

Peak Stage of Concern refers to percentile scores used

to provide a comparison of relative intensity of concerns at

each stage. Although an individual may communicate concerns

that are exemplary of many different stages, the stage of

predominant intensity is referred to as the Peak Stage of

Concern.

Cyberphobia is an intense fear of computers or any of

the associated devices.

Computer users will be categorized as follows:

A novice is a person who has some knowledge about

computing but whose experience is very limited either by type

of computing application or amount of actual time spent with

a computer or associated device.












An intermediate is a person whose profession is not in
the computing field, but through coursework and/or experience

has developed expertise with a variety of computing

applications.

A professional is a person whose profession is in the

computing field, such as a programmer, systems analyst,

operator, computing faculty member, or data base
administrator.

Assumptions
An assumption will be made that the respondents who

furnish information regarding their concerns about computing

on the open-ended statements of concerns and questionnaires

are providing accurate and honest expressions of their
sentiments.

Limitations of the Study
Stratified convenience samples will be used to provide

representatives of various categories of computer users.

However, the generalizability of the conclusions from the

study will be limited because of the inability to obtain a

true random sample of representatives of all categories of

computer users.

Summary
A description of the problem to be studied was presented
in this chapter. A succinct portrait of the necessity to

address the conditions that accompany technological

innovation was presented. A brief overview of the












development and application of the Stages of Concern model
was tendered as a framework for conceptualizing the change

process as it relates to computer users. Certain

definitions and limitations of the study were identified.

Chapter II contains a review of literature relevant to

the application of the Stages of Concern model as a feasible

approach to understanding change theory as it relates to
computer users. Chapter III includes a description of the

methodology which was followed in the development of the

instrument which was tested in this exploratory study. A

description of the design of the study and the procedures

used will also be included. The findings of the study are
presented in Chapter IV, and the conclusions and

recommendations of the investigator are reviewed in Chapter

V.


















CHAPTER II
LITERATURE REVIEW


Introduction
Technology has invaded almost every aspect of the human
existence in a modern society. This invasion has brought

with it tremendous changes in how people conduct themselves
in their personal and professional lives. Change is

frequently accompanied by a certain amount of concern. The

purpose of this study is to focus on the concerns of users of

computers through the employment of the Stages of Concern

model. More specifically, it is to determine if there are

Stages of Concern among computer users and, if so, to develop

a valid and reliable instrument that can be used to identify
them. The first section of this literature review focuses on

some of the primary contributors in the area of change

theory. The second section highlights the contribution of

the development of the Teacher Concerns Model and the

instruments that have been used to assess those concerns.

The third section is devoted to a description of the research
procedures employed to demonstrate the concept of sequential

stages of concern in the process of the adoption of an

innovation which is incorporated in the Concerns-Based












Adoption Model (CBAM). The fourth section is an exploration
of the extension and diversity of concerns-based theory. The

fifth section is a review of literature that is related to
the apprehension, anxiety, and concerns of computer users.

Change Theory Models
One of the best known researchers in the development of
change theory models is Havelock (1971). He proposed three
different structures for understanding change: the Social
Interaction Model; the Research, Development, and Diffusion

(R,D&D) Model; and the Problem Model.
The focus of the Social Interaction Model was on the

diffusion of user involvement from person to person

throughout the social system and understanding the five
decision phases through which an individual migrates
sequentially in the adoption process. "The diffusion of the

innovation depends greatly upon the channels of communication

within the receiver group, since information about the
innovation is transmitted primarily through the social

interaction of the group members" (Havelock, 1971, p. 29).
Rogers' (1962) conceptualization of social interaction has

been the one most used when studying innovation adoption
within this framework. The first phase is an initial
awareness of the innovation, followed by an interest in
learning more about it. A decision is made in the third

phase, evaluation, about the adoption of the innovation. The
last two phases are trial and then adoption.












Havelock (1971) identified Mort, Carlson, and Ross as
major advocates of the social interaction approach in
education. He further noted that the Model is based largely
on rural sociology tradition studies (Havelock, 1971).
The R,D&D Model depicts the change process as "an
orderly sequence which begins with the identification of a
problem, proceeds through activities which are directed
towards finding or producing a solution to this problem, and
ends with diffusion of this solution to a target group"
(Havelock, 1971, p. 39). This approach begins the study of
the change sequence at an earlier time then does the previous
model. The five basic assumptions upon which this model
(Havelock & Havelock, 1973) is guided are as follows: (a)
that a rational sequence exists in the formulation of an
innovation which should include research, development,
packaging, and dissemination; (b) that an enormous amount of
planning takes place over a considerable length of time; (c)
there is a division and coordination of labor that
corresponds with the first two assumptions; (d) there is a
passive but rational consumer who, if it is presented to him
or her in the proper manner, will accept and adopt the
innovation; and (e) a high initial development cost of the
innovation is acceptable because of the efficiency, quality,
and appropriateness for mass audience dissemination.
One of the best known adaptations of this model is that
of Guba and Clark (1976). It includes detailed descriptions












of objectives and activities which can be used to foster the
basic objectives of the R,D&D Model.
The Problem Solver Model provides the foundation for
group theorists and Organizational Development (OD)
proponents. This approach "includes studies which focus on
the efforts of a receiver in solving his own particular
problems" (Havelock, 1971, p. 69). The basic tenants of this
approach (Havelock & Havelock, 1973) include: (a) the user's
need is the major consideration in any planned exercise; (b)
diagnosis of user's needs is an essential aspect of the
change process; (c) a nondirective approach with users is
essential; (d) there should be complete utilization of
internal resources; and (e) commitment to change is likely to
be stronger if initiated by the user or there is at least
real user involvement.
Perhaps the earlier work of Lewin (1951) laid some of
the groundwork for this approach to change. The three major

stages in the change process that he outlined were

unfreezing, moving, and freezing. Lippitt, Watson, and
Westley (1958) expanded upon Lewin's change model by
redefining the stages of change. They also presented a
rational conception of the "change agent," a person who
possesses the skills required to assist a client with
problems encountered as change takes place. Havelock (1971)
suggested that the "change agent" or problem solver must act
in a collaborative manner in order to really be effective.












The OD strategy for planned organizational change falls
within the framework of the Problem Solver Model. The focus
of this approach is on group interaction rather than on the
individual (McGill, 1977). The basic assumptions about the
individual, groups, and organizational systems underlying OD
principles are generally in congruence with the theories of
McGregor (1967), Likert (1967), Argyris (1964), Herzberg
(1966), and Bennis, Benne, and Chin (1961).
OD strategies can be used to facilitate change in such
diverse settings as business firms, school systems, American
Indian tribes, labor unions, governmental units, service
organizations, and research and development laboratories
(French & Bell, 1978). An important OD study in school
systems is that of Schmuck, Runkel, Arends, and Arends
(1977), which noted the interdependence of people working on
tasks and the collaboration among individuals and groups that

takes place in that setting.
Clearly, educational institutions pose a most
challenging backdrop for the study of innovation adoption. A
basic mission of these institutions is to foster change by
aiding their charges to grow and develop. The ability to
carry out this goal in an optimal fashion has been the
subject of considerable research. Fullan (1982) draws
heavily on the literature regarding educational change in
order to aid the reader in understanding the complexities of
the process. One area of this research has focused on the












education of teachers. The next section presents the

development of a Model for recognizing and assessing concerns
of education students.

The Teacher Concerns Model

Fuller (1969) summarized her research regarding concern

about the benefit that students were obtaining from education

courses as follows. She defined "concerns" to include such
emotions as motivations, perceptions, attitudes, feelings,

and mental gymnastics indulged in by a person when confronted

with a new process or product. Fuller was aware that

students are more motivated to learn material that is of

interest to them and disinclined to learn that which is

considered to be less relevant. This phenomenon was confirmed
by her in a pilot study involving 100 education students who

were queried about the preparation they had received in

education courses. During the hour-long individual

interviews, 97 of the students made deprecating remarks about

a certain introductory education class. All of these

students were young undergraduates without teaching

experience. However, three of the students were quite

enthusiastic about the course. These students had

substantial teaching or similar experience and were

middle-aged. Two explanations were offered to account for

the difference in perception about the class. One

possibility is that the class was worthless. A second

possibility is that many of the students who enter education












programs are not prepared to benefit from them as they are

now taught, and the structure of the traditional teacher
education program may not be the most appropriate model for

educating student teachers.

Fuller conducted another study to explore what teachers
are concerned about and to ascertain if the concerns should

be conceptualized into a helpful framework. She found it

remarkable that there was so much consistency in the findings

of earlier investigators, particularly since they had

surveyed such diverse populations. Such concerns as class

control, feelings of self-adequacy, evaluations by

supervisors, and their teaching environment were noted.

However, these concerns were not the ones being addressed.
She hoped that her study would help educators to better

understand this discrepancy.

The study used counseling seminars conducted by a

psychologist. The psychologist met a group of six student

teachers for two hours each week during the students'

teaching semester. The following semester a group of eight
student teachers were involved in the counseling sessions.

This time there was co-counseling in order for the two

psychologists to insure that opportunities for expression

were not hampered. Yet another group of seven students was

similarly counseled during the third semester. The results

of these counseling sessions indicated that there were two

main categories of concerns: concerns about self, including












self-adequacy and self-protection; and concerns about pupils,

their learning and their progress. Self-concerns

predominated during most of the semester. Only towards the

end of the semester did concerns shift to the pupils.

An additional study was undertaken with 29 students.

They were asked to write about what they were concerned about

at the time. The groups were surveyed at approximately

two-week intervals during the course of the semester. Of the

29 students, 22 communicated concerns primarily about

self-adequacy, six students expressed both self and pupil
concerns, and one expressed pupil only concerns.

The aforementioned studies, along with a review of

research carried out by other investigators, led Fuller
to believe that there was sufficient evidence to support the

beginnings of a developmental conceptualization of teacher

concerns. She suggested three distinct phases: a

pre-teaching phase, an early teaching phase, and a late

teaching phase. In the first phase, students did not appear

to know what their concerns were. The concerns were

anticipatory or apprehensive, but rarely specific to

teaching. In the second phase, the concerns dealt with self.

Questions involving how to handle the job, how much

administrative support to expect, and working relationships

were paramount. There was a need to understand the

parameters of the task. Along with these covert concerns
were overt concerns such as ability to control a class,












knowledge of subject content, and the ability to contend with
evaluation. The third phase was evidenced by concern with
pupil progress and evaluation of one's own contribution to
that progress. These concerns could be clustered into four
distinct categories: concerns unrelated to teaching,
concerns about self in relation to teaching, concerns about
the task of teaching, and impact concerns. There was also
evidence to suggest that a person traversed sequentially
through the various concern stages from self, to task, to
impact. Although Fuller had been able to make headway in the
conceptualization process, she realized that the personal
counseling interviews were extremely time consuming and
required considerable effort by highly trained personnel.
As a result, the Teachers Concern Statement (TCS) was
developed. Because of its open-ended nature, the TCS does
not restrict teachers to specific concerns; instead one broad
question is asked: "When you think about your teaching, what
are you concerned about?" The teachers are provided with an
8.5" x 11" sheet of paper and given 10 minutes to respond.
The statements are scored by indicating one of six teaching
concerns categories or one non-teaching category (Fuller &
Case, 1972). This instrument enhances the process of
eliciting teacher concerns but is time-consuming to score and
requires significant preparation by the coder. Further
research by Fuller and others (George, 1978) resulted in the
development of an instrument called the Teachers Concern












Checklist (TCCL). It is a quick-scoring questionnaire that
consists of 56 items. A five-point Likert scale with ranges
of 0 (not concerned) to 5 (extremely concerned) is used to
assess the degree of concern about each item. It is scored
by summing the responses on certain items to obtain the five
scale scores.
Following the development of the TCCL, a Teacher
Concerns Questionnaire (TCQ) was developed and subsequently
validated, that would measure self, task, and impact concerns
of teachers about teaching. The TCQ consists of 15 items
which are also on Form B of the TCCL. Therefore, the
56-item TCCL-B can also be used to assess self, task, and
impact concerns.

Stages of Concern About the Innovation
During the 1969-70 academic year, staff members of the
Inter-Institutional Program of the Research and Development
Center for Teacher Education at the University of Texas at
Austin (UT R&D Center) realized that teachers and professors
who were experiencing change appeared to have the same type
of "concerns" that Fuller had identified relevant to
teaching. However, the concerns in this case were those
encountered when any type of educational innovation was to be
adopted (Hall et al., 1973). UT R&D Center staff members
realized that change encompasses technical problems involved
in the adoption of an innovation (including both processes
and products) and the personal needs of the individuals












involved. Through the 1970s, a series of studies were done

to test the model and related hypothesis about intervention

(Hall & Loucks, 1978; Hall & Rutherford, 1976). The result

of the work conducted by these researchers was the

development of a conceptual model referred to as the Concerns

Based Adoption Model (CBAM).

Certain assumptions were made based upon considerable
experience derived from involvement in innovations in

colleges and school settings. These assumptions are

fundamental to the CBAM in that they establish the

perspective for viewing the change process (Hall & Loucks,

1978):

1. change is a process, not an event, and it takes time
to institute change;

2. individuals must be the focus if change is to be

facilitated, and institutions will not change until their

members change;

3. the change process is an extremely personal

experience, and how it is perceived by the individual will

strongly influence the outcome;

4. individuals progress through various stages
regarding their emotions and capabilities about the
innovation;

5. the availability of a client-centered diagnostic/
prescriptive model can enhance the individual's facilitation
during staff development;












6. people responsible for the change process must work
in an adaptive and systematic way, and progress needs to be
monitored constantly.
The Concerns-Based Adoption Model addresses all of
these assumptions: the individual's concerns about the
innovation, the specific manner in which the innovation
is used, and the adaptation of the innovation to the
individual. The current study is based primarily on the
first of three diagnostic dimensions in the model, which is
the individual's concern about the innovation.
Based upon research conducted earlier by Fuller and
others, it was possible to identify seven distinct Stages of
Concern About an Innovation that an individual is likely to
encounter as he or she moves through the change process (Hall
& Loucks, 1978) (see Figure 1). As the study findings
emerged, it became clear that an individual can have concerns

in more than one stage at a time and that typically some
stages are more intense than others. A profile of concerns
could be talked about as the innovation is implemented.
Further, there is a hypothetical change in the shape of the
concerns profile as the change process unfolds. Initially
the stage 0, 1, and 2 concerns are most intense. Ideally,
once implementation actually begins to take place, Management
concerns, stage 3, increase in intensity and stages 0, 1, and
2 become less intense. In the final stages, 4, 5, and 6,
Impact concerns, can become more intense and Management
concerns abate.












6 REFOCUSING: The focus is on exploration of more
universal benefits from the innovation, including
the possibility of major changes or replacement
with a more powerful alternative. Individual has
definite ideas about alternatives to the proposed
or existing form of the innovation.
5 COLLABORATION: The focus is on coordination and
cooperation with others regarding use of the
innovation.
4 CONSEQUENCE: Attention focuses on impact of the
innovation on student in his/her immediate sphere
of influence. The focus is on relevance of the
innovation for students, evaluation of student
outcomes, including performance and competencies,
and changes needed to increase student outcomes.
3 MANAGEMENT: Attention focused on the processes
and tasks of using the innovation and the best
use of information and resources. Issues related
to efficiency, organizing, managing, scheduling,
and time demands are utmost.
2 PERSONAL: Individual is uncertain about the
demands of the innovation, his/her inadequacy to
meet those demands, and his/her role with the
innovation. This includes analysis of his/her role
in relation to the reward structure of the
organization, decision making, and consideration of
potential conflicts with existing structures or
personal commitment. Financial or status
implications of the program for self and colleagues
may also be reflected.
1 INFORMATIONAL: A general awareness of the
innovation and interest in learning more detail
about it is indicated. The person seems to be
unworried about himself/herself in relation to the
innovation. She/he is interested in substantive
aspects of the innovation in a selfless manner such
as general characteristics, effects, and
requirements for use.
0 AWARENESS: Little concern about or involvement
with the innovation is indicated.

Figure 1. Stages of Concern About the Innovation.

Adapted from Hall and Hord, 1987. (Original concept from
Hall, Wallace, & Dossett, 1973).












In order to assess user concerns, additional research
into the concept of Stages of Concern (SoC) focused on the
creation of a reliable and valid instrument; the SoC
Questionnaire is the product of that work (Hall et al.,
1979). The SoC Questionnaire provides the user with the
ability to quickly measure the Stages of Concern of teachers
and professors. The validation study took place over 3
years, in addition to the 10 years of research conducted by
Fuller. Several different formats and methodologies
initially were considered. "The resulting SoC Questionnaire
was tested for estimates of reliability, internal
consistency, and validity with several different samples and
11 different motivators" (Hall, George, & Rutherford, 1979,

p. 9).
The first pilot instrument designed to measure concerns
individuals might have about an innovation was developed in
the fall of 1973. It consisted of an open-ended concerns

statement and required a forced ranking. Variations of this
format, adjective checklists, interviewing, and use of Likert
scales were all contemplated. However, two instruments
ultimately surfaced by the spring of 1974 for measuring
Stages of Concern: the primary instrument (the SoC
Questionnaire) and the Open-Ended Concerns Statement
developed by Newlove and Hall (1976).












Open-Ended Statements of Concern
Newlove and Hall (1976) developed a direct and easy
method for ascertaining the concerns of users and nonusers
about an innovation called the Open-Ended Statement of
Concern About an Innovation. This format was the one
pioneered by Fuller and Case (1972). Individuals were
requested to respond to the question: "When you think about
[the innovation], what are you concerned about?" The purpose
was to develop a broad overview of the individual's concerns.
The instrument was intended to provide a fast and easy method
to assist a workshop leader or change facilitator with a
technique for identifying the concerns of his or her clients
in their own words, which meant a trade-off in comparison to
a more scientific instrument capable of "pinpoint" accuracy.

The developed instrument is composed of three pages (see
Appendix A). The first page provides information about the

question being asked. It also provides a means, other than a
name, for identifying the respondents. The second page
contains the question, an area for the response, and space
for scoring the response. The last page is used to collect
any additional data that are considered important.

The responses to the open-ended question are read
through completely in order to gain a general insight into
the affect, motivation, and needs of the respondent. The
analyst attempts to identify the concerns as self-, task-, or
impact-related, or unrelated to the innovation at all. If












desired, the content analysis can be made more systematic by
scoring the concerns according to the Stages of Concern
enumerated in Figure 1. Each idea is recognized as a
separate unit and is assigned a number that coincides with

the appropriate stage of concern. The numbers are not to be
averaged as that would provide a distorted picture of the
data and would be meaningless.
The Stages of Concern About the Innovation Questionnaire
The Stages of Concern Questionnaire, unlike the
Open-Ended Statement of Concern, does not rely on the
articulateness of the respondent. The instrument now has
been used extensively in concerns research involving many
different innovation adoption situations. The instrument
developers collected more than 500 statements of concern
about an innovation from college professors and teachers.
Those statements were subjected to a Q-sort procedure which
resulted in the identification of 195 statements that
represented the seven hypothesized stages of concern. The
statements became the items on a prototype questionnaire that
was administered to college professors and teachers who were
involved in implementing innovations.

Factor analysis using VARIMAX rotation was used on the
data (N = 366) which resulted in the identification of seven
factors corresponding to the seven hypothesized stages.
Subsequently a 35-item Stages of Concern Questionnaire was
developed comprised of those five items that loaded highest












on each scale. The items representative of each stage on the
questionnaire were selected in a way that enhanced the
likelihood that the instrument would have high internal
reliability. Respondents are asked to rate each item on a
zero (not true of me now) through seven (very true of me now)
scale as a means of indicating the degree to which that item
reflects their own present concerns. The scoring procedure
begins by adding the circled responses for the five items on
each scale. The total of these scores is the total raw stage
score (George, 1977). Additionally, percentile tables were
established which allow the stage scores to be converted into
percentile scores. Peak or predominant stages of concern,
and the relative intensity of the other concern stages, can
then be plotted from the percentile scores (Hall & George,
1980).
A number of innovation studies were conducted during the
mid-1970s to test the validity and reliability of the SoC
Questionnaire. Internal validity was examined using
Cronbach's alpha procedure on a large sample of data (N =
830) supplied by teachers involved with team teaching and
professors who expressed their concern about the innovation.
A subsample of teachers (N = 132) participated in a
test-retest of the questionnaire over a two-week period.
Alpha coefficients ranged from .64 to .83, and test-retest
correlations ranged from .65 to .86. This would indicate
that the instrument had internal consistency and stability
for each of the seven stages (Hall et al., 1979).












Since no other instruments for measuring the
conceptualized concerns existed, the validity of the scores

as measures of concern could not be established by using a
comparison method. An indication that the questionnaire did
measure concerns, as they were defined in the model, was

demonstrated initially through a check of the

intercorrelation of the scale scores on both the original
195-item prototype questionnaire and the 35-item SoC

Questionnaire. The scores correlated with each other in the
manner that would be expected if they formed a developmental
sequence as hypothesized. Guttman referred to this as the
simplex pattern. Each stage seems to be most like those
stages contiguous to it, and this suggests that the sequence
of concerns stages might well be developmental (Hall et al.,

1979).
Subsequently, researchers attempted to measure the

relationship between the estimated assessments of concerns

(based on either taped interviews or open-ended statements)
and raw stage scores. The findings of these studies were not
consistent. However, a relationship was shown between the

SoC Questionnaire raw stage scores and the estimates of peak
or prominent stage of concerns (based on interviews or
open-ended data). In one of the most rigorous of these
studies, findings indicated that six of the seven stages of
concern were supported by significant correlations (George,
1977; Hall & George, 1980).












Convincing evidence of the validity of the SoC
Questionnaire has been demonstrated through several
longitudinal studies. The findings of one study indicated
that significantly lower early stage concerns were reflected
in responses of teachers who had attended a five-week summer
workshop devoted to a new approach to reading instruction
than in the responses of teachers who had not attended the
workshop (but had received a one-day introduction to the
approach). In other studies, assessments of concern at
various times, prior to and after the introduction of an
innovation and innovation workshops, indicate that the SoC
Questionnaire measured a decrease in lower stage concerns and
an increase in higher stage concerns than might be expected
as a result of the experience (George, 1977; Hall & George,
1980).

Extension and Diversity of the Stages of Concern Model
Since the mid-1970s, the Stages of Concern dimension of
the Concerns-Based Adoption Model has been applied in a large
number of diverse change efforts in the United States,
Canada, Australia, Belgium, the Netherlands, and other
countries. Although Fuller envisioned concerns data as being
used in the development of more relevant educational
programs, CBAM has proven to be a framework that can provide
a structure for successfully implementing change in a broad
range of applications. Members of the original research

group, as well as others, have used and refined the basic












concepts developed earlier in the context of the educational
environment. Other researchers have extended certain
components of the model outside of the traditional
educational setting, including private sector settings.
Some recent studies incorporating the concerns-based
model include a description of the concerns that principals
have about facilitating change (Rutherford, Hall, & Newlove,
1982); an intervention used to aid in the implementation of a
new curriculum program in a Tampa public school system
(Nielsen & Turner, 1986); a conceptualization of strategies
in order to facilitate understanding and addressing personal
concerns of teachers (Marsh & Jordan-Marsh, 1985); a
presentation of a new concept for facilitating leadership
teams with certain characteristics (Hall & Hord, 1986); a
presentation of three vectors that can be used to determine
if the innovation has been established in the institution
(Hord & Hall, 1986); and a report on how the use of all of

the components of CBAM can provide the basis for facilitating
change at individual and institutional levels (Hall & Hord,
1987).
In addition, there are several current books founded
upon the principles of CBAM. The books include Taking charge
of change (Hord, Rutherford, Huling-Austin, & Hall, 1987) and
Change in schools: Facilitating the process (Hall & Hord,
1987). Further evidence of the diversity of the model is
represented by its extension into the field of nursing,












health, recreation, and the leadership component at the
United States Air Force Academy.

A 1983 study by Kolb resulted in the development and
application of a Nurse Concerns Questionnaire. Kolb had
extensive experience in nursing and nursing education. She
conjectured that a pattern of concerns existed among
prenursing students, nursing students, and practicing nurses
that was similar to the self, task, and impact concerns
identified by Fuller and later expanded upon by Hall and his
associates. Kolb used the same methods and procedures that
had been used in the creation of the earlier concerns
questionnaires. She was able to determine that there were
identifiable stages of nursing concerns and developed a valid
and reliable Nurse Concerns Questionnaire that could measure
those stages. The study that Kolb conducted allowed the
following research questions to be answered:

1. Are there identifiable Stages of Concern among
prenursing students, nursing students, and practicing nurses?

2. Can an instrument for assessing Stages of Concern
among nurses be developed which is both valid and reliable?

3. Is there a sequential relationship between
predominant Stages of Concern among nurses and phase of
professional preparation?
4. Are there significant differences in the focus of
concerns among prenursing students, nursing students, and
practicing nurses in relation to type of generic program?












5. What are the patterns of Stages of Concern among
nursing students who have had previous paid patient care

experience?

6. What are the patterns of Stages of Concern among
nurses in relation to selected facets of professional

practice?

In general, Kolb's findings supported the earlier
research about concerns theory. First, there was
considerable similarity between the seven stages of concern

identified by Kolb and those established by Hall and his

associates relative to the adoption of an educational

innovation. Second, a reasonably reliable and valid 35-item

instrument was developed following the procedures used in

prior concerns research. Third, there appeared to be a

sequential ordering of the earlier stages of concern that

coincided with professional preparation. However, the

pattern of additional years of experience paralleling higher

stages of concern did not seem to hold true.

The implications from this study are several. Kolb
advanced the following views: understanding and accepting

concerns differences among different categories of nurses

will aid nurse educators to cope with and resolve self

concerns of their students, and also allow them to derive

applicable instructional strategies; and understanding
concerns differences among nurses will promote cooperation

among nurses who are at disparate stages of concern. Kolb












(1983) also indicated that the findings of the study provided
additional support and knowledge of concerns theory. Changes
made to the original Stages of Concern definitions reflect
the nursing emphasis of the study (see Figure 2).
Barucky (1984) completed an extensive research study,
like Kolb, following the procedures and methodology advanced
by Fuller and Hall and associates in the development of the
SoC Questionnaire. However, the goal of Barucky's study was
to determine whether concerns theory applies to leadership
training in Officer Professional Military Education (PME).
His intent was to develop a valid and reliable questionnaire
that would assess the stages of concern about leadership
skill development among various categories of students
attending U.S. Air Force officer PME programs. The Barucky
study allowed the following research questions to be
answered:

1. Do identifiable stages of concern about development
of their personal leadership skills exist among participants
of Air Force pre- and postcommissioning PME programs?
2. Are these stages consistent with those found in
earlier concerns research?

3. Can a valid and reliable instrument for measuring
stages of concern be developed?
4. Do these stages of concern correspond to the degree
of leadership experience so that (a) respondents reporting no
leadership experience will have concerns reflecting "early"












0 CONTEXTUAL: Concerns are indirectly related to
nursing but are not specific to either the study
or the practice of nursing. The concerns
expressed have their origins within the context
of the individual's involvement in nursing.
1 INFORMATIONAL: Focus of concern reflects an
interest in learning more about what nursing
involves.
2 PERSONAL: Concerns are reflective of an
uncertainty about the demands of nursing and
about personal adequacy to meet those demands.
This includes concerns about personal status,
financial adequacy, and public opinion of self.
3 MANAGEMENT: Concerns are focused on process
and tasks involved in the provision of nursing
care. This includes concerns about organizing
time and resources in relation to the tasks to
be completed.
4 CONSEQUENCE: Attention focuses on the impact of
care in terms of patient outcomes within the
individual nurse's immediate sphere of influence.
This includes concerns about personal or
professional changes which can be made that will
have a positive effect on patient outcomes.
5 COLLABORATION: Focus is on coordination and
cooperation with other nurses and other health
care professionals, with the goal of improved
patient outcomes. This includes concerns about
collaboration between education and service.
6 REFOCUSING: Focus of concerns is on broader
professional issues. What changes can be made--
self, other nurses, and in nursing care. This
includes concerns about changes in career as an
alternative to the existing state.


Figure 2. Stages of Concern of Nurses.












stages to a greater extent then respondents with considerable

experience, and (b) those reporting considerable experience

reflecting "later" stages to a greater extent than those

reporting no experience?

5. Do differences among concerns stages exist among
respondents based on the type of PME program currently
attended?

6. Do differences in concerns stages exist among

respondents attending the same type of PME program, based on

amount of leadership experience?

Again, the results of the Barucky study, like the Kolb
study, generally supported the concerns theory findings

established in earlier research. First, there was similarity
between the seven stages of concern identified by Barucky and

those established through prior educational innovation and

nursing concerns research. Second, a reasonably reliable and

valid 35-item instrument was developed following the

procedures used in prior concerns research. Third, there

appeared to be a sequential ordering of the earlier stages of

concern that coincided with professional preparation. The

response pattern indicative of the developmental nature of

concerns stages was most strongly supportive of stages 0, 1,
2, and 6.

A major implication of this study was that concerns

theory can be extended to leadership training in officer PME
programs. There is also evidence to suggest that independent












concerns stages exist in the leadership framework. Again,
changes made to the original Stages of Concern definitions
reflect the U.S. Air Force leadership program emphasis of the
study (see Figure 3). Further, the stages of concern
identified in this study are very similar to those identified
in previous educational innovation and nursing concerns
research, and thus the benefits accrued from this knowledge
base are likewise considerable.

Researchers in Belgium and the Netherlands developed a
Flemish language 52-item version of the Stages of Concerns
Questionnaire and have used it in major educational change
efforts in those countries (Vandenberghe, 1983). Although
this instrument has a somewhat different focus, it also
measures seven stages of concern.

Marsh and Jordan-Marsh (1985) argued that understanding
the psychological dimensions of change, as well as the
sociological dimensions of change, offers additional

ammunition for helping teachers cope with personal concerns
in the innovation implementation process. Marsh and
Jordan-Marsh used psychological literature on decision
making, primarily Janis and Mann (1976) and Bandura (1981),
to aid in developing ideas that could lead to the development
of interventions to address personal concerns.
In addition, Jordan-Marsh (1985) adapted the Stages of
Concerns Questionnaire to reflect the concerns of
participants in an exercise program. The resultant












0 NON-LEADERSHIP CONCERNS: These concerns reflect
either lack of current concern for developing
leadership skills or the feeling that development
of these skills is in conflict with, or is
overshadowed by, other priorities.
1 INFORMATIONAL: These concerns reflect a need
for more general information about leadership.
They do not seem to imply a feeling of inadequacy
but rather an open desire to know more about what
leadership entails, what the role of a leader is
like in the air force, and the differing
environments that air force leaders will face.
2 PERSONAL: These concerns center on the demands
or requirements of a leadership role and reflect
some question/uncertainty about whether the
respondent can fulfill these demands adequately
or possesses necessary leadership skills. This
stage also includes concerns about how the
respondent will be perceived by others or whether
she/he will be liked, respected, or followed.
3 MANAGEMENT: These concerns relate to mastery of
the specific tasks, skills, or behaviors perceived
as necessary to be an effective leader. They also
include concerns about situational obstacles, such
as time, resources, or organizational support,
that may interfere with the mastery of these areas.
4 CONSEQUENCE: These concerns focus on the effect
of the respondent's leadership behavior on either
the welfare and needs of other people, or on
achieving the overall mission of the organization.

5 COLLABORATION: The focus of these concerns is on
coordinating and cooperating with others in an
attempt to refine leadership skills and/or to
improve the quality or positive effects of
leadership or leadership development programs.

6 REFOCUSING: These concerns revolve around the
improvements needed in either the general
philosophy or the practices involved in
leadership or leadership training. They include
those factors that detract from the quality of
Air Force leadership training and those changes
that may enhance the understanding and performance
of our leaders.

Figure 3. Air Force Office PME Program Participants' Stages
of Concern About Leadership Skill Development.












instrument, the Stages of Concern for Exercise (SOCE) scale,
was designed to better understand the high rate of attrition
which occurs in these programs. It can also be used as an
example to illustrate the adaptability of the original model
to a wide range of uses.

Concerns Among Computer Users
John R. Grey (in Darlin, 1985), Chevron Corporation
president, required that the organization's top executives
learn about personal computers even though they might never
be required to use them. Grey wanted them to understand how
the people who use computers feel. However, executives who
have this type of concern over employees' feelings are still
a rarity (Darlin, 1985). Anxiety and apprehension apparently
are not uncommon emotions for some individuals when computing
is at issue. In addition, depersonalization, loss of privacy
and fear are often associated with the use of computers
(Meier, 1985). Oborne (1985) wrote about the societal
implications of the technology as well as the impact in the
workplace. Brod (1984) interviewed people who work with
computers in a variety of ways, including clerical workers,
chief executive officers, computer programmers, and others.
He consistently found that computers were having a striking
effect on their professional and personal lives. Not only
were there identifiable stress reactions such as fatigue and
headaches, they seemed to be internalizing the principles by












which the computer works: accelerated time, a need for
perfection, and yes-no thinking patterns.
There have been a number of attempts made to measure the
attitudes of people towards computing. Nickell and Pinto
(1986) created the Computer Attitude Scale (CAS), a 20-item
five-point Likert scale questionnaire, to measure positive
and negative attitudes about the use of computers. Fleischer
and Morell (1985) used interviews of managers to collect data
to assess the impact of the change brought about by office
automation. Sproull, Zubrow, and Kiesler used a 61-item
attitudinal questionnaire (later a short instrument was
developed) to better understand the socialization experience
involved in computing. Turkle (1984) and Caporael (1986)
chose to study the man/machine computing relationship by
investigating the anthropomorphism that frequently takes
place. Gallo (1986) suggested that expectancy theory could
be used as a predictor of behavior when confronted with
computer technology.
However, there is one characteristic that all of the
procedures and methodologies reported in these studies have
in common, and that is that they approached the subject of
computing from a very narrow perspective in comparison to
that which is found in the concerns-based model. Therefore
only one, or at most a few dimensions, of the dynamics that
occur in computing was investigated. The focus of most of
these studies was on the apprehension, anxiety, and fear that












beset some people when they are confronted with using a
computer. Clearly this is an important aspect of the
computing experience, but it is only one aspect. Once the
concerns about "self" are addressed other concerns increase
in intensity. Concerns theory provides the systematic
framework necessary to venture beyond the personal anxiety
level, and therefore advance the understanding of the change
process that is encountered with the use of a computer. None
of the other research was based upon such an encompassing
structure. Consequently their limited approach to the
multidimensional aspects of computing renders them
insufficient as a framework for this study.

There have been occasions when the original SoC
questionnaire was used with teachers when microcomputers were
to be introduced into their curriculum (Cicchelli & Baecher,
1985, 1987; Wedman, 1986; Whiteside & James, 1985/1986). The
strategy was to replace occurrences of the word "innovation"
with microcomputer when it appeared in the questionnaire.
However, following the same procedure for this study,
although considered, was deemed inappropriate since most of
the respondents would not be teachers or others acting in an
instructional capacity. In addition, the data gathered in
Phase I (described in Chapter III) of this study clearly
indicated that the original SoC was not adequate to capture
the concerns of a wide range of computer users.












A Pilot Study
Tentative confirmation of the foregoing conclusions was
made during the spring of 1987 when this investigator
conducted a qualitative study designed to uncover the methods

that computing students use to obtain help in a laboratory
setting. Numerous observations in a microcomputer lab were
conducted. During those observations the investigator
attempted to become as inconspicuous as possible by

ostensibly working at one of the computers in the rear of the
room. Although this method supplied some information, it was
not a sufficient method to understand the scope of what
actually transpired. A series of in-depth interviews were
then conducted in order to more clearly understand what was
taking place. Three students, two females and one male, who
had never before used a computer, volunteered to be

interviewed periodically during the semester and report on

their perceptions, as they learned to use the computer. Six
other students, three females and three males, with moderate
to substantial amounts of computing experience were also
interviewed. The interview methods and the subsequent

analysis of the data followed that outlined by Spradley

(1980). Spradley's procedures involved such undertakings as
participant observation, taxonomic analysis, and componential
analysis.

As a result of using this methodology, this investigator
was able to uncover certain patterns of behavior. The












initial statements of the inexperienced students were either
unrelated to their own personal use of the computer or were
ones expressing apprehensions about their adequacy in the use

of the technology. The more experienced students' statements
tended to focus primarily on the computing resources such as

hardware and software or on the possibly exorbitant amount of
time involved in completing a task. They also reported
having a certain amount of anxiety when they had to make
major system changes and at the thought of taking certain

computing courses that had a reputation for being
particularly demanding. Totally unaware of the
concerns-based model, this investigator used the terms self

and task to identify the primary types of statements made.
Experience was cited as a factor when behavior was

categorized (Martin, 1987). This study lends encouragement

to the hypothesis that the concerns model is an appropriate
one for assessing concerns about computing.

Much of the research (Allen, 1986; Fritz, 1985; Martin,
1987; Martin & Martin, 1988; Meier, 1985; Nickell & Pinto,
1986; Sanders, 1986; Schubert, 1986; Sproull, Zubrow, &

Kiesler, 1986; Wallace, 1986; Wheeler, 1986) reviewed during
this preliminary phase cited evidence that a gender
difference exists relative to the amount of self concerns

reported. Females appeared to have greater anxiety than did
their male counterparts. The ramifications of this condition
could be considerable, affecting curriculum choices, career












choices, and advancement in employment. However, the teacher
can be the key to a successful computing experience (Adams &
Fuchs, 1986; Fritz, 1985) and therefore may be a mitigating

force in minimizing negative affects of the innovation
through thoughtful interventions.

Summary
In this chapter, the three different perspectives--the
Social Interaction Model; the Research, Development and
Diffusion Model; and the Problem Solving Model--as tendered
by Havelock, were used as a framework to present
representative literature regarding change. Each model
presents a view of the process from disparate viewpoints.
Although there is some overlap, the basic assumptions among
the three approaches are distinctive. This was followed by a
review of the evolution of concerns theory as a model for
change. The model was traced from its conception in teacher
preparation programs to its application in various
educational settings, nationally and internationally, nursing

education programs, the health field, and leadership skill
development in the Air Force.

Literature in the area of emotional reactions related to
computer use was also surveyed in this chapter. There is
evidence in the researcher that a gender difference exists
relative to the amount of "self" concerns reported. However,
most of the studies focused on the distress encountered when
personal involvement with the technology is considered. The












majority of the research takes a narrow view of the concerns
associated with the use of the technology. Concentration is
on fear, apprehension, anxiety, and other discomforts
suffered as part of the computing experience. This single
dimension viewpoint is predominant in the related literature.
This investigator's practical experience in computing
education and a study using college students provides some
credence to the concept that concerns patterns are similar to
those patterns found in earlier concerns research. Thus, a
certain amount of support is evident to imply that concerns
theory may be applicable to the computing experience and,
therefore, this study was proposed to determine if the Stages
of Concern Model is an appropriate conceptual model to assess
concerns among computer users and to determine if an
instrument for measuring the concerns can be developed.

















CHAPTER III
RESEARCH DESIGN


Organization
Problems to be investigated in this study included
determining whether the Stages of Concern Model is an

appropriate framework to assess concerns among computer
users; determining whether an instrument for assessing Stages
of Concern among computer users can be developed within the

framework; determining if there is a relationship between the
level of computing education and the sequence of concern
stages; determining whether there is a relationship between

the intensities of various concern stages and the computing
application experience level of an individual; determining

the associations between age/gender and SoC scores; and

determining what effect a major system or software change has

upon a computer user's concern stage and concern sequence.

Because a priority in this investigation was to
determine if the Stages of Concern Model is an appropriate
conceptual model to assess concerns among computer users, the
research design used in this study parallels the procedures
that were used in the prior related studies (Barucky, 1984;












Hall, George, & Rutherford, 1979; Kolb, 1983). In this
chapter, the procedures followed in the three phases that
comprise this exploratory study are described. Phase I
encompassed the identification of concerns that people have
about computer use. Phase II included the construction and
administration of a 78-item preliminary questionnaire
designed to determine the stages of concern of people about
computers. A final 32-item questionnaire was developed as a
result of the analysis of the data collected using the
preliminary questionnaire. Phase III included the
administration of the 32-item final questionnaire and an
analysis of the data. Each of the instruments developed in
Phase II and Phase III also included the construction of a
cover page, with directions for completing the questionnaire,
and a demographic information page. In addition, this
chapter includes a discussion of the data analysis
methodology, a description of the method used for data

collection within the context of the three phases, and a
description of the study population.

Phase I: Identification of Computer Concerns
This investigator, without prior knowledge of concerns
theory, had found evidence to suggest that students had
concerns that clustered into certain categories--primarily
those of self and task (Martin, 1987). However, the data
obtained from the interviews and observations that were
conducted in that pilot study, as described in the prior












chapter, were not sufficient to provide the broad range of
items necessary to develop an instrument for assessing stages
of concern about computing. An investigation similar to
those carried out by Newlove and Hall (1976), Kolb (1983),
and Barucky (1984) would be necessary in order to obtain a
greater variety of computing concerns data and determine if
the responses supported or refuted the earlier findings.
An open-ended statement-of-concern form was developed
that followed the guidelines provided in A Manual for
Assessing Open-Ended Statements of Concern About an
Innovation (Newlove & Hall, 1976). In the fall of 1987 and
spring of 1988, these forms were distributed by the
investigator to individuals at two different higher education
institutions located in the same city. Because of the high
incidence of part-time older students at one of the
institutions, the investigator was able to collect data from

mature adults who possessed varying amounts of experience.
Having this population to draw upon, along with the
traditional 18 to 21 age group at the second institution,
provided a greater range of individuals than might be
obtained from a single institution of higher education.
Professionals in the computer centers at both institutions
also provided information through the open-ended statement of
concern form. Participants were informed that their
responses were both anonymous and voluntary. No attempt has
been made or will be made to match responses with
individuals.












Some modifications were made to the original open-ended
statement form after voluntary comments made by some of the

respondents, when queried about the clarity of the form, were

assessed. The majority of the changes were made in the
explanatory cover paragraph to more fully reflect the
computing emphasis of the instrument. Further modifications
were made to the demographic information sheet when it was
decided to focus attention on all computer users and not just
students. The open-ended concerns instrument (Appendix B)

used to collect data closely resembles those used by Newlove

and Hall (1976), Kolb (1983), and Barucky (1984).

There were 222 usable responses returned from this
stratified convenience sample. The data collected through
this method were used to test for corroboration of
information gathered through observation and in-depth

interviews and substantially augmented the findings of the

pilot study (Martin, 1987) that was described in the previous

chapter. Further, the data provided the basis for the

concerns statements that were included in the instrument
developed for this study (Phase II and Phase III).

Results from the 222 sets of concerns data were analyzed
according to the specifications in the Newlove and Hall
Manual (1976). The analysis process produced sufficient
confirmation to support the view that concerns about
computing correspond with the self, task, and impact stages












that were defined in prior concerns models. In addition,
there appears to be ample evidence that concerns about
computing appear to progress in a sequential pattern from
self to task to impact. Those individuals who have the

greatest and most varied amounts of experience tend to have
concerns that cluster more to the advanced stages of the
model.

Phase II: Discussion of 78-item Instrument
The development of a valid and reliable instrument for
assessing the stages of concerns of computer users is central
to this exploratory study. The following is an explanation

of the procedures that were used in the development of the
instrument. As noted in the previous section, these steps
closely mirror those followed in previous studies (Barucky,
1984; Hall, George, & Rutherford, 1979; Kolb, 1983).

Questionnaire Construction and Modification Process

The first step, as previously indicated, was to develop
an adequate reserve of concern statements regarding

computing. There must be a sufficient number of items which
represent a broad range of concerns on the questionnaire, so
that individuals with varying amounts of computing experience

will find choices that are relevant to their circumstance.
The data for the concerns statements were obtained in the
investigation (Phase I described in the previous section)

and, to a much lesser extent, the observation/interview pilot
study (Martin, 1987). As in prior concerns studies, the goal












was to collect as great a variety of concerns statements as
possible in order to capture examples of the many types of
concerns that may exist among computer users.

Therefore, the second step was to identify all of the
concerns statements that could be derived from the 222
responses to the open-ended concerns statement instrument.
As a result of this procedure, 156 distinct and varied
statements of concern about computing were extracted from the
responses. Modifications were made to some of the statements
after consultation with one of the members of the R&D Center
for Teacher Education who participated in the development of
the original concerns questionnaire. Many of the changes

were structural in nature. For example, items stated in the
negative were reworded to be positive. Also, future tense

was eliminated because the items were to represent current
concerns about computing.

However, the possibility of the existence of another
phenomenon relating to the stages of concern began to

materialize. A number of the statements collected suggested
the possibility that the consequence stage might be comprised

of two separate components. One of the components reflected

the traditional concerns about the effect the individual had
on others. The following are typical of consequence concerns
that center on others.

I am concerned about the effect I have on others in
their use of computers.












My goal is to provide assistance with computer-

related work that is beneficial to the recipient.

In prior SoC studies, consequence concerns also centered
on the individual's effect on students, patients,

subordinates, or some other person or group of people. As

noted, the same type of consequence concerns surfaced in this

study but certain related items seemed to imply that another
type of consequence might also be present. The following are

typical of consequence concerns that center on self rather
than others.

Learning how to use computers provides me with

many benefits.

It is very rewarding to me when my computer work
is successful.

Further substantiation of the possibility of a
consequence/self concept surfaced in interviews with

experienced computer users. A certain status seems to be

accorded "computer people." One student spoke of the way she

was perceived by other students on her dormitory floor. She

had a certain distinction because of her computing major.

She enjoyed being viewed in such high regard. Her body

language and tone of voice clearly indicated that she found

her status rewarding even though she had to work much harder

than some of the other students.

One computing professional reported that he was
frequently in demand because of his computing expertise. He












felt that he had a special status because of his occupation.
There were many examples of computing expertise enabling the
individual to have increased status in relation to others.

Several discussions took place with two of the members
of the R&D Center for Teacher Education and with other
experienced researchers who had knowledge of this study. The
decision was ultimately made to include the consequence/self
items in the preliminary questionnaire. The concept is
reflected in the composition of the stage definitions. This
change and other changes were made to the seven, which became
eight, stages of concern in order to reveal the differences
uncovered when computing was the focus of concern (see figure
4). This was an evolutionary task which continued throughout
the data analysis process.

The third step was to select concerns statements that
were to be used in the 78-item preliminary questionnaire.
Once the 156 usable statements were determined, they were
Q-sorted by five individuals who had varying amounts and
types of computing experience and varying amounts of
familiarity with the Stages of Concern model. The
individuals were asked to sort the statements into the stages
of computing concern. An additional category was included
which was composed of those items which the sorters could not
fit in any other category. Those items that had the highest
rate of agreement as to which stage they represented were
included in the preliminary computing concerns questionnaire.











0 CONTEXTUAL: Societal issues and "Orwellian" fears.
These concerns are not directly related to the
individual's use of a computer. They have to do with
the use of computers in society. Emphasis is on
negative economic impact, influence on children, health,
undue dependencies on computers, and deemphasis of the
individual and human values.

1 INFORMATIONAL: These concerns suggest an interest in
having more information about computers in general or
about a specific aspect of computing. Emphasis is on
concerns about learning how computers can be used and
how they function.
2 PERSONAL: These concerns focus on implications for the
individual. There is an uncertainty or anxiety about
the demands of computing and the adequacy of his/her
ability to meet those demands. Emphasis is on concerns
about oneself, personal status, and the opinions others
have about them in relation to computing.
3 MANAGEMENT: These concerns focus on time constraints,
limited or inadequate instructional material, help
resources, data integrity, availability of resources,
and the steps required to complete a computing task.
4/S CONSEQUENCE (Self): These concerns focus on the effect
the individual's expertise with computers has on
himself/herself. Emphasis is on the individual's
personal or professional benefits available as a result
of having computing know-how.
4/0 CONSEQUENCE (Others): These concerns focus on the
effect the individual's expertise with computers or a
particular aspect of computing has on other people.
Emphasis is on concerns about the quality of the
computer-related work produced by the individual and
its impact on people evaluating or using the output of
the effort.
5 COLLABORATION: These concerns relate to coordination
and cooperation with others or a particular application
of the technology in order to have increased positive
effects of use.
6 REFOCUSING: These concerns focus on the extension of
the benefits of computer use in a more universal way,
including the possibility of major changes and
alternatives in the use of the technology. The
individual has definite ideas about alternatives to
the proposed or existing use of computers or a
particular aspect of computing.
Figure 4. Stages of Concern About Computing.












Discussions with several of the individuals who participated
in the Q-sort procedure led to further editing and
clarification of the statements. Further lengthy
consultation was entered into with the member of the R&D
Center for Teacher Education. This iteration of the
questionnaire was administered to 22 individuals who were
asked to critique and offer suggestions about the clarity of
the directions of the instrument. Their suggestions were
incorporated into the instrument.
Initially there were 76 items on the questionnaire.
Later, as a result of a second member of the R&D Center on
Teacher Education reviewing the items, other minor
modifications were made. This R&D Center member also
suggested that additional items that indicated certain types
of concerns, and had also elicited substantial agreement by
the individuals who had conducted the Q-sort procedure, be
included in the instrument. Those changes were made and two
additional items were added to the preliminary questionnaire.
Additionally the second R&D Center member made several
recommendations regarding techniques that could be used in
the design of the demographic information page of the
questionnaire to improve the chances of obtaining the desired
data. One technique that was suggested was to use categories
to be checked rather than open-ended questions for such data
as age and amount of experience. These suggestions were also
incorporated in the product.












Again, a small volunteer sample was administered the
questionnaire. Comments about clarity were requested. This
sample indicated that the directions and items were generally
clear, and although there was some discussion of the

instrument, no substantive modifications were required.
Item Placement
Location of the items on the preliminary questionnaire
was made by using a table of random numbers (Borg & Gall,
1979). However, a decision was made to shift some items due
to the possible effect their placement on the questionnaire
might have. For example, "computers have a way of making me
feel incompetent" was moved from the first item position to a
place later on in the document. Also, when it was discovered
that many items representing the same stage of concern were
clustered together, some exchange of items placement took
place. Even so, most items remained in the position to which
they were relegated through the randomization process.
Administration of the 78-Item Questionnaire

The next task was to administer the 78-item preliminary
computing concerns questionnaire (see Appendix). The
preliminary questionnaire was administered to 323
participants who had varying amounts of computer experience.
The amount of experience ranged from none at all to
extensive. Each respondent ranked how well an item on the
questionnaire reflected his or her feelings about computing
according to a seven-point Likert scale. The ranges for each












item were 0 (irrelevant) to 7 (very true of me now),
depending on how accurately the item represented the
individual's feelings about computing.

Almost all of the subjects were contacted in a group
setting. However, some of the subjects were contacted
individually and asked to return the questionnaire by mail.
Each of the subjects was provided the questionnaire, a
demographic information page, and cover sheet with
directions. In those cases when the subjects were in a group
setting, a brief statement (see Appendix) was given outlining
the purpose of the questionnaire, the voluntary and

anonymous nature of the instrument, and directions for
completing the questionnaire. All of the foregoing
information was also included in the cover sheet. Approval
for the use of human subjects was granted by the
Institutional Review Boards of the participating institutions
prior to the administration of the questionnaire. Where

students were concerned, classes from various disciplines
were chosen for inclusion in the study in order to gain
responses from individuals who represented nonusers of
computers and representatives of the various categories of
computer users: professionals, intermediates, and novices.
This stratified convenience sample is identified further in a
later section of this chapter.












Analysis of the 78-Item Questionnaire
The next step in Phase II was the factor analysis of
responses to the questionnaire. "Factor analysis refers to a
variety of statistical techniques whose common objective is
to represent a set of variables in terms of a smaller number
of hypothetical variables" (Kim & Mueller, 1978, p. 9). The

method used was factor analysis using VARIMAX and rotation
procrustes toward the hypothesized structure. The factors in
this instance are the hypothesized Stages of Concern of

Computer Users (indicated earlier in Figure 4). The desired
results of factor analysis is to discover the underlying
relationships among the variables so that the original data
can be reduced to a smaller set of factors. The data were
rotated to see what factors the items loaded on and whether

those factors compared with hypothesized stages of prior

concerns studies. As a result of this procedure, the four

items that loaded highest on each factor in the hypothesized

structure were included on the final 32-item version of the
questionnaire.
An analysis of the demographic data resulted in further
modification of the demographic information page of the

questionnaire. All of the background questions with one
exception were designed so that the respondent could place a

check in the appropriate blank. However, there was such a
diversity of responses to the employment question that it was
determined that the development of reasonable categories of
jobs was impossible. Again, these changes were done in












consultation with an experienced member of the R&D Center for
Teacher Education.

Phase III: Discussion of 32-Item Instrument
The form of the final 32-item questionnaire (see
Appendix E) that was administered in this exploratory study
was similar in construction to its 78-item predecessor. The
instrument is comprised of a cover page, with directions for
completion of the questionnaire, the questionnaire itself,

and a demographic information page. The instrument was
completed by 388 subjects. The participant sample was

composed of nonusers of computers and individuals with
varying amounts of computer experience.
The items on the questionnaire were randomly ordered
using a two-stage process. First, the eight stages were

randomly ordered. Then the items within the stages were
randomly assigned within each of the eight stages using a

table of random numbers (Borg & Gall, 1979). This was done

so that no two of the same stage items would be adjacent.

This resulted in each group containing one item that is
representative of each of the eight stages of concerns about

the use of computers. The groups of items were then randomly
ordered and placed on the pages of the questionnaire. There
was some adjustment of items so that statements about
consequence/self, consequence/others, and collaboration were
dispersed to a greater degree than strict randomization
allowed. It was hoped that this would encourage the













respondent to consider each item on its own merits rather
than developing a response pattern.

One group of subjects was asked if they would be willing

to complete the instrument again three weeks later to

ascertain if the responses to the items were consistent over

time. They were told to write the last four digits of their

telephone number or their social security number on the top
of the form if they were willing to complete the instrument

again. All but two of the subjects included four digits on
their forms. A group of 56 subjects completed the instrument

a second time.

Data Analysis

The final step was the analysis of the responses on the

final version of the questionnaire. The objectives of the

data analysis were to determine whether:

1. computing is an area of change that can be used to

test the generalizability of the Stages of Concern framework;

2. an instrument for assessing Stages of Concern among

computer users be developed within the framework which is

both valid and reliable;

3. there is a relationship between the level of

computing education and experience and the sequence of

concern stages;

4. there is a relationship between the intensities of

various concern stages and the computing application
experience level of an individual;












5. there are associations between age/gender and SoC
scores;

6. there is an effect when a major system or software
change has upon a computer user's concern stage and concern
sequence.

In order to respond to these research questions,
analysis methods were used that had been effective in prior
concerns studies (Barucky, 1984; Hall & George, 1980; Kolb,

1983; Van den Berg & Vandenberghe, 1981). To answer the

first two questions, factor analysis with VARIMAX and

procrutes rotation toward the hypothesized structure was used
to assess the responses of the subjects to the items on the
final questionnaire. The factors were then rotated toward
the hypothesized Stages of Concern of Computer Users. The

resulting factor matrix was examined to ascertain whether a
correlation existed between the items which had the highest

loadings and the items on the questionnaire previously
selected as being representation of the particular

hypothesized stage of concern. Ideally, a list of items
showing highest VARIMAX loads would reveal four items showing

highest loads on each scale. Correlation analysis is used to
obtain a measure of the degree of association that exists
between two variables (Hamburg, 1974). The Pearson r
correlation was used to scrutinize the relationship between
the items on the instrument and the raw stage scores. In












addition, Pearson r was also used to explore the relationship
among all of the eight raw stage scores. Each item should

correlate higher with the other items on the same scale than
with items on other scales (George, 1977). This can be
determined by correlating each item with the sum of the other
items on the scale to which it has been assigned and also
with the sum of the other scales. Each item should have a

higher correlation with its own scale than any other. In

addition, the individual item correlations between items on
the same scale should be higher than item correlations
between items on different scales. A table was constructed
showing these within-scale and between-scale correlations.

These treatments, along with the Q-sort procedure
described earlier, are methods that can be used to determine

construct validity of an instrument. Construct validity is

the extent to which a particular instrument can be used to
measure a hypothetical construct (Borg & Gall, 1983).

Further exploration of construct validity was undertaken
when an assessment was made to determine if the hypothesized
stages of concern are sequentially progressive. In order to

carry out this task, data were categorized according to stage

of concern and then compared with the amount of experience of
the computer user. However, care had to be taken because
experience alone is not the key to SoC development.

Content validity, the degree to which the items on the
instrument represent the content that the instruments are












designed to measure (Borg & Gall), were demonstrated through
the methods used to select the items for the questionnaire.
Attention was also paid to the estimation of reliability
of the instrument in terms of degree of consistency among the
items on the instrument and the degree of consistency over
time. The method used was Cronbach's Coefficient Alpha, a
general form of the Kuder-Richardson 20 formula which can be
used when items are not scored dichotomously. "Formula 20 is
considered by many specialists in educational and
psychological measurement to be the most satisfactory method
of determining reliability. This formula is being used to an
increasing degree to determine the reliability of
standardized tests (Borg & Gall, 1983, p. 285). In the case
of an instrument with multiple response possibilities where
each item and each response is given a different weight,
alpha is the appropriate method for computing reliability.
The data collected from the test-retest component of Phase
III were also used to analyze the degree of consistency of
responses over time.
Questions having to do with the sequential pattern of
the Stages of Concern, the effect of experience with computer
applications, and age/gender considerations, can be addressed
using analysis of variance (ANOVA). ANOVA was used to
examine differences in mean raw stage scores between groups
of people with varying amounts of computer experience to
determine if there is a sequential relationship between












predominant Stages of Concern of Computer Users and the
amount of computing experience. The sequential nature of
concerns are inferred based on the cross-sectional pattern of
SoC scores associated with demographic items. Percentile
scores were used to compare groups graphically and to explore
predominant or peak stage scores. Hall, George, and
Rutherford (1979) indicated that the use of percentile
scores provide a method for examining the data on a scale of
relative intensity.

Additional analysis included a cluster analysis of
respondents based on the raw scale scores to investigate
predominant patterns of concerns. These groups are described
not only using the SoC profile but also by their responses on
demographic items.

Question six concerning the effect of change in the
computing environment was addressed using the responses to
the question about impending change on the demographic page
of the questionnaire. The responses were correlated with SoC
raw scale scores to see if users with similar age/gender/
experience have different concerns when a change occurs.

Study Population
In the administration of the open-ended statement of
concern form, the 78 and 32-item questionnaires, a stratified
convenience sample was used in order to gain a representation
of individuals in each of the three categories of computer
users--professionals, intermediates, and novices as well as












nonusers of computers. An advantage of stratified sampling
is that there is representation of the various subgroups of a
population (Ary, Jacobs, & Razavieh, 1985). The majority of
the representatives of the novice and intermediate categories
were obtained from the student populations of two higher
education institutions located in the same urban center.
One of the institutions had approximately 2,500 students
who tended to be younger and more traditional, while the
second institution having approximately 7,000 students had a
large number of non-traditional (over 21 years of age)
students who were employed and attended classes part-time.
Each campus offered multiple sections of a course designed to
acquaint the students with basic computing concepts and the
three major genres of software: word processing,
spreadsheet, and data base. The questionnaire was
administered to all students taking this course at both
institutions. Most of the students in these sections are
novices. Each institution also had students enrolled in
courses designed for individuals who are majoring or minoring
in one of the computing disciplines. The questionnaire was
administered in classes that were most likely to have
students who had taken a minimum of three prior computing
courses. Additional classes were selected for administration
of the questionnaire because those students were likely to
have taken a minimum of six prior computing classes. This
selection process was carried out in order to obtain a as












great a cross-sampling of the population as possible. Most
of the students in these courses were intermediates. In

addition, classes of students from outside the computing
departments of the institutions were asked to complete the
questionnaire in order to gain greater heterogeneity in the

sample. For example, nursing, education, business, and
several liberal arts classes were included. Some nonusers of
computers were in attendance in these classes, as well as
those individuals who have learned to use the technology
elsewhere.

Students, staff, administrators, and computing
professionals from a third higher education institution were

included in the sample of those administered the final
32-item questionnaire. This institution has approximately
13,000 students and serves a diverse population. Most of the

students who participated were enrolled in classes that

utilize computers. The classes ranged from introduction of

computing concepts to more advanced and sophisticated

computing applications.

Obtaining concerns data from computing professionals
occurred using various methods. In one of the institutions,

there were many members of the computing profession who took
undergraduate and graduate courses. Therefore, data from
some representatives of the professional community were
collected via the classroom administration. The
questionnaire was also administered to a group of computing
professionals attending a computing seminar.












Summary
The focus of this chapter has been a description of the
procedures followed in the three major phases of this
exploratory study. Phase I involved the identification and
collection methods used to derive the various computing
concern statements to be used in the questionnaire. People
with various levels of computing expertise were sampled in
order to generate as many distinct items as possible.
Phase II highlighted the modification of items and
construction of a 78-item preliminary questionnaire. A
description of the many steps involved in the preparation of
the instrument itself, the cover sheet, and the demographic
page were included. Special emphasis was placed on the
Q-sort procedure, placement of items on the questionnaire,
administration of the instrument, and the method used for
data analysis.
Phase III centered on the administration of the final
32-item questionnaire and the methods used to analyze the

data. Validity and reliability issues were also discussed in
this chapter. A final section was devoted to a description
of the stratified convenience sample used in the study.














CHAPTER IV
FINDINGS



Introduction

The data analysis described in Chapter III was

designed to determine whether the Stages of Concern Model

is an appropriate framework to assess concerns among

computer users; to determine whether a valid and reliable

instrument for assessing Stages of Concern among computer

users could be developed within the framework; to determine

whether the intensities of the concern stages are dependent

upon the amount of experience a person has with a specific

computer application; to determine whether there is a

sequential pattern to the Stages of Concern which can be

associated with computing experience or education; to

determine if concerns about using computers are associated

with age and/or gender of an individual; and to determine

what effects a major system or software change has on a

computer users's concerns profile.

Because the priorities in this exploratory study were

to determine if the Stages of Concern Model is an

appropriate conceptual model to assess concerns among

computer users and to determine whether a reliable












instrument for assessing the computing concerns could be

developed, two separate sets of data were collected through

the use of two computing concerns questionnaires. Several

analyses were completed for each set of data. First, an

analysis of the responses collected through the

administration of the 78-item preliminary questionnaire

during Phase II of the study was conducted in order to

select the best items for the 32-item final questionnaire

to be used in Phase III; further, it was necessary to

ascertain whether the demographic page of the questionnaire

was sufficient to capture the data necessary to address the

remaining study questions. The results from this analysis

have been reported to show the evolutionary process which

culminated in the development of the final instrument.

Second, an analysis of the responses collected through

the administration of the 32-item questionnaire was

undertaken to corroborate the appropriateness of the items

selected for the final questionnaire and to answer the

other questions under investigation. The analyses of this

last set of data were considerably more comprehensive than

the analyses that were undertaken on the data collected

using the 78-item questionnaire since the latter results

were used to address all of the study questions.












Analysis of Responses to 78-Item Questionnaire

questionnaire Respondent Demographics

As noted in Chapter III, the sample in Phase II of

this study consisted primarily of students at two higher

education institutions in the same city. The sampling

process yielded a return of 323 usable responses out of 347

questionnaires that were disseminated (approximately 93%).

The high rate of return was probably attributable to the

personal administration of the questionnaire to a majority

of the respondents. The composition of the respondents is

reported in several ways in order to describe the diversity

of the sample. Having a diverse sample was important

because of the nature of the study questions.

Approximately 46% or 150 of the respondents were female.

There were 172 males. One respondent did not indicate a

gender. Although there was a considerable range in the age

category, 173 (or almost 56%) respondents were between 20

and 29 years of age. Table 1 shows the complete age

distribution. In Tables 2, 3, 4, 5, 6, and 7 the results

reported indicate the range and quantity of computer

related courses taken and types and duration of software

experiences various subjects possessed. In addition to the

software reported specifically, 51 respondents

(approximately 16%) reported experience with other types of

software which ranged from accounting to weapons-control












applications. Slightly more than 8% of the respondents had

never taken a computer related course. Approximately 13%

had taken one course. Roughly 28% of the sample had

received a moderate amount of instruction (two to four

courses) and a final group (almost 42%) had been involved

in an extensive amount of formal training. An examination

of the software experience tables reveals that word

processing is by far the most popular application followed

by the use of computer games. Over half of the respondents

(169 or 52.8%) use a computer on the job and an almost

identical number (167 or 52.2%) employ one for personal

use.


Table 1

78-Item Computing Concerns Questionnaire Frequency,
Percent. Cumulative Freauency. Cumulative Percent of
Responses Categorized by Six Age Groupings



Cumulative Cumulative
Age Frequency Percent Frequency Percent


Under 20 26 8.0 26 8.0

20 to 29 173 53.6 199 61.6

30 to 39 87 26.9 286 88.5

40 to 49 32 9.9 318 98.5

50 to 59 3 0.9 321 99.4

60 or over 2 0.6 323 100.0










Table 2

78-Item Computing Concerns OuestionnaireFrequency, Percent.
Cumulative Frequency, and Cumulative Percent of Responses
Categorized by Number of Courses/Seminars/Workshops Taken



Cumulative Cumulative
Courses Frequency Percent Frequency Percent


None 27 8.4 27 8.4

One 69 21.4 96 29.7

Two 42 13.0 138 42.7

Three 34 10.5 172 53.3

Four 15 4.6 187 57.9

Five 19 5.9 206 63.8

Six 9 2.8 215 66.6

More than six 108 33.4 323 100.0



Table 3

78-Item Computing Concerns Questionnaire Frequency.
Percent. Cumulative Frequency, and Cumulative Percent of
Responses Categorized by Four Levels of Word Processing
Experience



Cumulative Cumulative
Experience Frequency Percent Frequency Percent


None 36 11.1 36 11.1

Less than 6 mos. 109 33.7 145 44.9

6 to 18 mos. 50 15.5 195 60.4

More than 18 mos. 128 39.6 323 100.0











Table 4

78-Item Computing Concerr
Percent. Cumulative Frecq
Responses Categorized by
Experience


s Questionnaire Freauency,
ency. and Cumulative Percent of
Four Levels of Spreadsheet


Cumulative Cumulative
Experience Frequency Percent Frequency Percent


None 102 31.6 102 31.6

Less than 6 mos. 129 39.9 231 71.5

6 to 18 mos. 35 10.8 266 82.4

More than 18 mos. 57 17.6 323 100.0


Table 5

78-Item Computing Concerns Questionnaire Frequency.
Percent. Cumulative Frequency, and Cumulative Percent of
Responses Categorized by Four Levels of Database Experience



Cumulative Cumulative
Experience Frequency Percent Frequency Percent


None 94 29.1 94 29.1

Less than 6 mos. 122 37.8 216 66.9

6 to 18 mos. 47 14.6 263 81.4

More than 18 mos. 57 17.6 323 100.0


1












Table 6

78-Item Computing Concerns Questionnaire Frequency.
Percent. Cumulative Frequency, and Cumulative Percent of
Responses Categorized by Five Levels of Graphics Experience



Cumulative Cumulative
Experience Frequency Percent Frequency Percent


Did not report 1 1

None 137 42.5 137 42.5

Less than 6 mos. 105 32.6 243 75.2

6 to 18 mos. 34 10.6 277 85.7

More than 18 mos. 46 14.3 323 100.0


Table 7

78-Item Computing Concerns Questionnaire Frequency,
Percent. Cumulative Frequency, and Cumulative Percent of
Responses Categorized by Five Levels of Game Experience



Cumulative Cumulative
Experience Frequency Percent Frequency Percent


Did not report 1 1

None 62 19.3 62 19.3

Less than 6 mos. 81 25.2 144 44.4

6 to 18 mos. 31 9.6 175 54.0

More than 18 mos. 148 46.0 323 100.0












Another indication of computer experience was the

ability to program in one or more programming languages

(table not included). BASIC was the most popular

programming language as evidenced by the 203 respondents

(63%) who reported experience with the language. COBOL was

the second most popular language (126 or 39% of

respondents), followed by Pascal (98 or 30% of

respondents). The sample also reported how subjects

viewed themselves regarding computing competency

experience. A distribution of those responses is provided

in Table 8. Clearly most of the respondents considered

themselves either novices or intermediates. In Table 9 the

amount of time that an individual reported having used a

computer is displayed. The data in that table seems to

support the data provided in the previous table.

There was also considerable diversity in the

occupational backgrounds of the participants. Although

many subjects classified themselves as students, many of

those were employed as well. A sample of the occupations

listed included, accountant, computer operator, programmer,

electrologist, financial consultant, housewife, librarian,

pharmacy technician, pilot, radiologist, salesman,

secretary, teacher, and systems analyst.













Table 8

78-Item Computing Concerns Questionnaire Frequency,
Percent, Cumulative Frequency, and Cumulative Percent of
Responses Categorized by Five Levels of Computer Experience



Levels of Cumulative Cumulative
Experience Frequency Percent Frequency Percent


Did not report 3 3

Nonuser 32 10.0 35 10.0

Novice 106 33.0 141 43.0

Intermediate 125 38.9 266 81.9

Old hand 57 17.8 323 100.0


The last category of demographic data collected

explored the frequency, types, and time frame of change in

the computing environment. More than half of the

respondents (174 or 54%) reported that a change in

hardware, software, and/or procedures was currently taking

place or would take place soon in their lives.












Table 9

78-Item Computing Concerns Questionnaire Frequency.
Percent. Cumulative Frequency. and Cumulative Percent of
Responses Categorized by Nine Intervals of Time of Computer
Use



Cumulative Cumulative
Time Frequency Percent Frequency Percent


Did not report 3 3

Never used one 15 4.7 1 4.7

0 to 3 mos. 34 10.6 52 15.3

4 to 6 mos. 21 6.6 73 21.9

6 to 12 mos. 13 4.1 86 25.9

1 to 2 years 52 16.3 138 42.2

2 to 5 years 98 30.6 236 72.8

5 to 8 years 48 16.0 284 87.8

More than 8 years 39 12.2 323 100.0


Item Selection for Final Questionnaire

In order to select the items that would be included on

the final questionnaire the responses of the 323 subjects

on the 78-item instrument were assessed using a variety of

statistical techniques, including factor analysis. SAS

factor analysis using a VARIMAX rotation followed by a












Procrustes rotation were used in the analysis of the

questionnaire responses to aid in the determination of

existence of the hypothetical stages of concern. Kim and

Mueller (1978) state that one objective of factor analysis

is the reduction of a set of variables to a smaller number

of hypothetical variables. They also state that factor

analysis is a means of assessing data for possible data

reduction. Borg and Gall (1983) note that factor analysis

is a valuable tool in educational research when

interrelationships and commonalities among a particular set

of variables are to be explored.

Two rotation methods were used on the data in this

phase of the study. The results of each rotation were

compared to determine which of the items best represented

the concerns of computer users. However, according to Kim

and Mueller (1978),

Our advice to the user is that one should not be
unduly concerned about the choice of the
particular rotation method. If identification of
the basic structuring of variables into
theoretically meaningful subdimensions is the
primary concern of the researcher, as is often
the case in exploratory factor analysis, almost
any readily available method of rotation will
do the job. Even the issue of whether factors
are correlated or not may not make much difference
in the exploratory stages of analysis. (p. 50

Information in the SAS User's Guide: Statistics also

provides the point of view that statistically all rotation

methods are equally good. Further, "If two rotations give












rise to different interpretations, those two

interpretations must not be regarded as conflicting.

Rather, they are two different ways of looking at the same

thing, two different points of view in the common-factor

space. Any conclusion that depends on one and only one

rotation being correct is invalid (1985, p. 338)".

Although the results of VARIMAX rotation alone were

helpful in the analysis process, the addition of the

Procrustes rotation method made the data interpretation

easier. One outcome of the Procrustes procedure was a

factor matrix which lists the factor loadings for each item

on each of the factors or hypothesized stages (Table 10).

The matrix illustrates which items were most representative

of (loaded highest on) each stage (See Appendix D for a

listing of items by hypothesized stage). The entries on

the table are the factor loadings and express the

correlation between the individual items and the factor.

Sample size and number of variables being examined are

determinants of the significance factor. If the sample

size is 50 or more, a factor loading of .30 is considered

to be significant at the .01 level. When the sample size

is at least 300, a factor loading of .15 is significant at

the .01 level. The level for accepting factor loadings as

being significant decreases as the number of variables












tested increases (Hair, Anderson, Tatham, & Grablowsky,

1979). If the items loading highest on a factor were

predominantly the items selected as representative of the

hypothesized stage, then there was additional support that

the stages identified correspond to the hypothesized stages

of concern.

The entries on Table 10 illustrate that some items

load higher on one factor and lower on all others. Those

items became candidates for inclusion on the final

questionnaire. Item numbers 5, 25, 28, and 49 are examples

of items that loaded higher on one factor (stage 0) and

lower on all other factors. However, some items, such as

60, were less likely to be included on a subsequent

questionnaire because there is too much homogeneity in the

loadings, and the item did not relate distinctly to any one

stage. Loadings followed by an asterisk indicate a

corresponding item selected to represent that stage on the

final 32-item questionnaire.

Table 11 consists of the items from the 78-item

questionnaire with their highest and second highest loading

on each factor. From this table one can ascertain whether

those items clustering with each factor (stage) tended to

have the highest loadings and whether they load lower on

other factors. Also, one can see whether those highest

loading items tended to be from the stages as they were













Table 10

78-Item Questionnaire Factor Loadings on Eight Stages of
Concern by Seventy Eight Item Numbers



Factor Loadings on Each Stage
Stage Stage Stage Stage Stage Stage Stage Stage
Item 0 1 2 3 4s 4o 5 6


1 3 27 11 2 2 -13 2 7
2 4 2 -17 11 4 29 15 32
3 5 6 6 28 31 12 7 0
4 8 3 5 5 -18 2 7 46*
5 55* 5 1 5 -11 -17 2 3
6 1 -11 23 19 2 -14 -20 28
7 3 37 23 -12 2 -15 3 1
8 5 4 10 28 43* 21 2 -12
9 7 33 3 2 12 9 0 10
10 8 19 1 29 40* 22 8 7
11 1 1 51 3 -10 -23 -14 5
12 4 11 47 18 23 2 -11 -10
13 27 4 9 -19 -10 2 2 5
14 4 7 8 20 2 17 -15 43
15 38 2 3 -16 -12 2 16 10
16 -14 51* 7 4 9 5 6 5
17 3 6 9 7 2 2 7 33
18 28 0 18 4 -13 4 3 22
19 9 2 -12 40 27 36 -11 13
20 -14 1 8 3 -23 2 14 53*
21 7 27 5 6 2 18 16 16
22 9 11 52 -16 8 2 0 5
23 1 2 11 12 22 34 10 4
24 -13 10 40 2 10 3 10 2
25 58* 8 3 2 4 -13 8 4
26 3 1 59* 4 8 9 14 8
27 1 -14 5 3 8 19* 29 38
28 53* 5 6 4 2 1 9 3
29 2 -10 12 30 45* 26 20 -13
30 53 0 2 12 2 1 9 3
31 12 1 8 46* 36 25 -14 6
32 2 10 -15 18 9 39* 9 25
33 0 3 7 17 27 39 15 3
34 12 26 5 7 6 11 12 9
35 2 3 53* 12 7 -27 -10 7
36 3 4 29 16 20 -19 2 2
37 4 7 44 26 17 -25 8 2
38 7 1 22 5 4 -16 6 27












Table 10--Continued



Factor Loadings on Each Stage
Stage Stage Stage Stage Stage Stage Stage Stage
Item 0 1 2 3 4s 40 5 6


39 3 8 2 4 5 14 29 20
40 41 7 7 10 16 7 5 8
41 1 9 7 3 -13 23 25 43*
42 3 56* 2 -20 -15 4 14 6
43 7 9 5 10 -11 32* 15 42
44 4 5 7 35* 20 31 0 17
45 7 3 -10 4 -12 12 26 40
46 5 2 53 9 3 4 2 7
47 4 7 35 2 11 38 24 6
48 5 49 1 5 13 17 15 -11
49 61* 4 5 7 8 5 0 4
50 30 6 4 24 4 20 6 9
51 1 0 5 4 -16 19 21 47*
52 2 1 3 1 10 23 39* 15
53 2 8 37 4 12 42 26 9
54 3 0 1 2 0 17 36* 27
55 3 6 6 31 48* 27 16 -16
56 8 4 14 42* 38 18 2 -11
57 12 48 6 8 1 19 4 4
58 9 5 -11 10 10 34 28 15
59 1 42* 3 7 6 9 0 3
60 4 22 19 3 -23 1 9 0
61 3 4 6 7 1 3 37 24
62 3 1 3 16 21 15 32 9
63 17 7 18 21 15 29 2 3
64 3 6 0 10 13 37* 33 7
65 4 23 7 5 7 -16 2 2
66 4 4 -11 8 6 2 31 16
67 4 6 4 13 28 18 24 2
68 3 4 11 18 36 10 19 6
69 3 2 0 3 3 22 39 19
70 14 15 33 2 -24 4 -10 5
71 3 -10 42 11 7 6 2 12
72 1 2 52* 18 1 -17 9 2
73 5 7 11 28* 5 17 4 23
74 2 55* 0 9 -20 16 6 8
75 5 30 8 5 15 8 24 1
76 3 13 5 3 9 15 37* 10
77 37 0 4 18 8 23 9 2
78 8 3 5 2 7 25 43* 14











hypothesized. If these conditions were present, there was

an indication that eight somewhat independent factors did

exist, and that they corresponded to the proposed stages.

A look at Table 11 reveals that all of the items had

loadings well above .15. In fact, 70 of the 78 items

(approximately 90%) had loadings above .30. With the

exception of stage 4o all of the eight factors identified

are comprised primarily of items representing a single

stage. In stage 0 for example, all ten of the items

originally projected as stage 0 items were identified as

stage 0 items. However, even in stage 4o there are more

projected 4o items than any other items. A total of 51 of

the 78 items (65%) loaded highest on the hypothesized

factor.

According to Kim and Mueller (1978), "the most

commonly used procedure for determining the number of

initial factors to be extracted is a rule-of-thumb--the

rule known as the Kaiser or eigenvalue criterion

(eigenvalue greater than or equal to 1)" (p. 49). In this

case, there were seven eigenvalues greater than 1 and the

eighth was .96, making eight a reasonable number of

factors.












Table 11

Results From 78-Item Questionnaire: Numbers Loading
Highest. Highest Loading Value. Next Highest Loading Value.
and Hypothesized Stage by Eight Stages Of Concern


Items Second
Loading Highest Highest Hypothesized
Highest Loading Loading Stage


Factor Associated With Stage 0

49 .61 .07 0
25 .58 .08 0
5 .55 .02 0
28 .53 .09 0
30 .53 .12 0
40 .41 .16 0
15 .38 .16 0
77 .37 .23 0
50 .30 .20 0
18 .28 .22 3
13 .27 .09 0


Factor Associated With Stage 1

42 .56 .14 1
74 .55 .16 1
16 .51 .09 1
48 .49 .17 1
57 .48 .19 1
59 .42 .07 1
7 .37 .23 1
9 .33 .12 5
75 .30 .24 4s
1 .27 .11 1
21 .27 .18 5
34 .26 .12 1
65 .23 .07 1












Table 11--Continued


Items Second
Loading Highest Highest Hypothesized
Highest Loading Loading Stage


Factor Associated With Stage 2

26 .59 .14 2
35 .53 .12 2
46 .53 .09 2
22 .52 .11 2
72 .52 .18 2
11 .51 .05 2
12 .47 .23 2
37 .44 .26 3
71 .42 .12 3
24 .40 .10 3
70 .33 .15 2
36 .29 .20 3




Factor Associated With Stage 3

31 .46 .36 3
56 .42 .38 3
19 .40 .36 4s
44 .35 .31 3
73 .28 .23 3



Factor Associated With Stage 4s (self)

55 .48 .31 4s
29 .45 .30 4s
8 .43 .28 4s
10 .40 .29 4s
68 .36 .19 3
3 .31 .28 40
67 .28 .24 40
60 .23 .22 2


u












Table 11--Continued


Items Second
Loading Highest Highest Hypothesized
Highest Loading Loading Stage


Factor Associated With Stage 4o (others)

53 .42 .26 2
32 .39 .25 40
33 .39 .27 40
47 .38 .24 2
64 .37 .33 4o
23 .34 .22 4s
58 .34 .28 5
63 .29 .21 2

Factor Associated With Stage 5

78 .43 .25 5
52 .39 .23 5
69 .39 .22 5
61 .37 .24 6
76 .37 .15 5
54 .36 .27 5
62 .32 .21 4s
66 .31 .16 6
39 .29 .20 6



Factor Associated With Stage 6

20 .53 .14 6
51 .47 .21 6
4 .46 .07 6
14 .43 .20 6
41 .43 .25 6
43 .42 .32 40
45 .40 .26 6
27 .38 .29 40
17 .33 .07 6
2 .32 .29 5
6 .28 .23 3
38 .27 .22 5













The consistency or "internal reliability" of the scale

scores representing each stage on the instrument was

assessed using SPSS program, RELIABILITY. The formula is a

generalization of the Kuder-Richardson Formula 20 for

dichotomous items. The "corrected item-total correlation"

and "total scale alpha if item deleted" for each item are

indicated in Table 12. Also included are the total alpha

coefficients of internal consistency for each of the eight

hypothesized stages. The corrected item-total correlation

is important, because if the correlation is high, there is

an indication that the item fits the scale well. When

examining the values of "total scale alpha if an item is

deleted", one is interested in the items with the lowest

values; those items should be included to maintain a high

total scale alpha. Item number five, for example, has a

"total scale alpha if item deleted" value of .7536 which

means that if the item was deleted the total scale alpha

for Stage 0 would drop from .7785 to .7536. The total scale

alphas are generally higher for scales with more items.

With five item scales the alpha should be .60 or better,

based on previous concerns questionnaires (George, A.,

personal communication). The items with asterisks were

selected to represent their corresponding stage on the 32-

item questionnaire.












Table 12

Reliability Analysis from 78-Item Questionnaire
Stage of Concern. Corrected Item-Total Correlation.
Total Scale Alpha if Item Deleted, and Total Scale Alpha by
Seventy Eight Items


Corrected Total Scale
Item- Alpha Total
Total If Item Scale
Item Stage Correlation Deleted Alpha


.5150
.2830
.3651
.5590
.4936
.5202
.4051
.6449
.3606
.4122
.3544
.4684
.5852
.4708
.6523
.5823
.5277
.5931
.3361
.5990
.4538
.5469
.4917
.6015
.5302
.6165
.2870
.3515
.2538
.3858
.4503
.6089
.3741
.3583
.4350
.4855
.4320


.7536
.7767
.7716
.7459
.7541
.7504
.7673
.7373
.7725
.7658
.8291
.8137
.8027
.8133
.7940
.8041
.8076
.8011
.8262
.8022
.8022
.7928
.7985
.7882
.7946
.7875
.8169
.8109
.8172
.8076
.8020
.7869
.7727
.7740
.7663
.7600
.7667


.7785









.8252











.8144










Table 12--Continued


Corrected Total Scale
Item- Alpha Total
Total If Item Scale
Item Stage Correlation Deleted Alpha


37 3 .4100 .7689
44* 3 .5280 .7551
56* 3 .5211 .7555
68 3 .4408 .7655
71 3 .3384 .7758
73* 3 .4547 .7637 .7827
10* 4s .6389 .8324
19 4s .5232 .8503
23 4s .4996 .8520
29* 4s .7059 .8219
55* 4s .6985 .8228
62 4s .6784 .8255
75 4s .6148 .8348 .8547
3 40 .4008 .8599
8* 4s .5212 .8481
27* 40 .6287 .8357
32* 40 .6772 .8299
33 40 .6986 .8275
43* 40 .5997 .8392
64* 40 .6598 .8316
67 40 .6090 .8380 .8564
2 5 .6758 .9100
9 5 .5477 .9168
21 5 .6848 .9095
38 5 .3461 .9259
52* 5 .8215 .9012
54* 5 .7766 .9042
58 5 .7496 .9059
69 5 .7248 .9071
76* 5 .7898 .9032
78* 5 .7846 .9034 .9174
4* 6 .4983 .8768
14 6 .5612 .8723
17 6 .5472 .8732
20* 6 .6737 .8639
39 6 .6626 .8648
41* 6 .6861 .8631
45 6 .6416 .8663
51* 6 .7109 .8608
61 6 .5588 .8724
66 6 .5353 .8741 .8804












The corrected item total for all of the items is

relatively high. Only four of the items' values are below

.50, and 65% were 60% or greater. However, of particular

note is the level of the total scale alpha. All of the

stages had high alphas. The alphas exceeded those

generated in the prior stages of concern questionnaire

development research. Of the eight stages, six of the

alphas were above .80 and the alpha for stage five was

.9174. The high alpha levels were a major consideration in

the decision to use four items per stage. The use of four

items was deemed particularly beneficial because eight

stages were to be represented on the final questionnaire.

Fewer items also helps to reduce intercorrelations between

scales.

A composite of the results of four different item

analysis methods that were conducted are displayed in Table

13. The table was developed to help in the item selection

process. Initially all 78 items were assessed during the

analysis process. However, an initial screening reduced

the number to five, six, or seven representative items from

each stage based on the number of items that appeared

promising. Table 13 represents a summary of the findings.

Included in the table are the items for each scale along

with their corresponding stage. This was followed by a

ranking for each of the analysis measures used. Each item




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 E907Q3FCC_KPBTS0 INGEST_TIME 2013-11-16T00:42:37Z PACKAGE AA00014226_00001
AGREEMENT_INFO ACCOUNT UF PROJECT UFDC
FILES