The study of the selection of local experts for the diffusion of complex technology cluster innovations within an organi...

MISSING IMAGE

Material Information

Title:
The study of the selection of local experts for the diffusion of complex technology cluster innovations within an organization
Physical Description:
x, 141 leaves : ill. ; 29 cm.
Language:
English
Creator:
Rice, Judith A
Publication Date:

Subjects

Subjects / Keywords:
Employees -- Training of -- Data processing   ( lcsh )
Office practice -- Automation   ( lcsh )
Occupational retraining   ( lcsh )
Instruction and Curriculum thesis Ph.D
Dissertations, Academic -- Instruction and Curriculum -- UF
Genre:
bibliography   ( marcgt )
non-fiction   ( marcgt )

Notes

Thesis:
Thesis (Ph. D.)--University of Florida, 1994.
Bibliography:
Includes bibliographical references (leaves 134-140).
Statement of Responsibility:
by Judith A. Rice.
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 - 002012617
oclc - 32528658
notis - AKJ9955
System ID:
AA00011818:00001


This item is only available as the following downloads:


Full Text










THE STUDY OF THE SELECTION OF LOCAL EXPERTS FOR THE
DIFFUSION OF COMPLEX TECHNOLOGY CLUSTER
INNOVATIONS WITHIN AN ORGANIZATION










By

JUDITH A. RICE


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

1994













ACKNOWLEDGMENTS


There are many people without whose help this work would

not have been completed. I would like to take this opportunity to

express my gratitude to each of them.

First, I would like to thank each member of my committee, Dr.

Mark Hale, Dr. Russell Robinson, and Dean David Smith, each of

whom shared their wisdom at appropriate moments, gave

encouragement freely, and provided a collegial environment for the

exploration of the ideas that led to the research reported herein.

To.Dr. Elroy Bolduc, the chairman of this committee and my

advisor and role model throughout my studies in educational

technology at the University of Florida, I give my admiration,

respect, and eternal gratitude. His thoughtful mentoring has

provided me with questions, and his example has offered food for

thought. It is because of his help, encouragement, trust, and

expectations that my course of studies has been completed.

There are several people whose companionship, humor,

wisdom, critiques, support, anxieties, and camaraderie were an

integral part of my studies at U.F. For the friendships cemented

with Sebastian, Hugh, and Bob, their families, and others who
ii








joined us from time to time, I am truly grateful. The many hours

spent in friendly debate, mutual support and encouragement, and,

most important, mirth-making, made this period of study both

fruitful and memorable.

This study would not have been possible without the help of

Kimball Kendall, who administered the treatment and reported

data, the local expert subjects, the administration of Santa Fe

Community College, my good friend Dr. Dan McKinnon, as well as the

many other friends and colleagues who supported and encouraged

me in completing this study. It is with appreciation for their

efforts that I share this success with them.

Finally, I offer profound gratitude to my children, Maura,

Deirdre, and Brendan, for their love and support. It was their

unfaltering confidence in my ultimate success that obligated me to

persist, though sometimes it seemed an impossible task. It is to

them that I dedicate this work.














TABLE OF CONTENTS

Page

ACKNOWLEDGMENTS .................................. .................................... i i

LIST O F TABLES ........................................................... ......................... v i

LIST O F FIGURES ........................................................................... vii

A BST RA C T .............................................................................. ........................ i x

CHAPTERS

I PURPOSE, THEORY, AND RESEARCH QUESTIONS........... 1

Introduction......................... .. ..... ......................... 1
Statement of the Problem............................................. 2
Significance of the Study............................................ 5
Purpose of the Study............................ ................... 10
Instrumentation................................................. 11
Outline of Procedures........................... ....................... 12
Lim itations........................................................... 14
Definition of Terms........................................................... 14
Sum mary..................................... .......................... 18
Organization of the Document.............................. ........... 18

II REVIEW OF RELATED LITERATURE.......................................... 20

Innovation Diffusion............................................................ 21
Organizational Innovation and Change Process............... 31
Sum m ary............................... ......................... ............................ 59

III THE LOCAL EXPERT INTERVENTION MODEL......................... 61

Introduction..................................................... 61
System............................ .......... ......................... 63
Local Expert Program.................................................... 65
Client W ork Groups............................. ......................... 67

iv








Technology Transfer...................... 69
Summary ................... .... 70

IV M ETHODOLOGY............................................................ 72

Local Expert Profile............................ ................ ................ 72
Development of the LEP Instrument................................. 74
Study Procedures........................... ............................ 84
Data-Gathering Procedures................................... ............ 87
Statistical Procedures........................ ................................. 90

V RESULTS........................................................................................ 92

Research Questions.......................... .......................... 93
Summary of the Results............................................................1 09

VI SUMMARY, DISCUSSION, AND CONCLUSIONS......................... 11 0

The Study................................................ ........... 1 10
Results and Discussion................................................. .............. 11 2
Conclusions........................... ..... ................ ....... ............ 11 8
Recommendations for Further Research.............................121

APPENDICES

A LOCAL EXPERT PROFILE QUESTIONNAIRE............................. 124

B LOCAL EXPERT LOG SHEET......................................................... 128

C LOCAL EXPERT INTERVIEW SHEET........................................... 129

D MULTIMEDIA MENTORS DATA..................................................... 131

E LOCAL EXPERT DATA.............................................................. 133

R EFERENC ES............................................................... ...................................... 13 4

BIOGRAPHICAL SKETCH........................................... ........................... 141















LIST OF TABLES


TablePage

4-1 Analysis of Variance for LEP as Predictor
for MMPR....................... 79

4-2 Summary of Fit for Regression Analysis of
M M PR/LEP........................................ ............................................ 8 0

4-3 Effect Test of Subscores as Predictors of MMPR............ 83

5-1 Multiple Regression Analysis Effect Test Results
For All Variables in the Model.............................. ........... 107

5-2 Correlation Matrix of All Variables Used in the
Multiple Regression Analyses............................................. 108















LIST OF FIGURES


Figure Page

2- 1 Innovation-Decision Model............................ ............ ........ 23

2-2 Distribution of adopter types over time when
adopting an innovation .......................... ......................... 25

2-3 Gramson's framework for typing innovation and
change research. The four frameworks represented
are defined by the characteristics noted in the
headings of the framework table ........................................ 34

2-4 S-shaped cumulative curve and bell-shaped
frequency curve for adopter distribution ...................... 45

2-5 Technology Transfer Model ...................................................... 55

3-1 Local Expert Intervention Mode..................................... .......... 62

4-1 Histogram of the distribution of the LEP scores for
the group of randomly selected faculty and music
workshop attendees ................................ .......................... 76

4-2 Comparison of three leverage plots: One indicates
a positive significant relationship; the second shows
no significance in the relationship, and the third
shows a negative significant relationship.................... 78

4-3 Leverage plot of the results from a multiple
regression analysis of LEP as a predictor of MMPR........ 81

4-4 Leverage plots of the results from a multiple
regression analysis of LEP subscores, each as a
predictor of MMPR .............................. ............................. 82









5-1 Leverage plots of the results from a multiple
regression analysis of LEP as a predictor of LEPR......... 96

5-2 Whole-model leverage plot of the results from a
multiple regression analysis of LEP and status are
predictors of LEPR ................................................................... 98

5-3 Leverage plots showing the comparison of the
results from the analysis of the bivariate LEP/LEPR
test and the whole-model test with local expert
status and LEP as predictors of LEPR ............................... 99

5-4 Leverage plots showing the results from a multiple
regression analysis of local expert status as a
predictor of LEPR ..................................................................... 100

5-5 Leverage plots showing the results from a multiple
regression analysis of client status as a predictor
of LEPR ............................................................................................ 10 3

5-6 Leverage plots of the results from a multiple
regression analysis for all effects as regressors
in the model .................................................... ... 106













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

THE STUDY OF THE SELECTION OF LOCAL EXPERTS FOR THE
DIFFUSION OF COMPLEX TECHNOLOGY CLUSTER
INNOVATIONS WITHIN AN ORGANIZATION

By
Judith A. Rice

April 1994

Chairman: Elroy J. Bolduc
Major Department: Instruction and Curriculum

In organizations, technology adoption is defined as acquisition

of hardware and software, and sometimes adoption is followed by an

initial training period with the new equipment. Seldom is on-going

intervention planned or provided, although it is quite well known

that computer users need time and repeated interventions to master

software applications.

This study proposes the local expert intervention model for the

planned diffusion of technologies that require multiple, successive

adoptions. The model is based on the premise that the early

adopters of technology, the local experts, help their colleagues and

coworkers and that the local expert in a work group can be identified

by measuring those characteristics found in the early adopters and








opinion leaders. A questionnaire, the Local Expert Profile (LEP), was

developed for the study to rate 28 selected volunteers according to

five characteristics found in early adopters in innovation diffusion

studies: (a) venturesomeness, (b) cosmopoliteness, (c) connection

to social system, (d) experience with the innovation, and (e)

perception of the innovation. Status in the organization was also

measured.

The purpose of this study was to determine the relationship

between the local expert selection process and, given the local

expert treatment, their effectiveness in helping their clients with

technology questions and problems. The local expert model was

implemented for a period of 6 months at Santa Fe Community College

in Gainesville, Florida. During the period of the study and at the

conclusion of the study, the local experts were rated on three

criteria: (a) time spent helping others with technology, (b) number

of calls that their clients placed to other help sources, and (c) the

computer center rating. Multiple regression analyses were

conducted to determine the relationship between the subjects' LEP

ratings and their performance ratings. The LEP was found to be

significant at the .01 confidence level in predicting performance,

and the variable status in the organization was found to be

significant at less than .01 confidence level as a negative predictor

of performance.













CHAPTER I
PURPOSE, THEORY, AND RESEARCH QUESTIONS
Introduction

The systemic changes brought about by the proliferation of
information technology in communications, the media, and the
financial and business institutions have virtually transformed our
society. The change is irrevocable and irreversible. A simple truth
has emerged from our modern metamorphosis: Change is here to
stay. Further, the pace at which change occurs is escalating at
prodigious rates; each technological advance promises
exponentially more capabilities, all at lower costs, placing
increasing demands on individuals who must then tend to the
change diffusion process. These changes are by no means seen by
all as positive. Fullan (1982) believes that for some change
involves loss and anxiety. Rogers (1962) quotes Barnett on the
topic: "The disgruntled, the maladjusted, the frustrated, or the
incompetent are pre-eminently the acceptors of culture
innovations and change" (p. 10). For others, change is viewed as
progress and technological change as improvement. As quoted by
Commager (1950), Lester Ward, one of the first social scientists to
study change, believed that the same planning approach used to
bring about change in science should be applied to the management
of human change.








Human institutions are not exempt from this all-
pervading spirit of improvement. They, too, are
artificial, conceived in the ingenious brain, and wrought
with mental skill born of inventive genius. The passion
for their improvement is of a piece with the impulse to
improve the plow or the steam engine. The
origination and distribution of knowledge can no longer
be left to change or to nature. They are to be
systematized and erected into true arts. (pp. 213-
214)

Whether change brought about by new technologies is seen as
benevolent or malevolent, understanding, facilitating, and managing
the adoption of these innovations is becoming a necessary
component in the support systems of our modern institutions. If
computerization can be seen as an instance of technological
innovation in organizations, then the problem must be defined
within the context of innovation diffusion and organizational
change processes.
Statement of the Problem
Problem
The local expert intervention model (described in detail in
Chapter III) has been proposed by this study for the planned
diffusion of complex technology clusters, which Rogers (1983)
defines as those technologies that require multiple, successive
adoptions. The basis for the model is simple: People tend to call
on local experts first. If employees in an organization that has
committed to the adoption of computer technology can get more and
better help from those people they tend to call on most often for
assistance, the local experts, then the diffusion will be more
efficient, and users will be more satisfied. The unique components
of this model are the selection and formalization of the use of









local experts as advocates for and helpers with the dissemination
of technology. The success of this intervention model depends upon
the accurate identification and enlistment of these local experts.
Organizational support, training, and evaluation are components of
the model implementation but are not the foci of this study.
Background
In general, models for the adoption and dissemination of
innovation focus on either the characteristics of the organization,
the nature of the impetus for change, the change process, or the
individual adopter. The variables described in the organizational
models are the characteristics of the organization which promote
innovation, and those studied in conflict models are the forces for
and against change. The planned change models prescribe a
relationship with a change agent, the methods for diagnosing the
problems of the organization, implementation intervention
strategies, and the system for maintaining an on-going relationship
with the innovation (Dill & Friedman, 1979). The innovation
diffusion models describe the individual adopter characteristics
with respect to their social status, personality, and
communications behaviors, as well as focus on adopter perceptions
of the innovation. Rather than focus on the organization or system,
these models deal with individuals and their differences with
respect to adoption of innovations.
Most of the research using any of these models culminates
with the adoption and initial implementation of the innovation;
sustained use of the innovation has rarely been studied.









Furthermore, the bulk of the innovation diffusion or planned change
research is based on simple process or product innovations that are
used intact to accomplish known, specific results; few studies
have been done on the diffusion of complex technology clusters.
Those models that do extend beyond implementation into the
institutionalization of change view the process as linear, and
institutionalization is defined simply as continued use of the
innovation.
In an article on the institutionalization of change, Fullan and
Miles (1992) state that "change is a journey--not a blueprint" (p.
749). Bennis (1966) suggests that change is a process, not an
event. But none of these models reflect the problems associated
with continuous learning and improvement or with participatory
adoption and implementation as is required with the adoption of
complex technology cluster innovations, and none deal with use of
the innovation at different levels of competency or efficiency.
A review of the literature on intervention strategies for the
diffusion of innovations reveals that there is a profound connection
between the perceptions or concerns of the individual adopter and
the effectiveness of the intervention (Hall & Loucks, 1978;
McCarthy, 1982). Individuals within an organization committed to
adoption of technology often determine the extent to which they
will "buy in" to a complex technology cluster. A critical element in
this decision process, according to McCarthy (1982), is the
individual adopter's knowledge about and recognition of the
personal need for the technology. Because individuals progress at









different rates toward the acceptance of change (Rogers, 1962) and
because they pass through many stages of concern as they are
learning about the innovation (Hall & Loucks, 1978), both the
content and the timing of interventions are critical for individuals
learning complex technologies.
According to Rogers (1983), opinion leaders influence the
adoption decision, positively or negatively, of their peers through a
communications network. When the organization favors the
adoption, the opinion leaders will be the early adopters of the
innovation. By communicating information about the innovation and
by demonstrating its use, these early adopters, or opinion leaders,
influence the later adopters with respect to the innovation. Thus,
the key elements for a positive decision on the part of the later
adopters, knowledge of the innovation and perception of its value,
are present. It is at this point that the early adopters are called
upon by their colleagues to help them learn to use the innovation.
This is especially true of computer technology innovations being
adopted by individuals within organizations. Bannon (1986)
identifies these people whom he calls "local experts" as an
extremely valuable resource within the organization. Bikson and
Eveland (1991) agree but allow that there is no easy way to
identify these people and no assurance that the ones who are
helping know enough to really be helpful.
Significance of the Study
Tight resources demand accountability. Technology is very
costly, and, at best, it is difficult to demonstrate cost









effectiveness by way of increased productivity with the
implementation of computer technology. There are really two
questions that must be addressed when evaluating the
implementation of computer technology. They are as follows:
1. Are we doing the thing right?
2. Are we doing the right thing?
The first question addresses using the tools efficiently and
competently, and the second addresses choosing the correct tools
for improving the outcome or the effectiveness of the task being
performed.
The inefficient or incorrect use of technology most certainly
causes many wasted hours and lost work. For instance, retyping a
document created on another computer rather than converting it to
a compatible text file format is a waste of time, yet it is a
common occurrence. Making changes to multiple spreadsheets in a
business system rather than linking them together so that changes
to one will be propagated to all is inefficient; it invites errors and
takes more time to handle. Using tab stops rather than the column
feature in a word processor renders the document unchangeable
without hours of manual reformatting. Failure to take advantage of
the print preview feature in most modern applications encourages
wasted time and much wasted paper. All of these are examples of
problems created by the inefficient use of the applications.
Selecting the correct tool or procedure for the task will
greatly increase effectiveness. For example, using a word
processor to collect data makes it impossible to manipulate or









retrieve the desired data in different formats or sequences, forcing
redundant work to be performed. While the word processor is an
effective tool for creating a document, it is ineffective for
creating a data base. Failure to make backup copies of work, a
relatively simple procedure, is known to have caused many lost
hours in the workplace.
Even more wasteful is the lack of vision for using technology
for its intended higher purposes. For example, many college
instructors who have access to computer technology view it as a
better typewriter or grade book, which it is, but fail to see its
potential as a portal to global information or as a medium for the
delivery of knowledge to students. Upper management is another
example of those who typically see the computer as a clerical
machine but whose work would be greatly enhanced by the use of
decision support software, spread sheet graphs, communications
software, and database manipulation. As Bikson and Eveland (1991)
have stated,

When organizations provide training, it is typically
for beginning-level use. Very little effort is directed
toward users continued learning, or even to providing
back-up assistance while they try to come up to speed
on new tools. Consequently, even high-level
professionals may remain amateur users of tools
closely implicated in the tasks for which they are
responsible. (pp. 245-246)

Sophisticated applications give users a distinct edge in the
competitive marketplace, regardless of the nature of the business
being conducted. Clearly, there is a need for an implementation









plan for more extensive training on these increasingly
sophisticated applications.
In most organizations, the technology adoption decision is
followed by planned activities involving acquisition and
installation of the hardware and software. Formal interventions,
such as seminars or workshops, are planned (if at all) to get
employees started. Then, from that time forward, they are
expected to get everything they need to know from books or
manuals, even though it is well documented that most people do not
read manuals to increase their knowledge of computer technology
(Carroll, 1985; Duffy, Curran, & Sass, 1983; Scharer, 1983). This
process is followed by high expectations from the organization's
leaders who have risked considerable resources to gain cost-saving
advantages that technology is believed to bring. Although it is well
known that there is a steep learning curve involved in becoming
proficient with computer applications, very little attention is paid
to the on-going interventions required to assist individuals in this
endeavor. Bikson and Eveland (1991) refer to the knowledge, skills
and technical resources necessary to take advantage of computer-
based tools as "humanware" (p. 245). They suggest that once the
hardware is acquired and the software is chosen, the development
of the humanware must be an on-going activity.
The pitfalls are obvious to anyone who has attempted to
integrate computers into any process: Inadequate investment in
humanware leads to inappropriate or inefficient use of computer
technology. This results in extra time and work required to









complete tasks, which adds rather than subtracts costs from
existing budgets. Since computer technology requires multiple 4-
adoptions over time with later adoptions building on the success of
earlier ones, any existing problems are magnified when more tasks
are targeted for computerization.
The magnitude of the anticipated change, the substantial
resources needed to introduce and infuse technology into an
organization, and the sizable commitment required from
individuals within the organization point to the need for a plan for
continuous learning. The plan must be a organizational change plan -
which includes an intervention and support system at the individual
level. If employees are to reach competence in the use of
technologies, then they must be provided with an environment that
encourages learning and nurtures those who are willing to help
others in the pipeline.
With microcomputer saturation being the eventual goal of
most organizations, providing this continuous support for
individuals presents another potential drain on resources for the
organization. Staffing an organization with enough technology
experts to provide continuous training and support at the desk top
is not an option for most.
The local expert intervention model suggests a solution to
this training dilemma. The proposed model builds on an existing
personnel resource, the local expert. Since the local expert
interactions occur to some degree whether or not they are
supported by the organization, any significant improvement in this









naturally occurring process will be extremely cost effective. By
enriching their technological expertise, the organization ensures
that the local experts will not only be in a better position to help
their colleagues, but will also convey to their peers a more
positive perception of the innovation. The fact that the local
experts are integrated into the working population of the
organization means that there is continuous desk-top support
available to those learning the new technologies.
The local expert intervention model embodies the qualities of
both efficiency and cost effectiveness and offers a solution to the
technology training dilemma that most organizations face. The
model is an "in-house" solution which has potential for longevity
and stability.
Purpose of Study
The purpose of this study is to determine the relationship
between the local expert selection process and, given the local
expert treatment, their effectiveness in helping their clients with
technology questions and problems.
Research Questions
This study is designed to answer the following questions:
1. Do individuals in the local expert program who score high
on the Local Expert Profile (LEP) provide more and better help to
their clients than those in the program who score low on the LEP?
2. Do local experts provide assistance to their clients
irrespective of the difference in their status in the organization?









In order to answer these questions, the following research
questions have been posed:
1. What is the relationship between the local expert profile
(LEP) and the performance rating (LEPR) of the local expert?
2. What is the relationship between the status of the local
expert and the LEPR?
3. What is the relationship between the status of the work
group clients and the LEPR?
Additionally, each of the variables which comprise the LEP
will be analyzed to determine the relationships between them and
the local experts' performance:
4. What is the relationship between prior experience with
the technology and the LEPR?
5. What is the relationship between venturesomeness and the
LEPR?
6. What is the relationship between cosmopoliteness and the
LEPR?
7. What is the relationship between perception of the
innovation and the LEPR?
8. What is the relationship between connection to the social
system and the LEPR?
Multiple regression analysis is the statistical procedure used
to address each of the questions above.
Instrumentation
The Local Expert Profile (LEP) (see Appendix A) was
constructed by the investigator and is based on similar surveys









from two studies (Hamilton & Thompson, 1992; Sachs, 1976), both
of which used the Rogers (1962) adopter profile to identify the
characteristics of early adopters of technology and innovators in
an educational setting. The questions used from these studies
were those which identified the characteristics that correlated
positively with innovativeness and opinion leadership in one or
both studies. The LEP was administered by the computer center
training coordinator and scored by the investigator.
The 25-question instrument is in multiple-choice format in
three sections. The characteristics surveyed by the questionnaire
are (a) status in the organization, (b) perception of the innovation,
(c) venturesomeness, (d) connection to the social system, (e)
cosmopoliteness, and (f) experience or expertise with the
technology. Status of the local expert is held as a separate
variable, with the other five comprising the profile (LEP).
Outline of Procedures
Subjects
The subjects participating in this investigation were staff,
faculty, professional specialists, technicians, and administrators
at Santa Fe Community College (SFCC). Twenty-eight (28) work
groups of no more than 15 and no fewer than 4 members each were
identified. A work group consisted of individuals who worked in
proximity with each other, regardless of their position in the
organization. A local expert subject was selected from each work
group. Some of the subjects were known to be early adopters of









computer technology; others were simply enlisted by their
supervisors to participate in the local expert program.
Treatment
The local expert intervention model was implemented by the
computer center training coordinator at SFCC where a new local
area network (LAN) and a new digital telephone system were
recently installed. The treatment consisted of (a) biweekly
meetings, (b) visitations from computer center trainer, (c)
computer center hot line support, and (d) advanced applications
workshops. The period of the study was 6 months, beginning in
April and ending in October, 1993.
Data Gatherina
The LEP was administered to all local experts at the first
meeting. It was scored and recorded by the investigator. All
clients' positions in the organization were coded and recorded with
the local expert identification number.
The local expert performance rating (LEPR) was derived from
three metrics: a local expert log of transactions recorded and
collected at 2-week intervals during the study (see Appendix B), a
computer center log of requests for help recorded during the last
month of the study using the electronic mail bulletin board
application, and a self-report of effectiveness and enthusiasm
recorded during individual interviews conducted by the training
coordinator with the local experts at the conclusion of the study
(see Appendix C).









Limitations
One limitation of this study is that the local expert
intervention model was implemented and studied in only one
organization. Although data were gathered from subjects in
another organization to validate the LEP instrument, the
individuals who were tested and rated did not get the same local
expert treatment as the subjects in the study.
Another limitation of the study is that the members of the
treatment groups were not selected randomly. Work groups were
determined by proximity, and subjects who acted as local experts
had to volunteer from those groups. Additionally, the subjects'
effectiveness may have been influenced by subtle differences in
the work groups with respect to technology access or preexisting
expertise.
A third limitation is the overall duration of the study. A
longer implementation of the model may have yielded more
dramatic or more conservative treatment effects and could have
reduced the possibility of Hawthorne effects.
Definition of Terms
The terms used in this study are adopted from the literature
and are defined below for the purposes of this study.
Innovation. The most widely used definition for innovation is
Everett Rogers' (1983): "an idea, practice, or object that is
perceived as new by an individual or other unit of adoption" (p. 11).
In this context, the innovation appears to be a single product or
process which can be adopted intact.









Innovation diffusion. This study uses this term extensively
in several contexts. Rogers (1983) defines innovation diffusion as
"the process by which an innovation is communicated through
certain channels over time among the members of a social system"
(p. 5). Dill and Friedman (1979) use the term to define the type of
research that Rogers does and differentiate between innovation
diffusion and other traditions in the change and innovation
literature. Both of these definitions are consistent with the way
in which the term is used in this study.
Complex technology cluster. This phrase is used by Rogers
(1983) as well as herein to describe a technology innovation which
requires multiple adoptions and multiple interventions in order to
diffuse it. The adoption of complex technology clusters requires
that each instance of adoption builds on the previous adoption. A
general purpose computer, such as a personal computer on the desk
of an employee, would qualify as a complex technology cluster
innovation.
Reinvention innovation. Rogers (1983) refers to an innovation
which is customized or configured by the adopter, rather than used
intact, as a reinvention innovation. Not only is the general purpose
computer itself considered to be a reinvention innovation, but the
applications typically used on a personal computer are also
reinvention innovations, as they are designed to be configured for
individual preferences.
Early adopter. The early adopter shows a propensity to
embrace new ideas earlier than most. On a normal distribution, the









early adopter falls between the first and second standard
deviations with respect to how soon they will adopt new
technologies. According to Rogers (1983), these individuals are
more venturesome, have more cosmopolitan information sources,
have a more positive perception of the innovation, have a higher
degree of connection to the social system, and have a higher status
in the organization than later adopters.
Opinion leader. The opinion leader is an individual within a
social system or organization who influences opinion in the
direction of the goals and biases of the organization. The opinion
leader is looked to by colleagues as a trusted person to emulate.
Local expert. Local expert is a term which defines a function
within an organization. The person who performs this function is
both the early adopter of technology and an opinion leader in
matters of technology. By definition, these local experts can only
exist in systems that have a protechnology stance.
Efficiency. The use of this term relates to the competent use
of technology. When a user exploits an application or feature to its
fullest potential and does so without errors and false starts, then
the application is being used efficiently.
Extensiveness. The use of this term relates to the extent of
the use of technology by an individual. If a client has a personal
computer but uses it only to retrieve e-mail, while hand writing a
document or using a calculator with notepad for figuring the
budget, then the technology is not being used extensively. The term









may also be used to indicate the numbers of individuals using a
particular application or technology.
Effectiveness. Effectiveness is defined by the success of the
outcome being pursued. If a technology application produces better
results or produces them faster, then the use of the technology was
effective in achieving the goal.
Connection to social system. This variable label connotes the
extent to which people know the other members of their social
systems (organizations), and, therefore, the likelihood that their
opinions will be sought when questions of adoption of innovation
arise.
Cosmopoliteness. This term is a variable label which Rogers
(1962) uses to describe the communications habits of individuals
who seek and value information from outside of their circle of
peers and colleagues. Examples of cosmopolite sources are media,
journals, conferences, and distant colleagues.
Experience with the innovation. This variable label is used to
indicate the extent to which individuals have used computer
technology and the degree of expertise that they see themselves as
having.
Perception of the innovation. This term is used to connote
the degree of positive or negative expectations that individuals
hold for technology.
Status. This term is used in this study to refer to the
position that the individual holds in the organization.









Venturesomeness. This term is a variable label which Rogers
(1962) uses to describe the propensity of an individual to embrace
change, take risks, and, therefore, adopt new ideas, products, and
procedures.
Summary
Chapter I provided a rationale for the proposal of the local
expert intervention model for the dissemination of complex
technology cluster innovations as well as a rationale for the study
of the characteristics of the local experts. The rationale includes
the statement of the problem, the background, the significance of
the study, a statement of the purpose of the study, and the
research questions to be answered. It also presents an outline of
the procedures used in the research, the description of the
instrument developed for the study, the limitations of the study,
and an operative definition of key terms.
Organization of the Document
Chapter II is a review of two traditions in the study of
change: (a) innovation diffusion theory with respect to individual
adopter characteristics and (b) organizational change and
innovation literature, including complex organizational change
theory, conflict change theory, planned change theory, and
organizational innovation diffusion theory. Chapter III describes
the local expert intervention model proposed herein.
Chapter IV describes the development and validation of the
LEP and elaborates on the procedures used to conduct the research.
It also provides a description of the statistical analyses used to






19


address the research questions. Chapter V presents an analysis of
the data, and Chapter VI summarizes the findings, discusses the
implications of the findings, and suggests questions for further
research.













CHAPTER II
REVIEW OF RELATED LITERATURE

This chapter will present a review of the literature which
forms the theoretical base for the model proposed herein and for
the study itself. The chapter is organized as follows:
1. innovation diffusion
a. decision model
b. adopter characteristics
c. opinion leadership
2. organizational innovation and change process
a. organization and conflict models
b. organizational innovation diffusion models
c. planned change models
Any study about innovation diffusion must begin with a
review of the contributions of Everett Rogers. As the dominant
theorist for this study and for the proposed model, Rogers'
extensive work on the characteristics of the adopter as well as his
work in organizational innovation and change is reviewed in the
first section. In the second section, a chronology of organizational
innovation and change theory is presented using a framework for
typing the theories according to research traditions.









Innovation Diffusion
The most definitive research done on the characteristics of
individuals in the innovation process was done by Rogers beginning
in the mid-1950s and continuing through the mid-1980s. The first
edition of Diffusion of Innovations (1962) included a synthesis of
some 400 publications on innovation diffusion and presented a
theoretical framework for the study of diffusion. In the second
edition, which he coauthored with Shoemaker, Communications of
Innovations: A Cross-Cultural Approach (1971), they analyzed over
1,500 studies. In his most recent edition of Diffusion of
Innovations, Rogers (1983) used meta-research techniques to
synthesize 3,085 empirical innovation diffusion studies. From
these works, he derived generalizations about the diffusion process
itself and about individuals involved in that process.
In an attempt to respond to criticism about his earlier
research which focused only on the individual as the adopter unit,
Rogers, in his 1983 edition, revised his theoretical framework to
include organizational innovation diffusion theory. This model, like
those of many others, does not effectively deal with the
implementation or intervention processes but, rather, focuses
solely on the decision to adopt. His organizational model
specifies organizational units as adopters and organizational
characteristics as determining factors in the adoption decision
process.









Decision Model
Rogers (1983) defines diffusion as "the process by which
(1) an innovation (2) is communicated through certain channels
(3) over time (4) among the members of a social system" (p. 10).
His proposed model (see Figure 2-1) depicts five stages of the
adoption-decision process which occur within communications
channels: knowledge, persuasion, decision, implementation, and
confirmation. The variables which influence the decision are (a)
prior conditions of the system state, (b) adopter characteristics,
and (c) perceptions of the innovation. Outcomes are specified as
adoption or rejection and allow for reversal of the decision process
through the implementation and confirmation phases.
The innovations in the Rogers (1983) analysis are treated as
intact, simple innovations--some process, most products. The
characteristics of the innovation are analyzed with respect to the
perception of them by the potential adopter. The five
characteristics are (a) relative advantage, (b) compatibility, (c)
complexity, (d) trialability, and (e) observability (pp. 15-16). As in
the other early change research, these studies do not examine the
characteristics of the diffusion of complex technology cluster
innovations, and Rogers acknowledges that as a weakness in the
current body of innovation diffusion research.
The communications channels in Rogers' (1983) model are
both interpersonal networks among adopters and mass media, such
as television, journals, and newspapers. Rogers looks at the
homophily and heterophily of participants in the communications


















C 0



if


I I i
8. -. 0.




I f u 4
CA

~ iihCu
g ila
I-~ Ca)









network; homophilous groups find it easier to communicate, but
heterophilous groups are more likely to communicate about an
innovation. For example, change agents are usually heterophilous
with the target group but are effective in communicating necessary
information about the innovation. Additionally, the media, also
seen as heterophilous with the group, is accessed by individuals in
the system to obtain information about the innovation.
Rogers' (1983) model has the element of time, not usually
seen as a variable in change models. His thesis is that innovations
which are viewed as appropriate will be adopted by all or most
individuals in the system, but the rate of adoption will vary
according to the characteristics of the adopter (see Figure 2-2).
Finally, Rogers' (1983) social system component is loosely
defined as "a set of interrelated units that are engaged in joint
problem solving to accomplish a common goal" (p. 24). This may be
an organization, a profession, a local society, or a global society.
The characteristics of the system are not examined in the
innovation diffusion research literature except to acknowledge
that diffusion takes place within its boundaries. The prior
conditions specify the variables that may be present.
Adopter Characteristics
Using innovation diffusion research primarily from
agriculture studies, Rogers (1962) classified individuals into five
adopter categories based upon "personal characteristics, salient
values, communications behaviors, and social relationships" (p.
185). The classifications are innovators, early adopters, early































Innovators
Innovators Early Early
Adopters Majority
2.5% 13.5% 34%
S- 2sd -sd


Late
Majority Laggards
34% 16%
i 7 + sd


Figure 2-2. Distribution of adopter types over time when adopting
an innovation (Rogers, 1983, p. 247).









majority, late majority, and laggards. If the innovation is
perceived as relevant to some task being performed by the culture's
participants, the distribution of adoption over time will be a
normal curve, with each adopter type representing one standard
deviation. The classifications indicate the time, relative to the
others in the group, that it takes for an individual to adopt an
innovation. Rogers emphasizes that adopters actually fall on a
continuum but that the characteristics which distinguish them
from each other create five discrete categories which may be used
in comparisons.
In his meta-analysis of the research, Rogers (1983)
developed generalizations about characteristics held by the five
adopter categories. Specifically, he defines 31 generalizations
about adopter characteristics which form the theoretical base for
the selection of the local experts in this study. He summarized
the observations about the variables under three headings: (a)
socioeconomic status, (b) personality variables, and (c)
communication behaviors.
The socioeconomic characteristics which correlate positively
with innovativeness in Rogers' overall analysis are education,
social status, literacy, upward social mobility, members of larger-
sized units, commercial orientation, favorable attitude toward
credit, and specialized operations (1983, pp. 251-252). The
personality characteristics which correlate positively with
innovativeness in Rogers' work are venturesomeness, positive
attitude toward change, less fatalism, greater achievement,









greater aspirations, less dogmatic, more able to deal with
abstractions, greater rationality, and greater intelligence (1983,
pp. 257-258). The communication behaviors that correlate
positively to innovativeness are greater social participation,
connection to the social system, cosmopoliteness, change agent
contact, mass media exposure, communications channel exposure,
greater information seeking about the innovation, greater
knowledge of the innovation, higher opinion leadership, and more
belonging to interconnected systems (1983, pp. 258-259).
Several researchers (Hamilton & Thompson, 1992; Sachs,
1976) have used Rogers' findings to identify characteristics of
innovators and early adopters among faculty at their institutions.
Because the research reported in Rogers' (1983) book occurs in
different societies and in many different fields, there were
characteristics which were meaningful in some of those studies
that would not be meaningful in the more homogenous populations
in Sachs' and Hamilton's institutions.
Sachs (1976) developed a questionnaire based on Rogers'
adopter profile to determine which of 18 variables were correlated
with innovativeness as practiced by faculty at Michigan State
University. He polled the subjects who took part in an innovator
support program called Educational Development Program. He also
polled a random sample of other faculty who had not taken part in
this program. Analyzing data from the questionnaires, he found
significant differences among innovators, early adopters, and
noninnovators with respect to innovativeness, integration with the









social system, cosmopoliteness, information seeking about the
innovation, and opinion leadership and was able to differentiate
further between institution-supported and unsupported innovators,
particularly with respect to opinion leadership.
Other adopter characteristics measured in his study were
age, social status, size of teaching load, specialization of teaching
responsibility, fatalism, social participation in the system,
knowledge of the innovations, information seeking about
instruction, and membership in integrated system. None of these
was found to be significantly different between the innovator and
noninnovator subjects of his study.
In their study of the characteristics of the early adopters of
a state-wide educational electronic communications network,
Hamilton and Thompson (1992) developed a questionnaire based on
11 characteristics from Rogers's adopter profile: (a) education
level, (b) social status, (c) social participation, (d) cosmopolitan
outlook, (e) mass media use, (f) personal communication, (g) degree
of innovation information seeking, (h) attitude toward change, (i)
attitude toward risk, (j) aspirations, and (k) attitude toward
fatalism. They administered the questionnaire to 35 subjects, all
educators at different levels and from different locations
throughout the state of Iowa. They also asked their subjects to
rate their perception of the innovation based on Rogers' five
categories which characterize the innovation. Their results
verified Rogers' claims; they showed that the subjects were
significantly similar in socioeconomic status, personality









characteristics, and communication behaviors. Their results also
showed a high rating on the perception of the innovation among the
early adopters both before and after the adoption process took
place. In conclusion, they recommended that implementers of
electronic networks begin by seeking out the early adopters,
thereby enhancing the diffusion process.
Bikson and Eveland (1991), in presenting their technology
transfer framework, identify two adopter attitudinal
characteristics that have been linked in recent studies to
implementation success: (a) diffusion status, which is a person or
group's conception of its role in diffusing a technological
innovation, and (b) change orientation, which is the extent to which
participants view the change as positive, problem-solving, and
achievable. The correlates in Rogers' characteristics are opinion
leadership and positive perception of the innovation, both of which
Bikson and Eveland find to be high predictors of implementation
success.
Opinion Leadership
Rogers (1983) also developed a composite profile of opinion
leaders based on their personality characteristics, their position
within the social system, and their communications behaviors.
According to Rogers and Shoemaker (1971) early adopters, more
than other adopter types, have "the greatest degree of opinion
leadership in most social systems" (p. 184). Since later adopters
look to the opinion leaders for expert advice about the innovation,
the position of the organization is a key influence in the rate of









adoption. When the organization or system supports the innovation,
the opinion leaders support the adoption. Likewise, when the
system is anti-innovation, the opinion leaders will support the
system value to reject the innovation.
Sachs' findings (1976) confirmed the value of social system
support in the rate of innovation diffusion. The innovative faculty
who were supported by their departments did not seek outside
support from the Education Development Program and were
considered to be opinion leaders in their departments. The faculty
who sought support from the Education Development Program did
not perceive their innovative work to be valued by their
departments or peers; they appeared to be working in relative
isolation. In a later publication, Sachs (1993) stresses that the
research must continue to focus on the role of the opinion leader in
the diffusion of complex innovations.
Both Hamilton and Thompson (1992) and Rogers (1983) found
higher social status to correlate positively with innovativeness,
early adoption, and opinion leadership. Sachs' (1976) subjects
were all of similar status, so his results found no significance
with respect to status. It is not clear whether higher status, as
measured by Rogers (1962), will predict more effective help from
the local experts, and, in fact, early observations of people helping
others with technology suggest no such correlations.
Bannon (1986) makes a strong case for the value of the early
adopters of technology as opinion leaders and help sources for their
peers. He points out that these individuals, when supported rather









than being penalized, offer on-going help and security for less-
skilled colleagues. He suggests that the characteristics of these
local experts that make them the chosen opinion leaders in their
peer groups are organizational rank, technical expertise,
sociability, reciprocity, accessibility, availability, organization
role, and shared experiences (with peers).
The interest of this investigator in the adopter profile as put
forth by Rogers (1983) is to identify those characteristics that can
be seen in the local expert. The variables chosen, perception of the
innovation, venturesomeness, connection to the social system,
cosmopoliteness, and experience or expertise with the technology,
are based upon the findings of Rogers as refined by Sachs (1976),
Hamilton and Thompson (1992), and Bikson and Eveland (1991).
Additionally, the relationship of the status in the organization of
both the local experts and the clients is of interest to the
investigator. Of those Rogers' generalizations (pp. 260-261)
chosen for the identification of the local experts, all have been
supported by at least 73% of the research studies that measured
those adopter variables.
Organizational Innovation and Change Process
In the discussion of innovation and change, it is important to
make a distinction between the two. Rogers' (1983) definition of
innovation is the most widely used: "an idea, a practice or object
that is perceived as new by the individual or other adopter unit" (p.
10). Lindquist (1978) defines change as "the modification of,
deletion of, or addition to attitudes and behaviors existing in a









person, group, organization, or larger system" (p. 1). Zaltman,
Duncan, and Holbek (1973) describe the difference between the two
terms: "The innovation, then, is the change object. Change, on the
other hand, is the alteration in the structure and functioning of a
social system. All innovations imply change. Not all change
involves innovations since not everything an organization adopts is
perceived as new" (p. 158).
Research in organizational change was formalized at the turn
of the century in the early days of social science. As reported in
"The Roots of Change" (Bennis, Benne, & Chin, 1961) there were two
schools of thought that emerged: the "automatic adjustment"
school that was based on the law of nonintervention and the
"planners" school that was based on the law of radical intervention

(p. 7).
The noninterventionists, who were an extension of the "social
atomism" philosophy dating back to 1800 (Stanley, 1961),
relegated themselves to observer roles. They sought to understand
the system by documenting its response or reaction to the change
process. An early proponent of this school of thought was William
Sumner. As quoted by Bennis, Benne, and Chin (1961, pp. 8-9) from
an earlier work by Henry Commager (1950), Sumner's advice to his
fellow sociologists was simply to live and let live. The automatic
adjustment school was largely theoretical rather than practical.
The planners believed that the knowledge gained from the
study of change should be used to develop models for effecting
change in organizations and systems. An early proponent of this









school of thought was Lester Ward, also quoted by Bennis, Benne,
and Chin (1961, p. 8) from Commager (1950). Ward believed that
social scientists should use the scientific method to gain social
knowledge with which to induce change, also viewed as progress, in
individuals and systems.
The automatic adjustment theory has long since given way to
planning. In his essay on the mid-twentieth-century collapse of
the noninterventionist theory, Stanley (1961) argues that if social
scientists do not develop models for intervention, other forces,
both good and evil, will step in to fill the void. Because of the
profound effects of the technological revolution in modern
industry, the idea of planning for growth and change is no longer
seen as academic, rather it is viewed as necessary and beneficial
(Counts, 1961).
The review of organizational change and innovation literature
revealed a plethora of theories and models. It also revealed a lack
of empirical research to test the theories (Dill & Friedman, 1979).
The interest of this researcher was to trace the evolution of
planned change and innovation diffusion theories in order to
identify the components of the models that should form the theory
base for a new model that can be applied to the adoption of complex
technology cluster innovations.
This review included a search for a typology for placing the
traditional models into an historical and theoretical perspective.
Gamson's Typology (see Figure 2-3) as explained by Dill and
Friedman (1979), was chosen because it allows the grouping of





















Prescriptive/ Loci of
Common Unit Descriptive Dependent Impetus
Frameworks of Analysis in Intent Variable for Change


Complex Organization Descriptive Rate of New Not Considered
Organization Technology
Addition

Conflict Power Blocs Descriptive Natural Histor Power
of the Differentials
Organization

Diffusion Indivdiuals or Descriptive Trajectory of Communications
Adopting Units Innovation Channels and
Change Agents

Planned Social or Prescriptive Effectiveness System Leaders
Change Work Group of Intervention Facilitated by
Strategies Consultants


Figure 2-3. Gamson's framework for typing innovation and change
research. The four frameworks represented are
defined by the characteristics noted in the headings of
the framework table (Dill & Friedman, 1979, p. 423).









models into frameworks for comparing theory and research
methodologies. Gamson uses four frameworks to categorize the
traditional theories on organizational change. The four frameworks
are (a) complex organization, (b) conflict, (c) innovation diffusion,
and (d) planned change. The frameworks group theories based on
the unit of analysis, the prescriptive or descriptive nature of the
model, the locus of impetus for change, the dependent variable to
be studied in researching the model, and one not shown in Figure
2-3, the type of research typically conducted. The typology was
constructed for Gamson's classes so his students could compare
strategies for conducting empirical research on the nature of the
relationships between the variables. Dill and Friedman (1979)
point out that Gamson uses the predominant characteristics of the
change model to type them, recognizing that none of the models are
purely one type or the other. In fact, the more contemporary the
model, the more likely there is overlap of characteristics.
These four parallel traditions have persisted into modern
innovation and change literature: the prescriptive models for
planned change on the one hand, and the descriptive complex
organization, conflict, and innovation diffusion theories on the
other. In developing an intervention model for technology
diffusion, the need for understanding the individual's role, the
role of the organization in the adoption of computer technology,
and factors that pose resistance to implementation of technology
led the researcher to review models from each of these
frameworks.









Complex Organization and Conflict Models
The complex organization model is descriptive, attributing
innovativeness in organizations to variables that characterize the
system. The goal of the theorists who propose complex
organization models is to identify those characteristics so that
organizational innovativeness can be better understood. The most
common variables analyzed in models of this type are size,
complexity, decision style, diversity of jobs, and age of company.
The researchers in complex organization theory conduct ex post
facto correlational studies to validate identified organizational
characteristics that are present in innovative organizations. The
practitioner, usually an external change agent, uses the knowledge
gained from an analysis of the particular organization in order to
initiate and manage the change process.
Typically, complex organization theories do not model the
role of the individual adopter in the change process nor do they
consider the nature of the change process itself. Also, neither the
characteristics nor the perception of the innovation are treated as
variables. Moreover, the state of change in an organization is
considered to be a variance from the norm, with a return to
stability marking the completion of the innovative period.
The conflict framework is similar to the complex
organization model except that the primary emphasis is on the
nature of conflict and change within the organization. These are
also seen as political or problem-solving models that view an
innovation as a potential solution to conflict. Typically, the change









process itself and the characteristics of the innovation are not
treated as variables. The research is descriptive, primarily case
studies or longitudinal studies. The goal of the researcher is to
understand the nature of resistance to change, and the practitioner,
a change agent, seeks to deal with these factors in promoting the
innovation process.
Hage and Aiken's (1970) proposed model is a good example of
an early complex organization model. They identify seven
organizational characteristics, four structural and three
functional, that influence the rate of adoption of innovation. The
structural variables are complexity, centralization, formalization,
and stratification, and the functional variables are production,
efficiency, and job satisfaction. Their research findings indicate
that complexity and job satisfaction correlated positively with
innovativeness, and the other variables correlated negatively.
Although their model includes an analysis of the stages of the
innovation process, they do not examine differences in the
correlations of these variables from one stage to another.
Wilson's (1966) theory focuses on the levels of complexity
and diversity of the organization as being the chief factors in the
initiation and implementation success of innovation. He points out
the dilemma of characteristics which correlate positively for the
initiation stage and negatively for the implementation stage.
Organizational complexity, for example, increases the likelihood of
innovation proposals, but decreases the likelihood of successful
implementation. Wilson also introduces the idea of economy of









incentives (p. 196). He argues that the real cost of innovativeness
is in the adding or redistribution of incentives.
Burns and Stalker (1961, pp. 119-122) propose an
organizational conflict model, which includes the characterization
of the organization as mechanistic (stable) or organic (unstable).
The state of the organization determines the relative difficulty or
ease in implementing an innovation, and it dictates the activities
that will be necessary to implement the change. Their model does
not deal with stages of the innovation process.
Harvey and Mills (1970, p. 182) propose a conflict model in
which internal and contextual organizational variables affect the
organization's ability to impose particular solutions on particular
problems. They suggest that organizations are political systems
and that dissatisfaction with existing solutions causes a problem
which promotes a search for new alternatives. Innovations may be
the solutions to those problems. They elaborate on Burns and
Stalker's idea of mechanistic and organic categories in typing
solutions. Their model does not examine the stages of innovation
nor does it deal with structural characteristics of the organization
as they affect innovation.
Zaltman, Duncan, and Holbek (1973) expand on these prior
theories by integrating the organizational and conflict variables
into the stages of innovation and by more fully developing the
innovation process component. The stages of innovation are
categorized as initiation and implementation (pp. 58-62). The
substages of initiation are knowledge-awareness, formation of









attitudes, and decision. The substages of implementation are
initial implementation and continued-sustained implementation. In
their model, they examine the characteristics of the organization
and of the mediators as those variables interact with the initiation
and implementation processes. The organizational characteristics
that they examine are complexity, formalization, centralization,
interpersonal relations, and dealing with conflict. They elaborate
on Wilson's theory which states that characteristics which are
favorable for initiation may be unfavorable for successful
implementation.
Borrowing heavily from Rogers and Shoemaker (1971),
Zaltman et al. (1973, pp. 10-50), acknowledge the importance of
the attributes of the innovation and the perception of the
innovation itself. They listed a series of attributes, such as cost,
communicability, and so on, and identified the substages at which
those characteristics are relevant. They also incorporate the
conflict model theories of Havelock (1973) in their resistance-
adoption model (Zaltman et al., 1973, p. 95). They identify the
source for resistance at the various substages and suggest both
social system and individual resistance processes at work during
the phases of innovation adoption.
In formulating their theory, Zaltman et al. (1973) synthesized
and extended the work of many others who came before them. They
incorporated aspects of complex organization and conflict models
and adopted ideas from innovation diffusion theory. Their work









follows the trend toward increasing complexity and a focus on the
internal aspects of the organization.
In a departure from this trend to focus more and more on the
internal structures and on the individual, Rosabeth Kanter (1983)
argues for a complex macro organization model. She posits that
organizational design, goals, hierarchical relationships, and
interorganizational relationships are the critical elements that
define the context for technology use. She pays considerable
attention to the innovation type, product or process, and
acknowledges past research in organizational characteristics
which generally hold true for innovativeness. Her chief
contribution to the literature is that she recognizes that
technology innovations change the nature of innovation process.
The reasons she states are that technology innovation is uncertain,
is knowledge-intensive, may be controversial, crosses boundaries,
changes work relationships and hierarchies, and that conditions for
adoption are different from conditions for invention (1983, pp. 17-
19).
Change theorists who propose conflict or problem-solving
models are Havelock (1970) and Lindquist (1978). Havelock (1973,
p. 18) describes the relationships between the various players, who
he calls resources, linkers, and clients, in the organizational
change process. Havelock (1970) sees the organization as a
complex system, with individuals and groups having power to block
or facilitate the flow of information about the innovation. In his
guide for change agents (1973, p. 7), Havelock describes the









innovation process as being initiated with a "disturbance" or a
problem, followed by a decision to do something and a diagnosis of
the problem, a search for solutions, a selection of a solution, and
satisfaction or dissatisfaction with the application of the solution.
He describes the process as cyclical, with each iteration
addressing a different problem or the same problem from a
differing perspective.
Lindquist (1978) studied innovation and change in seven
different colleges over a period of 4 years to gather data before
formulating his "adaptive development" model of innovation. He
elaborates on Havelock's model proposing five characteristics
which are key to successful implementation of innovation: linkage,
openness, leadership, ownership, and rewards. His adaptive
development model is cyclical, like Havelock's, and proposes
research and development and needs assessment components, each
connected to linkage, all of which lead to open development and
decision and then to supported implementation which then cycles
back to research and development and needs assessment.
Both conflict models and complex organization models
provide much needed insights into organizational conflict,
resistance to change, interrelationships among groups, and the
roles of individuals in the change process. It also provides
information about organizational characteristics, such as size, age,
stability, and diversity which tend to nurture or stunt innovative
activities.









The weakness of these theories is that they do not fully
address the importance of the attributes of the innovation itself
nor do they examine the characteristics of individual adopters or
their perceptions of the innovation. Further, the assumption in all
of these models is that the change process is a disruption of the
norm and that the driving objective is always to return from the
change state to stasis. Finally, the research is of little use in
orchestrating change; it is either ex post facto or case studies and
is entirely descriptive. Organizational characteristics, rather than
the change process or the individual, are the dependent variables.
A notable exception to the above criticisms of complex
organization theory is Clark and Staunton's (1989) theory of
innovation and organization. In discussing the differences between
traditional organizational models and the new technology models,
they point out that "many studies have imposed an objectification
on innovations so that an innovation is treated as a 'thing' which is
detached from its contexts and pathways" (p. 8). They suggest that
the nature of the innovation is a key factor in the discussion of
innovation diffusion. They argue that "there is a strong tendency to
equate innovations with equipment and to neglect the knowledge
which is embodied in other dimensions" (p. 8). They propose a
typology that characterizes the innovation event as being on a
continuum from radical to incremental rather than as being a
dichotomous occurrence.
They also raise an objection with the traditional literature
as to the characterization of the innovation being induced by









change agents. They believe that additional attention should be
given to the unintended or reinvented innovations initiated entirely
by individuals from within the organization.
Another difference between the early organizational
innovation models and the newer technology transfer models is the
view of the organization. Clark and Staunton (1989) argue that the
view of the "innovation as a detour which will be followed by a
return to normalcy" (p. 10) is a viewpoint that is no longer, if ever
it was, operative. Rather, they see continuous improvement as the
operative change plan. Bikson and Eveland (1991) agree that the
view of successful innovation diffusion being followed by
organizational stasis is no longer a valid paradigm. They argue
that "there is no end to implementation processes while the
technological state of the art is rapidly advancing" (p. 231).
Another very important observation that Clark and Staunton (1989)
made about organizations is that the impetus for change has been
seen in the past as arising from external or internal conflict or
need, whereas now the market presents technology opportunities
which establish need for change.
Innovation Diffusion
Innovation diffusion theory's roots can be traced to the turn
of the century European father of sociology, Gabriel Tarde. In his
1903 work, The Laws of Imitation, he explored the reasons why
some new ideas are adopted while others are not. He recognized
that the cumulative rate of adoption for an invention was an s-
shaped curve over time, and he made some observations about the









spread of innovations within a social setting which have been borne
out in later research (see Figure 2-4). Interest in diffusion
research was rekindled in the United States in the early 1940s
when Ryan and Gross, as reported by Everett Rogers (1983),
published a model for diffusion research in their study on hybrid
corn. Most of the research conducted between 1943 and 1962 was
likewise in the field of agriculture, although, as Rogers (1983, p.
46) emphasizes, diffusion research is by definition a type of
communications research because new ideas are adopted only after
those ideas have been communicated to other individuals.
The work of Everett Rogers, discussed in the first section of
this chapter, spans both organizational innovation diffusion and
characteristics of the individual adopter. His work on the
individual adopter is the seminal research which forms the theory
base for the proposed model and for this study. Many contemporary
diffusionists, whether in planned change, organizational or conflict
theory, or innovation diffusion, rely heavily on the results of his
research.
One such individual is Gene Hall whose work in the field of
organizational innovation and change spans from 1973 to the
present. Hall, Dossett, and Wallace (1973) introduced the Concerns
Based Adoption Model (CBAM), a three-pronged model derived from
empirical research on innovation diffusion in educational
institutions. The CBAM treats the concerns of the individual in
learning to use the technological innovation as central to the
adoption and implementation processes.




















100%

90% -

80% -

70% -
/ Cumulative
/ S-shaped curve
60% -

50% /

40% -

S%\ -Bell-shaped
30% / \ frequency curve





0 /-I "
Time


Figure 2-4. S-shaped cumulative curve and bell-shaped frequency
curve for adopter distribution (Rogers, 1983, p. 243).









The CBAM suggests that there is a sequence of seven Stages
of Concern (SoC) through which individuals move: awareness,
informational, personal, management, consequence, collaboration
and refocusing (Hall et al., 1973). Hall depicts innovation adoption
and implementation as a process which is facilitated by trained
adoption agents who customize their interventions on the basis of
the assessed personal needs and motivations of the individual
adopters. By being sensitive to the concerns of users and by seeing
the use of an innovation as a developmental process, change agents
are expected to be able to reduce the threat which change poses to
individuals. Levels of intervention corresponding to these stages
of concern, with specific tactics and strategies, are proposed in a
later report (Hall & Rutherford, 1983).
The other two components of CBAM measure the Level of Use
(LoU) of the innovation and the Innovation Configuration (IC). The
level of use instrument developed for CBAM measures the behaviors
associated with the use of the innovation (Hall, Loucks, Rutherford,
& Newlove, 1975). This component is tied to the Stages of Concern
with the assertion that since the adoption of innovation is a
developmental phenomenon based on the characteristics and
concerns of individuals, there will be different levels of
competence with the innovation. Furthermore, they address the
cycles of use, or repetitions and refinements, that are inherent in
the adoption of complex innovation. In a report (Hall, Loucks,
Rutherford, & Newlove, 1975) of research conducted on the levels
of use of innovations, two preuse phases were identified, and









effective change agent activities with respect to client diagnosis
were identified. In a 1981 article, Hall and Loucks discuss the
ways in which innovation use is evaluated. They presented the
third component of CBAM, the Innovation Configuration (IC), which
addresses the question, "What is the innovation?" Each of the
components has a measurement instrument developed to allow the
change agent to assess and evaluate the characteristics and
qualities of the three elements of the model. In an attempt to
address the lack of assessment and evaluation of the effects of the
innovation on student outcomes, Hall and Hord (1986) offer the
measurements of the SoC, LoU, and IC components of CBAM and
suggest that these can be used to determine if an innovation has
become institutionalized.
In a recent paper summarizing the research of the 1970s and
1980s, Hall (1991) proposes an approach for the 1990s which
includes creating conditions in the system that support innovation,
redefining the innovation itself, recognizing participants'
involvement in the process, and allowing time for institutional
change. Hall's concerns about the lack of innovation and change in
education echo the concerns of sociologists for our institutions and
organizations in general. He believes that educational innovation
depends on cooperation and collaboration and the use of a holistic,
systemic, planned approach.
In his article on educational staff development, Marsh and
Jordan-Marsh (1985) further categorize the types of concerns of
individuals in the change process as (a) organizational/political/









professional, (b) decision making/commitment, and (c) self-task
and says that each of these "clusters" must be addressed in order
for the change to occur. Both Hall and Rogers focus on the
decision-making process as key to the adoption, but Marsh and
Jordan-Marsh (1985) look at the process from the perspective of
the change agent. They define the task of the change facilitator as
providing "conditions necessary for the most adaptive coping
patterns" (p. 62). In an article on institutionalization, Miles (1983)
suggests that even though individuals are mandated to adopt an
innovation, their involvement in the change decision will
determine the extent to which the innovation is used and,
therefore, institutionalized.
In addressing the needs of the individual faced with adoption
of innovation, Bandura (1982) argues that unless people feel they
are capable of effectively performing their work tasks with the
innovation, their self-efficacy expectation, or sense of mastery of
the task, will inhibit their successful adoption of the innovation.
He further states that this takes place over time and that people
need multiple instances of activities directed at learning the
innovation. He suggests that traditionally held ideas of incentives
to adopt are much less effective than what the individual believes
about future outcomes and consequences and that building self-
efficacy is even more important. Bandura (1977) suggests that the
most effective way to build self-efficacy is through mastery
accomplishments and experience; the least effective way is
through the verbal persuasion, such as workshops and seminars,









which are the predominant way that we approach staff
development.
The major contribution of innovation diffusion research is in
the analysis of the characteristics of the individual adopter with
respect to the successful diffusion of innovation. The researcher's
interest is in describing the characteristics of the adopter, the
interactions of the adopter with the innovation, and the
communications networks which facilitate the diffusion of the
innovation. The research is descriptive, and the practitioner,
usually a change agent, uses the knowledge of the individual to
foster positive perception of the innovation and, hence, the
adoption of the innovation.
Hall takes it a step further when he prescribes intervention
strategies for aiding the change process. In fact, Bernice
McCarthy (1982) superimposed Hall's CBAM on her 4MAT model,
arguing that his stages of concern should be addressed in an
intervention model which allows the learner to pass through the
information-gathering, personalization, skill-building, and inquiry
("what if") stages. She presents this model as an intervention
spiral which fosters continuous improvement while providing
activities that interact with the different learning styles of
individuals. Her acknowledgment of the need for continuous
learning and of the differences in the rate at which individuals
progress is mirrored in the work of Rogers (1983) and many others.
Although her 4MAT intervention model is situated in the domain of
learning theory, it is worth noting as a viable intervention strategy









for technology diffusion that is clearly based on the theories of
Rogers and Hall.
Planned Change
The fact that change is with us constantly, as Hereclitus
said, is the one thing that does not change. Change process may be
purposive or unintentional or both. It may be seen as positive or
negative or some of each. The introduction of technology
innovations into an organization, though, implies purposive planned
change.
Planned change models have in common five basic
components. They are (a) the establishment of a relationship with
an external change agent, (b) diagnosis of the problems of the
organization, (c) an implementation intervention strategy of
varying intensity, (d) a system for maintaining an on-going
relationship with the innovation, and (e) the disengagement of the
change agent (Dill & Friedman, 1979). They are, by nature,
prescriptive and purposive. The research tradition is for action
research with experimental manipulation of the variables, usually
the interventions. This framework, and the research tradition
embodied by its proponents, is clearly central to this study.
Early proponents of planned change, Ronald Lippitt, Jeanne
Watson, and Bruce Westley (1958), see change as a process. Their
emphasis is on the change agent who they define as external to the
organization. The change process that they propose is based on
Lewin's phases of unfreezing, moving, and freezing, which suggest
that the organization must be shaken from its normal stability,









then moved through change to a new stability. Lippitt et al. (1958,
pp. 129-130) refined Lewin's work to propose their five phases of
change: (a) development of a need for change, (b) establishment
of a change relationship, (c) working toward change, (d)
generalization and stabilization of change, and (e) achieving a
terminal relationship. Their model proposes the blueprint for
introducing, initiating, and implementing change. They also focus
on the training of change agents.
Bennis, Benne, and Chin (1961), define planned change as "a
deliberate and collaborative process involving change agent and
client systems. These systems are brought together to solve a
problem or, more generally, to plan and attain an improved state of
functioning in the client system by utilizing and applying valid
knowledge" (p. 11). Valid knowledge, as the term is used here,
refers to the applied social science research conducted on the
dynamics of change which identifies and studies the variables and
interrelationships inherent in the process.
A great deal of emphasis in this definition is placed on the
nature of the collaboration of the client system with the change
agent. This collaboration or trust, contends Benne (1961), is
necessary to overcome the resistance to change which exists in all
individuals and organizations. The change agent in this definition
may be an outsider or may come from within the client system.
Chin (1961) further elaborates on the change agent as the
practitioner of change. He sees the role of the change agent as









being dependent on the nature of the change model adopted by the
change agent.
Bennis et al. (1961) point out that the system model (complex
organization model) assumes organizational stability as the norm
and sees change occurring only as a result of external stress
factors. The change agent in this model is separate from and
usually is external to the system. A developmental model (planned
change), on the other hand, assumes constant change, with
improvement, development, growth, and then decay being the stages
through which the organization continually moves. The change
agent may be part of the client system or may be brought in from
the outside to assist the organization in initiating the change.
They state that the nature of the research is empirical and
analytical, recognizing groups and organizations as well as
individuals as subjects of study.
Like the two types of models discussed in the previous
section, early theories of planned change do not treat the
innovation as a key factor in the change process. In fact, it is the
evolution of innovation itself into complex technology clusters
that has drawn attention to this shortcoming in the organization
innovation and change theory.
Later models for managing computer technology innovation
continued to expand the emphasis on the change process and on the
innovation itself (Bikson & Eveland, 1989, 1990; Johnson & Rice,
1987; Tornatsky & Fleischer, 1990). The later models also focus
more on the individual's and workgroup's role in the innovation









process. Johnson and Rice (1987) particularly note the role of the
communications network and the opinion leader, as defined by
Rogers and Shoemaker (1971), in the spread of the use of the word
processor. In fact, almost all models proposed from the mid-1980s
and afterward benefit from Everett Rogers' work in innovation
diffusion.
The limitations of all of these models persist to some degree.
The first limitation is the view of the innovation. When it is
considered at all, it is almost always seen as a constant, a thing to
be adopted intact and forever after used in lieu of a previous thing.
None of the models reviewed to this point have considered the
characteristics of the complex technology cluster. In other words,
the nature of the innovation does not change the model for infusion.
The second limitation is the view of the change process. It is
seen as linear, beginning with awareness, moving through adoption
and implementation, and then stopping with routinization. Even
when the model views continuous change as the norm in an
organization, a particular innovation diffusion process is seen as
having a beginning, middle, and end.
The third limitation is the lack of attention paid to the
interventions and the interactions of the interventions with the
innovation, the individual, and the organization. Even when the
innovation is a complex technology, which is known to require an
extended learning period, types of interventions are neither
proposed nor studied.









Finally, the models that stress one or another of the
innovation and change components to the exclusion of the others,
whether the organization, the individual, the change process, or
part of the change process, present an incomplete picture of the
organizational innovation and change process.
Bikson and Eveland (1991, p. 232) propose a conceptual model
for technology transfer (see Figure 2-5) that addresses three of
the four limitations mentioned above. They have incorporated the
recurring ideas seen in theories from each of the four traditions
and have constructed a view of technology transfer. Their model
shows the interactions of the technology with the organizational
context and the implementation process as they create
organizational consequences. They have suggested that technology
transfer is change of a particular kind, dictated by the nature of
the innovation and that planned change models may need to be
developed that address different types of innovations.
The properties of technology innovations in this model are
(a) validity and efficacy, (b) scope, testing, and scale, (c)
adaptability, and (d) packaging. Each of these properties addresses
the level of complexity and generalizability of the technology. The
model treats the technology not as one innovation but as a complex
information technology cluster. There is no expectation that the
technology will be uniformly adopted nor is there the expectation
that it will be viewed as universally effective. Bikson and Eveland
(1991) argue that "this element of variability has proved both a
blessing and a curse. Information technology can be applied to a




































Fire 2-5. Technology Transfer Model (Bikson & Eveland, 1991, p.
232).









great many parts of human activity, and there is little inherent in
the tools themselves that conditions how they are to be used."
They continue, "The burden falls on the adopters to create patterns
of effective use" (p. 233). Thus, Bikson and Eveland have entered
into the realm of individual adopter differences with respect to
accepting, modifying, or rejecting the innovation.
The properties of the involved organizations are divided into
two levels: the firm level and the work group level. While not
discounting the influence of organizational characteristics, such as
size, resources available, innovativeness, and dependence on
technology, they view job design and work group variables as being
better predictors of successful innovation diffusion. Bikson,
Gutek, and Mankin (1987) found that when job design creates the
need, information technology diffusion is more complete.
Although they acknowledge the role of the individual in the
successful diffusion of innovation, they fall short of exploring the
roles of individuals of varying technological capability with
respect to the diffusion process. Also, noticeably missing in their
model is the change agent. The closest they come to acknowledging
the role of the consultant is to propose the technology vendors and
developers as interested parties and, therefore, part of the
organizational whole (Bikson & Eveland, 1991, p. 239). They
further elaborate on the boundary spanners-- those individuals who
span work groups or even organizations in bringing technology
assistance, training, and support to the individuals adopting the
technology. Their view of the role of the change agent in the









decision process at the organization level, work group level, or
individual level, however, is not clear.
The properties of the implementation process are (a) planning
change, (b) supporting change, and (c) committing to change. As
they point out (p. 243) and as the literature reveals (Bannon, 1986;
Bikson & Eveland, 1986; Bikson, Gutek, & Mankin, 1981, 1987;
Bikson & Eveland, 1990; Johnson, 1985), there is growing evidence
that the implementation process significantly influences the
outcomes of technological innovation efforts.
In planning for change, there are three variables found to be
significant. The first is the reason for change. Bikson and Eveland
argue (1991, p. 243) that the innovation initiatives should
represent both need (established internally) and opportunity
(established externally). This departs from the traditional models
that argue that a problem is the impetus for change. The second
variable is the thoroughness of the effort. They argue for a
generalized, organization-wide plan that also allows for more
detailed implementation plans later. The third variable they
present is that positive outcomes have been programmed into the
implementation process.
In supporting change, Bikson and Eveland call for user
involvement and user training. They make an excellent case for the
importance of training.

The importance of good training for effective computer
use is often acknowledged but seldom enacted. The
discrepancy between anticipated learning needs and
common organizational practices led us to coin the
term "humanware," designating the knowledge, skills,








and technical resources necessary to take advantage of
computer-based tools. Underinvestment in humanware,
relative to hardware and software, is often responsible
for the underuse of the latter. (p. 245)

In their discussion of commitment to change, Bikson and
Eveland talk about the "diffusion status" of the work group, that is,
their view of their role as early or later adopters in the diffusion
process. They posit that the later adopters have more to gain
because the organizational structures are there to support their
adoption, whereas the earlier adopters may be laying the
groundwork for the organizational structure. They also suggest
that the change orientation of the group is key to the successful
innovation diffusion.
Bikson and Eveland's model is a major contribution to the
field. It does not follow a single research tradition in
organizational innovation and change, rather it draws on
characteristics from each of the four frameworks. The elements
of organizational characteristics, implementation process, and
attributes of the innovation are developed well beyond previous
models in complex organization, conflict, and planned change
theory. Their model reflects the work of the innovation
diffusionists to some degree, but it is the least well developed
component of the model.
The single biggest strength of the Bikson and Eveland model
is that it is a comprehensive model designed specifically for
complex technology cluster innovation diffusion. The proposition
that technology clusters are different from single product or
process innovations and that this type of innovation requires a









different type of plan for change is well supported in the literature
(Clark & Staunton, 1986; Hall & Hord, 1986; Hall & Loucks, 1981;
Rogers, 1983). The nature and complexity of the innovation, as
well as the magnitude of change which follows its diffusion, is
reflected in the complexity and specificity of the model.
The technology transfer model advances planned change
theory to the threshold of the model proposed herein, the local
expert intervention model. Like Bikson and Eveland's model (1991),
it draws on the elements of each of the four frameworks. Unlike
their model, it incorporates an intervention strategy as an integral
component, and it defines the role of the individual with respect to
the adoption and implementation of the technology cluster. Rather
than a conceptual model, it is a planned change model whose core is
the on-going intervention strategy.
Summary
This chapter presents the literature of organizational
innovation and change and of individual characteristics associated
with adoption of innovations. The evolution of organizational
change models, as exhibited in the discussion of the four
frameworks of complex organizational, conflict, innovation
diffusion, and planned change models, is apparent. Also apparent is
the lack of research on the diffusion of complex technology
clusters within an organization.
The need for a model which addresses both the diffusion of
complex technology cluster innovations and the nature of
continuous change and improvement in organizations was the






60

inspiration for the literature review, and it has illuminated the
path for the development of the local expert intervention model.
The proposed model contains elements of each of the traditions in
organizational change, with a heavy reliance on the results of
Everett Rogers' study of the individual in innovation adoption.
Chapter III discusses the proposed model in detail and shows the
relationships between the components and the reviewed literature.













CHAPTER III
THE LOCAL EXPERT INTERVENTION MODEL
Introduction

The proposed local expert intervention model for the diffusion
of complex technology cluster innovations can best be described as
an adaptive, cyclical model (see Figure 3-1). It supports the
continuous improvement of the use of existing technology and the
continuous diffusion of new applications and new technologies. It is
based primarily on the research of Everett Rogers (1962, 1983;
Rogers & Shoemaker, 1971) and those who followed his traditions in
research and, secondarily, on the work of the planned change
theorists, particularly Bikson and Eveland (1991).
The model components are the system or organization, the
local expert program, the client work group, and technology
transfer. The system provides the environment; the local expert
program provides the intervention strategies; the work group
system fosters the interventions; and technology transfer is the
activity in which the participants are engaged. Continuous
evaluation of each of the components fosters continuous
improvement. The cycles indicate process, evaluation, and
feedback.
The model capitalizes on a naturally occurring phenomenon--
users helping each other (Bannon, 1986; Bikson & Eveland, 1991;









































Figured 3-1. Local Expert Intervention Model.









Hamilton & Thompson, 1992). By strengthening this help network,
the model provides opportunities for individuals to receive
assistance from many sources and provides for continuous personal
growth by the participants, whether local experts or clients. Once
implemented, the various components of the model interact freely
with each other, allowing both the local experts and the clients to
give feedback which validates or changes the interactions, the
training, and/or the technology. It is driven by continuous
evaluation of the local expert program, the client interactions, and
the technology itself. The model is as variable as work groups
within an organization; each instance of its implementation is
customized to meet the specific requirements of the work group.

System
The system describes the environment of the organization.
For effective dissemination of computer technology, the
organization must be pro-innovation and aggressive in its efforts
to foster use of the technology (Bikson & Eveland, 1991; Lindquist,
1978; Zaltman et al., 1973). Furthermore, it is imperative that the
leadership of the organization be committed to continuous change
and subscribe to Clark and Staunton's (1989) view of change as
improvement rather than anomaly. In order for the early adopters
of technology, or local experts, to function as opinion leaders, the
organization must support the adoption of technology (Rogers,
1983).
In an introduction to a recent symposium on people and
technology in the work place, Laumann, Nadler, and O'Farrell (1991)









point out that it is the "leaders who make or allow the organization
modifications necessary to accommodate the change and the
transition process. They create the mechanisms for employee
participation, training, incentive schemes, organization design, and
structure, as well as technology choice and development" (p. 6).
Lindquist's (1978) "adaptive development" model makes a similar
argument for system support through "linkage, openness,
leadership, ownership, and rewards" (p. 240).
Practically, the organization must include an internal
computer center training unit whose responsibility it is to
administer and evaluate the local expert intervention program. The
minimum support required for the successful implementation of
the program is as follows:
1. The selected local experts are released to attend special
seminars, workshops, and meetings.
2. Their role as local experts is sanctioned by their
supervisors.
3. Their new task is included in their job descriptions.
Any additional support in the way of incentives, such as
conference attendance, awards of new technology, formal
recognition, release-time, and stipends, are extremely valuable as
well. Ideally, system support activities will increase as the
program is found to be effective in disseminating technology
humanware, and the additional support will, in turn, promote
greater success.









Local Expert Program
The implementation of the local expert intervention model
consists of (a) selection of appropriate individuals, (b) regular
meetings, (c) visitations from computer center trainers, (d) special
computer center hot-line support, and (e) advanced applications
workshops.
Good local experts in technology have the characteristics of
the early adopter and opinion leader (Hamilton & Thompson, 1992;
Rogers, 1983; Sachs, 1976). They are generally eager to try new
technologies, are knowledgeable about them, believe that
technology adoption will be a positive experience, seek information
about technology from several sources, and find themselves helping
others frequently. These characteristics are measured by the LEP
which may be used to either confirm the choices already made or to
screen possible candidates to find the most suitable local experts.
They must serve voluntarily and, hopefully, with enthusiasm. They
must also have time, or perceive that they have time, to perform
the tasks.
The local expert meetings should occur on a regular basis,
although regular meetings may be replaced with visits from the
computer center trainers. During the meetings and the visitations,
the role of the local expert, the problems of the various work group
clients, and the plans for the introduction of new technologies are
likely topics of discussion. The local experts should be encouraged
to view each other and the computer center as resources to help
them solve their clients' problems. As Bikson and Eveland (1991)









stated, the success of the local experts is due, in part, to their
positive perception of their role in the dissemination of knowledge
and information about the technology.
Furthermore, a hot-line (telephone number) and/or an e-mail
address installed just for local expert queries allows them to
command priority in solving problems. More rapid response to
technology questions is the promise.
The workshops to be offered are over and above the
workshops offered to the general population. Generally, the local
experts are excellent candidates for advanced training on existing
applications and should be the first to be given training when new
applications are adopted. With this training as well as an
awareness of their clients' concerns (Hall & Loucks, 1978), they
will be able to intervene more effectively when problems arise.
Local experts are given information to distribute about
organizational technology plans and coming events. They are also
given resource materials, such as manuals or self-help tutorials,
so that they can become a local resource for their clients. By
nature, they are more cosmopolite in character than their clients
(Rogers, 1983), which means that they prefer to get information
and knowledge from many sources. Having these resources
available increases their ability to gain knowledge and pass it
along to their clients.
It is important for the computer center to continue to have
regular dialogue with the local experts. It is also important for
local experts to be in contact with each other. A combination of









group meetings and seminars, directories of local experts cross-
indexed by areas of expertise, e-mail and telephone
communications, and individual meetings among local experts and
between the computer center and local experts are all means of
sustaining this contact. This network of expertise provides the
communication channel, as in Rogers' (1983) model, for the
successful diffusion of the various instances of technology
innovation.
Local experts are engaged in continuous evaluation of the
support that they receive from the computer center and from the
organization. They participate fully with the computer center staff
in the needs assessments, planning, and evaluation of the
effectiveness of the use of technology and are the intermediaries
in the information flow to and from the clients and computer
center staff in the execution of this evaluation process. It is
essential that this be a continuous feedback loop, as the local
experts are the liaisons between the end-users and the trainers.
They are aware of both the technological goals of the organization
and the individual needs and problems of their client work groups.

Client Work Groups
The clients are chosen because of proximity with the work
group. The work groups are often engaged in similar or related
tasks, but not always. The work groups may be comprised of people
in similar positions and ranks or different rank and position. In
taking advantage of the aspects of naturally occurring help in the









work place (Bannon, 1986), the one thing that does matter is that
the people work in proximity with the local expert.
The client intervention takes on several characteristics. It
is, at times, initiated by the local experts. If a new feature or
resource is introduced into the system, the local experts may be
asked to take the role of instructor to help disseminate
information and training on the new feature. The local experts may
also be asked to assist the trainers in general workshops which are
sometimes held for everyone to teach introductory skills and
concepts on new applications.
Most of the interventions, however, are initiated by the
clients. The local expert responds to requests for help. These
interventions are similar to the transactions that would be taking
place were there no local expert program in place (Bannon,1986;
Bikson & Eveland, 1991). The difference is that the local experts in
the program have opportunities to gain far more expertise and more
information and, therefore, to be more helpful. Other requests for
help may come from other local experts not familiar with a
particular application or from the clients of other local experts.
Client-initiated interventions may be seen as positive signs
that clients are tackling new technologies. Even though introducing
new applications or features will temporarily increase the
numbers of interventions, it should result in more extensive and
efficient use of those technologies.
Clients are asked, periodically, to evaluate the effectiveness
of the technology transfer. Additionally, they are asked to evaluate









the technology itself as to the ease of use and the appropriateness
of its employment to the task at hand. All evaluations are
formative and are used by the local experts and the computer
center to improve the program.
Technology Transfer
An assumption made by this model is that the decision to
adopt technology has been made by the organization because it
believes that its use will produce more effective outcomes. The
effectiveness of an application of technology to a task, however,
cannot be realized unless it is being applied expertly, or at least
competently.
The proof of the value of the model itself is in how
efficiently and extensively the technology is being used.
Individuals will always differ in the degree and ease with which
they use technology (Rogers, 1983), but progress can be expected if
proper interventions occur at appropriate times (Hall, Dossett, &
Wallace, 1973; McCarthy, 1982). The progress of individuals over
time with respect to increasing their expertise in the technologies
which they use is one criterion with which to gauge its successful
transfer. Another is the increased number of applications and
features that individuals adopt over time.
Because the use of technology usually redefines and combines
tasks, the change expected in the end product attributable to the
use of technology is the outcome which must be assessed. If the
technology is being used expertly and extensively and no
improvement is made in either the time required to complete the









task or in the final outcome or product produced by the task, then
effectiveness through the use of technology has not been achieved.
If, on the other hand, the introduction of technology or more
efficient use of the technology improves the outcome, then
effectiveness has been achieved.
Summary
The success of the local expert program depends, in part, on
the organization's overt support. Having made the commitment to
adopt technology, the system, through the computer center,
supports the local experts with advanced training so they can not
only be efficient users of the technology, but can also help their
peers become efficient users of the technology. Through incentives
and recognition, the system is able to influence the opinion of the
local experts, who, in turn, become an adoption model for their
clients. The local experts reciprocate with information and
evaluations from their clients so that the system can monitor the
technology transfer, the technology choices, and the overall
effectiveness of the innovation in improving the organization's
processes.
The expressed interest of the organization's leadership in
humanware acquisition provides both resources and incentives for
local experts to help and for clients to learn. The more clients
learn about technology, the more they tend to use it. The
efficiency and extensiveness of technology use is evaluated so that
the interventions can be modified. The effectiveness of the
technology in achieving better outcomes is evaluated so that the









technology selections can be revisited. Successful technology use
is likely to trigger more adoptions, which then calls for more
interventions from local experts and more local expert support
from the computer center.
The interactions between the clients and local experts are
particular to the work group itself. Each work group, being engaged
in its own processes and, perhaps, having different technology
requirements, is able to find help of the right kind at the right time
from its own local expert. The work group clients also have a link
back to the computer center and are able to communicate directly
with it when appropriate. The satisfaction levels of the clients
with the technology, the local expert program, and the computer
center are used to evaluate and to improve the program.
In addition to playing an important role within the work
group, the local experts are key in the successful diffusion of
technology into the organization's processes. In evaluating its end
products and services, an organization must define the processes
as well as the role of technology in those processes. Because local
experts are in a position to understand the role of technology in the
organization, when they are supported by the system, they will be
able to lead their work group clients through the adoptions of
applications of complex technologies.













CHAPTER IV
METHODOLOGY

As stated in Chapter I, the purpose of this study was to
determine whether the variables measured by the LEP accurately
identify the individuals who can best serve as local experts for the
dissemination of complex technology clusters within an
organization. Additionally, the study was designed to determine the
relationship of status in the organization to the performance of the
local expert.
The steps taken by the investigator to answer the questions
were to (a) select local expert subjects for the study, (b) implement
the local expert intervention model, (c) record LEPs, (d) collect
client data, (e) measure and record the performance (LEPR) of local
experts, and (f) perform statistical tests on those data.
The purpose of this chapter is to describe the following: (a)
the procedures used to develop, score, and validate the local expert
profile (LEP), (b) the study procedures, (c) the data-gathering
techniques, and (d) the statistical procedures used in the study.
Local Expert Profile
Researchers indicate that certain personality, communications
style, and socio-economic characteristics are present to a higher
degree in the early adopters of innovation and are not present in
later adopters (Hamilton & Thompson, 1992; Rogers, 1983; Sachs,









1976). The Local Expert Profile (LEP) was developed by the
investigator (see Appendix A) using questionnaires from both Sachs'
(1976) and Hamilton and Thompson's (1992) studies in order to
identify the local experts in an organization. Both studies used the
Rogers' (1962) adopter profile to identify the characteristics of
early adopters of technology and innovators in an educational
setting. The questions used from these studies for the LEP were
those which identified the characteristics that correlated positively
with innovativeness and opinion leadership in one or both studies.
The personality variables tested by the instrument are percep-
tion of the innovation and venturesomeness. The communications
variables are connection to the social system, cosmopoliteness,
and experience or expertise with the technology. Each of these
variables was found to correlate positively with early adoption in
both the Sachs (1976) and Hamilton and Thompson (1992) studies, as
well as in the analyses of research studies in Rogers' (1983) work.
Not included in the LEP but included on the LEP questionnaire
was the socio-economic variable, the status of the local expert in
the organization. As stated in Chapter II, the status of the
individual, one of the positively correlated variables in both Rogers'
(1983) and Hamilton and Thompson's (1992) research, was left out
of the profile. In the investigation that preceded the development of
the proposed model, it was observed that people prefer to seek help
from those who know more about computers than they but who are in
nonthreatening positions in the organization. It was also observed
that the vast majority of the most enthusiastic users of computers









are in clerical or technical positions rather than in administrative
or faculty positions. This observation refutes Rogers' findings that
the higher the position an individual holds in the system, the more
likely the individual is to be an early adopter and an opinion leader.
Since the local expert is defined as both an early adopter and opinion
leader, the observations with respect to status present a conflict
with the literature which this study intends to resolve.
The 25-question LEP instrument is in multiple-choice format
in three sections. In the first set of questions, there are four
categorical answers; in the second set, a Likert scale ranges from
strongly agree to strongly disagree; and in the third set, there are
three possible responses: frequently, seldom, and never.
The responses for questions were assigned a value based upon
the hypothesized characteristics of the local expert, with the
highest value being given to the most favorable characteristics.
Each of the five characteristics being measured was scored by
computing the sum of the squares of the response values of the
questions relating to the variable. The five variables were then
adjusted so that they were equally weighted in the overall profile.
The LEP is computed as the sum of the five subscores. The highest
possible score on the LEP is 750; the lowest is 23. It is
hypothesized that the higher the score, the higher the probability
that the individual will perform well as a local expert.
Development of the LEP Instrument
The first version of the LEP questionnaire was given to 8
randomly selected faculty at Santa Fe Community College and a









group of 12 music teachers and music students from throughout the
state of Florida who attended a multimedia summer workshop. Some
of the subjects from both groups were known to be knowledgeable
and enthusiastic about computer technology, and some were known
to be less enthusiastic and/or less knowledgeable.
The questionnaire was given to both groups in order to refine
the questions and to analyze the distribution of adopter types as
differentiated by the LEP. The results showed a normal distribution
(see Figure 4-1). The high score was 415 and the low score was
187; the mean was 284.8 and the standard deviation was 59.7.
Comments from the subjects led to revision of questions in order to
make them clear and easy to answer.
The LEP was revised and the current version was then
administered to a group of 24 faculty at Gulf Coast Community
College. The group consisted of participants in a multimedia
mentors program as well as an equal number of randomly selected
individuals from the remaining faculty population. The multimedia
mentors program participants were selected because of their
willingness to learn the technology tools and to help their fellow
faculty learn the tools. The program consists of workshops in using
multimedia authoring systems and release-time for the development
of multimedia projects to be used in a learning environment.
Ultimately, the mentors are assigned to new teams to act as the
multimedia resource person.
The LEP questionnaires were administered by the coordinator
of the multimedia mentors program and were scored and recorded by













Number of Cases















228 276 324 372 420
Local Expert Profile (LEP)


Mean
Std Dev
Std Err Mean
upper 95% Mean
lower 95% Mean


284.80
59.68
13.34
312.73
256.87
20


Figure 4-1. Histogram of the distribution of the LEP scores for the
group of randomly selected faculty and music workshop
attendees.









the investigator. In a separate activity, the coordinator of the
multimedia mentoring program rated each of the 24 faculty with
respect to their potential (or lack of potential) as multimedia
mentors.
The range of LEP scores was 624 (high) to 363.5 (low), and the
mean score was 479.3. The rating scores ranged from 200 (high) to
10 (low) with a mean rating score of 74.7 (see Appendix D for
scores).
A regression analysis was conducted to test the null
hypothesis that the predictability of multimedia mentor
performance ratings (MMPR) from LEP scores is zero. The decision
rule was set at .01 for rejecting the null hypothesis.
The regression analysis results are reported in analysis of
variance and summary of fit tables and in leverage plots. Leverage
plots show data points, the response mean line, the regression line,
and confidence curves drawn at the .05 level of confidence. If the
slope parameter of the regression line is significantly different
from zero, the confidence curves will cross the mean line, showing
significance.
Figure 4-2 gives three examples of leverage plots. The first
plot depicts a significant positive relationship. The slope of the
regression line shows a positive relationship between the predictor
(X) and response (Y) variables, and the confidence curves clearly
cross the mean response line, indicating significance.
The second plot in Figure 4-2 shows a lack of significance
between the predictor (XX) and response (Y) variables. The







Significant and Positively Correlated


150-

100-

50 -" -
50

0 -1 7--r-
300400500600700
X Leverage



Not Significant

150-

100

50

0
60 80100 130
O( Leverage


regression line
confidence curves
- mean response line


Significant and Negatively Correlated

150-
100 -

50

0

0 5 10 15
)0( Leverage


Figure 4-2. Comparison of three leverage plots: One indicates a positive
significant relationship; the second shows no significance in
the relationship; and the third shows a negative significant
relationship.









confidence curves do not cross the mean response line and the slope
of the regression line is nearly nonexistent.
In the third example, significance is shown, but the direction
of the relationship is negative rather than positive. In this example,
as the predictor variable (XXX) increases, the response variable (Y)
decreases. Again, the confidence curves cross the mean response
line, so the relationship is significant.
The analysis of variance results, reported in Table 4-1,
indicate that the predictability of the multimedia mentor ratings
from the LEP is significant at the .01 level (p<.0001). The R square
value (see Table 4-2) of .889 indicates that almost 89% of the
increase in MMPR is attributable to the LEP score, and only 11% is
independent variance. The leverage plot (see Figure 4-3) is a
pictorial display of the regression analysis which describes the
relationship between the two variables. The steep slope of the
regression line and the confidence curves' dissection of the response
mean line indicate high significance.

Table 4-1
Analysis of Variance for LEP as Predictor for MMPR


Source DF Sum of Squares Mean Square F Ratio Prob>F

Model 5 70313.460 14062.7 28.5380 0.0000
Error 18 8869.874 492.8
C Total 23 79183.333








Table 4-2
Summary of Fit for Regression Analysis of MMPR/LEP


Rsquare 0.887983
Root Mean Square Error 22.19844
Mean of Response 74.16667
Observations (or Sum Wgts) 24



The high significance in the findings indicates that similar
populations who score high on the LEP and are given the same
treatment and incentives could be expected to perform well as
multimedia mentors. While this particular program is unique to Gulf
Coast Community College, it is very similar to the local expert
intervention program being implemented in this study. Even more
important, the characteristics that determine an individual's
suitability for both programs is believed by the investigator to be
very similar, if not the same.
To examine further the relationship of the variables,
experience, venturesomeness, cosmopoliteness, connection to the
social system, and perception of the innovation to the performance
rating, a multiple regression analysis was conducted. This
statistical technique reveals which of the composite subscores,
when analyzed with respect to the correlations of all of the other
variables, are significant in predicting the MMPR.
The leverage plots in Figure 4-4 show the relationships
between the individual regressor variables and the MMPR, while














LEP



200-


150-


100'



O- -T------------ -------------------
50


0
350 400 450 500 550 600 650
LEP Leverage






F Ratio = 28.530
P < .0001
R Square = .887983



Figure 4-3. Leverage plot of the results from a multiple regression
analysis of LEP as a predictor of MMPR.










Experience

200-
150
S100- ....
------ ------
50 -
0
2550 75100 150
Exp Leverage


Venturesomeness

200-
S150

0 100

0
40 60 80 100 120
Ven Leverage


Social System

200-
150-


50 .


40 60 80 100 120
SS Leverage


Figure 4-4. Leverage plots of the results from a multiple regression
analysis of LEP subscores, each as a predictor of MMPR.









holding constant the contributions of the other variables. The closer
the slope of the regression line is to zero, the lower the
significance of the relationship. Also, the confidence curves that do
not cross the response mean line indicate the degree of significance
(or lack thereof). Table 4-3, the effect test results, shows for each
variable the probability that it, independent of the other variables,
added predictability to the MMPR.

Table 4-3


Source Nparm DF Sum of Squares F Ratio Prob>F

Exp 1 1 2034.6164 4.1289 0.0572
Ven 1 1 3926.5968 7.9684 0.0113
Perc 1 1 5495.3718 11.1520 0.0036
Cos 1 1 352.3426 0.7150 0.4089
SS 1 1 6904.2396 14.0111 0.0015




Not all of the variables, when partialed out or taken into account
with all of the others, showed significance. As can be seen in the
leverage plots in Figure 4-4, both perception of the innovation and
connection to the social system were significant. But the other
variables, experience, venturesomeness, and cosmopoliteness, did
not add significance of their own.


6 ct est~ -02 Su siores oi fLua


cffExt Tnrt f


b P di f MMPR









The whole model analysis is significant at the .01 level, so the
lack of significance in the partial regression coefficients indicates
multiple collinearity. This means that not all of the variables add
independent predictability for MMPR. Rather, some of the variables
are redundant in their prediction of the MMPR. Preliminary
indications are that several variables, perception of the innovation
and connection to the social system, could give us almost as reliable
information about the propensity of the adopter to be successful as
a multimedia mentor as the overall profile gives us.
The question that could not be addressed by these data was the
status question. The subjects from Gulf Coast Community College
are largely faculty, and their clients are faculty. Since there is no
difference in status among either subjects or clients, the
relationship of status to performance could not be explored.
It was concluded from this pilot investigation that the LEP
should be used as the selection instrument for the population of
local experts in the main study and that it should be the focus for
the investigation into the selection process.
Study Procedures
This study was conducted to determine the effectiveness of
the selection instrument, the Local Expert Profile (LEP), in
identifying local experts and to determine the role of local expert
and client status in the successful execution of local expert
interventions. In order to determine the effectiveness of the LEP,
the relationships between the variables which comprise it and the
performance of the local experts are also being studied.









Population
The subjects participating in this investigation were staff,
faculty, professional specialists, technicians, and administrators at
Santa Fe Community College. Twenty-eight (28) work groups of
between 5 and 15 members each were identified. A work group
consisted of individuals who worked in proximity with each other,
regardless of their position in the organization. Most work groups
were intradepartmental, but a few were interdepartmental. A
local expert subject was chosen from each work group. Local expert
subjects were selected from all levels and positions from within
the organization. The subjects chosen to perform as local experts
were either those known to be early adopters of computer
technology, those who were pinpointed by their supervisors, or those
who simply volunteered. All subjects agreed to perform the role of
local expert and to participate in the study.
The local experts were given an identification number to be
used in the study. Their clients were identified and given the same
number as a client group number in order to match data gathered
from the clients with the local experts.
Local Expert Intervention Program
The local expert program was implemented by the training
coordinator in the Technology Resource Center (TReC) at SFCC where
a new local area network (LAN) and a new digital telephone system
were recently installed. The applications being implemented with
these new technologies were telephone voice-mail, electronic mail,
electronic fax mail, mainframe access through the LAN, shared









printing, and other related technologies. The treatment was carried
out for 6 months.
The treatment consisted of (a) introduction to local expert
program, (b) advanced features training, (c) computer center hot line
support, (d) information clearinghouse activities, and (e) networking
with other local experts.
The local experts were introduced to the program by the
investigator during their first meeting. A computerized
presentation was used for the introduction. It stressed the role of
the local expert as well as the positive outcomes expected by the
college from the program. Other higher-level administrators were
involved two times during the treatment period in the process. They
reiterated the expectations of the administration and congratulated
the local experts for taking part in the program.
The advanced training consisted of six seminars held over the
course of the 6-month study period, one per month. Sessions were
conducted to teach more esoteric features of Windows, cc:Mail,
Macintosh-Windows compatibilities, network file-sharing, network
printing and voice-mail.
The Technology Resource Center established a special e-mail
address for local experts to communicate problems and information
to the training coordinator and to the support staff of the Center.
The Technology Support Help Desk personnel were given the names of
the local experts and instructed to give them priority in helping to
solve any problems that they may have. In addition, the local









experts were invited, repeatedly, to call or e-mail any of the staff
of the Center anytime they had a question or problem.
The local experts served as an information resource to both
their clients and to the Technology Resource Center. During their
biweekly meetings and by e-mail, they were given information as to
the progress of the implementation of the network and telephone
systems. They were asked to pass this information on to their
clients and to be aware of and respond to problems caused by
changes being made. In addition, the local experts brought theirs and
their clients' questions to the biweekly meetings. In turn,
information as to the needs for client training was communicated
from the local experts to the Technology Resource Center.
Local experts were encouraged to network with the other local
experts at the college. They were each given a list that included the
names, locations, and phone numbers, as well as the particular
expertise and interest of all local experts. The biweekly meetings
also provided opportunities for communications among the local
experts. Although they are referred to as biweekly meetings, some
were canceled because of conflicting activities or simply because
there was not enough information on an activity to warrant taking
the time of the participants. Communications to that effect, when
appropriate, were sent from the Technology Resource Center.
Data-Gathering Procedures
Data were collected during the course of the study for three
purposes: (a) to determine the LEP of the local experts, (b) to
determine the client variables which may help to explain results,









and (c) to determine the effectiveness of the subjects in performing
their roles as local experts.
LEP
The LEP, described in detail in the first section of this
chapter, was administered to all subjects at the first meeting of
local experts. It was scored and recorded by the investigator using
the name and identification number of the local expert. The scores
used in the analysis of the data were subscores for the five
variables comprising the LEP and an overall LEP score.
Client Status
After clients were assigned to groups, their position code was
recorded with the local expert identification number. The mean
scores for client status were computed and recorded for each client
group.
LEPR
To determine the effectiveness of the local expert, three
measures were used. First, the subjects were instructed to log the
name, date, time, and nature of all transactions with their clients
(see Appendix B). The logs were collected at the biweekly meetings.
The interaction frequencies were recorded by counting one
interaction for every 15 minutes of interaction time that occurred.
The average number of hours per week was computed and recorded
for each local expert.
Second, the computer center help desk log provided
information as to all requests for help placed by the local experts
and their clients during the last month of the study period. The help









desk is a service available to all employees. The local experts were
encouraged to call the help desk to resolve problems that they or
their clients had. The clients were encouraged to contact their local
experts first to resolve problems. The frequency of calls initiated
by local experts to the computer center was recorded as one point
per call. An extra point was added if the call was made on behalf of
a client. A point was subtracted if a problem was reported to the
help desk by a client which could have been resolved by the local
expert.
Third, the local expert training coordinator interviewed the
local experts at the end of the study period to assess their skill
levels in the technologies being implemented, their attitude toward
the local expert program, and their feeling of effectiveness in
performing their role as local expert (see Appendix C). The scores
ranged from 57 at the high end to 20 at the low end.
The hours per week and the self-reported effectiveness rating
were adjusted to be similarly weighted, and the help desk component
was increased by three to add significant value to the scores. This
third subscore was generally very low, under five, and, without
adjusting it upwards, it did not impact the overall score. Although
tripling it gave it more clout, it did not change the results of the
regression analysis. In fact, when looking more carefully at the
LEPR subscores, it was seen that each of the scores could be used
singly as the response variable with nearly the same results. The
three sets of data were then summed to give the local experts a
performance rating (LEPR).









Statistical Procedures
There were eight research questions posed in Chapter I. In
order to discuss those questions, a bivariate or linear regression
was used to determine the relationship between the LEP or predictor
variable and the LEPR or criterion variable. Multiple regression
analyses were conducted to determine the relationships between
LEP, local expert and client status, and LEPR. Finally, a multiple
regression analysis was used to explain the relationships of the
subscores which comprise the LEP to each other and to the local
expert status and the LEPR. When multiple regressors are used in
the model, a partial regression coefficient is computed for each
variable in order to take into account the correlation of multiple
predictor variables with each other.
Regression analysis is useful where coefficients are interval
variables, such as the LEP and the subscores and the LEPR. In
regression analysis, the correlations between two or more variables
are used to predict the performance on one measure from the
performance on another measure (Smith & Glass, 1987). The results
from the regression analysis may be used to determine if the
criterion or dependent variable can be predicted from the
independent or predictor variable.
The statistics were executed by the SAS Institute's JMPS
statistical applications software package. Statistics reported for
regression analyses are E ratio, p value, and B square. If the
reported E ratio is greater than the E critical value as computed
from the "Distribution of F Table" (Kennedy & Bush, 1985, pp. 565-




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 EAM58BRJS_31JJ4L INGEST_TIME 2012-09-24T13:25:51Z PACKAGE AA00011818_00001
AGREEMENT_INFO ACCOUNT UF PROJECT UFDC
FILES