LIBRARY
Michigan State
University

 

 

 

This is to certify that the

thesis entitled

TELEMEDICINE AS A SUPPLEMENTARY INFORMATION
SOURCE FOR PHYSICIANS, NURSES AND MANAGERS

presented by

STEVEN LEE BECK

)

has been accepted towards fulfillment
of the requirements for

MA degree in TELECOMMUN ICATION

 

Major profe 1'

Date gr//lL/7>

0-7639 MS U is an Affirmative Action/Equal Opportunity Institution

 

 

 

 

MSU

 

 

 

RETURNING MATERIALS:
PIace in book drop to

 

LJBRARJES remove this checkout from

4—15— your record. FINES will
be charged if book is
returned after the date
stamped beiow.

 

jUN 9 3 “Na

P / (_. '

  
  

  

 

 

 

TELEMEDICINE AS A SUPPLEMENTARY INFORMATION SOURCE
FOR PHYSICIANS, NURSES AND MANAGERS

BY

Steven Lee Beck

A THESIS
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of
MASTER OF ARTS

Department of Telecommunication

1987

ABSTRACT

TELEMEDICINE AS A SUPPLEMENTARY INFORMATION SOURCE
FOR PHYSICIANS, NURSES AND MANAGERS

BY

Steven Lee Beck

As the financial and political environment changes in
health care, the need for continuing medical education is
becoming more apparent. While the technology presently exists
to deliver healthcare information simultaneously to
participants all over the world, the technology will not
flourish simply because it exists. Little has been done to
study those attitudinal factors which affect the
decision-making process of health care professionals.

Users from the 125 member hospitals of the Michigan
Hospital Information Network were surveyed to find those
attitudinal factors associated with telemedicine in general,
speaker and program traits which are perceived to be
important, and user input about their present health care
network. Attitudinal factors were tfluni correlated and
regression equations formed to predict the two dependent
variables of satisfaction and behavioral acceptance.

The two best predictor factors for satisfaction were
found to be utility and users interest, both in the
individual program and telemedicine in general. The best
predictors for behavioral acceptance were found to be program
quality, sociability and the speaker's knowledge & ability to
teach the subject matter.

To my mother and sisters who never doubted me and to my wife
Linda for the countless number of hours she spent with me on
this project and all of those that lead to it. I could not

have completed it without any of them.

iii

ACKNOWLEDGEMENTS

I wish to thank Drs. Chip Steinfield and Kent Creswell for
the help and guidance that they provided me throughout this
project. They kept me honest and on my toes on every step of
this endeavor.

I also wish to thank Marlene Hulteen of the Michigan
Hospital Information Network in Lansing, David Levine from the
Hospital Information Network in Princeton, New Jersey and all
the fine people from those two organizations who lent me
financial as well as moral support on this project.

I also wish to thank the graduate students and faculty in
the Department of Telecommunication at Michigan State
University whom I hounded for answers to countless numbers of
questions until the lights went on. To all of you, thanks.

iv

List of
List of

QUARTER
I.

II.

III.

TABLE OF CONTENTS

Introduction ............................

Overview and use of telemedicine ........
Definition of problem ...................
Research variables ......................
Research questions ........ _ ..............
Procedures ..............................
Summary of findings .....................

Implementation ..........................

Theoretical Rationale ...................

Conceptualization of variables ..........
Behavioral acceptance ...................
Research hypothesis #1 ..................
Diffusion and adoption ..................
Research hypothesis #2 ..................
TV as a learning source .................
Research hypothesis #3 ..................
Michigan Hospital Association ...........

Summary .................................

Procedures and Methods ..................

Data collection .........................
Respondents .............................
Survey instrument .......................
Response motivation .....................

Population ..............................

Operationalization of research variables

V

Statistical procedures ... ........... . ......... 39

IV. Results ....................................... 41
Factor loadings ........................ . ...... 41
Regression models ..... ..... . .................. 55
Summary ....................................... 58

V. Conclusions and Recommendations ............... 60
Review of procedures .................... ..... 60
Conclusions and discussion .................... 61
Recommendations ............................... 65
Directions for further research ............... 65

APPENDICES: Appendix A: Questionnaire dispersement .. 67

Appendix B: Questionnaire ............... 69
Appendix C: Follow-up letter ............ ‘79
Bibliography .......................................... 80

vi

E

4.1
4.2
4.3
4.4
4.5
4.6
4.7
4.8
4.9

LIST OF TABLES

RAGE
F.A. of Attitudinal Variables Towards HIN .... 42
Factor Analysis of Attitudinal Variables ..... 43
Factor Analysis of Speaker Traits ............ 45
Factor Analysis of Satisfaction Variables .... 46
Factor Loading Summary ....................... 48
Summary Table for Variables .................. 49

Predictor and Criterion Variable Relationship. 51
Pearson Correlation Coefficients ............. 54

Regression Summaries ......................... 56

vii

LIST OF FIGURES

Behavioral Acceptance ........ ..

External Variables in Predicting Behavior

Composite Model of Behavioral Acceptance

viii

RAGE

..............14
...16

31

CHAPTER I
Introduction
W

Technological innovations are commonplace in the
approaching information age. The word approaching looms large
in this sense because we have yet to work out many logistical
problems which impede progress towards universal information
availability. Two industries that are perhaps unequaled in
new product innovation are telecommunications and medicine.
It would only seem natural then that the two should
incorporate their resources to embark on a new field. This
field is commonly known as telemedicine. While telemedicine,
in it's present form, got started in 1969 with the Logan
Airport/Massachusetts General Hospital link using a
black-and-white broadband television system, telemedicine can
be traced back to 1906 with the development of the
electrocardiogram (Conrath, et al., pp. 24-25). This paper
however, will deal with the application of telemedicine in the
state of Michigan which began it's programming in July of 1984
under the direction of the Michigan Hospital Association.

H E I J 1' .

The uses of 'telemedicine range from long-distance
diagnosis to how to maintain floor maintenance (Healthcare
Information Network, 1987). For purposes <xf this study
however, the scope will be limited to the analysis of
information directed toward physicians, nurses and hospital
managers-administrators because they receive the bulk of
programming from the Healthcare Information Network, a

national health care information network headquartered in
Princeton, New Jersey, and because these three groups are of
primary interest to the network.

W

The concentration of hospitals will in all likelihood
increase as urbanization continues. This, along with
increasing advancements in the medical profession, compound
the problem of dispensing medical information to personnel
around the country and especially in outlying areas where
travel to training seminars is long and difficult.

Teleconferencing, electronically mediated communication
among three or more people in two or more separate locations,
has been deemed most appropriate to deliver this vital
information. This alternative form of information delivery is
capable of delivering a wide range of programming via
satellite at a relativly inexpensive per viewer cost.
Teleconferencing has been deemed a communication wonder by
researchers like Johansen, Vallee, and Spangler (Neumann,
1985), but alters the preexisting communication patterns of an
organization utilizing the technology. Certain perques like
travel and the ability to interact with fellow healthcare
professionals may be lost, or certainly diminished, by use of
telemedicine. The problem exists as to how a technology
should be utilized in an organizational structure with such
historically unpermissive norms as the healthcare field.

Many researchers, including Victor Ferkiss, feel that it
is their job to keep us from being the technology-driven
society that we seem to have become. French sociologist,
Jacques Ellul, asks, do we analyze a new communications
technology with the same cost-benefit fervor that we do other
technological advancements? (Dizard, 1982) Researchers
speaking specifically on teleconferencing point out:

In itself, the installation of an
appropriate teleconferencing technology
does not necessarily guarantee effective

use.
(Olgren and Parker, 1983, p. 61)
Another researcher warns:

It is a serious mistake to assume that
teleconferencing will be used just because
it is available. There is a good deal of
evidence to the contrary.

(Elton, 1982, p. 53)

This assumption of use was realized by programmers and
executives of the Michigan Hospital Association, a private
state-wide health care network, in May 1986. Although member
hospitals numbered 125, rising costs and lack of programming
resources forced MHA into thoughts of a merger with another
network. Merger proceedings started one month later, in June
1986 and the Healthcare Information Network took over the
teleconferencing portion of the MHA's information duties
(Hulteen, 1987). Since then, users have enjoyed an increase
in the number of available programs from the national network
located in Princeton, New Jersey while still being able to
watch the more popular programs produced out of the Lansing
bureau. Increasing pressures, both internal and external, are
mounting t1) step-up efforts in) make continuing education
mandatory for several healthcare professions, mainly
physicians and nurses. In Michigan, and several other states,
it is currently mandatory for physicians to amass continuing
medical education units (CMEs) and that state's board of

nursing is proposing legislation to make similar rules for
it's nurses. While this controversy may be considered a
separate issue, there may be need for a teletraining system to
be implemented on a grander scale. This necessitated the need
to study the acceptance of telemedicine and it's impact on
the healthcare community.

W

The two variables that are most important to this study
are behavioral acceptance and satisfaction. Behavioral
acceptance is defined as the use of an innovation or
technology which is sponsored by an organization.
Satisfaction is defined as how a user feels about an
innovation or technology after utilizing it.

The predictor variables used in this study were divided
into two catagories. One catagory included personal variables
such as: age, sex, job type, experience and marital status.
The other catagory included attitudinal factors such as
compatability, credibility, sociability, program quality,
ability to provide feedback and speaker quality.
Compatability is defined as whether the use of a new
technology is consistant with currently held beliefs and
practices among individuals within an organization.
Credibility is defined as how users feel a technology ranks
with the available alternatives. Sociability is defined as
how sensitive, flexible and personal a technology is to it's
users.

Researchmcdel

The predictor model used for this study was developed in
1980 by Fishbein & Ajzen. The model was a refinement of their
"theory of reasoned action" approach. The theory is based on
the assumptions that humans will make use of information

available to them in a systematic and rational manner. People

The predictor model used for this study was developed in

1980 by Fishbein & Ajzen. The model was a refinement of
their "theory of reasoned action" approach. The theory is
based on the assumptions that humans will make use of

information available to them.ixiaa systematic and rational
manner. People will consider the implications of their
actions before they engage in a given behavior. These
considerations are formed by attitudes, beliefs, and those
feelings that a user's behavior would be consistant with
those of their fellow workers. Fishbein & Ajzen believe
that behavior can be explained. by a series of steps
beginning with external variables such as demographic
information, then their beliefs and attitudes about
performing time behavior, their Intentions tx> perform the

behavior and finally performing the actual behavior.

WW

#1. Is there any significant relationship between
individual variables such as age, cmganization tenure, job
type, sex auui marital status, euxi users' satisfaction of
telemedicine?

#2. Is there any significant relationship between
individual variables such as age, cuganization tenure, job
type, sex, and marital status, and the behavioral acceptance
of telemedicine?

#3. Is there any significant correlation between
attitudinal variables such as credibility, sociability, ease
of use, and production quality, and the behavioral
acceptance of telemedicine?-

#4. Is there any significant correlation between
attitudinal variables such as credibility, sociability, ease
of use, and production quality, and users' satisfaction of

telemedicine?

#5. What individual and attitudinal factors will best
predict behavioral acceptance of telemedicine?
#6. What individual and attitudinal factors will best

predict users satifaction of telemedicine?

It is inmortant txa clarify that satisfaction does not
necessarily lead to behavioral acceptance. In most cases
satisfaction will lead tx> further behavior. Satisfaction
may lead to acceptance but a user may be satisfied and still
not accept telemedicine as a: supplementary information
source. Respondents may be satisfied by indicating good or
excellent quality to a measure, but may still disagree that
telemedicine is a viable alternative to face-to—face

education.

Procedures

After discussion with network managers, it was decided
upon that the user groups which will best serve this project
are physicians, nurses and hospital administrators/managers.
It was felt by Healthcare Information Network President,
David Levine, that these three groups comprise the majority
of the Michigan Hospital Information Network users. Three
questionnaires were sent to each of the 125 member hospitals
of the MHIN., to be completed by one respondent from each of
the user groups at each hospital throughout Michigan, for a
total CH? 375 questionnaires dispersed” The questionnaires
were sent out by the Lansing bureau of HIN to the
teleconferencing coordinators at each member hospital with
instructions on tune to select 21 random sample. This was
done to uphold a contractual agreement that no one ever be
given or sold.ea list of the nember hospitals. Three weeks
after the questionnaires were sent out, a fellow—up letter

was sent to

the coordinators as a reminder to prompt them to return the
questionnaires. In an effort to enhance the return rate of
the surveys, a final follow-up phone call was made to each
member hospital.

The final count on the returns was 136. Of these, 15
were not filled out, with the explanation that physicians at
those hospitals did not view any of the programs and 121 were
returned completed for a return rate of 34%.

: . E E' 1.

Personal and attitudinal factors towards behavioral
acceptance and satisfaction do not differ significantly
between the three target groups.

Only one personal variable, experience, was a significant
predictor of behavioral acceptance and satisfaction.

There are significant correlations between the
attitudinal factors <xf delivery, utility, interest, and
satisfaction while sociability is significantly correlated
with behavioral acceptance.

There is a statistically significant relationship between
attitudinal factors such as program quality, speaker quality,
and behavioral acceptance.

Implmentatinn

The findings of this study point out areas of interest to
the users which should be evaluated by network managers and
programmers. While satisfaction does not necessitate
acceptance, the factors which increase user satisfaction could
do no worse than to create a more favorable experience by
those users who may not be able to get this information via an
alternative source. While program suggestions which were made
by users are an easy way to implement the aim of user
concentration, it is necessary to examine the factors which

user concentration, fix: is necessary tx: examine tfimz factors
which affect users attitudes. The findings of this project
enable programmers to get a better understanding of which
production factors need additional attention, how personal
variables affect behavioral acceptance and which attitudinal
factors should be addressed by possible changes in the aim
and delivery of healthcare information. The changes in aim
and delivery of information may prove to be the best way to
enhance behavioral acceptance of telemedicine. Programmers
and managers will be able to utilize information which
details how users receive the network and it's programming.
Programmers can now concentrate on several factors which
will lead.tx> a more personal and compatable acceptance by
users. Managers will now be able to eliminate some of the
guesswork in determining who their users really are. This
should lead to the elimination of programming geared to a
non-existant audience anui lead tx> additional programming
time for those audiences who really need it.

'These findings will allow managers to better meet the
need for the types of pmogramming created by the changing
political environment. Regulating agencies in Michigan such
as the board of nursing and the American Medical
Association, are helping to draft legislation for mandatory
continuing education. This study gives the decision makers
the viewpoints of the users who have opinions and

suggestions on how to meet these growing needs

CHAPTER II
THEORETICAL RATIONALE
Marxian

Much has been done to study the use of mediated
communication as an alternative to face-to-face instruction.
Most of the studies have dealt with mediated messages as the
sole training device and not as a supplement to conferences,
trade shows or other face-to-face alternatives. While this
is useful to some professions, the healthcare field is being
inundated with new technologies, techniques, drugs,
treatments, modern patient methods and changes in law and
policy. :n: is difficult for most, and impossible for some
healthcare professionals, to get the most up-to—date
information possible. There is no one delivery system that
will answer all the concerns necessary. Likewise, the study
of continuing education for these healthcare professions has
no one theoretical base.

The main theoretical base for this chapter is derived
from:

-Behavioral acceptance models developed by
Fisbein & Ajzen. These models serve to explain
the theoretical linking of beliefs to behavoirs
auui the determining factors which. affect the
formation of attitudes and beliefs.

-Attitudinal research by McGuire which suggests
that the receiver go through a five step

process if persuasion is to succeed

—Diffusion and adoption research by Rogers who

10

outlines five characteristics of innovation.
Each characteristic must be consistent with
preexisting norms to provide any relative
advantages.

-Research by Chu and Schramm who generalize

about research done on the effectiveness of
television as a learning source.

-Olgren & Parker et al., whose research leads to
suggestions on how teleconferencing instructors
can enhance learning and retention of

participating students.
Dependenuariahles

The dependent variables in this study are:
Behavioral acceptance - Usage cnf telemedicine
as a supplementary information source.
Satisfaction - 'Subjective evaluative response
towards an telemedicine based on previous experiences with

telemedicine use.

Individual variables such an; demographic information
should play a very small roll in 'explaining either
acceptance or satisfaction but may lead to a better
understanding of user motivation.

Age - Although not expected to account for much
variance, age may still be an important variable in this
organizational framework.

Job type - A level of organizational status
which. should. provide 21 better' explanatory rationale for
differences in predispositions which affect behavioral

acceptance.

11

Organizational ‘tenure -.Another variable which
should be influential in the formation of attitudes which
are formed and affect relationships between dependent and

independent variables.

Attitudinal variables, which have been shown to be
significant in earlier research, are formed by beliefs which

are derived by many factors.

The totality of a person's beliefs serves as
the information base that ultimately determine

his attitudes, intentions and behavior.
(Fishbein & Ajzen, 1975 p. 14)

It is important to determine those factors which affect
attitude formation for predicting further steps in the

behavioral acceptance model. These factors are:

Program quality - Perceived importance of
technical quality as well as the style, length and quality

of the speaker.

Ability and importance of being able to
provide feedback — Subjective response to the users
capacity to provide their opinions and attitudes towards
programming practices and specific teleconferences.

Other attitudinal variables include:

Compatability'- An attitudinal measure of how
consistant a technology is with an organization's
preexisting set of values within a corporate culture.

Credibility - An attitudinal measure of a
technology’s perceived ranking with its alternatives and the

perceived believability of that technology.

12

Sociability - The degree of personalization,
sensitivity and socializing afforded users when utilizing a
technology.

Ease of Use - The degree of complexity,
convenience, comfort and flexibility that respondents

perceive in a technology.

Determining these attitudinal factors and the relationship
of other variables mentioned should lead tn) a. greater
understanding to what ultimately leads to behavioral
acceptance of telemedicine. By analyzing how this
acceptance process differs from group to group, whether age,
tenure, or job type, will allow managers and programmers to
alter methods or practices to help reach target audiences

and aid in behavioral acceptance.

MW

Early researchers assumed 51 more) direct relationship
between a person's attitudes and feelings and some
measurable behavior. Attitude concepts have only recently
been understood. as being complex concepts to measure.
Earlier research by Cartwright (1949), Smith (1947), and
Katz & Stotland (1959), led to Rosenberg and Hovland's
(1960) three-componant view (ME attitudes (Fishbein, 1980).
It is this research which has led to current measurement of
attitudes and the determination of the factors which are
most important to a theoretical understanding of behavioral
acceptance.

Administrators can cnflgz order rmnv technological systems
to be installed. They cannot make physicians, nurses,
dieticians, housekeepers, hospital managers, anyone for that
matter, accept a technology especially one which will alter
preexisting communications models. This realization leads

to the study of behavioral acceptance.

13

The behavioral acceptance models utilized for this study
were developed by Fishbein & Ajzen during their studies of
behavioral prediction which began in the late 19505 (Ajzen &
Fishbein, 1980, p. 5). From their years of research dealing
with the predicting of behavior in laboratory and applied
settings, they have developed several behavioral theories
and models. The "theory of reasoned action" is an approach
that Fishbein and Ajzen introduced in 1967 (Fishbein, 1967).
This theory is based on the assumption that humans will make
use of information available to them in a systematic and
rational manner. Fishbein & Ajzen argue that people consider
the implications of their actions before they decide whether
or not to engage in a given behavior. These considerations
of implications are formed by attitudes, beliefs and
normative components. They further ‘theorize: that people
consciously weigh the positive and negative affects of their
actions before engaging in that behavior (Fishbein & Ajzen,
1980). Simply stated: If the outcome from performing a
behavior is perceived as receiving a positive outcome, the
person should feel good about that behavior. Figure 2.1 is a
model which represents the determining factors of a person's
behavior. This model allows us to get a more comprehensive
account of the causes of the underlying behavior (Fishbein,
1980, p. 7-9).

Figure 2.1 shows how behavior can be explained by a
series of steps in a sequence which begins with a person's
beliefs. Beliefs that certain outcomes will result from a
particular behavior. Beliefs that 21 particular behavior
should be ignored or acted upon in order to comply with
preexisting norms. These beliefs lead to attitudes,

intentions and eventually, to performing the behavior.

14

 

The person's belief that
the behavior leads to
certain outcomes and his LI"
evaluations of these
outcomes

 

Attitude toward
the behavior

 

 

 

 

 

 

 

 

 

Relative importance
of attitudinal and
normative
considerations

Intention—v Behavior

 

 

 

 

 

 

 

 

 

 

/\

The person's beliefs

that specific individuals
or groups think he

should should not Lu Subjective Norm
perform the behavior and
his motivation to

comply with the specific
referents

(Arrows indicate the direction
of influence)

 

 

 

 

 

 

 

FIGURE 2.1

Factors Affecting Behavioral Acceptance

(Ajzen & Fishbein, 1980, p. 8)

One point that makes this model different from many others
is the emphasis on the person's attitude toward a particular
behavior rather than their preexisting attitudes and
prejudices towards objects, people cu: institutions. This
model does run; however factor ix: external variables and
explain their effects, though they may be indirect, on
behavior. Fishbein & Ajzen believe that there is not
necessarily any relation between any given external variable

and behavior. They further theorize:

Some external variables may bear a relation
to the behavior under investigation while
others may not, even when a relationship is
discovered, it may change over time and from
one population to another.

(Fishbein & Ajzen, 1980, p. 9)

15

Further behavioral research by Fishbein & Ajzen which
does include external variables :hi the explanation of
acceptance has some theoretical roots in the early attitude
measurement done by Thurstone. Thurstone applied
psychometric methods to the problem of attitude measurement.
Thurstone defined attitude as "the affect for or against a
psychological object" (Thurstone, 1931). He devised a scale
which :measured. attitudes ‘towards opinion statements and
assumed that differing degrees of favorableness or
unfavorableness toward the attitude object could be
evaluated and compared to other scores. In his scaling, the
attitude score represents a person's evaluation of an object
implied by a set of their beliefs, intentions, or actions.

Thurstone stated the implications of this measure:

It is quite conceivable that two men may
have the same degree or intensity of affect
favorable toward a psychological object and
that their attitudes would be described in
this sense as identical but that they arrive
at their similar attitudes by entirely
different routes.

(Thurstone, 1931, p. 263)

This development made it clear that although a person's
attitudes toward an object can affect behavior, there is no
necessary direct relationship between attitude and any given
behavior. Fishbein & Ajzen created another model which
helps to describe the indirect effects of external variables
on behavior, figure 2.2. These indirect influences are
shown here to have some effect on a person's behavioral
beliefs, outcome evaluations, normative beliefs, motivations
to comply, or how strong the relative weights of the

attitudinal and normative components are perceived to be

16

(Fishbein & Ajzen, 1980, p. 84-85). The relationships
between the external variables are treated almost as
supporting evidence in this theory of reasoned action. An

example used by Fishbein & Ajzen includes this model to
explain an observed relationship) between education and
smoking. It is assumed that there will be fewer smokers
among higher educated individuals because of their knowledge
of the health risks of smoking. Education may be a variable
but it is highly unlikely that a direct relationship would
exist between smoking and education.

These external variables should not be overlooked however

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

in the molding of a person's beliefs. Beliefs play a major
role in persuasive communication. If messages are to be
EXTERNAL VARIABLES

Qemogrgphic Variables f

f ' 5:? Beliefs that the

SOCFP‘ 1°“ . St t behavior leads to '

°i°i°°mm ‘ “5 certain outcomes ‘h‘ Attitude toward
Religion ' the behavior
Education ' f ‘
Evaluation of 4

Attitudes towards the ““9“” may. fl

t$11993: importance —— .

Attitudes toward eedh- of attitudinal Intention Bebav1or

people and

Attitudes towards 3911,15 that specific‘ normative .

institutions eferents think I J componants

should or should not

Personality traits perform thfi b9h!Vi°f Subjective I

Intro-Extraversion Motivation to IR?” .

“é“r°t§°isi , comply with the ‘

Authoritarianism specific referents

Dominance

Light Arrows- possible exlanations for observed
relations between external variables and behavior

 

 

 

Heavy Arrows- stable theoretical linking beliefs to
behavior

FHHWEJLZ

Role of External variables in Predicting Behavior
(Ajzen & Fishbein, 1980, p. 9)

17

successful in the process of attitude change, it is
necessary to neasure Innv a person receives information and
how messages are comprehended. Research by McGuire (1968,
1969) suggests that attitude change involves a: series of
five steps: attention, comprehension, yielding, retention,
and action.

The receiver must go through each of these

steps if communication is to have an

ultimate persuasive impact, and each depends

on the occurrence of the preceding step.

(McGuire, 1969, p. 173)

To fully understand how to utilize a persuasion model, it
is necessary to apprehend the effects of these external
variables on behavior. From figure 2.2 you can see how
external variables like demographics, attitudes, and
personality can be used for possible explanations to the
evaluations and motivations that are in fact stable
theoretical relations linking beliefs to behavior.

Until very recently, the nature of an individuals work
and job type have been somewhat overlooked. Research by
Steinman and Shulman (1978) theorized a link between
contextual factors, such as work type, needs, group
favorability, and group norms, and puedispositions toward
video-conferencing. Organizational teleconferencing
research by Downs and Mohr (1979) contends that potential
users of teleconferencing will choose to give the technology
a fair trial in order to assess its costs and benefits
accuratelyu Users predispositions towards teleconferencing
were explored In; Svenning (1982). Independent variables
such as: Background, attitudes, communication requirements,
perceived needs, organizational norms, work characteristics,
perceived attributes, and expected impacts, were all
included in the regression equation. Svenning and

Ruchinskas (1981) explored the influence of job type in

18

predicting behavioral acceptance (n? teleconferencing. Job
type was found to be a significant predictor of
teleconferencing use :hi their study. These findings
however, have not been explored and replicated in other
studies to help explain behavioral acceptance among users.

Innovation and change has long been the answer to a
change in perceived need (Brief, et al., 1976). Although it
has been stated that findings of a job type linkage to
predicting teleconferencing use (Rice, 1984, I). 235) were
somewhat unexpected, the adoption of innovations answers the
gaps between the expected and the actual that March and
Simon talked about (1958). The study of needs and
expectations of new technologies answering these needs has
been studied in almost every study done on acceptance.

From this theoretical base, we can infer the theory of
reasoned action to explain a rational thinking and reasoning
process by potential users. Utilizing the second behavioral
acceptance model developed by Fishbein & Ajzen, it is
necessary tn) determine which external variables, beliefs,
motivations and attitudes are important in predicting
behavioral acceptance. Within this model are demographic,
attitudinal enui personality' variables. This Ibehavioral
acceptance model also includes factors of compliance with
normative zuui accepted practices. While IN) direct causal
relationship can be inferred, understanding how these
variables affect the formulation of attitudes will aid in
the prediction of behavioral acceptance.

Based on the review of this literature, a hypothesis can

be formed:
WM
Only two personal variables, job type and organizational

tenure, vfiJJ. explain. significant ‘variance 1J1 behavioral

acceptance and satisfaction.

l9

E'EE . i i .

The theoretical framework for the study of the diffusion
process was presented by Rogers in 1983. The five main
characteristics of innovation he listed were; relative
advantage, complexity, observability, trialability and
compatibility (Rogers, 1983, pp. 210-240). All five of
these innovation characteristics must be consistent with an
organizations preexisting communications system for
successful diffusion to take pflece. Teleconferencing has
proven it's ability to provide many of the relative
advantages over present training practices such as time and
money saved from sending staff members to seminars. Too
often, economics and the inability to replace large amounts
of participating staff members, mean only a select few
personnel travel to these training sessions. This makes the
possibility of telemedicine training attractive and somewhat
compatible. What has not been researched as often however
is whether or not people in the medical field feel that the
training they receive through teleconferencing is thorough
enough and as equally beneficial to their medical practices
as other forms of training they may receive. Legislation
currently mandates that physicians participate in continuing
medical education in order for them to continue practicing
medicine. Physicians are among the first to admit however,
that abuses exist in the current system of monitoring their
continuing education efforts. :n: is common practice for
physicians to get off a plane in some tropical paradise,
sign in to receive the credit and spend the rest of the time
working on a tan or lounging around a pool (Rothert, 1987).
The luxury of being able to participate in this kind of
abuse will naturally' lead. some jphysicians tx: say that
teleconferencing is not compatible with their current

continuing medical training practices.

20

As Acker summarizes in a paper entitled Tim;

- - ea - -. - '. ..' ‘ ..- o - '09 o- a' '..
WWW, much of the current
research on teleconferencing has focused on the relative
advantages rather than exploring whether teleconferencing
may change the culture of an organization. Cost
effectiveness, increased participation auui time saving nay

not be good enough reasons for adoption to take place.

If the complexity of teleconferencing could
be better explained, if system users could
establish a low—risk background, and if
teleconferencings' advantages could be made
observable through concrete demonstration,
the technology might be viewed as compatible
with the culture of more organizations.
(Acker, 1985, p. 211)

The key word here is compatibility. Compatibility, as
defined by Rogers, is every organization possessing a set of
values within a corporate culture. The members of an
organization must decide whether the use of a new technology
is consistent with currently held beliefs and practices and
whether the relationships among individuals in the
organizationtion will continue undisturbed. This analysis
will take place before they will decide to accept or reject
a new technology (Acker, 1985, p. 211).

Many users have been conditioned to concentrate on the
technology itself rather than how to deal with the loss of
interpersonal and other forms of communication no longer
present without face-to-face interaction (Williams, 1978).
Participants might alter their patterns of agreement and
disagreement. People who would normally speak up might be
too apprehensive because of the presence of the technology.

Face-to-face interaction on the other hand, "Encourages the

21

intrusion of interpersonal feelings, which is of benefit to
the person arguing his/her personal beliefs" (Williams, p.
126). ther researchers such as Chapanis and associates,
have tested a wide range of media types against face-to-face
communication. Almost without exception, the messages given
through mediated form were shorter and involved less
interaction among participants than did face-to-face
(Chapanis en: al., 1971,72,74,76). These shorter messages
may be a result of the lack of time given participants
during mediated communication. Shorter messages may also be
caused by 21 fear of the technology and it's intrusion on
interpersonal communication. If either of these suggestions
are legitimate, psychological blockades may prevent even the
most logical argument for a technology from being accepted.
Barriers which are harbored and reinforced by disgruntled
early users may be extremely difficult to surmount. .

One of the bdggest reasons for physicians reluctance in
adopting teleconferencing may be the fear that they will
lose one of their longest standing' perques - travel.
Doctors have long ibeen given. red. carpet treatment at
conventions and seminars where they build up their social
contacts and can get away from the office (Ribble, 1985).
When subjects were given the opportunity to give reasons for
a teleconferencing system failure, of a system not yet
developed, the top responses were:

-The hospitals were unwilling to pay for the
service

-Doctors were unwilling to attend.tx> "watch
television"

-No one used it

-The faculty were unwilling to spend the
time to produce top quality programs

-The loss of face-to-face contact was
disappointing

(Ribble, 1985, pp. 317-18)

22

While the nature of this study is somewhat realistic,
subjects were more likely to comment on a non-existing
system because they did not fear any consequences to anyone,
it is interesting to study these responses. fflne top four
reasons given for failure of the system were all related to
persuading people to accept the system. Social
psychological research says that this resistance is rooted
in fear which. is the primary' motivation of inhibitory

behavior.

The fear of losing tangible benefits through
the acceptance of high tech communication
without compensatory high touch pmivileges
may be the most important inhibitory factor

affecting teleconferencing.
(Ribble, 1985, p. 317).

The fear that exists in most people is rooted in the
apprehension of what change will bring us. Political and
social pressure are often catalysts for change to avoid a
degradation of services that often occurs when technology
displaces the present work process (Cooley, 1981). The
medical profession has fallen under such pressures and has
been attempting to integrate a system of cooperation between
machine and man for the betterment of the patient. This is
evident in the areas of medical diagnosis, patient
monitoring and laboratory automation that has already taken
place (Child, et al., 1985). These changes, as well as many
others in the healthcare field, have been consciously
rejected by personnel because of their perceived rebellious
nature against established medical practices ( Child, et
al., p. 422).

While these changes in treatment and decision making may

be of greater consequence, the adoption of a new

23

supplementary training method is often greeted just as
rudely, as innovators of teleconferencing systems have found
out.

The role of initially adopting a teleconferencing system
usually falls on the shoulders of hospital administrators.
These administrators are often not prepared for the amount
of dependence on others they will become and even less
prepared to determine the effect that such a change will
have cut present communication practices and rmmnm. Their
dependence (n1 technicians, external organizations, program
coordinators, consultants anui many' others will increase
greathu These fears, along' with those of Ibudgetary,
underutilization and others they already face, strike even
more fear and reluctance into these administrators (Ribble,
1985).

Several hospital networks have arisen to take some of the
fear and reluctance away. There are still too many complex
issues and misinformation for many administrators to deal
with” David Shively, Director' of the .Association of
Hospital Television Networks, on the current state of
teleconferencing:

Until proponents stop telling administrators
what the technology is and start emphasizing
that it can have a positive impact on the
bottom. line, they' will be rewarded. with

study indifference.
(Shively, 1984, p. 23)

Administrator's fears of reliance on other factors should
not keep a good learning source from being adopted. It is
seriously doubtful whether such a source even could be kept
out of a medical facility if in fact it were a universally

accepted method of education.

24

Potential users of a new technology must evaluate a given
technology to determine whether the innovation is consistant
with preexisting norms before the diffusion and adoption can
sucessfully take place. Five characteristics are utilized
in the evaluation process, they are: relative advantage,
complexity, observability, trialablity and cmmpatibility.
While many outside factors (such as political environment
and fear of reliance on others) also affect the decision
making process,’ before adoption can take place, all five
characteristics must be consistant with widely shared norms.

Based on the review of this literature, a hypothesis can
be formed:

Wm

Attitudinal factors, such as compatability, credibility,
sociablity and ease of use will be significantly correlated

with satisfaction and behavioral acceptance.

W

In summation of the instructional television literature
presented by Chu and Schramm in their book W
Waugh—Salts, they raise several
generalizations worth further study. On the subject of
whether students actually learn from television, research by
Schramm (1962) indicates that overall, instructional
television is more effective for primary and secondary
students tfluui for college students. Other studies have
replicated this research. and findings have generally
indicated that there is little difference in overall scores
of those students who received instruction through mediated
sources verses those students who were taught face-to-face.
Schramm indicates that although younger students are more

open to newer forms of instruction, the older a student

25

gets, the less likely they are to accept less conventional
forms of instruction. Schramm summarizes that the more
complex: the material gets, the ‘more students feel the
necessity for immediate feedback or the ability to ask
questions (Schramm, p. 7). Research by
(Monson, 1978; Olgren & Parker 1983; Bronstein, Gill, &
Koneman, 1982) on television as a teaching tool, resulted in
a list of suggestions for instructors to improve teaching
behaviors in order to aid in facilitating a successful

teleconferencing experience. These suggestions include:

-advanced organization and preparation of
written materials for students

-providing students printed materials before
each class

—previewing what is to be covered in each
class

-repeating and summarizing main points
throughout class

-speaking to individuals, using their names,
not to the whole audience

-using various instructional techniques,
e.g., question and answer periods, case
studies, role playing

-providing a summary at the end of each
session

-personalizing comments to sites or
individual students

~utilizing queStions to keep students
involved

-pausing to ask students for feedback about

the class

(Shaeffer & Roe, 1985)

26

It can be argued that many of these suggestions apply to
face—to—face or any type of instruction. While many of
these suggestions would be helpful in teaching under any
circumstances, they are especially important in a
teleconference situation.

A study done in 1984 by Haaland and Newby, utilizing the
work of Murray (1979), examined perceived teaching behaviors
by students in an audio teleconferencing course.' Although
audio - only was involved, the questionnaire that Haaland
and Newby designed concentrated on 5 "dimensions associated
with instructional behaviors." The five are: clarity,
enthusiasm, organization, pacing, and student participation.
The attempt in this study was to find out what facilitates
effective teaching in the absence of visual stimuli.
Haaland and Newby's comparison of both methods of
interaction found that the instructional method "had no
effect (n1 the students overall ratings (ME the course and
instructor" (Haaland and Newby, 1984).

This study was replicated with some modifications by
Shaeffer and Roe in 1985 at the University of Wyoming.
Video teleconferencing was added tt) the study which
included. audio and *videotaped inserts along' with class
discussions. Results from tflua questionnaire <distributed
indicated that teleconferencing students gave higher ratings
to the instructor than the on-campus students. The six main
areas (of study: overall impression, clarity, enthusiasm,
organization, pacing and student participation, all yielded
higher scores for the teleconferencing instructor than for
their face-to-face counterpart (Shaeffer and fun» 1985).
While there may be several other reasons for the differences
in the ratings given the instructors in the Shaeffer, Roe
study, an attempt was made to shed some light on the
necessity of a teleconferencing instructor to prepare their
lessons somewhat differently than they would for a

face-to-face lecture. :n: is possible that time instructor

27

may have been aware that they should be fostering more
attention t1) the teleconferencing students. This emphasis
on ndssing stimuli for the teleconferencing students may
have somewhat bored the students receiving their lessons
face-to-face. While only speculation, this is aa possible
explanation for some of the effects found in experiments
which compare mediated and face-to-face communications.

Early research by Chu and Schramm lead to a further
concentration (n1 television as ea teaching tool. Schramm
indicated that instructional television is more effective
for younger students than for college students. The more
complex the material gets, the less likely people are going
to accept a mediated form of delivery for the subject.

Later research summarized that there was basically no
difference in the scores of individuals which participated
in mediated teaching anui their face-to-face counterparts.
From this work came several suggestions on how to overcome
the loss of face-to—face elements when teaching via mediated
method. These suggestions included previewing and
summarizing the material whenever possible, preparing
written materials on the subject matter for students,
pausing for questions and feedback and personalizing the
message more than you would do in face-to—face
communications. Some of the suggestions made would not be
possible when more than one site is being utized but Olgren
and Parker et al., propose to alter the instructors teaching
methods to fit the situation.

This theoretical base was used to begin a study on
behavioral acceptance of telemedicine by the three main user
groups of aa healthcare information network. The Michigan
HOspital Association was one of only three such healthcare
networks :uu the 'United. States delivering information ‘to
hospital personnel.

Based on the review of this literature, a hypothesis can

be formed:

28

WW
Attitudinal factors such as program quality and speaker
quality will be significantly correlated with satisfaction

but not with behavioral acceptance.

If 1. H . J E . .

To fully understand the particular setting in which this
study was to take place, several people were interviewed to
get information on how the telemedicine system has evolved
and how users attitudes and opinions are utilized to Hake
any necessary changes.

The first of these interviews was with Ms. Marlene
Hulteen, Director of member services for the Michigan
Hospital. Association. Ms. Hulteen jpointed cnn: that a
feasibility study from the user standpoint was never done.
The only early research that was done was a cost analysis by
the accounting firm of Ernst and Whinney in 1982. From that
research, 125 of the 206 hospitals surveyed were willing to
invest in an MHA telemedicine system. The member services
divisicmi of MHA. Communications Systenn a wholly owned
subsidiary of the MHA Mutual Insurance Company, acted as the
program producers of the system. The inaugural program, a
thirty minute bi-monthly intelligence report, was launched
in July of 1984. In addition, member services was
contracted to provide seven other hours of programming per
month. The one person in charge of producing these programs
had rm) television experience and IN) research tx> let them
know the wants and needs of their target audience. After
nearly a year of "floundering", needs assessments,
interviews and other types of user information was gathered
on the system and it's ability to support itself through
pay-per-view programs the way the parent company had
intended. It became very clear that users were willing to

watch these programs but not if they had to pay to do so.

29

Users felt that since they were already paying a monthly fee
for the service, they should not have to pay anything extra
(Hulteen, 1987). These findings parallel those experiments
and observations by owners and managers of cable television
operations (Baldwin, 1983). Because of costs and the amount
of programming needed to fill the schedule, MHA had to rely
(Mi other healthcare programmers like IMTL .Aren, vanderbilt
and others. In or around May of 1986, competition had grown

to a point where MHA had trouble competing with the other

networks at their present rates. In their first year of
operation, MHA's telemedicine system lost about $150
thousand. Their second year they cut their losses to about

$75 thousand but the state of Michigan had become saturated
with information vendors-(Hulteen, 1987). In June 1986, the
marketing and programming areas merged and fell under the
care of Marlene Hulteen. In October-November of 1986, the
investors looked for a buyer of their system. It seemed
natural for the Healthcare Information Network, who already
supplied some programming to MHA, to incorporate them into
their preexisting national healthcare network. It was under
this new structuring that the system could finally be used
for it's intended purposes of educating people in remote
areas, rural areas and ancillary personnel who never got out
of the hospital to gain the necessary continuing education
(Hulteen, 1987).

The central focus of the healthcare network has recently
been undergoing a major shift according to HIN Vice
President David Levine. Levine has stated that the network
is IN) longer focusing the majority (n5 their programming
efforts on pmysicians but rather (n1 other users who have
fewer information alternatives to choose from. In the past,
said Levine, sponsors of these networks were necessary to
help offset rising costs. The early backers tended to be
medical supply houses or drug manufacturers. These sponsors

insisted on programs that appeal to a group with the power

30

to influence purchasing decisions of their products. Almost
without exception, the target audiences were physicians. Now
that the technology is becoming more affordable, an
increasing number of smaller companies can afford to back
programming 1km: other healthcare personnel (Levine, 1987).
As more user research. is (done ‘however, HIN and other
healthcare networks are finding that physicians are not the
group that utilize the system. Doctors have opportunities
to travel to conventions and gatherings whereas nurses and
other personnel do not have many chances for continuing
education. Since the merger of MHA and HIN, no one has done
an in-depth study of users to find.cnn: if the increase in
resources has benefitted the target groups. Several major
programming decisions are going to be made in the fall and
much. of the information gained fem“ similar academic
research to predict behavioral acceptance can be utilized by
Levine, heads of the Healthcare information Network and the
Michigan Hospital Information Network, to aid them in these

decisions (Levine, 1987).
Model

Figure 2.3 represents a revised version of Fishbein &
Ajzen's behavioral acceptanCe model. Several of the external
variables have been modified for the study of behavioral

acceptance within the structure of the Healthcare Informaion

Network. Socioeconomic status and religion are IN) longer
relevant 1J1 this study. Neither' are tflua variables of
personality as defined by Fishbein & Ajzen. Since the three

target groups are::hi the healthcare profession, 23 certain
type of "people—oriented" personality is assumed. Additions
have been node ix) the external variables in time areas of
attitudes towards target. People's attitude towards time
infringement and the surrounding political environment are

especially important in 21 field. like Ihealthcare. The

31

availability and quality of information of alternatives are

also additions to the model. If potential users have an

abundance of quality alternatives, this may greatly affect

their behavioral acceptance decision. Another belief factor
which is present in the new model is fear of lost perques

and/or lost aternatives for funding. One could argue that

this is already accounted for in the model under subjective

norms but these fears or motivations are key in formulating
subjective norms.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

EXTERNAL VARIABLES
<3 1. I! . I]
298' se?‘ Beliefs that the
Eccupation behavior leads to
M::?f;fn::atu5 certain outcomes Attitude towar
Edubation the behavior
Evaluation of
Attitudes towards the outcomes .
Ia'git . Relative
, . importance
AFtlt9°ef5.t°ward P of attitudinal
time in ringnent
arid
Attitudes towards -
institutions Beliefs that specif' normative u
Attitudes towards referents think I componants
political should or should no
. erform the behavior , , _
envrronment p Subjective Satisfac-
, , Norm tion
Airgrnazjyes Motivation to
Availability and comply with the
perceived qua'ity of spec1f1c referents
alternatives
Fear of lost
perques,
alternatives

 

Light Arrows- possible exlanations for observed
relations between external variables and behavior

Heavy Arrows- stable theoretical linking beliefs to
behavior

FIGURE 2.3

Composite Model of Behavioral Acceptance
(Adapted from Ajzen & Fishbein, 1980)

32

Summary

Factors which are anticipated to be most important to

behavioral acceptance of telemedicine include:

Perceived attributes

Expected consequences

Compatibility with current communication models
Perceived educational needs

Ease of user operation

Quality of content and subject matter

Program length

Production quality

Scheduling Convenience

These expected relationships between the independent and
dependent variables are outlined below in the following

summary table: Satiaflyfl;fiuiBehavicral acceptance
W

Age

Sex

Marital status

Job type

+ -+ 0

++

Organizational tenure — -

Satisfacticn Behavicral acccpnancc

Z . 3° 1 . 1]

Program quality ++

Speaker quality ++ +
Credibility + ++
Sociability ++ +
Compatability ++ ++
++= strong positive relationship

+ = positive relationship

— = negative relationship

0 = no expected relationship

33

These factors along with job status anui other external
variables such as experience, age, job type and
availability of information alternatives, should be of great
use in predicting behavioral acceptance. Telemedicine is
defined for use in this study as the channel by which
continuing medical information flows. The importance of the
dissemination of medical information is not a question.
What is however, is whether or not this new technology can
answer the needs of the healthcare profession. The current
political environment which tflUr; profession. operates in
forces them to withhold the bottom line - profit margins.
Whether or not this technology can deliver economically is
one major reason why this study is so important.
Legislators and governing bodies like the board of nursing
are attempting to make major job decisions before assuring
the nursing profession that there is sufficient means in
which to update their personnel. This study will point out
possibilities to aid their cause and that it helps
programmers and. managers make more informed decisions.
These decisions should at least be made with an

understanding of the people that the decisions will affect.

CHAPTER III
Procedures and Methods
Qiremieu

The following chapter will define the Operationalization of
the research variables for this study, describe the context in
which the research questions were introduced, distinguish the
variables, describe the research methodology, provide the
rationale for the measurement tool and briefly state the
statistical analysis procedures.

Winn

The survey population was chosen randomly from each of the
125 member hospitals. Because of the aforementioned problem
of confidentiality, teleconferencing coordinators at each
hospital were sent a memorandom (found in appendix A) giving
them instructions on the procedures of random selection. They
were asked to distribute one of the questionnaires to a
physician, one to a nurse, and one to a hospital
administrator/manager. Then they were instructed to: make a
list of all the physicians that have seen at least one M.H.A.
or H.I.N. program or teleconference. From this list choose
the fifth (5th) name and give them a questionnaire to
complete. Make a similar list of all nurses that have seen at
least one program or teleconference and choose the eighth
(8th) name and give them a questionnaire to complete. Make a
third list of all the administrators or managers that have
seen at least one program or teleconference and choose the

34

35

third (3rd) name on the list and give them a questionnaire to
fill out. Then they were told that if any of the three
refused to complete the questionnaire, they were to go back to
the list, start with the name immediately after the refused
party and choose either the next 5th, 8th, or 3rd name
(depending on whether they were choosing another physician,
nurse, or manager).

Respondents

After discussion with network managers, it was decided upon
that the user groups which will best serve this project are
physicians, nurses and hospital administrators/managers. It
was felt by Healthcare Information Network President, David
Levine, that these three groups comprise the majority of the
Michigan. Hospital Information iNetwork 'users. Three
questionnaires were sent to each of the 125 member hospitals
of the MHIN., to be completed by one respondent from each of
the user groups at each hospital throughout Michigan, for a
total of 375 questionnaires dispersed.

Wt

The survey is 22 questions spanning ten pages, including
cover letter and comments page, (found in Appendix B). The
questionnaire is divided into several sections to increase
flow and continuity. Section one deals with viewing habits
and includes questions 1 through 6. Section two, questions
7-17, is aimed at gathering attitudinal information about
telemedicine in general and more specifically the Heathcare
Information Network. Section three includes questions 18-22
and inquires about the statistical data necessary for the
study.

 

36
E I . l'

Because of the length of the questionnaire, 10 pages with
the cover letter, several steps were taken to enhance the
return rate. The teleconferencing coordinators were instructed
to go back to the three respondants and collect the
questionnaires in person. It was felt that personal urging
from a familiar party would greatly enhance completion. If
respondants were not finished, the coordinators were to give
them a few more days but urge them to complete the
questionnaire as soon as possible. Upon collection of the
surveys, the coordinators were to return them in a
self—addressed stamped envelope which was provided for them.
Approximately three weeks after the questionnaires were sent
out, a follow—up memorandom (found in appendix C) was sent out
to all 125 hospitals to thank those who had already
participated and to remind the others to comply as soon as
possible. Approximately two weeks later, a personal call was
made to each coordinator who had not included their return
address (there were about 25 who did).

W

136 of the 375 questionnaires sent out were returned. 15
of the 136 were sent back with notes indicating that the
coordinators felt that their input would not advance this
research because they were relativly new members of the
Hospital Information Network. Breakdown of the returns by
jobtype indicates that 56 of the 121 questionnaires were
completed by nurses (46.3%), 43 by hospital managers (35.5%)

and 22 by physicians (18.2%). Further return breakdown
indicates:

-86% of the respondents were married
-68% of the respondents were female

37

These statistics make any analysis of the variable marital
status impossible and sex somewhat difficult. The final return
rate was computed at 34%.

Researchmethodologx

The research methodology selected for this study involves
correlation, factor analysis and. regression. we are
interested in testing the relationships of several variables,
to determine the factors most important in user satisfaction
and to determine the factors most important in the decision of
healthcare personnel whether or not to perform the behavior of
acceptance of telemedicine as a supplementary information
source.

: . 1' l' E 1 . l]
The dependent variables in this study are measured by:

Behavioral acceptance - Responses tn) direct
questioning of the number of health care programs
respondants viewed in the previous year.

Satisfaction - Measured by responses to 11
attitudinal statements about telemedicine. These
responses were then factor analyzed to determine
similarities, then reliabilities were run on these
factors to determine the strength of the factors.

Independent research variables:
Job type - Defined in this study by selection of

respondents as either a physician, nurse or
hospital administrator/manager.

 

38

Organizational tenure - Measured fur years
of on-the-job experience.

Age - Interval measurement taken by direct
questioning and although not expected to account
for much variance, may still be an important
variable in a health care setting.

Program quality - The audio and video quality of
the network as well as program length, pace, tOpic
and scheduling procedures, use of handouts and
quality of speakers. Measured by asking
respondents to rate several factors on a Likert
style scale as poor-excellent or three points in
between then factor analyzed to determine similar
variable loadings.

Ability and importance of being able to
provide feedback - The measure of the role of
interactivity in teleconferencing and the ability
for respondents to provide program input.
Measured by asking repondents to rate factors
relating to feedback on a five point Likert scale
then factoring the variables to determine likely
relationships.

Other attitudinal factors include:

Compatability - The ability of telemedicine to
deliver information to personnel without being
perceived as interfering with preexisting
communications structures. Measured by responses
to statements on a Likert scale about telemedicine
and. measuring attitudes to adjectives on a
semantic differential scale.

39

Credibility - Measuring lunv respondents view
telemedicine with other sources of information.
Measured. by asking respondents to rank 'the
importance of telemedicine with other information
sources as well as respondents attitudes on

credibility on a semantic differential scale.

Sociability - The degree of personalization,
sensitivity and socializing afforded users when
using telemedicine. Measured by factor analysis
of responses tx> attitudinal adjectives (n1 a

semantic differential scale.

Ease of Use — The degree of complexity,
convenience, comfort and flexibility that
respondents feel that there ii; in telemedicine.
Measured by factor analysis of responses to
attitudinal adjectives on a semantic differential

scale.

3 . . J E J . E l

A factor analysis was done on the independant variables
to reduce them tx: attitudinal factors about telemedicine,
about the Healthcare Information Network and about speaker
traits, for ease of analysis. 'The factors were them
submitted to reliability tests to determine internal
reliability. The 11 measures of the dependent variable,
satisfaction, were subsequently reduced to three factors or
indices of satisfaction. fNua attitudinal factors derived,
were then renamed and computed by adding their scores and
dividing the sum by the number of variables per factor.

A Pearson's correlation was then.inni on the attitudinal
factors and the interval level personal variables of age

and organizational tenure. In addition, these attitudinal

4O

factors were correlated via Pearson's, with the 13 derived
factors of satisfaction.

A chi-square or "goodness of fit" test was done on the
remaining personal variables to test the scores with those
which were expected from the sample.

An analysis of variance or "anova" was done on the
personal variables to test for significant differences among
group means.

Regression analyses were then done (M1 the dependent
variables to determine the variables most important in

predicting satisfaction and behavioral acceptance.

CHAPTER IV

Results

QiLemieu

This chapter will restate the research questions and
present the results from the statistical analysis which will
prove or disprove the hypotheses.

W

#1 Only two personal variables, job type and organizational
tenure, will explain significant variance in behavioral
acceptance and satisfaction.

#2 Attitudinal factors, such as compatibility, credibility,
sociability and ease of use will be significantly
correlated with satisfaction and behavioral acceptance.

#3 Attitudinal factors, such as program quality and speaker
quality will be significantly correlated with
satisfaction but not with behavioral acceptance.

Besults_nf_Eactor_LQading

Variables were reduced by factor analysis to create factors
which will serve as attitudinal factors, independent
variables, as well as satisfaction factors and dependent
variables.

The findings will be examined by how they relate to the
research hypotheses. The first step of the analysis is to
examine how each of the factors were formed. The factors were

41

W”*'_‘~‘“T~—.TT
1

div:

42

divided into two main types, attitudinal and satisfaction.
Ell'l l' J E I

Table 4.1 shows the 15 variables used to measure attitudes
about the Healthcare Information Network, which were reduced
to 5 factors.

EABIABLE DELIMERXEEKKKEJD:IECSLCINLHHUJEEDIINIEBESI
information quality.66

speaker quality .86

speaker ability .81

program learning .81

video quality .85

audio quality .90

topics .74

time per topic .70

timeliness of topic .83

broadcast time .75

provide feedback .72

ability to ask questions .70
interest in cont. ed. .80
pace .74
% of variance

accounted for: 37.5 11.6 9.7 7.7 7.2

TABLE 4.1
Attitudes Towards HIN Factor Analysis

The first factor formed was the attitudinal factor of
delivery. Delivery accounts for 37.5% of the 73% of the total
variance accounted for enui includes the ‘variables of
information quality, speaker quality, speaker ability and
program learning; The second factor formed was pmogram
quality which includes the variables of video and audio
quality. The factor includes only two variables but accounts

43

for 11.6% of the variance accounted for by the five factors.
Topic selection was the third factor extracted and includes
the variables topics, time per topic and timeliness of the
topic. Topic selection accounts for 9.7% of the variance and
includes three extremely similar variables. The final two
factors formed were user input and user interest. User input
includes the variables of broadcast time, ability to provide
feedback and ability to ask questions. The latter two factors
differ in that providing feedback is a way for users to give
their input on future topics, speakers and ways to improve the
network and ability to ask questions is an expression for the
need for clarification. User interest includes the variables
of users interest in continuing education and pace of the
programs. While this loading seems somewhat odd on the
surface, both variables are related to some level of interest.
User input accounts for 7.7% and user interest 7.2% of the
total variance accounted for by these five factors.

EABIABLE UIILIIX. EEBCEEIIME. .SQCIABILIIX

inexpensive .70

reliable .58

comfortable .73

easy .75

effective .81

convenient .81

simple .64

useful .89

personal .86

sensitive .78

sociable , .76

formal .76

% of. variance

accounted. for: 40.4 15.3 10.0
TABLE 4.2

Attitudinal variable Factor Analysis

44

An additional 12 attitudinal variables were factor analyzed
and loaded into 3 factors. These factors are related to users
attitudes towards telemedicine. The factor loadings are shown
in table 4.2.

Eight of the twelve attitudinal variables loaded on the
first factor, utility. Utility accounts for 40.4% of the
variance accounted for by the three new factors. Utility is
composed of the variables inexpensive, reliable, Comfortable,
easy, effective, convenient, simple and useful. It is
interesting to note that the variable inexpensive loaded with
variables associated with ease of use. While on the surface
these variables seem somewhat related, it appears that
respondents feel that they are the same. The second
attitudinal factor, perceptive, is made up of the variables of
personal and sensitive. .Perceptive accounts for 15.3% of the
total variance accounted for. The third factor of sociability
is composed of the variables sociable and formal. Sociability
accounts for 10.0% of the 65.7% of the variance accounted for
by the three factors. The factors perceptive and sociable
loaded with some interesting results. It is normally expected
that the variables of sociable and sensitive will load
together. It is interesting to note that the health care
professionals who responded feel that there is a significant
difference between the two variables. The nature of the work
they do may allow them to observe that telemedicine can be
sensitive to the feelings of users and their problems and
still not be sociable or vice versa.

Table 4.3 shows the 11 variables measuring speaker traits.
The variables loaded into 5 factors which account for 50.5% of
the total variance accounted for by the factors of speaker
traits. The first speaker trait factor to load was
expressiveness which is made up of the variables style and
humor. Convenience was the second factor to load and includes
the variables of pauses for questions, use of handouts and the
viewers location. The third factor loading was teaching
ability and includes the variables of speech clarity and

45

knowledge of the subject matter. Factor four, speaking aids,
is made up of the variables of graphics and speaker pace.
Respectability was the last speaker trait factor to load and
includes the variables of prominence and setting (background
and surrounding where the speaker is). A few of the
respondents used the comments page of the questionnaire to
make suggestions that the Healthcare Information Network
consider paying more attention to speaker quality and many of
the variables which loaded highly into these five factors.

EABIABLE EXBBEXNSanflflflflfliIEAQHBIXiflfiflflUEEiRSECIBIX
style .88

humor .63

pauses for questions .85

handouts .52

your location .47

speech clarity .95

knowledge .42

graphics .77

speaker pace .28
prominence .65
setting .35
% of variance

accounted. for: 16.9 12.8 8.8 7.5 4.5

TABLE 4.3
Speaker Trait Factor Analysis

Unlike any of the previous factor loadings, none of the
five factors of speaker traits load as strong and clearly
account for the majority of the variance. Expressiveness
accounts for the most variance of the five factors, 16.9%,
while respectability accounts for 4.5% of the total variance
accounted for by the five factors. Four of the five factors
related to speaker traits include only two variables.

46

In table 4.4 the 11 variables which were used to measure
the dependent variable, satisfaction, loaded into'3 factors.
The first factor to load was usefulness and includes the
variables of the usefulness of the Healthcare Information
Network, information gained through telemedicine being equal
to information gained by face-to-face methods, respondents
recommendation of telemedicine and the ability to replay
programs at a later date being helpful. Usefulness accounts
for 35.7% of the 61.2% of the variance accounted for by the
three factors of satisfaction. All variables make sense and
loaded very highly. Statistically, usefulness is one of the
strongest factor loadings in terms of variance accounted for.

YABIABLE LEEEIHAEEEi.EQQNQMlC IMMEDIAQX
HIN useful .80

telemedicine equal to f-t-f .78

recommend telemedicine .66

replay helpful .65

decreased travel .73

resources spent elsewhere .57

not willing to watch tv .70

good source of information -.57

interact with colleagues .80
learn more with live speaker .81
attention to live speaker .70
% of variance

accounted. for: 35.7 14.6 10.9

TABLE 4.4
Satisfaction Factor Analysis

The second factor extracted was economic, which accounts
for 14.6% of the variance and includes the variables of
decreased travel, resources spent on telemedicine better spent
elsewhere, users willingness to attend meetings to watch
television and telemedicine being a good source of

47

information. The statements were recoded for ease of
analysis so all the results can be compared. The variable,
good source of information, does not on the surface seem to
fit with the rest of the included variables and was dropped
from the factor after reliability tests were run. The final
satisfaction factor, immediacy, is composed of the variables
interaction with colleagues, respondents ability t1) learn
more from a live speaker and that it is easier to pay more
attention in) a live speaker. Again. the 'variables ‘were
recoded for ease of analysis. Immediacy accounts for 10.9%
of' the 'variance accounted for In; the three factors of
satisfaction. Immediacy deals with interactivity
behaviors, which do fit 'together' if you theorize that
healthcare professionals feel the need to be in some amount
of personal contact with others in their field and desire
immediate reinforcement cfif their attitudes aumi knowledge
gained.

The four sets of factors were then subjected to
reliability ‘tests ix) assess their internal reliability.
Variables were deleted from the factors where an improved
alpha level would be attained upon deletion of the variable.
Table 4.5 is a summary of the included variables in the
final scales.

ATTITUDINAL FACTORS
EESLQL V ' l ' ' i l l
Delivery Information quality, speaker quality,
speaker ability and program learning
Program quality Video quality and audio quality

Topic selection Topics, time per topic and timeliness of

topic
User input Ability to ask questions and ability to
provide feedback
Interest Interest in continuing education and pace
Utility Inexpensive, comfortable, easy, effective,

convenient, useful and simple

Sociability Formal and sociable

48

Expressiveness Style and humor

Convenience Pauses for questions, use of handouts and
location

Teaching ability Speech clarity and knowledge of subject
matter

Speaking aids Use of graphics and Speaker pace

Respectability Prominence and setting

SATISFACTION FACTORS

Eactcr Variables included in the final ecale
Usefulness Telemedicine is equal to face-to-face,

recommend and HIN is useful

Economic Decreased travel, resources spent elsewhere
and no watch television

Immediacy Attention to live speaker and interaction

with colleagues

TABLE 4.5

Factor Loading Summary

Factor scores were created by adding the variables in
each factor and dividing by the number of variables in that
factor. A summary table for the factors including the mean,
standard deviation and alpha levels appears in table 4.6.

The factor with the highest mean score is utility.
Utility had a mean score of 5.498 out of 7 which indicates
that respondents feel that ease of use is very important for
telemedicine. Other factors which displayed. high :mean
scores were teaching ability, expressiveness and smeaking
aids. Users also felt that the speakers ability to teach,
the ability for the respondents to express their views on
the subject matter, and the use of handouts and grahpics
were also important to them to enhance the telemedicine
experience. Three factors: respectability, speaking aids and

interest displayed low alpha levels which indicates that the

49

MABLAELES EAN SiDi ' ALREA
Delivery 3.517 .640 .876
Program quality 3.971 .737 .913
Topic selection 3.824 .584 .792
User input 3.213 .755 .664
Interest 3.659 .786 .490
Utility 5.498 .956 .880
perceptiveness 3.789 1.209 .674
Sociability 3.385 1.312 .512
Expressiveness 4.070 .586 .715
Convenience 3.612 .740 .636
Teachability 4.686 .419 .583
Speaking aids 4.017 .559 .441
Respectability 3.372 .726 .340
Usefulness 3.703 .659 .754
Economic 3.482 .815 .638
Immediacy 3.471 1.051 .750
TABLE 4.6

Summary Table for Factors

scales used to measure these factors may not be as strong as
scales designed to measure other factors.

The results of the study will be examined individually by
hypothesis.

Results

Hypothesis #1- Only two personal variables, job type and
organizational tenure, will explain significant

variance in behavioral acceptance and satisfaction.

The summary tables provided in table 4.7 indicates that
there are only 3 significant relationships, at the .05

level, between the predictor and the criterion variables.

50

mean
s.d. Relationship Between Job type & Dependent Variables
Usefulness 3.787 3.739 3.612
.586 .687 .662
Economic 3.621 3.434 3.472
.6024 .8526 .8675
Immediacy 3.227 3.455 3.616
1.151 1.050 .999
Behav. Acceptance 1.772 2.232 2.348
1.109 .1.464 1.541
Experience & Dependent Variables
Setiefecticn 5 c; less .d;lQ 11-15 16;;Q 21-25 25;
Useful 3.722 3.551 3.787 3.712 3.796 3.592
.422 .632 .706 .789 .617 .595
Economic 3.527 3.538 3.272 3.439 3.631 3.814
.642 .772 .843 .945 .845 .603
Immediacy 3.791 3.769 3.227 3.59 3.263 3.222
.940 .971 .893 1.385 .977 1.063
Behav. Accept 2.50 1.807 2.545 1.636 2.789 1.666*
1.381 1.20 1.66 1.255 1.397 1.118
mean
s.d. Age & Dependent Variables
Satjfifiggtjgg 25’3 Under 26-34 35—43 4 - 4 ijrfii
Useful 3.00 3.919 3.616 3.548 3.94*
.000 .520 .620 .791 .606
Economic 2.666 3.707 3.525 3.229 3.58*
.000 .771 .6831 .848 1.09
Immediacy 4.0 3.56 3.525 3.25 3.5
.000 .907 1.143 1.085 1.154
Behav. Acceptance 3.00 2.333 1.95 2.218 2.307
.000 1.613 1.244 1.337 1.797

 

 

g.-

~.v

 

 

51

Sex & Dependent Variables

Satisfaction Male Fem 1e
Useful 3.622 3.729
.570 .697
Economic 3.473 3.485
.734 .854
Immediacy 3.250 3.572
1.032 1.051
Behav. Acceptance 2.263 2.156
1.519 1.410
mean
s.d. Marital Status & Dependent Variables
Useful 3.393 3.746 3.466
.573 .608 1.483
Economic 3.272 3.533 2.866
.442 .816 1.215
Immediacy 3.727 3.466 3.0
.467 1.103 .707
Behav. Acceptance 2.363 2.152 2.60
1.747 1.412 1.516
* = analysis of variance pS . 05
TABLE 4.7
Relationships between Predictor and Criterion
Variables

Levels (fl? behavioral acceptance do run; differ by job
type as expected but do by respondents' level of experience.
Age is another personal variable that provides the other
significant relationship with. behavioral acceptance, in
particular the older and younger age groups. The middle age
groups do not show significantly different levels of
behavioral acceptance. While job type was expected to
sccount for significantly different levels of behavioral
acceptance, the relationship between experience and
behavioral acceptance was expected. Review of past research

did not reveal any indication that either sex or marital

52

status would show significantly different levels of
behavioral acceptance, especially the variable sexg It was
theorized that :Ui a healthcare environment however, where
dedication to jobs is commonplace and family obligations may
prevent the use of more expensive, time consuming
alternatives, the variable of marital status may account for
significant differences :hi the levels (ME satisfaction and
behavioral acceptance. As mentioned earlier, the fact that
86% of the respondents are married make further analysis of
this variable nearly impossible.

Age is significant when related to the satisfaction
factors of useful and economic. Both of these significant
relationships are understandable and exhibit a good deal of
face validity. The highest mean scores for the factor
useful, are for ages 55—64 and 26-34. One thing that both
of these age groups have in common is time. Younger people
have less of it while older people have more and do not have
the patience to waste it. There were not enough respondents
in the 25 and under age category, which is understandable
given the educational requirements of the job types, and the
middle two age groups have less mobility or are somewhat
"stuck" with whatever information source which allows them
time to fulfill other obligations such as family. The
highest mean score for the factor economic is 26-34. This
is very understandable as this is the group that has the
least money ix) spend.Organizational tenure ix; significant
when related to behavioral acceptance. Somewhat
surprisingly is the finding that the highest mean score in
the relationship of the two factors is the level of 21—25
years of experience. While it is understandable for people
to want to keep up to date on new innovations, past research
indicates a trend towards the fear of technology as we get
older. It. is possible that this tenure group has a
reasonable level of job anxiety and is fearful of upwardly
mobile younger' people ‘which is Luxxi tC) explain similar

phenomenon.

53

We can therefore conclude that research hypothesis #1 was
not statistically proven to be completely correct.' One other
personal variable, age, did create a significant difference
when related to behavioral acceptance and the satisfaction
factor economic. Job type and organizational tenure could only
manage to produce 1 statistically different set of variances.

Hypothesis #2- Attitudinal factors, such as compatibility,
credibility, sociability and ease of use
will be significantly correlated with
satisfaction and behavioral acceptance.

Hypothesis #3- Attitudinal factors, such as program
quality and speaker quality will be
significantly correlated with satisfaction

by not with behavioral acceptance.

Because factor loadings were somewhat different than
expected, factors such as compatibility and credibility
either did factor or loaded with other variables which make
logical sense. The variables associated with ease of use
factored with other variables such as inexpensive and
effectiveness. This factor was renamed utility.

Table 4.8 is a correlation matrix which indicates the
significance, strength and direction for each correlation.
From the table, 26 sets of correlations an be derived, 16
between attitudinal factors and the dependent variables, 2
between speaker trait factors and the dependent variables and
8 sets of correlations when the dependent variables were
correlated vnxxi each. other. The satisfaction factor
usefulness, is correlated with: delivery, topic selection,
interest, utility, perceptiveness and the dependent variables
economic and immediacy. Economic is correlated with user
input, interest, utility, perceptiveness, respectability and
the dependent variables usefulness and immediacy. Immediacy

is correlated with delivery, interest, utility, perceptiveness

Coeff

E:

Delivery

Program quality

Topic selection

User input

Interest

Utility

Perceptiveness

Sociability

Expressiveness

Teaching ability

Speaking aids

Respectability

Usefulness

Economic

Immediacy

pS.001

Pearson Correlation Coefficients

UsefulnessflconnminlmmedimBehaLment

.427

*‘k‘k

.132

.158

*

.091

.332

***

.440

***

.169

*

.035

-.061

-.049

.070

-.008

.354

***

-.377

***

54

-.O33

-.035

-.011

-.147

'k

.269

**

.471

*‘k‘k

.202

**

-.062

-.138

.032

-.O96

-.235

**

.354

*‘k‘k

-.406

***

TABLE 4.8

-.163

*

-.067

-.250

.135

-.320

***

-.346

***

-.313

***

-.006

.108

.054

.057

.139

-.377

***

-.406

*‘k*

-.128

-.261

**

-.079

”kit

-.132

-.093

.131

.053

.272

**

-.016

.058

-.169

it

-.024

.163

*

.094

-.156

'k

55

and the dependent variables usefulness and economic.
Behavioral acceptance is significantly correlated with program
quality, topic selection, sociability, speaking aids and the
dependent variables usefulness and immediacy.

The factors interest, utility and perceptiveness were
significantly correlated with all three satisfaction factors.
Statistically, utility proved to be the best factor when
correlated with satisfaction. All three of the correlations
between satisfaction and utility were at the .001 significance
level and all correlations are moderate in strength. The
strongest correlation is between utility and the satisfaction
factor economic.Immediacy and usefulness are the strongest
dependent variables when correlated with other dependent
variables. Both are significantly correlated with behavioral
acceptance but both correlations are weak. No one factor was
significantly correlated with all of the dependent variables.

Among the surprising results of the correlation analysis
was the lack of significant correlations between satisfaction
ans sociability, speaker aids and satisfaction. It would seem
on the surface that these attitudinal factors would correlate
with satisfaction but were noticably absent from the analysis.

W

In order to test the possibility of a linear relationship
and to determine which factors will act as predictors for the
Variance in Usefulness
Explained by Regression Model

% of total
luring—Variable Beta I Sig W
Utility .445 4.95 .000 19.8
Delivery .34 3.94 .000 30.8
Interest .223 2.6 .063 38.2
Topic selection -.206 -2.13 .035 38.2

F=l4.891

Significance=.000

r‘ nan. I. uxwxmxnsm

Indsnrlariable
Utility
Interest
Delivery
Respectability

56

Variance in Economic

Explained by Regression Model

% of total
Beta 1: Big explainedmr
.472 5.33 .000 22.2
.307 3.66 .000 31.6
-.305 -3.595 .000 39.5
—.224 -2.807 .006 44.2‘
F=19.077

Significance=.000
Variance in Immediacy

Explained by Regression Model

% of total
IndeeJariahle Beta 11 Sig explain“:
Interest -.328 -3.461 .006 10.7
Utility -.302 -3.33 .002 19.9
User input .211 2.34 .000 24.2
Perceptiveness -.246 -2.77 .004 29.8
Topic selection -.231 -2.523 .013 34.2
F=9.89O

Significance=.000
Variance in Behavioral acceptance
Explained by Regression Model

% of total
IndenJariahle Beta I Sig. explainedJar
Program quality -.387 -4.178 .000 14.9
Sociability .384 4.531 .000 29.7
Teaching ability .180 2.124 .036 32.8

F=15.806

Significance=.000

TABLE 4.9

Regression Summaries

57

dependent variables, a stepwise regression was run. Table 4.9
is a summary of the regression output.

All of the regression equations proved to be good
predictors by the amount of variance that the predictor
variables accounted for and the significance level of the
equations.

The best predictor variables for the satisfaction factor
of usefulness are: utility, delivery, interest and topic
selection. The total variance of usefulness explained by the
factored variables is 38.2%. Each of the included variables
appears fits into the equation theoretically and the equation
has a reasonably high F value which serves to support the case
that the included variables are good predictors.

The satisfaction factor economic, is best predicted by
the variables: utility, interest, delivery and respectability.
The total factor variance accounted for by the variables is
44.2%. Once again the variable utility, was the first included
in the equation and accounts for 22.2% of the total factor
variance accounted fOr In] the equation. The regression
equation for the factor economic, is the strongest in terms
of F values and total variance accounted for. Two of the
variables, delivery and respectability, are somewhat
negatively correlated which indicates that a change in the
delivery or respectability score will create a negative change
in a users economic satisfaction level.

The regression equation for the dependent variable of
immediacy, includes the factors: interest, utility, user
input, perceptiveness and topic selection. The total variance
in immediacy accounted for by all the variables is 34.2% and
all factors in the equation except for user input are
negatively correlated. The negative betas may be explained by
the constant change in the need for immediate information
gathering. A.change in any of the factors which make up
immediacy, for the exception of user input, will bring about a
negative change in the immediate need for the information.

Theoretically this does make sense because if topics, speakers

58

or other similar attitudes about the information to he
delivered changes, this will naturally affect the level
ofsatisfaction. Since these changes in perception are often
negative in nature, the shift in satisfaction will be negative
as well.

The regression equation for behavioral acceptance
includes only 3 factors: program quality, sociability and
teaching ability. While there are only three variables in the
equation, total variance accounted for is 32.8%. Each of the
included variables accounts for a fair amount of variance with
program quality accounting for 14.9% while sociability
accounts for 14.8% of the total variance accounted for in
behavioral acceptance by the three variables. Program quality
is somewhat negatively correlated meaning that a change in
program quality will mean levels of behavioral acceptance will
change in a negative direction. This finding is not
surprising because of how sophisticated our viewing
expectations have become. We expect that quality will be
equal to the broadcast quality that we receive on our
television sets at home. Any deviation from these
expectations may cause the user ix) seek alternative

information sources.

Summary

Hypothesis #1 was somewhat disproved by this research
project. No significantly different relationship between job
type and the dependent variables of behavioral acceptance and
satisfaction was uncovered. Organizational tenure did create
a significant difference when related to the dependent
variables.

In analizing hypothesis #2, attitudinal factors such as
delivery, topic selection, user input, interest, utility,
perceptiveness and respectability displayed marginal or
significant correlations with the satisfaction factors. Three

of the factors, interest, utility and perceptiveness are

 

59

significantly correlated with all three of the satisfaction
factors. Program quality, topic selection, speaking aids and
sociability were the only attitudinal factors that were
significantly correlated with behavioral acceptance.

Utility and interest proved to be the best predictor
variables as they appeared in all three of the regression
equations for satisfaction variables. The variables uncovered
by the regression equations accounted for at least 32% of the
total variance accounted for in each of the four equations,
with the equation for the factor economic accounting for 44%.

Hypothesis #3 was statistically disproved. The only
significant correlation which involves program quality is the
correlation with behavioral acceptance. Teaching ability did
not correlate with any of the dependent variables.

The factors uncovered by the analysis did prove logical
and significant in explaining and predicting levels of

satisfaction and behavioral acceptance.

CHAP TE R V

Conclusions and Recommendations

Oxerxiew

This final chapter' includes several conclusions drawn
from the research gathered for this study, past studies
related to this research and their practical limitations.
This chapter will conclude with some recommendations for
further research tn) better aid pmogrammers, managers and
researchers in gathering a better understanding of how to
meet the potential applications of telemedicine and similar
technologies.

W

This study has explored the application of a new
technology in 21 state-wide network settingu Two important
outcomes have been measured; satisfaction and behavioral
acceptance. An attempt has been made to predict those
factors which are believed to be most important for
increasing levels of these two vital outcomes.

Data was collected by distributing a 10 page
self-administered.rmxtl questionnaire. Questionnaires were
sent out to physicians, nurses and managers in the 125
member hospitals of the Michigan Hospital Information
Network. The attitudinal variables of telemedicine as an
information source, program and speaker quality and
information on respondents attitudes towards the network,

were reduced to several factors for ease of analysis.

60

T‘s—a. _

61

Analyses of variance were done on the personal variables
which were likely to influence satisfaction and behavioral
acceptance. differences in personal variables, such as job
type, Multiple regressions were run to provide analyses of
possible linear relationships between the predictor
variables and the criterion variables. From the regression
analyses, predictor models were constructed. to aid in
determining which factors are most important in increasing

levels of satisfaction and behavioral acceptance.
: J . I i' .

The analyses that were done provide managers and
programmers vital information about the attitudes of the
three main user groups of the Healthcare Information
Network. Research by Ajzen & Fishbein concludes that
attitudes of users must be understood before taking the next
step :hi predicting behavioral acceptance. This research
allows HIN managers to see what variables are most important
to the formation (ME users attitudes. The attitudinal
factors which statistically proved tx> be correlated mmst
highly with levels of satisfaction are: interest and
utility. Two key variables make in) the factor interest,
user interest in continuing education and pace. It cannot
be assumed that because a certain amount of continuing
education is required, people will find their experience
with telemedicine enjoyable. The material must be presented
in a way that users maintain an interest level. The other
variable which makes in) interest, pace, is related to the
first variable. People in the health care profession have
to deal with rising costs, slashed. budgets, decreasing
hospital populations and.ai changing political environment
where lawsuits often dictate policy and treatment. The last
thing that users want to do is to sit in a conference room,

or even in their own home, and view a program which is too

._ _.~_4_____ .._..—_- “M_- .nm

62

long and dragged out. These personnel are not used to
sitting around in meeting rooms or waiting for something to
attract their attention. Programmers must keep this time
schedule in mind. Speakers should be instructed on how to
convey their subject manner in a quick and concise manner.
The most important factor formed by this research was
utility. Telemedicine must be inexpensive, comfortable,
easy to use, effective, convenient, useful and simple if
personnel are to utilize the technology to its full
potential. Health care professionals have had equipment
that has tried to make their job eaSier since medicine was a
declared a science. It cannot be assumed that these
personnel will treat telemedicine, a tool cmeated for the
same purpose, any different than a stethoscope or any other
tool. For telemedicine to fully succeed, it will need to

become almost second nature to users like any other

instruments they use. Users will have to be trained how to
use the networks to get the most out of them. At present it
seems like this is not being done. Greater efforts to train

users how to get the most out of the programs they view is
almost as important as creating programs which are full of
useful information.

As important as interest and utility were found to be in
telemedicine, neither factor is aa significant predictor of
behavioral acceptance. Both are excellent predictors of
satisfaction levels, which will at least serve to reinforce
behaviory but not 1x3 predict additional network usage.
Other factors which help to predict satisfaction are
delivery, topic selection, user input, perceptiveness and
respectability of the speaker. Delivery and topic selection
appeared in two of the three satisfaction regression
equations with delivery proving to be a slightly better
predictor. The appearance of delivery in the regression
equations indicates the importance of information quality,

speaker quality, speaker ability and the learning potential

63

of the program. This factor serves to reinforce the belief
that people are not satisfied with simply being presented
the information. The information mmst 1x3 presented with
quality and by a quality instructor or speaker. Several
comments were made by users on the questionnaires that these
deliver methods needed some additional attention. Topic
selection is another factor which makes a great deal of
sense but and the way it loaded was exactly as predicted.
Topic selection is a good predictor of two factors of
satisfaction, usefulness and immediacy. Any factor dealing
with topic selection, time allotted to topic and timeliness
of the topic should affect levels of satisfaction. Perhaps
somewhat surprisingly however was the fact that topic
selection did not enter the regression equation for
behavioral acceptance» One could, theorize that topic
selection would affect how many of the programs a user views
:hi a year Inn: many other factors also affect behavioral
acceptance.

Since tflma Healthcare Information.‘Network; has been in
Michigan for such a short time, less than a year, many of
the member hospitals are very new to. the network. About
fifteen hospitals returned their questionnaires blank with
the explanation that they were so new to the network, they
did not feel they could comment on the programming or about
telemedicine in general until they had a little more
experience. One explanation for some of the factors which
were expected t1) help predict behavioral acceptance, like
tOpic selection, utility, delivery and interest, is the
possibility that there were not that many program offered to
users. People may have seen a good number of the programs
offered to them that pertained to their specialty already.
It is possible that while satisfaction levels could have
been higher, until more programs are made available to their
specialty, they may not be able to view more programs.

Those factors which were included in the regression

64

equations were program quality, sociability and teaching
ability. One fact run: mentioned. earlier 1J5 that the
regression equation for behavioral acceptance accounted fore
the least amount of variance of the four predictor models.

While the equation for behavioral acceptance is
significant and its F value is fairly substantial, it only
includes three factors. Sociability was not expected to be
included in the equation but serves as proof that health
care personnel will not utilize a technology unless it
conforms with a few of the standard normatives of its users.
Users feel that levels of formality and sociability will
affect their decision to accept an information source or
whether to seek an alternative source. This is a fairly
large finding since some of the earlier research tends to
indicate that behavioral acceptance will not be affected by
this factor. Somewhat surprising was the finding that
technical program quality is negatively correlated with
behavioral acceptance. One possible explanation for this is
that users feel that pmoductions whioh are pmoduced with
production quality in mind are not seen as important as
speaker and information quality. This finding is somewhat
significant in that it indicates that healthcare personnel
feel that the topic and the ability of the speaker to relate
the subject matted? is :more important. to 'their decision
whether or not to use telemedicine rather than one of its
information alternatives.

Health care professionals apparently will view programs
if the technical quality is perceived as less than broadcast
quality. The danger exists that when an alternative becomes
available with equal or superior topics and speakers, users
may abandon telemedicine if the alternative also displays
superior production quality. Health care networks finished
second to in-service training when respondents were asked to
rank-order six information alternatives. Televised

presentations of information can perhaps better illustrate

65

to participants concepts and techniques which may run: be
presented that way via alternative sources provided they are
high quality programs. Several of the respondents noted
that there were audio problems with some of the programs and
they simply would no "waste their time trying to figure out
what the speakers were saying". Users expect more than
"talking heads" which is an expression that means that
simply showing a person standing or seated, talking about a
subject just will not do anymore.

Despite the fact that some of the findings which were
extracted from tflua statistical. procedures were somewhat

unexpected, they should prove very useful.

Recommendations

The attitudinal factors which are most important to users
will allow researchers, programmers and network executives
to concentrate on how to better enhance levels of
satisfaction and behavioral acceptance. Several key factors
need to be addressed before a substantial gain in either
satisfaction or behavioral acceptance can occur; The
network has grown to a large enough size to integrate their
resources and. has a rmnfli larger' pool of speakers and
instructors to choose from. These speakers must be
instructed on how to best utilize the technology. The
formal, somewhat cold nature of network communications must
conform more to what health care personnel are used to. All
technologies have to relate to potential users for
behavioral acceptance to take place and it the importance of
this happening with health care professionals can not be
overstated. These personnel are used to dealing with people
and deserve the same treatment in their continuing
education. The network must be as dedicated to this cause
as physicians, nurses and hospital managers are to theirs or

telemedicine will never realize its full potential. Network

66

programmers and managers must work on enhancing levels of
satisfaction as well as behavioral acceptance. As mentioned
earlier, while satisfaction is run: directly connected to
behavioral acceptance, it will at least serve to reinforce
behavior. Even though superior technical quality was not an
important predictor of behavioral acceptance, the
availablity of alternatives should warrent caution in
overlooking program quality. Respondents feel that ttmdc
and speaker quality are more important to their decision to
use telemedicine than 'technical quality Inn: users make

become fickle.

[' . E E 1 1

Several attitudinal factors emerged that were moderately
correlated with user satisfaction. One concept which was
not explored was the role that opinion leaders and peer
pressure play in.tflu3 decision process. It would also be
helpful to compare levels of satisfaction and behavioral
acceptance within each. hospital across the various job
types.

One lesson learned through this project is that time of
year will affect return rate. Several people contacted
through this study indicated that participation would have
been much easier in the late fall or winter when staff size
and responsibilities are at more manageable levels.

This study took place within a network structure that has
undergone massive changes iJi'the last year: One important
factor that was somewhat lacking in this study was
stability. VWKHI so many changes occur within the network
structure, uneasiness and confusion often occurs. Several
hospitals sent their questionnaires back because they were
too new tx> the systemn By surveying users vflua have been
involved with the network for a longer period of time,

results may be more like what was expected.

APPENDIX A

67

TO: ALL MHIN TELECONFERENCING COORDINATORS
FR: MARLENE HULTEEN, EXECUTIVE PRODUCER, MHIN

RE: TELEMEDICINE SURVEY

The Michigan Hospital Information Network, the Healthcare
Information Network, and Michigan State University are now
working together as part of our continuing effort to deliver
our members the kind of programming and healthcare information
most useful to their needs. Your cooperation is vital to the
success of this project.

Very shortly, probably as soon as next week, you will be
receiving three (3) questionnaires from the Healthcare
Information Network from their main office in Princeton, New
Jersey. We are asking that you distribute one of these
questionnaires to a physician, one to a nurse, and one to a
hospital administrator/manager. Because this a survey being
done as part of a masters thesis and it is necessary that we
get unbiased data, the method by which you choose the three
people to give the questionnaire to is very important. Make a
list of all the physicians that have seen at least one MHIN or
HIN program or teleconference. From this list choose the
fifth (5th) name and give them a questionnaire to complete.
Make a similar list of all nurses that have seen at least one
program or teleconference and choose the eighth (8th) name and
give them a questionnaire to complete. Make a third list of
all the administrators or managers that have seen at least one
program or teleconference and choose the third (3rd) name on
the list and give them a questionnaire to fill out. If any of
the three refuses to complete the questionnaire, go back to
the list, start with the name immediately after the refused
party and choose either the next 5th, 8th, or 3rd name
(depending on whether you are choosing another physician,
nurse, or manager.

A few days after you hand it out, please go back to the
three people and collect the questionnaire. If they are not
finished, give them a few more days but please urge them to
complete it as soon as possible. After you have collected the
questionnaires, please return them to:

STEVE BECK
3305 WOOD ST.
LANSING, MI 48906

68

To allow time for the responses to be analyzed, it is
important that you get these questionnaires sent back as soon
as possible. The results of this survey will be made
available to us to allow us to make programming decisions and
any necessary changes to make the Healthcare Information
Network and the Michigan Hospital Association even stronger.

To give you some background information on the person who
is designing and analyzing the questionnaire, here is a short
biography:

Steve Beck is the head video teaching assistant in the
department of telecommunication at Michigan State University.
His interest in this project is his concern with quality of
continuing education and his interest in it's application in
medicine, the latter of which was spawned by his mother who
has been a registered nurse for 32 years.

If you have any questions about the distribution of the
questionnaire or any other part of this project please do not
hesitate to call me. Your help in this project is very
important and very appreciated.

 

APPENDIX B

 

69

June 5, 1987

Telecommunications and medicine have combined to help close the information
gap through a process called Telemedicine. More research is necessary
however to find ways to deliver better continuing education to health care
professionals. This survey is being done in cooperation with the Healthcare
Information Network and the Michigan Hospital Information Network to find out
what policies, regulations and programs are important to physicians, nurses,
dietitians, and hospital managers/administrators.

Please take a few minutes of your time to fill out this questionnaire and promptly
return it to your teleconferencing coordinator. Your answers will allow us to

in-poinf areas which need special attention and provide us with the necessary
Information we need to make the Healthcare Information Network exactly what
you need to keep up-to-date in your specialty.

You may be assured of complete anonymity. Your name will never be placed on
the questionnaire. This insures that no one will be able to tell how you
answered any of the questions. Completion of the survey is strictly voluntary
and returning it constitutes consent of the respondent. If you do not wish to
complete this questionnaire, please return it to your teleconferencing
coordinator.

The results of this research will be made available to managers and
programmers of the Michigan Hospital Information Network and the Healthcare
Information Network. If you would like to receive a summary of results, please
write or call me to keep your name anonymous. Please mm put your name
and address on the questionnaire.

I will be happy to answer any questions that you might have. Please write or call
if I can be of any assistance.

Sincerely,

Steven L. Beck

Project Coordinator
3305 Wood St.
Lansing,Michigan 48906

(5137) 372-7594

'70

TELEMEIDEUNE SELDRVEV

How many Michigan Hospital Information Network (M.H.I.N.) or Healthcare
Information Network (H.I.N) programs have you viewed in the last year?
(circle one number)

1. 5 OR LESS
2. 6-10
3.
4
5

1 1-15
16-20
MORE THAN 20

. What reason(s) do you watch programs on the M.H.I.N.-H.I.N.?

(circle all numbers that apply)

PQNP’WPPN.‘

CONVENIENCE
QUALITY OF INFORMATION

SUBJECT MATTER OF PROG RAM(S)
QUALITY OF TEACHING

LOW COST

REQUIRED BY DEPT. HEAD OR SUPERVISOR
CURIOSITY

SAVES TIME

OTHER (please specify)

 

Do you usually watch M.H.I.N.-H.I.N. programs: (circle one number)

$09!“?

ALONE
WITH A FEW COLLEAGUES
IN A LARGE GROUP

OTHER (please specify)

 

. Where do you usually watch M.H.I.N.-H.I.N. programs? (circle one number)

1.
2.
3.
4.
5.
9.

IN A CONFERENCE ROOM

IN A LOUNGE

IN YOUR OFFICE

IN A DESIGNATED ROOM IN THE HOSPITAL
AT HOME

OTHER (please specify)

 

How many "live" programs or teleconferences have you attended? (give
specific number)

71

6. Have you ever phoned in a question during a "live” program?
1. YES —How many times? 2. NO

6a. Why not? <--
1. TOO SHY

2. DIDN'T HAVE A GOOD QUESTION
3. DIDN'T HAVE TIME

4. COULDN'T GET THROUGH

5. OTHER (please specify)

 

 

7. Please rate the following factors of the Michigan Hospital Information
Network on a 5 point scale with 1 being poor, 2 fair, 3 satisfactory, 4 good,
and 5 excellent. Poor Excellent
1 2 3 4 5
(circle number below)

Video quality 1 2 3 4 5
Audio quality 1 2 3 4 5
Quality of information 1 2 3 4 5
Topics selected ‘ 1 2 3 4 5
Time allotted per topic 1 2 3 4 5
Timeliness of topic selection 1 2 3 4 5
Quality of speakers 1 2 3 4 5

Speakers ability to present topic understandably 1 2 3 4 5
Your interest in continuing medical education 1 2 3 4 5
Pace of programs 1 2 3 4 5
Program length 1 2 3 4 5
Scheduled broadcast time of teleconferences 1 2 3 4 5
Ability to ask questions 1 2 3 4 5
Usefulness of programs for learning 1 2 3 4 5

Ability to provide feedback on
programs, speakers, and topics 1 2 3 4 5

 

72

8. How many times a year do you go to medical conferences or other
face-to-face seminars to get information in your specialty, for CME-CEU

credit and for general knowledge?
(circle one number)

1. 5 OR LESS

2. 6-10

3. 11-15

4. 16-20

5. MORE THAN 20

9. Please rate the following factors in terms of your acceptance of telemedicine
with 1 being unimportant and 5 being very important.

Speed of information dissemination

Ability to see operations/policies in practice
via videotape

No need to travel

Having written materials on subject
while viewing program

Ability to phone-in during programs for
clarification of issues or questions

VERY
UNIMPORTANT IMPORTANT

1 2 3 4 5
(circle number below)

1 2 3 4 5

1 2 3 4 5

10. Please rank the following in order of importance as an information source of
industry advancements with 1 being least important and 6 being most

important. (number each source)
__ Magazines
_ ln-service training
__ Hospital memos
__ Healthcare networks
_ Co-workers

Conventions

73

11. For the following statements, please answer on a scale from 1 to 5 with 1
being that you strongly disagree, 2 that you disagree, 3 neutral, 4 that you
agree, and 5 strongly agree.

 

STRONGLY STRONGLY
DISAGREE AGREE
1 2 3 4 5
(circle number below)
Televised medical education (telemedicine)
is a good source of information. 1 2 3 4 5
I usually pay more attention to a live speaker
than I do a televised speaker. 1 2 3 4 5
All factors being equal, information gained from
an M.H.I.N.-H.I.N. program is as useful as
information gained through a class, a conference
or other face-to-face method. 1 2 3 4 5
Maintaining face-to-face interaction with colleagues
is vital in gathering information on my profession. I 2 3 4 5
I generally learn more from a "live' speaker than a
televised speaker. 1 2 3 4 5
I would recommend telemedicine to my colleagues. 1 2 3 4 5
I fear that telemedicine will decrease my ability
to travel to conferences to meet with
fellow colleagues. 1 2 3 4 5
Resources spent on telemedicine may be better
utilized elsewhere. 1 2 3 4 5
It is helpful to be able to ”replay" the information
back at a later date if needed. 1 2 3 4 5
I do not wish to attend meetings to watch television. 1 2 3 4 5
I find most of the programs on the M.H.I.N.-H.I.N.
informative and useful. 1 2 3 4 5

12. Using the adjectives below, please indicate how you feel about using
telemedicine as an information source for your profession. For example, if
you felt that telemedicine is an effective source of information you would

mark:

INEFFECTIVE

 

EFFECTIVE

Generally, using telemedicine as an information source in my profession

is:

INFORMAL

EXPENSIVE

FORMAL

INEXPENSIVE

 

UNSOCIABLE

SOCIABLE

 

UNRELIABLE

RELIABLE

 

UNCOMFORTABLE
DIFFICULT

IN EFFECTIVE
INCONVENIENT
IMPERSONAL

INFLEXIBLE

COMFORTABLE
EASY
EFFECTIVE
CONVENIENT
PERSONAL

FLEXIBLE

 

COMPLEX

INSENSITIVE

SIMPLE

SENSITIVE

 

NOT USEFUL

USEFUL

75

13. Please rate the following speaker traits in terms of importance to you as an
effective speaker with 1 being unimportant and 5 being very important.

VERY
UNIMPORTANT IMPORTANT
1 2 3 4 5
(circle number below)

Speech clarity 1 2 3 4 5
Knowledge of subject matter 1 2 3 4 5
Speaker's prominence in their specialty 1 2 3 4 5
Speaking style 1 2 3 4 5
Use of humor 1 2 3 4 5
Use of graphics and special effects 1 2 3 4 5
Pace of speaker 1 2 3 4 5
Speaker's setting (background, location) 1 2 3 4 5
Your location (as you view speaker) 1 2 3 4 5
Pauses for questions and clarification 1 2 3 4 5
Use of handouts 1 2 3 4 5

14. Do You feel that Continuing medical education units (CMEs-CEUs) should
be mandatory for health care personnel?
(circle one number)

1. YES
2. NO
9. UNDECIDED

15. Do you think telemedicine (televised medical programs) can provide proper
continuing education for health care personnel?
(circle one number)

1. YES
2. NO
9. UNDECIDED

'76

16. Do you think telemedicine can be used to teach non-clinical classes for

credit?

(circle one number)

1.
2.

9.

YES
NO—Why not?

 

 

UNDECIDED

17. What types of programs would you like to see Mill on the Michigan
Hospital Information Network?
(circle all that apply)

PSPPNP’PIPP’N?

LEGISLATIVE UPDATES

MEDICAL RESEARCH REPORTS
LEGAUBUSINESS ADVICE

PATIENT CARE INFORMATION

PATIENT HOW-TO'S (for patient viewing)

CASE STUDIES

JOB REFERRAUADVANCEMENT INFORMATION
DRUG INFORMATION
SELF-IMPROVEMENT
OTHER (please specify)

 

Finally, we would like to ask a few questions about yourself for statistical

purposes.

18. Are you a: (circle one number)

1.
2.
3.

PHYSICIAN
NURSE
HOSPITAL MANAGER/ADMINISTRATOR

19. Age: (circle one number)

NP’P‘PP’NT‘

25 OR UNDER
26-34
35-44
45-54
55-64
65-74
75 AND OVER

77

20. How many years have you been out of school and working in the health care
profession?

(circle one number)

5 YEARS OR UNDER
6-10 YEARS

11-15 YEARS

16-20 YEARS

21 -25 YEARS

MORE THAN 25 YEARS

P’S’IPP’N.‘

21. What is your marital status?

1. SINGLE

2. MARRIED

3. DIVORCED
4. SEPARATED
5. WIDOWED

22. Are you a:

1. MALE
2. FEMALE

Next page please

 

78

Please use this page, and others if needed, to make any further comments,
suggestions, and-or criticisms that you may have about programming or
practices of the Healthcare Information Network, the Michigan Hospital
Information Network, this questionnaire

or telemedicine in general.

THANK YOU VERY MUCH FOR YOUR TIME AND EFFORT. WITHOUT YOU.
THIS PROJECT COULD NOT BE COMPLETED.

please send questionnaires to:

Steve Beck
Project Coordinator
3305 Wood St.
Lansing, MI 48906

APPENDIX C

79

MEMORANDUM
TO: All MHIN Teleconferencing Coordinators
FROM: Marlene Hulteen
DATE: June 23, 1987

RE: Telemedicine Survey

Two weeks ago, you received three questionnaires from our
Office along with instructions on the dispersion of the survey
to selected health care personnel.

Many of you have already returned the questionnaires to the
project coordinator in the envelopes provided. Thank you very
much for your prompt attention. Some of you have yet to
return the completed surveys. I urge those of you who have
not returned them to try to convince those people whom you
have handed out the questionnaires to, how important their
input is. Please try to get the completed questionnaires back
as soon as possible, no later than the June 26th deadline.

In order to better understand the needs Of our users, the
completion of these questionnaires is extremely important. We
need you for the implementation of this vital step.

If you have any questions or need additional
questionnaires, please feel free to contact me immediately.
Thank you very much for your cooperation on this important
project.

MH/sl

BIBLIOGRAPHY

Acker, Stephen R, "The Teleconferece Trialability Lab:
Fostering The Diffusion Process Through Organizational
learning." ha Teleconferencing and Elegtrnnie
C . I' IX E 1' I' I 1 J . I H
Restore. Ed. Lorne A. Parker and Christine H. Olgren.
Madison, Wi.: Regents of the University of Wisconsin
System Press, 1985.

Ajzen, 1.. & Fishbein. M. W
W. Englewood Cliffs, New
Jersey: Prentice-Hall, 1980.

Awbrey, Susan. Personal Interview. 18 May 1987.

Becker, Ann DeVaney. "A Survey and Evaluation." In Jenznel
W. Summer 1978 .

Baldwin, T., and McVoy, D. Ceble_gemmnnieetiene. Englewood
Cliffs, New Jersey: Prentice-Hall, 1983.

Brief, A., Delbecq, A., Filley, A., & Huber, C. "Elite
Structure and Attitudes: An Empirical Analysis of

Adeption Behavior. Administrationmim. v.8, 1976.

Cartwright, D.. "Some Principles of Mass Persuasion: Selected
Findings of Research on the Sale of United States War
Bonds." Humanjelations. 1949.

Child, J., et al. "The Quality of Employment in Services." In
W. Ed. T. Forester.
Cambridge, Mass: MIT Press, 1985.

80

81

Chu, Godwin C., and Wilbur Schramm. Learning Erom Beleyjsjcn°
What the Research Says. Washington, D.C.: NAEB, 1975.

Clark, Sally L.. "Practical Applications Of Incorporating
Teleconferencing into Information Technology Management."
In WWII
WWW. Ed. Lorne
A. Parker and Christine H. Olgren. Madison, Wi.:

\. 1m’-va'_.‘hL-'-.u' - I,

Regents of the University of Wisconsin System Press,
1985.

‘L: . 5".“ 5",;
.
.

Conrath. David W. et al. W
Technology—inJedicine. Dedham, MA.: Artech House.
1983.

Cowan, Robert. W.
Reston, VA.: Prentice-Hall, 1984.

Dizard, Wilson P. Jr.. Ihe_Qcming_lnfcrmaLicn_Age. New York:
Longman Communication Books, 1985.

Downs, G., & Mohr, L. "Toward a Theory Of Innovation."

W, V- 10, 1979.

Elton, M. C. J. W
Meetinge. New York: American Management Association,
1982.

Fishbein, M.. A consideration Of Beliefs and Their Role in
Attitude Measurement." In W

and_Meaenrement. Ed. M. Fishbein. New York: John Wiley,
1967.

82

Flournoy, Don M.. "Distance Teaching: Telecommunications in
Education and Training." In W
- 00' on“ 0' . '00 :90 ' . '00 ‘ 00- 02"
and_fiuman_EaQL915. Ed. Lorne A. Parker and Christine H.
Olgren. Madison, Wi.: Regents of the University of
Wisconsin System Press, 1985.

Frazier, Hilda A.. "Video-Teleconferencing in Graduate Nursing
Education-The Wave of the Future." In Ieleccnferencing
I E] I . C . |' IM 2 J' I'
W. Ed. Lorne A. Parker and
Christine H. Olgren. .Madisqn, Wi.: Regents of the
University of Wisconsin System Press, 1985.

Haaland, B.A. , and W.G. Newby. "Student perception of
effective teaching behaviors: An examination of
conventional and teleconference-based instruction." In
Madison, Wi.: Center for Interactive Programs, 1984.

Hulteen, Marlene K.. Personal Interview, 18 May 1987.

Katz, D., & Stotland, E. "A Preliminary Statement to a Theory

Of Attitude Structure and Change." In Eeychclcgy;_A_$Lndy
cf_a_ficience. Vol. 3. Ed. 8. Koch. New York: McGraw-Hill,

1959.
Lavine, David. Telephone Interview, 13 May 1987.

Lazer, 3.. et al. W
W. White Plains, N.Y.:
Knowledge Industry Publication, 1983.

83

Liguori, Cathy. "A Video Teleconferencing Success Story." In

1 E . l E] l 3 . I' III
WW. Ed. Lorne

A. Parker and Christine H. cngren. Madison, Wi.:
Regents of the University of Wisconsin System Press,
1985.

McGuire, W.J. "Personality and Susceptibility to Social
Influence . " In W
Research. Eds. E.F. Borgatta, & W.W. Lambert. Skokie,
Il.: Rand McNally, 1968.

McGuire, W.J. "The Nature of Attitudes and Attitude Change."
In W (2nd ed., Vol. 3).
Eds. G. Lindzey; & E. Aronson. Reading, MS.:
Addison-Wesley, 1969.

March, J., & Simon, H. Organizaticne. New York: John Wiley &
Sons, 1958.

Neumann, David R.. "Three Factors Affecting Successful
Teleconferencing: Facilitators, Attention, and
Information Understanding and Retention." In

I] E . I E] I . 3 . I' III
MW. Ed. Lorne

A. Parker and Christine H. Olgren. Madison, Wi.: Regents

of the University of Wisconsin System Press, 1985.
Prowse, A“ ChmfitflL "The Alberta Hospital Association's
Teleconferencing Network: A Growing Success Story." In

I] E . I E] l . 3 . l' Ill
WWW. Ed. Lorne

A. Parker and Christine H. Cngen. Madison, Wi.:
Regents of the University of Wisconsin System Press,
1985.

84

Pryor, Brandt W.. "Increasing User Acceptance of
Teleconferencing: A Model for Understanding and Changing

User Attitudes and Behaviors." In Teleconferencing_and

O I
- o o ‘0. o ‘ o co ‘
.0 CI“ 0 . O .-- o. O -

and_Hnman_Eactcrs. Ed. Lorne A. Parker and Christine H.
Olgren. Madison, Wi.: Regents of the University of

Wisconsin System Press, 1985.

Ribble, Judith G. "Resistance to Teleconferencing in Medical
Settings."' Int In Ieleccnferencing and Electrcnic
C O I. IEZ E J. I! I 1 J I i Ii
Eactcre. Ed. Lorne A. Parker and Christine H. Olgren.
Madison, Wi.: Regents of the University of Wisconsin
System Press, 1985.

Rogers, E.. Diffusion_of_lnnoyations. 3rd ed. New York:
The Free Press, 1983.

Rosenberg, M.J., & Hovland, C.I. "Cognitive, Affective, and
Behavioral Components of Attitudes." In AL_t_i_t;_u_d_e
WW. Eds. C.I. Hovland & M.J.

Rosenberg. New Haven, CT.: Yale University Press, 1960.

Rothert, Marilyn L.. Personal Interview, 18 May 1987.

Ruchinskas, J. and Svenning, L. "Fermatiye Eyaluaticn_IQL
E . . I I J I' : . I' J 2 . I'
I2QhnQlflgi&&L_Ihfi_JEUELJZE_¥idflflzflflnffilfiflfling-" Paper
presented at the International Communication Association,
Minneapolis, May 1981.

Schramm, Wilbur. Men. MessagesL_and_Mediai_A_LQQk_aL_Hnman
Ccmmunicaticn. N.Y.: Harper and Row, 1973.

85

Shaeffer, James M., and Robert Roe. "Effective Teaching

Behavoirs As Perceived By Students In a Face-To-Face and

Teleconferencing Course . " In meccnferencinLanc
- .0 00m 0' . 00 49.0 ' . 00 ‘ 000 0-.-
and_Hnman_£achre. Ed. Lorne A. Parker and Christine H.
Olgren. Madison, Wi.: Regents Of the University of

Wisconsin System Press, 1985.

Short, John., (N: al. The Sccial Psychclcgy cf
Ieleccmmunicaticne. London: John Wiley and Sons, 1976.

Shulman, A. and Steinman, J. "Interpersonal Teleconferencing
In an Organizational Context." In W

Ielecommunication.$eryices. Eds. M. Elton. W. Lucasa, and
D. Conrath. New York: Plenum Press, 1978.

Smith, M.B. "The personal Setting Of Public Opinions: A Study

of Attitudes toward Russia." Public_Oninion_Quarterl¥,
11, 1947.

Svenning, Lynn L. "Predispositions Toward a Telecommunication
Innovation: The Influence Of Individual, Contextual, and
Innovation Factors. on Attitudes, Intentions, and
Projections Toward Video-Conferencing." Diss. Southern
California 1982.

Thurstone. L. L.. "The Measurement of Attitudes." figurnal_efi
AbnOrmal_and_Social_£S¥cholog¥. 1931.

Williams, Ederyn. "Experimental Comparisons Of Face-to-Face
and Mediated Communication; A Review." In Eeychclcgical
Bulletin., v. 84, NO. 5, 1977.

Williams, Ederyn. "Social and Psychological Factors." In

Journal_of_flcmmunicatien. Summer 1978.