A Recoucmumzm MODEL OF MESSAGE ~ .

mnuoe - BEHAVIOR Ramomm

Thesésforthe DegraeofMA,
MICHIGAN ST ATE UNIVERSITY
1975

IIIIIIIIIIII III IIIIIIIIIIIIIIIIIIIIII

00649 1009

 

 

 

 

MSU

LIBRARIES
m

 

 

RETURNING MATERIALS:
Place in book drop to
remove this checkout from
your record. FINES wiII
be charged if book is
returned after the date
stamped below.

 

 

 

 

 

ABSTRACT

A RECONCEPTUALIZED MODEL OF MESSAGE-
ATTITUDE-BEHAVIOR RELATIONSHIPS

BY

Robert D. McPhee

This thesis is directed to the development and test
of a reconceptualized model of message-attitude-behavior
relationships. .The model is a transformed version of one
used by Don Dulany to explain verbal conditioning effects.
It posits four kinds of beliefs which subjects might hold,
as predictors of their intentions to behave. The model
also suggests certain communication variables.

The power of the model was tested in a pretest post-
test study using subjects from five introductory communica-
tion classes. Three out of four specific components of
the model are significant determinants of behavioral
intention, and the model as a whole explains a highly
significant amount of the variance in both pretest and
posttest results. But the predictive equations using the
model do not have the stability, power, or immediacy pre-
dicted by the model. Hypothesis involving communication

variables received questionable support.

Robert D. McPhee

In the conclusion, implications are stated and sugges-
tions are made for a more decisive test of the relationships

posited by the model.

A RECONCEPTUALIZED MODEL OF.

MESSAGE-ATTITUDE-BEHAVIOR RELATIONSHIPS

BY

Robert D. McPhee

A THESIS

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

1975

TABLE OF CONTENTS

I,

/"/.I

0II

Page

CHAPTER I: INTRODUCTION AND THEORY . . . . . . . . . 1
CHAPTER II: DESIGN AND OPERATIONAL PROCEDURES . . . 14
CHAPTER III: DATA ANALYSIS AND RESULTS . . . . . . . 22
CHAPTER IV: CONCLUSIONS . . . . . . . . . . . . . . 34
APPENDIX I: DULANY'S THEORY . . . . . . . . . . . . 37
APPENDIX II: DERIVATION OF AN ALTERNATIVE MODEL . . 41
TABLES . . . . . . . . . . . . . i . . . . . . . . . 44
SAMPLE QUESTIONNAIRES . . . . . . . . . . . . . . . . 60
91

BIBLIOGRAPHY

ii

To my parents,
to all my teachers,

and to Dale.

iii

ACKNOWLEDGMENTS

I owe much to too many people to mention here. In
particular, I must thank Tom Florence, Norm Fontes, Ginny
McDermott, Bob Craig, and Dave Seibold for various cooperation
in conducting this research, and for various comments and
advice about questionnaire construction, analysis, and the
thesis itself. Scott Poole and the persons mentioned in the
dedication gave me patient moral support through trying times.
I am very grateful to Dr. C. DaVid Ralph and Dr. Gordon
Thomas for their assistance and patience. And above all, I
am grateful to Dr. Donald P. Cushman for his long intellectual
guidance (trying at times to both of us), of which this thesis

is an imperfect and minor product.

iv

CHAPTER I

INTRODUCTION AND THEORY

In a recent review of the literature, Seibold (1974)
finds various and inconsistent effects of messages of atti-
tudinal and behavioral variables. This lack of cumulative
findings in the literature seems to indicate the need for a
revised model of message-attitude-behavior relationships.
One such model can be derived from the work of Dulany (1961,
1962, 1963, 1968) in verbal conditioning; it is the purpose
of this thesis to present and report an experimental test of
that model.

A model designed to account for attitude-behavior
inconsistencies has previously been derived from Dulany's
work by the social psychologist Martin Fishbein (1967)--
indeed, his model gave impetus to the present derivation.
Fishbein's model has been extensively tested and supported
in the social psychological literature, and has received
as well some attention in the communication literature (see
Mortensen and Gereno, 1973, part I; Holdridge and Lashbrook,
1973; and Seibold, 1974, for reference in the communication
literature). This derivation of a new model from Dulany's
work was undertaken for two reasons. First, Fishbein

1

reinterprets Dulany's theory in terms derived from the
traditional components of attitude--cognitive, conative, and
affective. These are primarily psychological variables,
representing 'states of mind' of subjects. However, Dulany's
original variables were conceived in terms of information and
propositional form (see below for clarification) and were
thus very close to communication variables in their under-
lying logic. Second, Fishbein's model is relatively unsugges-
tive about (a) message variables, and (b) variables, which
might ease or hinder the communication process. Dulany's
original logic is much more suggestive in this regard, and
its approach has been incorporated into the present research.
Later in this chapter, the communication variables suggested
by Dulany's approach will be described.

In this thesis I shall explain the model of message-
attitude-behavior relationships which I have derived from
Dulany's theory, and describe the results of an experiment
which allows a test of the theory. The body of the thesis
will be divided as follows. The rest of Chapter One will
be devoted to a summary of Dulany's theory and derivation of
a message-attitude-behavior model from it, and to the state-
ment of hypotheses. Chapter Two will include a discussion
of the design and operational procedures used in the study.
Chapter Three will contain a description of the data analysis
and a discussion of its results, and Chapter Four will draw

conclusions about those results.

Summary of Dulanyfs Theory

 

Dulany's original theory was proposed as a result of
work done in verbal conditioning. His theory is rather revo-
lutionary for that domain--he believes that "conditioned"

verbal responses are actually under the conscious control

 

of subjects, and are made for reasons which subjects can
voluntarily report. Dulany's theory, loosely stated, consists
in two main assertions. The first assertion is simply that

we can predict subjects' responses if we ask them how they
intend to respond. The second assertion, more complex, goes
back a step and deals with the prediction of subjects' inten-
tions to respond. Here Dulany asserts that we can predict

subjects' intentions by knowing how they interpret the rein-

 

forcement situation, what they understand to be going on when
the experimenter reinforces one of their responses. In par-
ticular, to predict intention we must at least know:

a) whether the subject thinks a particular response
will be reinforced;

b) whether the subject likes or values the reinforce-
ment;

c) what the subject thinks the reinforcement means--
for example, does the experimenter want him to do
just those behaviors that are reinforced, or not?
and

d) whether the subject cares what the experimenter
wants.

According to Dulany, each of these factors is present in a

subject's mind as a proposition of a certain form--for example,

 

a) might be represented as "If I do response X, I will be
reinforced." By asking subjects whether they agree with that
statement, and others, we can predict their intentions.

Experimental evidence adduced by Dulany (1961, 1962,
1968) indicates that he can successfully predict behavior from
intention, and intention from the four factors mentioned above.
(See Appendix I for a complete and formal examination of

Dulany's theory.)

Derivation of an Alternative Model

 

If we want a model that predicts behavior in a natural
context, rather than in a conditioning laboratory, some parts
of Dulany's theory must be changed. First, we take a step
similar to Dulany's: we presume that behavior is under
intentional control, so that Behavioral Intention is a good
predictor of behavior. (Ajzen and Fishbein (1973) argue for
this presumption, given certain situational constraints--for
example, the subject must have the ability to accomplish what
he intends, he must not change his mind, etc.) The real
problem becomes, then, the prediction and explanation of
Behavioral Intention.

My model has been derived by drawing up, following
Dulany, a list of factors which (a) might influence intentions,
and (b) can be represented as propositions of a particular
form. The model contains five kinds of factors.

1. First, a subject might decide to do something

because the consequences of his act would be rewarding or

 

valuable. This factor I shall call a Belief about the
Extrinsic Value of the act--BEV for short. This factor is
represented by propositions of the form "Doing X would be
rewarding (or punishing)." Thus, if we ask someone if he
agrees with a statement of that form, we gain information
useful in predicting Behavioral Intention to do X.

2. A subject might also decide to do something just
because it would be fun or pleasant, rather than because of
the later useful consequences of doing it. This factor is
called a Belief about the Intrinsic Value of the act--BIV.
It is represented by propositions like "Doing X would be
pleasant (unpleasant)."

3. Further, a subject might decide to act because
he thinks it his moral duty or personal obligation. This
factor has been called by Fishbein a Personal Normative
Belief--abbreviated NBP. It corresponds to propositions
such as "I should do X" or "I have a moral duty to do X."

4. Finally, a subject might decide to act because

others expect him to. This factor is called a Social Norma-

 

tive Belief, NBS. It corresponds to propositions of the
form "(Some group of significant others--fami1y, teachers,
etc.) expect me to do X."

5. The fifth factor has a different status from the
first four. It is the extent to which a subject feels com-
pelled to comply with each of the above factors. This factor

is called Motivation to Comply--MC. It is measured separately

for each of the above components. For example, if a subject
thinks that his parents expect him to attend class, and if

he wants very much to live up to their expectations, he is
likely to attend class. On the other hand, if he doesn't

care what his parents expect, their expectations are unlikely
to have much weight in determining his behavior. Indeed, this
factor, MC, functions as a weight applied to each other factor
before the data are analyzed.

The presumption is that these factors are related to,
and allow us to predict, Behavioral Intention. Dulany, and
Fishbein after him, used a multiple regression form for data
analysis because regression emphasizes the prediction of the
dependent variable (here BI). But each of the first four
factors is weighted, or multiplied, by MC, before it is
entered into the regression equation. This weighting allows
us to compensate for subjects who think one factor very
important (or very unimportant).

The final model is, then, in the typical regression

format,

BI = b (MC1XBEV) + {b2 (MC

1 ,xBIV) + b3(MC3xNBP) + b4(§MC4ixNBSi)

2

Note that there may be more than one social normative belief
term, if there are several groups of significant others rele-
vant to a behavior.

If our theory is correct, several things should be

true of this model. First, the independent variables--the five

factors--should account for a good deal of the variance in
Behavioral Intention. Second, each factor in the model should
be important under some conditions--otherwise, we could just
drop it from the model. (Note that if the theory suggests
that each factor should be important, and one isn't, that
finding casts doubt on the validity of the theory.) Third,

if the factors are really the immediate determinative causes
of Behavioral Intention, no gthg£_variable should be able to
change BI independently--if BI is changed, that change should

reflect a change in one or more of the factors.

Relation of the Model to Communication

 

At this point a discussion of the implications of
this model for communication is in order. The discussion is
clearer if the implications of this model are contrasted
with the corresponding implications of Fishbein's model.

In a study which used messages to alter the intentions
of subjects, Ajzen (1971) used messages in two ways. First,
he used messages (not explicitly recognized as such in his
study) to establish the "motivational orientation" of the
subjects--essentially by telling them either that their own
self-interests or their mutual interests with others should
control their behavior. These messages affected the relative
causal importance (beta weights) of the different components
of the model, though effects on MC are not reported. Second,

Ajzen used messages to affect the specific attitudes and

beliefs held by subjects--that is, to alter the value, rather
than the importance, of the causal components. In both cases,
message variables were dichotomous--messages were written to
stress one component or another, one behavior or another. In
addition, these messages are provided as part of the instruc-
tions in playing a Prisoners' Dilemma Game--they provided the
only information subjects had about how to play the game.

In contrast, the model provided here presents a
rationale for generalizing beyond the special situation of
Ajzen's study, and provides a more sophisticated message

variable-~amount of information.

 

In the model presented here, messages may similarly
be designed to affect the value and importance of one or
another component, but a more powerful and accurate message
variable, suggested by Dulany's logic, is the amount of
information in each message encoded in each relevant form
mentioned in the model by subjects. Thus, a message aimed
at affecting BEV may still have information which is encoded
by the subject so as to affect NBP or NBS. This variable
can be operationalized either objectively, according to the
experimenter's view of the message, or subjectively, accord-
ing to estimates by naive subjects.

In addition, it is completely consistent with the
logic of this theory to expect the effect of information of

a certain propositional form in the message to be inversely

dependent on the amount of that type of information already
held by the subject. Thus,
--if S already has information that others expect
him to do X, telling him that others expect him
to do X may have little impact on his intentions.
--if S already has information that others don't
expect him to do X, telling him that others expect
him to do X may have reduced impact due to his
opposed belief.
This confounding effect of outside information available to
S has been conceptualized by Woelfel (1973) as the inertial
mass of a concept. He reasons that the more information we
have about a concept or relationship, the harder it is to
change our minds about that concept or relationship. Thus,
we will expect that the amount of change in a belief produced
by a message will be (a) directly related to the amount of
information in the message which has the same propositional
form as the attitude and (b) inversely related to the amount

of information the subject already has relative to that propo-

sitional form.

Operational Definitions

 

Eight variables, including terms in the model and
communication variables, require explicit operationalization.
These operational definitions will be stated in the next
chapter, but two of them, because of their subtle connections
to theory, require some consideration now.

In both Dulany's and Fishbein's work, the variable

"motivation to comply" has been measured only for some of the

10

components in the model. This procedure has resulted in
measurement problems (cf. Ajzen and Fishbein, 1969, p. 257;
Schwartz and Tessler, 1973, p. 229), which might well be
corrected if subjects could be indeed to give comparative
ratings of the importance of all components. The operationa-
lization was designed to do this, by asking, subjects first
to rank the variables by importance, then to rate the impor-
tance of each variable on a semantic differential scale. In
using a rank-order question first, it was hoped that the
conmarison among components would become salients to subjects
Inhen they answered the immediately following questions. This
liope can be empirically checked by examining the variance in
:seven-point ratings and the correlation between ratings and
rankings .

The most important communication variable to be used
iJI this study is the amount of information in the message
Ioearing on each propositional form in the model. To opera-
tionalize this variable, I simply asked an independent sample
of subjects to rate the messages (one message per subject)
as to the percentage of information in the message stating
or supporting each propositional form. This procedure, to be
completely valid, would require the assumption that subjects
can objectively classify and measure amounts of information
in a message. Such an assumption may be false-indeed, much
research has shown that a subject's prepotent mental set is

as important in determining a message's effect as the message

11

itself. I make the weaker assumption that (a) various "mental
sets" are fairly normally distributed in my population and
samples, so that (b) the means of various subjective estimates
will be accurate in determining at least ordinally the amount
of information in the message. The value of this operational
definition can be checked by examining the covariation in
semantic differential ratings and rank orders, of subject

estimates.

Hypotheses

 

Hypotheses related to this model can be presented
in two areas: relative to the model itself as a predictor
of behavioral intention, and to the effects of messages on
variables in the model.

As Schwartz and Tessler point out, if a model of the
form proposed is to be accepted, it should contain all the

significant immediate determinants of BI, and only such

 

immediate determinants--i.e., there should be no irrelevant
or ineffectual components. In particular, communicative
influences should change behavioral intention only by changing
some other variable(s) in the model.

These requirements are reflected in the following
hypotheses:

H l. The relationship between BIV and BI is

necessary (for at least one treatment

condition), contingent (on MC), stochastic,
irreversible, and coextensive.

12

H 2. The relationship between BEV and BI is
necessary (in at least one condition),
contingent (on MC), stochastic, irrever-
sible, and coextensive.

H 3. The relationship between NBP and BI is
necessary (in at least one condition),
contingent (on MC), stochastic, irrever-
sible, and coextensive.

H 4. The relationship between NBS and BI is
necessary (for at least one condition,
at least one group of significant others),
contingent (on MC), irreversible, stochastic,
and coextensive.

H 5. The relationship between a linear combination
of the components MC x BEV, MC x BIV, MC x
NBP, and MC x NBS, and BI is necessary,
sufficient, deterministic, irreversible, and
coextensive.

Hypotheses also can be advanced in the area of message
variable influence. The theory does not demand any particular
quantitative relationship between message information (here-
after termed, after Woelfel, message mass, or MM), information
previously held (or previous mass PM), and components of the

model. The simplest available relationship is Woelfel's

prediction of a linear relationship between belief change

and the ratio of new information to old, g%.
H 6. The relationship between MM/PM for any
component (including MC as a multiplier)
and change in that component, is necessary,
sufficient, deterministic, irreversible,
and sequential.

Our last hypothesis is that, for a message to affect BI, its

mass ratio, MM/PM, must affect the components of the model.

13

H 7. The relationship between message con-
dition and change in BI is substitutable,
contingent (on the effect of MM/PM upon
the model's determining variables,
including MC), stochastic, irreversible,
and sequential.

Note that in each of these last two hypotheses we contend
that MM/PM may affect MC, by bringing to the subject's
attention the fact that a certain propositional form is

relevant and that he Should comply with it. A diagrammatic

model of these hypotheses is as follows (Figure 1):

MC

MMEv<:

..
.‘
‘fi
.
...‘.Q
‘Q
--
-n
g

 

 

MC ”7”,“
Arms» ..... ”My” ......
4. MC """""""""""
MMN “=A_~_-. ___________
B NBS —————

Figure l.--Hypothesized relations among major variables in
model.

In this diagram is a causal influence, is
a contingent causal influence, and the solid arrow in
is a contingency-producing influence: in each case, MC makes

the relationship a contingent one.

CHAPTER II

DESIGN AND OPERATIONAL PROCEDURES

In this chapter I shall first briefly sketch the
experimental design used in testing the model presented in
the first chapter. Then the questionnaire used in the research
will be described in depth. Finally, the method used to derive

message variables will be reported.

Design

The study was constructed as a pretest-posttest con-
trol group design with three message treatment groups and a
control group. Subjects were drawn from five undergraduate
communication classes taught during Spring term, 1974.

Between April 29 and May 3, 1974, pretests were
administered in the five classes. The pretest form included
an announcement that a lecture on family communication would
be given by Professor Donald Cushman on May 22 and 23; after
this announcement subjects were asked to respond to a series
of questions about this lecture and their thoughts about
attending it. Two weeks later, between May 15 and 21, a
second set of questionnaires, consisting of a message treat-
ment and questions including those used in the pretest, was
distributed in the five classes.

14

15

There were three message conditions besides the control
condition (in which no message beyond the neutral announcement
was used.) Messages were constructed with essentially the
same informational content but designed to support three
different reasons for attending the lecture. The messages
argued either that (a) consequences of going to the lecture
would be pleasant and rewarding, or (b) each student had a
moral obligation to attend the lecture, or (c) each student's
family would expect him to attend the lecture. Subjects
were randomly assigned to treatment or control conditions
by randomly distributed treatment-instrument packages.

The lectures were indeed held on May 22 and 23, and
a record was kept of which students attended the lectures.

194 students completed the first questionnaire, and
178 completed the second; in all, 103 subjects usably res-
ponded to both the pretest and the posttest; these subjects
were fairly evenly distributed among all four conditions.

(Of these subjects, 16 actually attended the lecture.)

The Questionnaire

 

A cover sheet provided an announcement of the place,
time, and subject matter of Professor Cushman's lecture, and
indicated that the questionnaire sought information about
student reactions to the prospective lecture. The question-
naire itself was 9 pages long for time 1, with a one-page

message and one more page of questions for time 2. The

16

items are discussed below in the approximate order of their

appearance on the questionnaire.

Behavioral Intention.

Subjects indicated their answers to the question,
"Professor Cushman of the MSU Department of Communication
intends to give a public lecture on the topic "Family
Communication." will you attend that lecture?" Answers were
marked on a seven-point scale ranging from "Definitely Not"

to "Definitely Yes."

Belief about Extrinsic Value.

Subjects responded to the question:
Attending a lecture by Professor Cushman of the MSU Department
of Communication on the topic "Family Communication" would

be . . .
Very Quite Slightly Neutral Slightly Quite Very

Punishing: : : : : : : :Rewarding

Belief about Intrinsic Value.

Subjects response to the same question, on a seven-
point scale ranging from "unpleasant" to "pleasant."

In order to duplicate Fishbein's procedures, subjects
also responded to this question on a scale ranging from "good"
to "bad." They were also asked of their certainty on these

and other answers.

17

Personal Normative Beliefs.
The following question tapped students' beliefs
about their normative obligations to attend:

The next question concerns whether any moral obli-

 

gations which you personally feel toward yourself or others
will affect your decision whether to attend a lecture on
familyacommunication.

Do you think that attending the lecture is something

you ought to do or something you should not do?

 

 

 

Obligation No Obligation Strong
Not to Either Obligation
Attend Way To Attend

l 2 3 4 5 6 7

Social Normative Beliefs.
Three questions tapped social normative beliefs for
three possibly influential reference groups: friends, pro-

fessors, and family. Subjects were asked, "Regardless g:

 

your own personal views, would each of the following kinds
of people feel you had a moral obligation to attend such a
lecture? On the average, would each group think that this

is something you ought to do, or something you should not

 

do?" Responses were registered on scales like that used

for personal normative beliefs.

Motivation to Comply (Rankings.)
Students were asked to rank items corresponding to

the questions cited above--"what would be rewarding to me,"

 

18

"what I feel I should do," "what my best friends would say

 

I should do," etc.--"according to their importance £2.20” in

 

deciding whether to attend a lecture 9g Family Communication."

Motivation to Comply (Ratipgs).
Subjects were then asked to rate each of the six
items they just ranked on seven point scales ranging from

"Very Important" to "Very Unimportant."

EXEOSUIB 0

Subjects were asked how much they knew (including
information from all possible sources) about the following
objects: family communication, public lectures, MSU Depart-

ment of Communication, and Professor Cushman. For example:

 

Know Do Not Know Know
Nothing Know Fairly Extremely
At All Very Much Well Well
FAMILY o o o o o
COMMUNICATION ° ' ' '
l 2 3 4 5 6 7

In addition, subjects were asked whether they had
attended any prior lecture specifically about family communi-
cation, and about whether they had taken courses "in which

family communication was a major topic of discussion."

Attitudes Toward Related Objects.
Subjects rated the four 'objects,‘ family communica-

tion, MSU Department of Communication, public lectures, and

19

Professor Cushman, on semantic differential scales ranging
from punishing to rewarding, good to bad, pleasant to un-

pleasant.

Demographics.

 

Subjects indicated their class standing, sex, marital
status, number of children, whether communication was their

major, and their date of birth (for matching purposes).

Family Tie.

 

Subjects responded to the question,"How close do
you feel to your family?" on a seven point scale from "Very

close" to "Very distant."

Attitude Toward the Object.

 

The "object" most relevant to the behavior studied
here was determined to be "instructing people in techniques
of family communication." In an operational technique vali-
dated by Schwartz and Tessler (1973), subjects were presented

with five objections to family communication instruction.

 

They were asked "how much merit" they found each to have as

an objection to "classes about family communication." A

 

sample question:

No Some Very Much
Merit Merit Merit

After learning family communication
techniques, people become overly
self-conscious about their communi-
cation with their families.

20

Questions Added in Time 2

 

In the second wave of questionnaires, subjects were‘
asked six additional questions. They were asked:

--whether they filled out a time 1 questionnaire

--whether they had heard about the lecture from any
outside source

--whether they had talked with anyone about the
lecture

--whether they knew of any reason why they could
not attend the lecture

--HOW often they generally attended public lectures.

In addition, as a manipulation check, subjects in treatment
groups were asked, immediately after reading the message,

to state its main point.

Message Variables

 

As was mentioned before, the three treatment messages
were designed to suggest either that (l) attending the lecture
would be rewarding, or (2) subjects had a moral obligation
to attend, or (3) subjects' families would expect them to
attend. But what we wanted the messages to express and
argue for, may not come across in the actual messages, as
they are interpreted by naive readers. I therefore sought
from naive readers an estimate of the relative impact of the
three messages.

Students in undergraduate communication classes were
each presented with one message and a questionnaire. Before
reading the message, they were informed of my research pur-
pose, the subject of the message, and six possible reasons

for attendance:

21

--because attendance would be pleasant.

--because attendance would be rewarding.

--because they had a duty to attend.

--because their friends would expect them to attend.

--because their best professors would expect them

to attend.

--because their families would expect them to attend.
The students read the messages, then were asked to rank the
six reasons on the basis of their relative importance in
each message, and to rate the messages on the amount of
information relevant to each reason they contained, on a
seven-point scale from "all" to "none" of the information.
Their responses were averaged for each message to provide
an estimate of the amount of information in each message
supporting each reason. (In Table IV below, estimates of
message mass derived by this technique are reported.)

In this chapter we have examined the design and

operational definitions used in this experiment.

IIII

 

CHAPTER III

DATA ANALYSIS AND RESULTS

In the next two chapters, I shall describe the data
analySis done to test the hypotheses and its conclusions,
then make a few more general statements about conclusions that
might be drawn from the data, and finally, in Chapter IV,
draw conclusions about the study as a whole. The first five
hypotheses are so closely related to Dulany's use of his own
model, and to Fishbein's and Schwartz and Tessler's analyses
of it, that the initial analysis is practically dictated by
their prior examples. The analysis relevant to the last
two hypotheses is somewhat more tentative, for a number of

reasons.

The First Four Hypotheses

 

The first four hypotheses, as stated in Chapter I,
are as follows:

H l. The relationship between BEV and BI is
necessary (for at least one treatment
condition), contingent (on MC), stochastic,
irreversible, and coextensive.

H 2. The relationship between BIV and BI is
necessary (in at least one condition),
contingent (on MC), stochastic, irrever-
sible, and coextensive.

22

23

H 3. The relationship between NBP and BI is
necessary (in at least one condition),
contingent (on MC), stochastic, irrever-
sible, and coextensive.

H 4. The relationship between NBS and BI is
necessary (in at least one condition),

contingent (on MC), irreversible, stochastic,
and coextensive.

Generally, the most important implication of each of these
hypotheses is that each component of the model is an important
predictor of behavioral intention in at least one treatment
condition, at least one point in time. This implication,
and the hypotheses themselves, can be tested by examining
the results of regression analyses performed on the data,
as the model itself strongly suggests and as Dulany and
Fishbein have done in the past. Using this approach, the
hypotheses would be confirmed if, for every component, in at
least one treatment condition its regression coefficient were
significantly different from zero. That would mean that,
inside the treatment group, that component of the model is
important and useful. The full results of the analysis are
given in Table 1. (Here, and throughout, a<.05 is set for
significance.)

Regression equations were estimated for the model as
a predictor of Behavioral Intention at time 1 and 2. Within
each time-period, overall regressions for the whole sample,
as well as regressions for each treatment or control group,

were calculated. Three different forms of the equation were

24

calculated in each case-one using Motivation to Comply as
measured by semantic differential ratings, one leaving out
the MC factor altogether (to provide information about Fish-
bein's assertion that the factor can be dropped from the
theory without loss), and one using MC as measured by
ranking-~the ranking of propositional forms that was re-
quested of subjects before they filled out the semantic
differentials. (The reason for computing alternate forms
will become more apparent in the sequel.)

As can be determined from Table I, the components
corresponding to Belief in the Extrinsic Value of the act,
Belief in the Intrinsic Value of the act, and Social Normative
Beliefs are all, at least sometimes, important predictors
of Behavioral Intention. (BEV is significant in 6 cases,
BIV in 6, and NBS in 4.) (The probability of each of these
patterns of findings is a good deal less than .05.) In no
cell does Personal Normative Belief as a component have a
beta weight which differs significantly from 0. Thus,
Hypotheses 1, 2, and 4 may derive support from this data,
while Hypothesis 3 is not supported by the data.

Other results of the analysis, though, seem to lend
support to the inclusion of NBP in the model. First, the
component corresponding to NBP does in several instances have
a fairly large beta weight, though it is not statistically
significant (e.g., .2117 or .2389 in the NBP group, time 1,

.3201 in the control group, time 1). Second, this hypothesis,

25

like H1, H2, and H4, requires the assumption that, if NBP
SQElB be triggered as a determinant of BI, it wag triggered--
that is, that the NBP group manipulation was successful.
There is some reason to doubt this, as shall become more
apparent in our discussion of the message variables below.
Third, and most important, the NBP component is yegy'impor-
tant in predicting change in behavioral intention, as becomes
apparent when we examine Table 2. The third component of the
model is in at least one case (MC form, Attitudinal group),
the most important and the only significantly predictor in
the model. It may therefore be unwise to drop this component
from the model: it is surely distinguishable from the other
components of the model and from behavioral intention, and
sometimes makes significant contributions to explanation;
at least further examination of the issue is indicated.
Hypothesis 5: The relationship between a linear

combination of the components MCx

BEV, MCxBIV, MCxNBP, MCxNBS, and

BI is necessary, sufficient, deter-

ministic, irreversible, and
coextensive.

The fifth hypothesis concerns the adequacy of the model as a
whole. Are the components powerful and immediate predictors
of Behavioral Intention? Two ways of testing this hypothesis
were suggested in Chapter I. First, we can examine the
multiple correlation coefficients to see whether the model,
in various forms, accounts for a substantial proportion of

the variance in BI. This will give us some index of the

w IIIIIII

'
I
\

 

 

26

predictive power of the model. Second, we can see whether
the model mediates the affects of other variables that are
strongly related to BI--whether, for example, a variable that
can cause change in BI must "first" cause change in one or
more of the model's components.

With respect to the first test, the results are mixed.
(See Table I.) In the overall sample, and in the Attitudinal
and NBP message treatment groups, the model does explain a
significant, and sometimes very substantial, proportion of
the variance. In the NBS and Control conditions, though, the
model fails to explain a significant proportion of the vari-
ance.

It is not extremely difficult to find a possible
explanation of these findings. Given a very small sample
size in some of the groups (due to mistakes in questionnaire
completion), we might expect great instability in the corre-
lations, due to sampling error. This would account for the
extremely high R2 found in the two treatment groups (Attitud-
inal and NBP), too. Of course, this means that the beta
weights fround in the groups are also unstable. We might
also note that our coefficients are in no case as large as
those found in laboratory experiments. On the whole, the
hypothesis seems to have been supported, but less strongly
than we would have desired.

To perform the second test of the hypothesis, I

examined the relations of certain variables: external to the

27

model, with behavioral intention. The statistical hypothesis
implied by H5 is that the partial correlation between any
outside variable and BI, when controlled for the values of
components of the model, is 293 significantly different from
0. Results of tests of this hypotheses are presented in
Table 3. The "raw" (uncontrolled) correlations of all var-
iables presented are different, significantly, from O. In
26 cases (17 variables, 9 at both time 1 and time 2), con-
trolling for the components of the model reduced the corre-
lation substantially, so that it was not significantly
different from 0. These cases fulfill the hypothesis. But,
in five cases, controlling for the components of the model
did not substantially enough reduce the original correlation
to make the partial correlation not significantly different
from O; indeed, in one case partialling increased an already
significant correlation (Reasons, time 2), and in another
case partialling doubled the value of an insignificant corre-
lation, making it significantly different from O (Attitude
toward family comm., #2, time 2). Thus, altough the model
mediates the effects of a wide number of variables, five
cases prove that it is not a sufficient immediate determinor
3f BI.

The two "Reasons" variables deserve some comment here.
They are two-valued variables. The first indicates whether
or not the subject said he had a reason why he could not

attend the lecture. The second indicates whether or not he

28

had a schedule conflict--whether he was too busy or had an-

 

other meeting at that time, or had to be out of town--and so
could not attend the lecture. Before partialling, their
correlations with BI are -.29 and -.35, respectively; after
partialling, the correlations are each approximately -.35.
The indication is that either of these variables, insofar
as they indicate schedule conflicts, explains a fairly large
and independent portion of the variance in BI. What they
explain, the components in my model cannot explain--thus,
schedule conflicts, etc., set an upper limit in the explana-
tory power of the model. Note also that this variable is
usually automatically controlled in a laboratory, usually by
the subject's sheer presence to participate in the experiment.
In Table 2, a more inclusive test of the necessity
and sufficiency of the model is undertaken. There, I attempt
to predict change in BI on the basis of change in the com-
ponents. If H5 is completely correct, change in BI should
take place ggly because of change in the components of the
model, and the strong relation present at times 1 and 2 should
also be present in the 'change equation.‘ Once again, support
for the model is uneven; in particular, in the NBP message
group, strong relationships present at time 1 and time 2
disappear when we look at the change relationships. (In the
regression equations reported in Table 2, I have added one
more variable to the equations--Behavioral Intention at time

1. This was done to remove the often-present spurious effect

29

due to relation between change in the model's components and
the initial point at which we began to study change--time 1.
This procedure thus removes bias in the estimation of the
regression coefficients (Werts and Linn, 1970; Cf. also
Harris, 1962).) On the whole, then, it would be unwise to
conclude on this evidence that the model is an immediate,
anecessary, and sufficient determinor of behavioral intentions.
The idea that a relationship exists is supported; the pro-
position that it mediates the effects of all other variables

is simply untrue.

The Communication Hypotheses

 

The hypotheses read as follows:

H 6. The relationship between MM/PM for any com-
ponent, including MC as a multiplier, and
change in that component, is necessary,
sufficient, deterministic, irreversible,
and coextensive.

H 7. The relationship between message condition
and change in BI is contingent (on the
effect of MM/PM upon the model's determin-
ing variables, including MC), stochastic,
irreversible, and sequential.

These hypotheses cannot be tested directly, since no
prior mass measurements for each component of the model were
made. However, comparatively indirect tests can be made,
using four measures of exposure to relevant attitudinal
objects as indirect indicators of prior mass.

Message Mass (MM), for each message treatment condi-

tion, was measured by asking students in 3 Communication 100

classes to determine, for each message, how much of the

30

information in the message supported each component propo-
sition in the model. Thus, there are four variables involved:
mass of the message relative to eggh propositional form; and
each of these variables takes on four values, one for each
treatment group, plus a value of 0 for the control group.
Two different sets of questions, alternative measures of
message mass, yielded values of message mass in the three
messages that correlated. .98+ (Spearman rank-order correla-
tion for the values for all three messages, all six propo-
sitions dealt with (including expectations of friends,
professors, and family under social normative beliefs.))
This is an indication that the measurement of "mass" in
various messages relative to various propositional forms is
at least ordinally highly reliable. (See Table 4)

Prior Mass (PM) measured at time 1 in four different
ways, for purposes of these tests, as exposure to: family
communication, the Communication Department, public lectures,
and Professor Cushman. Message Mass relative to each propo-
sitional forms was divided by each of these exposure items
to yield four indices of MM/PM for each propositional form--
a total of 16 MM/PM variables. These are the message vari-
ables used to test H6.

The findings in Table V seem to confirm H5, at least
for the second and fourth components of the model. There is
evidence of a relation between the mass ratio, relative to

those components and change in those components. A closer

31

examination of the data, however, revealed that this relation-
ship obtained between mass ratios relative to all components,
and change in each of these two components. This can be
seen by comparing pairs of columns in Table V; the second
column gives the average correlation between component
change and mass ratio, no matter with respect to what com-
ponent mass was measured. In short, the finding is
spurious--it exists because of a sharp contrast between
the control group and the other groups (between no-message
and message conditions), for change in components 2 and 4.
When the control group is removed, findings are so mixed,
for all components, as to be inconclusive.

Hypothesis 7 can be examined by looking at Table 6.
There are no major differences between group means on
Behavioral Intention, so the typothesis cannot be tested
as stated. On the other hand, if we assume that the mess-
ages were not sufficiently different to produce significant
differences in intention, then the data is consistent with
the hypothesis. Does this finding cast doubt on the results
of H6, since mass ratio is sometimes significantly related
to component change, which is in turn related to change in
BI? No; the correlations involved are fairly small, so
that, unless there were an independent effect of MM on BI,
no raw effect could be expected. I tested for an indepen—

dent effect, and found none whatsoever.

32

Additional Comments

 

Two additional comments, based on the results of the
analysis, seem worthy of mention here. First, it might be
instructive to compare, in Tables 1 and 3, the results when
Motivation to Comply is measured in different ways, or not
included in the model at all. No clear pattern of super-

iority 25 inferiority emerges for a model that includes MC,

 

or that measures it one way rather than another. If there
is any noticeable trend, it is that social expectations
(NBS) have a greater role in explaining BI when MC is included
in the model and is measured by rank-order. Moreover, it is
slightly more often that the model, and its components, attain
significance when MC, measured by ranking, is included in the
equation.

However, the intent of the MC operationalization was

to have subjects recognize the comparative differences between

 

the importances of components when they marked the semantic
differentials, rating MC. The experimental results seem to
me to indicate that this purpose was not accomplished in the
questionnaire instructions, and thus that MC as a variable
deserves more study, with an eye to clearer and more valid
measurement. In particular, when we are studying, or de-
signing a message to influence, the perceived social expecta-
tions on a subject, we ought to take the motivation to comply

with these expectations into account.

33

The second point I would make relates to the message
manipulation used in the experiment. The results, and the
measures of message mass, show clearly that the messages
were not sufficiently different in content and aim to distin-
guish among the experimental groups--in every case, the mass
of the message was taken by subjects to be focussed on the

proposition that attending the lecture would be rewarding--

 

BEV. Given this lack of difference within the manipulation,

it may not be surprising that treatment groups displayed

no systematic differences in their responses to the messages.
Clearly, a reprication is indicated, using more powerful

and more distinct messages. It is possible that the measure

of message mass introduced above will provide a means, unavail-
able up to now, of insuring the "validity" of a manipulation

by pretesting. Also, a message which more clearly focuses

on NBP than those actually used, will provide a fairer test

of Hypothesis 3.

CHAPTER IV

CONCLUSIONS

This thesis has included the description of a 'new'
model of message-attitude-behavior relationships, and a test
of that model. The model is new in the sense that it is
rederived from Dulany's original theory, using, I believe,
sounder principles of theory-building than Fishbein has used
previously.

The first five hypotheses dealt with the model per g3
as a predictor of Behavioral Intention. The statistical
tests showed that subject beliefs about the extrinsic reward
he might derive, the intrinsic pleasure he might feel, and
the expectations of others were all influential in determining
subject intentions. (That is, H1, H2 and H4 were confirmed.)
While the model as a whole showed a fairly high correlation
with behavioral intention, it proved not to be as substantial
a predictor as expected, nor to mediate the influence of
all external variables. Thus, H1, H2, and H4 are accepted,
H3 and H5 rejected with partially extenuating conditions. A
qualification must be placed on these findings, though. The
pattern of findings is uneven and unpredictable; while they
differ from chance, they do so in no clearly recognizable

34

35

direction. The problem here may be caused by how sample
size sampling error weak manipulations, or all three. I
cannot find, as Fishbein did in several studies, a straight--
forward explanation for the fluctuations in beta weights.
(The fluctuation is duplicated in the ppstandardized
regression coefficients.)

- Another perspective on these findings comes when
we compare them to selected findings of Dulany, Fishbein
(et. al.), and Schwartz and Tessler. In his experimental
study of verbal conditioning and propositional control,
which was probably most influential on the development of
Fishbein's and my models, Dulany (1968) found that his com-
ponents accounted for 77% fo the variance in BI (p. 237).
In their review of several studies using Fishbein's model,
Ajzen and Fishbein (1973) find many multiple correlations
(R's) in the range .80-.95-two exceptionally low R's have
values .385 and .594--the second of which exceeds the values
found for all regressions run on the total groups (R ranged
from .49+ to .58 for the overall group) and the average R
was .808. These studies ranged from tightly Specified
experiments to very broad-ranging surveys. Schwarz and
Tessler (1973) found a multiple correlation of about .50 in
their study about organ donation. On the whole, the present
study does not duplicate these various stronger findings.

The communication hypotheses--that message mass,

relative to prior mass, would affect the model components

36

and, through them, intentions, were not supported but were
consistent with the results of the study. The manipulations,
as measured by message mass relative to various components,
were not easily distinguishable, and may thus have failed
to produce the variance needed to confirm the hypothesis.
At any rate, the findings which seemed to confirm H6 were
seen to be produced instead by the strength of the message-
no message contrast-a spurious effect. H7 is consistent
with the data; since the contrast among message treatments
was not clear, no differences among the groups were observ-
able.

Among the contributions of this study to future
research are: its introduction of an operational definition

of message mass relative to various model components, which

 

seems to have utility in predicting effects of messages on
attitudinal variables, and the discovery of a new proposi-
tional form, in Dulany's terms, which might be phrased as
"Schedule conflicts prevent me from performing act X."
Among the problems which should be corrected in
future studies are: lack of multiple indicators for com-
ponent variables, low sample sizes in treatment cells, weak
manipulations, and model components which may not have been

phrased so as to seem independent to subjects.

APPENDIX I

DULANY'S THEORY

Dulany's theory was formulated as a new approach to
the field of conditioning, especially verbal conditioning,
which has been dominated by behavioristic paradigms. Behav-
iorists usually argue that learning occurs under the control
of functional reinforcers--rewards or punishments--with or
without subject awareness of whether he is being conditioned
o of the principle governing the conditioning--i.e., the rule
of behavior he is being taught. Dulany argues that theories
taking account of subject awareness are advances over simple
behavioristic theories for several reasons. First, such a
theory can explain why certain verbal statements by experi-
menters are in fact reinforcing in certain situations and not
in others, while other verbal statements are not reinforcers.
Second, conditioning affects like speed of learning and
overall increase in accuracy have a quite wide variance over
subjects. In some experiments, nearly all of this variance
can be explained by reported subject awareness of the rule
being taught or of a "correlated" (functionally similar)
rule. In at least one experiment, conditioning which did not
induce subject awareness of a fitting rule governing rein-
forcement could produce no significant increases in accuracy

37

38

of response over chance (Dulany, 1961). These results led
Dulany to formulate a "theory of propositional control"
(1962, 1968). The theory has seven main principles.

1. Mental contents--subjectively received informa-
tion, of which we are aware--are encoded as propositions.

2. The effect of information depends on the form of
the proposition in which it is mentally encoded.

3. There exists a class of behaviors-~conscious acts--
which are entirely under the control of the subject, and thus
are determined by information held by him.

4. Certain propositional forms are particularly
relevant to the determination of behavior:

a. RHd--Hypothesis of the Distribution of
Reinforcement--of the form, "Response
Class X is followed by reinforcement."

b. BH--Behavioral Hypothesis--of the form,
"Response class X is what I am supposed
to do."

c. BI--Behavioral Intention--of the form,
"Response class X is what I am trying
to do."

d. RHs--Hypothesis of the Significant of a
Reinforcer--of the form, "Occurrence
(nonoccurrence) of the consequence
meant that I had just done what I was
supposed to do (not to do, or neither).

e. RSv--Subjective Value of a Reinforcer--
of the form, "Occurrence (or nonoccurrence)
of the consequence felt pleasant (or
neutral, or unpleasant)."

f. MC--Motivation to Comply-~of the form,
"Whatever I am supposed to do (or not to

do), I want (or want not, or neither) to
do."

39

5. The determinative relations among these proposi-

tional forms are given by the following equations:

WO(RHd X RHS) = BH

wl(RHd x RSV) + w2(BH x MC) = BI

6. For conscious acts, BI determines behavior, pro-
vided-that the actor is able to do what he intends. Such
acts are said to be under propositional control--that is,
they are under the control of certain propositions of which
we are aware and about which we can report.

7. The research paradigm implied is experimental--
subjects are asked to perform fairly simple acts, different
reinforcement schedules are administered, and questions asked
during the series of responses are used to get six classes
of propositional answers from subjects. These answers are
cast into a regression format, used to predict BI and
behavior.

I should note that the brevity required of this sum-
mary of Dulany's thought forces me to ignore certain nuances.
For instance, Dulany would merely say in (1) above that a
mental content "can be represented as a proposition." (1968,
p. 342). But a reading of Dulany (1968) makes it clear that
the interrelations and dynamics of the components of his model
follow strictly the patterns of natural implications of these
propositional forms (thought not a formal logic of any appar-
ent sort). If mental contents can be represented as proposi-

tions and follow the logic underlying those propositions,

40

then we are justified in saying that they are encoded as

propositions.

APPENDIX II

DERIVATION OF AN ALTERNATIVE MODEL

There are three changes in Dulany's model required
to render it applicable to message-attitude-behavior study.
One of these is along the lines suggested by Fishbein; the
others depart from his model.

1. The term BH--the subject's hypothesis about what
he is supposed to do--must be split because of our shift of
focus from the conditioning laboratory to the external, social
world. The subject no longer is oriented (we hope) to res-
pond to what the experimenter is conditioning him to do;
instead, he must adjust his activities to (a) different
groups of significant others, and (b) the demands of his own
moral code, if such demands are present. This means that
the BH term of the model is transformed into several com-
ponents, from "I am supposed to do X," to (a) "My family
(or some other group of significant others) expects me to
do X," a propositional form which we shall call a social
normative belief, or NBS; and (b) "I (morally) should do X,"
a propositional form which we shall call a personal normative
belief, or NBP. The relevance of personal normative beliefs

or the expectations of any particular group of significant

41

42

others is relative to act X, and there is one social norma-

tive belief for each relevant group of significant others.
2. Similarly, RHd-~the hypothesis of the distribu-

tion of reinforcement, of the form, "if I do X, consequence

Y will ensue,’ is transformed because, for a given social
act, it may be intrinsically pleasurable or it may be done

on account of its desired and rewarding consequences.
Therefore, RHd becomes (a) a Belief about the Intrinsic

Value of the act--BIV—-of the propositional form "Doing X
would be pleasant," and (b) a Belief about the Extrinsic
Value of the act, or BEV, of the propositional form "Doing

X would be rewarding."

3. Both for Ajzen and Fishbein (1971) and for
Schwartz and Tessler (1973), the most troublesome term to
operationalize has been MC--motivation to comply. The main
advance made here is to note that, like MC, RSv--the sub-
jective value of the reinforcer--is a subjective measure of
a 'motivator,‘ the reinforcer, although the reward is a
direct one rather than due to meeting a given set of expec-
tations. What is intended, in fact, is a measure of the
comparative motivations to obey one propositional form rather
than another. If this is true, we can measure MC for all
propositional forms so as to induce 85 to rate their compar-
ative importance in determining BI. Thus MC is transformed

into a comparative variable, measured for each component of

the model.

43
The resulting new model is:

B’bBI=bl(MCXBEV) + b2 (MCXBIV) + b3(MCXNBP) + b4 (MCXNBS)

ABBREVIATIONS FOR TABLES

*=P<.05; **=P<-Ol; ***=P<-001

BEV stands for findings related to the Belief in
Extrinsic Value component of the model (rewarding-punishing).

BIV stands for findings related to the Belief in
Intrinsic Value component (pleasant-unpleasant).

NBP stands for findings relevant to the Personal
Normative Belief component.

NBS stands for findings relevant to the Social Norma-
tive Belief component.

MC, MCR, and (NoMC) stand for different fgpmg of the
equation model on a 1-4. MCR stands for the equation, calcu-
lated using Motivation to Comply with each component, measured
by rank-orders. MC stands for the equation with the MC
factor measured by semantic differential. (NoMC) stands for

the model, excluding the MC multiplier.

In Table IV, column 1 was determined by averaging

judge rank-orderings of the amount of information relevant

 

to each model component, in each message. Column 2 is
equivalent to 6 minus the number in Column 1. Column 3 was
calculated by averaging judges' semantic differential ratings
of message mass for each component.

44

45

 

memo. owed. mmmo. mmha. mmz
hmaoo. Nome. ammo. mmma. mmz
hvoo. mvoa. mmao. hwwa. >Hm
NmmN. «romwm. mHvN. rewvom. >mm
«Ravmmm. rrrwvmm. ADS OZV
mmmo. Homa. mvmo. mmmH. mmz
moooo. tho.l Hmao. wwao. mmz
memo. «mmmm. mmmo. anmm. >Hm
«5mm. «mmmm. mmna. «vmmm. >mm
*«amomm. kkkmmvm. US
mmmo. Reammmm. ammo. areovhm. mmz
maooo. mvao.l oomo. hmva. mmz
mmmmo. hmma. mnvo. «Rmvmm. >Hm
hmmm. «Rammmm. mama. Raemmmm. >mm
kkkmovm. kkrvomm. mUE
Umcflmamxm cmcflmamxm
oocmflhm> unmamz muom m mocmflnm> unmaoz mpom m ucocomfioo Euom
Hmaoauaaea . N Hmaoflbaaaa . N
N UEHB H UEHB

 

mflmmamcfl coflmmmummm OHQHDHDS mo muHSmom

H mqmfle

46

 

mmooo. mnmm.l ammo. omom. mmz
ooooo. omao.| moooo. whoo.| mmz
omao. omm~.| mmvnm. «mhmv. >Hm
mmvm. «mmnn. omvhm. avmmv. >mm
«Hamm. «Ramwoo. A02 ozv
mvoo. mmmNH.I mooo. homo. mmz
hHoo. mmvo.| mane. amowo. mmz
Homo. Hamm. mnmma. *HmHm. >Hm
wmmm. mmmv. movm. mem. >mm
Hmam. Panoam. Oz
mmvoo. mmmo. ommo. oamno. mmz
Naooo. unvao.l mmooo. mammo. mmz
ooooo. mmmoo. mnmm. RROOHm. >Hm
monm. Rmmmm. mnmm. vam. >mm
rwvmm. «RHhHm. mUE
ommmmoz Hmcflcdpfluufi mcfl>flmoom msouw
cocflmfimxm confimamxm
mocmanm> usmaoz mumm m mocmaum> ucmfloz Opom m ucmcomaoo Ewom
Hmcoflwflaea . m HEEOHAaaaa m
N oEHB H oEflB

 

COUCHDGOUIIB mqmfia

47

 

mmmoo. mnva. ammo. mmam. mmz
hmmo. hmmoo. ammao. mmmao. mmz
mmmmo. mmhm. mmmo. mmmm. >Hm
oaav. mmmm. mmmmv. mmom. >mm
kamahv. «armonm. ADE OZV
mmmo. ahmm. ammo. mvvo.l mmz
maao. mamm.| vmao. Naam. mmz
mamo. Namm. mnom. «ammmm. >Hm
mmmmv. «avhm. nmam. mnoo. >mm
«#«mmmm. «Rommmv. Us
mmaN. «mmmv. mama. «omom. mmz
mmooo. mmmmo.| homo. mmmm. mmz
mmmmo. wham. mnma. mama. >Hm
mmvm. nmmm. mama. mnmm. >mm
kkhommv. *«kmovm. mUZ
ommmmoz amaaom o>aumEHoz aOGOmuom mCa>amoom msouw
cocamamxm Umcamamxm
mocmaum> uzmaoz muom m moacanm> unmaoz muom m pcocomeou Euom
amaoauaaea . m anneadaeea m
m 08a9 a oEaB

 

UOSGHHCOUIIB mamfle

48

 

oaao. ahva. ammo. amam. mmz
mmvo. mhma. waooo. mmmvo.| mmz
mooo. wmmo.| mmoo. amoa. >Hm
mmhm. mamm. oahm. mmhm. >mm
mmmm. mmam. A02 ozv
mmmo. mmma. wmoo. mmbmo. mmz
mmao. mmnmo. mmoao. mmmo. mmz
mahoo. mama. ammo. mmmm. >Hm
mmmm. maam. ahva. nmmm. >mm
mmmm. mmam. OE
mmmo. noma. mhmo. hmmm. mmz
anao. mhao. ooooo. mmmo.| mmz
mnoo. mhma. vamo. mamm. >Hm
hamm. mmmm. mama. mmmm. >mm
mmmm. mamm. muz
ommmmoz moaaom o>aumEuoz amaoom mca>aooom msouo
cocamamxm cocamamxm
mocmanm> unmaoz muom m mocmaum> unmaoz muom m ucocomEOO Euom
amcoabaaaa . N amaoNDaaaa N

N maaa

N asaa

 

UODCHDCOUIIH mamfie

.mamaucouomwao UaucmEom >9 OUHDmmoE Mameou ou c0aum>auo¢ mammfiuoz
.mmcaxcmu ma OOHSmmoE hamEou on coaum>auoz mcmozuvmz

49

 

 

aoo.vmnkac ao.vmuaa mo.vQNe
mamo. mmmm. mmoo. ommo. mmz
mooo. maao. amaa. mmmm. mmz
mvmo. mmmm. omoo. moaa.l >Hm
mamo. mmao. mmho. ovmm. >mm
aama. owom. A02 ozv
maoa. ammm. mmoo. mmmo.l mmz
maao. mmma.l ammo. aomm. mmz
mamo. vmam. mamo. mmmv.l >Hm
mmmo. maao.| vnma. *mamm. >mm
mmom. mmmm. Oz
mmmo. mmmm. «mac. moma. mmz
mmao. ammo. mmmo. mhma. mmz
ammo. mmvm. mmoo. mmmmo.| >Hm
mmao. mmoo. nmam. «wavm. >mm
mhha. mmmm. m0:
QDOHO aoaucoo
conamamxm Dosamamxm
mocmaum> usmamz muom m mocmanm> unmaoz muom m poocomEoo Euom
amcoaaaeea . N anneauaeea . N
m oEaB a oEaB

 

cmscaBCOUIlh mamde

50

TABLE II

Regression Equations for Change
in Behavioral Intention

 

 

2 Additional
Group Form(MC) Variable R Beta Variance
(Component) Weight Explained
Overall
MCR .3393***
BI, Time 1 -.4505*** .1674
BEV Change .3271*** .0668
BIV Change -.0191 .00032
NBP Change .2297** .0461
NBS Change .2527** .0587
MC .2905***
BI, Time 1 -.4236*** .1674
BEV Change .1873 .0261
BIV Change .0459 .0431
NBP Change .0604 .00497
NBS Change .1958* .0488
(No MC) .2720***
BI, Time 1 -.4321 .1674
BEV Change .2175* .05436
BIV Change .0348 .0010
NBP Change .0917 .0082
NBS Change .1262 .04095
Additudinal
Message MCR ‘ .5420**
Group BI, Time 1 -.6708*** .4031
BEV Change .2169 .0346
BIV Change -.2412 .0585
NBP Change .1776 .0361
NBS Change .1052 .0094
MC .5762**
BI, Time 1 -.7981*** .4031
BEV Change .1332 .0129
BIV Change -.3082 .0774
NBP Change .3582* .0799
NBS Change .06496 .00215
(No MC) .5935***
BI, Time 1 -.67918 .4031
BEV Change .3223 .1487
BIV Change -.2069 .0329
NBP Change .0838 .0048
NBS Change -.0075 .0037

51

TABLE II--Continued

 

 

2 Additional
Group Form(MC) Variable R Beta Variance
(Component) Weight Explained
NBP Message
Group MCR .2176
BI, Time 1 -.3375 .0453
BEV Change .3072 .881
BIV Change .0382 .0029
NBP Change .3147 .0452
NBS Change .2164 .0421
MC .2389
BI, Time 1 -.3207 .0453
BEV Change .1822 .0296
BIV Change .1433 .0669
NBP Change .0862 .0378
NBS Change .2846 .0592
(No MC) .1797
BI, Time 1 -.2709 .0453
BEV Change .1470 .02498
BIV Change .1563 .0338
NBP Change .1434 .0435
NBS Change .1932 .02998
NBS Message
Group MCR .4539*
BI, Time 1 -.2830 .0857
BEV Change .4553* .1765
BIV Change .2395 .1740
NBP Change .1980 .01545
NBS Change .0872 .0021
MC .4895*
BI, Time 1 -.2258 .0857
BEV Change .2942 .2652
BIV Change .3663 .0983
NBP Change .0982 .0232
NBS Change .1474 .0171
(No MC) .4139
BI, Time 1 -.2169 .0857
BEV Change .0596 .0028
BIV Change .4287* .2138
NBP Change .0564 .0026
NBS Change .3223 .10897

52

TABLE II--Continued

 

 

2 Additional
Group Form(MC) Variable R Beta Variance
(Component) Weight Explained
Control Group
MCR .3968
BI, Time 1 -.4495 .2558
BEV Change .2440 .0236
BIV Change -.2002 .0032
NBP Change .2466 .0359
NBS Change .3858 .0787
MC .3717
BI, Time 1 -.3692 .2558
BEV Change .2176 .0013
BIV Change -.2530 .0171
NBP Change -.1506 .0294
NBS Change .2364 .0684
(No MC) .3497
BI, Time 1 -.4928* .2558
BEV Change .0434 .0006
BIV Change -.1088 .00598
NBP Change -.1559 .02937

NBS Change .2657 .05796

 

 

Amvm.vmmo. Avam.vovo. Aaoo.vamm. Ammm.vmmo. Aama.vmmo. Avoo.vmma. .umoo .800
on .uum
a .OII
Aomm.vmmo. Ammm.vmmo. Aaoo.vnam. Amoo.vmmm. Aaoo.vmmm. aaoo.vmmm. .umoo .500
on .uu¢
N .0}
Amvm.vmmo.l Amam.vmmm.u Ammo.vmma. Amma.vwmo. Amma.vmmo. Amao.vmma. .U.m op
. mcsuapum
. m .OII
Avmo.vmma. Ammo.vama. Amva.vawo. Amma.vvno. Ammo.vmaa. Amoo.vwam. .O.m op
Otspaaufi
.EEOU
Amvm.vmmo.: Amma.vomo.l Amoo.vmma. Ammm.vmmo. Amma.vmmo. Aaoo.vnma. .Emm on
3 mcsuauum
5
Ammv.vbao. Amma.vaoo.l Aoao.vn>a. Ammo.vama. Ammo.vaaa. Aaoo.vovm. cmEnmso op
chamomxm
.udoo
Amoa.vamo.| Aomw.vaao.| Aamo.v>ma. 8800 on
musmomxm
Euom mos now Show Us MOM EHOM mu: Mow anm 02 MOM
mcaaaouucou mcaaaouucou coaDMaonuou mcaaaouucoo mcaaaonpcou coaumaouuou
amauumm awauumm 3mm amauumm amauumm 3mm
HH MEHB H MZHB magmanm>

 

masocomaoo accoz How mcaaaouucoo .mmanmaum>
accuoaxm spaz Hm mo mQOaumaouuoo amauumm cam 3mm

HHH mamdB

54

 

a .0:
Ammv.vmoo.l Ammm.vamo.n Aaoo.vomm.| Avmm.vmmo.| Aamm.vmmo.| ono.vmma.l .oca um
0» .uu<
N .0:
Ammm.vmmo. amam.vavo. Aaoo.vmmm. Aomm.vvmo.| Aomm.vmmo.| Aaoo.vomm.l cmszmso
0p .uu¢
N .o-..
Aamv.vvoo.l Anmv.vaoo.l Aaoo.vmmm. Ammo.vvma. Avmo.vmma. Aaoo.vmvm. cmESmsO
0» .uu¢
a .OII
Amma.vvno.l Amma.vomo.a Avoo.vmom. Aomo.vmma. Ammo.vmaa. Aaoo.vmvm. cmfinmso
ou .uufi
m .01:
Amaa.vaoa. Amam.vmmo. Ammo.vnva. Ammm.vmmo. Ammo.vmaa. Aaoo.vmmm. .ooa .Qsm
ou .uud
m .011
.ooa .nsm
0p .uu<
N .0!
Aoaa.vmoa.| Aawm.vmmo.l Amoo.vmmm. Aomw.vmoo.| Ammm.vmmo. Aaoo.vmmm. .ooa .Qsm
on .pp¢
m .OII
Amma.vmmmo. Aaoa.vmoa. Aaoo.voam. Ammo.vmaa. Avmo.vmma. Aaoovmmm. .umoo .500
OD .upm
EH0m mo: How EHOM OE HOm Snow mu: MOM EHOM Us How
mcaaaowpcou mcaaaouucoo coapmaouuou mCaaaouucoo mcaaaouucou coaumaouuou
amauumm amauumm 3mm amauumm amauumm 3mm

HH mzaE H mzae cancanm>

 

CODCHBCOUIIHHH mamda

55

.ouoN Eonm pcmuommac

maucmoamacmam on on Ummcsm ma coaumamuuoc OED .HOBWH no mo. ma am>Oa moccoamacmam Una ma
.coaumamuuoo map ma Baccamacmam OHOE can .ma pa HmaamEm onpllao>oa cocooamacmam map
ma Amommzucoumm Gav ccooom mcu “unmaoammmoo COaumaounoo on» ma Hogans umuam one

 

Aaoo.vmmvm.l Aaoo.vamm.l Aaoo.vmmm.n

Aaoo.vmm¢m.l Aaoo.vm¢m.l Aaoo.vamm.l

Ha Enom
.chmmom

a
Shownlom ow

no: chmmom

 

Edwin
Ammo.vmma.| Aoam.vmao.| Amao.vmma.| .ooa om
ou .upm
m .Onu
Ammm.vmmo.| Awaa.vmoa.l Amao.vmma.| .oma 0m
Op .upm
Show moz MOM EHOw Us How {EACH MOZ_Hom..EHOH.US.HOM.
moaaaouucou mcaaaONDGOU coaumaonuou mcaaaoupcoo mcaaaouucou coapmamuuou
amauumm amauumm 3mm awauumm amapumm 3mm
HH mEHB H MZHB magmaum>

 

GOSCHDQOUIIHHH mamfle

.+mm. ma conga cam OBD mcESaoo cwmzuon GOHDMaOHHoo HOUMOIxcmu cmEHmomm one

 

56

 

mm a In: II: II: amooeuummz
m.a m.a m.a v.4 mmouauummz
N.N m.N m.N N.N maaeamuummz
N.o a.a N.o m.m meamaumuummz
N.N o.N N.N N.m amz
N.a +N.a N.s N.a >Hm
a.N o.m N.N N.m >mm
m>aumEHoz amaoom
N.a II. In- In- amuoaunmmz
N.N a.a m.a N.N mmoumuummz
N.N o.m N.N N.m maaewmuummz
N.a N.a o.a o.m magmaumnnmmz
N.N a.m N.m o.N amz
N.a 4.4 0.4 o.N >Hm
o.m N.m N.N N.m >mm
o>aumEHoz aOCOmpmm
N.N II. II. In: amuoeIImmz
N.a o.a a.o a.m mmouauummz
N.a o.N m.a m.a saaemmnummz
a.o N.o :a.o +a.m mammaumunmmz
moN MoN O.m Com. mmz
o.N IN.4 N.N N.a >Hm
OoV Oov mom HoN >mm
amaaaabauua
Aoamom .um 5v mmmz mumEaumm xcmm mmmz Enom aOGOauamomoum
mmmum>m mo mmeaumm xcmm mafia: o.m mo mumEaumm w msouw ucoEummHB

 

mmmz omMmmmE mo macaumEaumm

>H mam49

57

TABLE V

Correlations Between (Message Mass/Exposures)
and Model Component Change

 

 

COMPONENT EXPOSURE TO FAMILY COMM. EXPOSURE TO PUBLIC LECTURES
Component- Average r Component— Average r
Specific for all Specific for all
Message Mass Measures Message Mass Measures
Mass r Mass r
BEV Change
With Control Group
MC .1339 .1228 .1323 .1216
MCR .0950 .0890 .1480 .1404
Without Control Group
MC .0217 .0819
MCR -.0364 .0952
BIV Change
With Control Group
MC .3059*** .2698 .0950 .0920
MCR .3893*** .3514 .1553* .1517
Without Control Group
MC .2035* —.l760
MCR .2043 —.1525
NBP Change
With Control Group
MC -.O380 -.0330 -.0090 -.0046
MCR -.0054 -.0233 -.Ol67 -.0326
Without Control Group
MC -.l3l9 -.1557
MCR -.1325 -.l42l

NBS Change
with Control Group

MC .0899 .0959 -.O47l -.0454

MCR .1981* .1970 .0300 .0165
Without Control Group

MC -.O674 .343l***

MCR -.0812 .2774**

 

58

TABLE V--Continued

 

 

COMPONENT EXPOSURE TO COMM. DEPARTMENT EXPOSURE TO DR. CUSHMAN
Component— Average r Component- Average r
Specific for all Specific for all
Message Mass Measures Message Mass Measures
Mass r Mass r
BEV Change
With Control Group
MC .0620 .0567 .1500* .1354
MCR .0240 .0209 .0651 .0600
Without Control Group
MC -.0509 .0847
MCR -.0509 .0232
BIV Change
With Control Group
MC .1156 .0944 .1428 .1239
MCR .2188* .2091 .2793** .2501
Without Control Group
MC -.O748 -.0699
MCR -.O338 .0260
NBP Change
With Control Group
MC .0116 .0089 .0325 .0280
MCR .0019 -.0145 .1435 .1276
Without Control Group
MC -.1016 -.0252
MCR —.l2OO .0150
NBS Change
With Control Group
MC .1053 .1021 .0896 .0944
MCR .2300** .2388 .1600* .1607
Without Control Group
MC -.0668 -.0526
MCR -.1261 -.0867

 

59

 

a.muoom mmcmgo amou may
on cocoa coon m>m£ .BDCac may
Ga o.> cam .mBOH m umuam can
ca o.mm mo mucmumcoo moaamomv

 

ceaucouca
mmm.m mmm.m mnm.m hoa.m amHOa>mnom
Ca omcmsu
mv.om mm.mm mm.mm mm.vm o mm.a m.a m.a mosxmmz
ma.am mm.mm mo.mm ah.mm o mm.a m.a m.a szmmz
am.mm oo.mm oo.mm ah.mm o m.m m.m m.m muzxmmz
mm.mm mm.mm mm.mm oo.nm o m.m m.m m.m szmmz
mm.mm mm.mm mm.mm ma.mm o m.a m.a m.a mozx>am
oo.om mm.mm mm.mm vo.mm o m.a m.a m.a sz>am
mm.mm mm.mm oo.vm mm.mm o m.m m.m m.a mozx>mm
oo.vm oo.mm mm.mm «o.mm o m.m o.m m.a Ozx>mm
msouw msonw msouo macaw mmmmmmz omwmmoz mommmoz ommmmoz
aouucoo mmz mmz .cuud aonucou mmz mmz .cuud

"How omcmnu cmoz

"mmmz mommmmz

uucocomEou mom

 

mQSOHO mmOHom mcmozalomcmno ucmcomfioo cam mmmz ommmmoz

H> mamfifi

60

SAMPLE QUESTIONNAI RES

Note: Following are examples of time I and time II
questionnaires, and of the questionnaire used to obtain
meaSures of message mass. The time II questionnaire had
four forms: questions with p2_message manipulations, or

with a message manipulation aimed at attitudes, or at

 

personal normative beliefs, or at social normative beliefs.

 

 

The time 2 questionnaire presented includes an attitudinal
message; the two alternative messages appear immediately
after that questionnaire. The sample of the message mass
questionnaire also contains the attitudinal message--two
other forms of it contained the other two messages, with

no other differences.

: 6 1
MICHIGAN STATE UNIVERSITY

College of Cbnmunication Arts East Lansing ' Michigan 0 48824
Department of Cannmicatim

On Wednesday, May 22 and Thursday, May 23, 1974 the Department of Commi-
cation will sponsor a lecture by Prof. Donald P. Cushman on Family Communication.
This lecture will focus on the students role in the family. 'l'ne lectures will
be held at 7:00 P.M. in Room 504 South Kedzie Hall. All students who are inter-
ested are invited to attend.

In the next few minutes the Department of Commmication would like to ob-
tain some information from you regarding your reactions to such a lecture ser-
ies. Three tines in the next five weeks we will seek similar informatics: on
this subject. We are seeking this information for statistical purposes only.
Your answers are entirely confidential. No one will be identified by name.
Please do n_<_>__t Lu: your name gn_ this questionnaire. It is very iuportant to you
and to the_' Department that you answer each and every question a accurately as
possible. Please read each question twice before answering it. Upon couple-
tion of the questionnaire please go back and check to see that you have answer—
ed all the questions to the best of your knowledge.

Thank you.

Joseph Woelfel
Chairman
Department of Commication lectures Series

62

Professor Cushman of the MSU Department of Communication intends to give a pub-
lic lecture on the t0pic "Family Communication". Will you attend that lecture?

 

Even
Definitely Chance Definitely
Not Either Way Yes
1 2 3 4 5 6 7

ttending a lecture by Professor Cushman of the MSU Department of Communication
on the topic "Family Communication" would be...

Very Quite Slightly Neutral Slightly Quite Very

punishing : : : : : : : : rewarding
l 2 3 4 5 6 7

 

 

How certain are you of this judgement?

 

 

Very
Certain Uncertain
1 WI, 2 iv 3 u 5 6 7 TV
good : : : : : : : : bad
1 2 3 4 5 6 '_ 7

How certain are you of this judgement?

 

 

Very
Certain Uncertain
l 2 3 u 5 6 ‘7‘
unpleasant : : : : : : : _ : pleasant
l 2 3 4 5 6 7
How certain are you of this judgement?
Very
Certain Uncertain

 

63

The next question concerns whether any moral obligations which you personally

 

 

feel toward yourself or others will affect your deciEion whether tofiattend a
lecture on family communication.

Do you think that attending the lecture is something you ought to do or some-
thing you should not do?

 

 

Obligation No Obligation Strong
Not To Either Obligation
Attend Way To Attend

1 2 3 I 4 5 ' 6 J 7

How certain are you of this judgement?

Very
Certain Uncertain

 

 

64

The next few questions ask how other people would react if you discussed with
them whether you should attend a public lecture on the topic of Family Commu-
nication.

Regardless of Lour own personal views, would each of the following kinds of
people feel_ you had a moral obligation to attend such a lecture? On the aver-
age, would each group think that this is something you ouggt to do, or something
you should not do?

 

 

 

 

 

 

Obligation No Obligation Strong
Not To Either Obligation
Attend 'Way ~ To Attend
a) YOUR BEST
FRIENDS . : : : : : :
WOULD SAY. l 2 3 1+ 5 6 7
b) YOUR BEST
PROFESSORS : . : : . . .
WOULD SAY l 2 3 1+ 5 6 7
c) YOUR
FAMILY : : : : : : : :
WOULD SAY I 2 3 u s 6 7

Below is a list of five factors which peOple often take into account in making
decisions. Please ran___I_<_ these factors according to their importance to y_g_ in
deciding whether to attend a lecture on Family Communication. Please _rank the
five factors in order of importance-- 7 " for most important, "2" for next most
important, and so forth.

What would be pleasant for me
What would be rewarding for me
What I feel I should do

What my best friends would say I should do

 

What my best professors would say I should do
What my family would say I should do
f-Toz-r certain are you of these rankings?

Very
Certain Uncertain

 

l 2 3 1+ 5 6 7

65

Now consider each of the five factors by itself. How important to you will each
of these factors be in deciding whether to attend a lecture on Family Communica-
tion?

 

 

 

a) What would be pleasant for me:

Very Very
Important : : : : : : : : Unimportant
1 2 3 N 5 6 7

 

How certain are you of this judgement?

Very
Certain : : : : : : : : Uncertain

 

b) What would be rewarding for me:

 

Very Very
Important : : : : : : : : Unimportant
l 2 3 u 5 6 7

 

How certain are you of this judgement?

Very
Certain : : : : : : : : Uncertain

 

c) What I feel I should do:

Very Very
hmmtmm : : : : : : : :thmmth
1 2 3 B 5 6 7

 

How certain are you of this judgement?

Very
Cmtmn : : : : : : : :mwmtfin

 

d) What my best friends would say I should do:

 

Very Very
Important : : : : : : : : Unimportant
l 2 3 H 5 6 7

 

How certain are you of this judgement?

Very
Certain : : : : : : : : Uncertain

 

66

e) What my best professors would say I should do:

 

Very Very
hmmtam : : : : : : : :[mnmmtmu
1 2 3 4 5 6 7

 

How certain are you of this judgement?

Very
Certain : : : : : : : : Uncertain

 

f) What my family would say I should do:

Very ‘ Very
Important : : : : : : : : Unimportant
l 2 3 N 5 6 7

 

How certain are you of this judgement?

very .
Certain : : : : : : : : Uncertain

 

67

The next few questions concern how much you feel you know about certain things.
In answering, consider any information you have gain ed from guy source, includ-
ing personal experience, from talking to other people, or from media like books,
How much do you feel you know about each of the fol-

newspapers or television.
lowing?

 

 

 

 

 

Know Do Not Know Know
Nothing Know Fairly Extremely
At A11 Very Much Well Well
a) FAMILY . :
COMMUNICATION 1 2 3 5 6 7“
b) PUBLIC . :
LECTURES l 2 3 5 5 7
c) MSU DEPART- :
MENT OF'COM— l 2 3 5 6 7
MUNICATION
d) PROFESSOR : : g : . :
CUSHMAN 1 2 3 5 6 7
(Exposure)

Have you ever attended a lecture or seminar about the specific topic of family

connumi cation?

Yes No

Have you ever taken a course in which family communication was a major topic of

dis cussion?

Yes No

68

Family Comnmni cation is . . .

 

 

 

 

Neutral
Punishing : : : : : : : : Rewarding
l 2 3 4 5 6 7
Good : : : : : : : J : Bad
1 2 3 1+ 5 6 7
Unpleasant : : : : : : : : Pleasant
l 2 3 1+ 5 6 7

The MSU Department of Communication is. ..

 

 

 

 

 

 

 

 

 

 

 

 

Neutral
Punishing : : : : : : : : Rewarding
1 2 3 4 5 6T7 7
Good : : : : : : : : Bad
I 1 2 3 u * 5 5 ‘7
Unpleaant : : : Pleasant
l 2 3 u S 6 7
Public Lectures are...
Neutral
Punishing : : : : : : : .1 : Rewarding
l 2 3 4 5 6 7
Good : : : : : : : : Bad
1 2 3 1+ 5 6 7
Unpleasant : : : : : : : : Pleasant
1 2 3 1+ 5 6 7
Professor Cushman is. ..
Neutral
'nishing : : Rewarding
u, 7
Good : : Bad
4 '7
Pleasant : _ : Unpleasant
1+ 7

69

(Demographics 8 Miscellaneous)
Please answer the following questions about yourself:

1) Is Communication your Major field? Yes

 

 

No
2) What is your class standing? __ Freshman __ Senior
SOphomore __ Graduate Student
Junior __ Special
3) What is your §_e_>_<_? __ Male
Female

4) What is your marital status? __ Single, never married
Single, formerly married
__ Engaged
Married

5) Do you have any children? Yes If yes, how many?

*

No

 

6) How close do you feel to your family?

Average

Very Close : : : : : : : L : Very Distant

7) What is your date 31: birth?

 

 

Month _Day Year

70

The following are some reasons that professors and others have given for 3gp in-
structing people in techniques of family communication. How much merit do you
find in each of these statements as objections to classes about family communi-
cation techniques?

 

The statement has

No Some Very Much
Merit Merit Merit
After learning family communication tech-
niques, people become overly self-conscious
about their comunication with their
families. . l 2 3 1+ 5 6 7

 

 

After learning family communication tech-
niques, family members often disturb pat-
terns of communication inside the family,
creating 931 communication problem. 1 2 3 4 5 6 7

 

 

Other family members often resent attempts
by college students to change family commu—
nication patterns according to what they

learn at school. 1 2 3 I4 5 6 7

 

 

People who have learned family connumica-
tion techniques develop rigid and unreal-
istic ideas about the ways their families
should conununicate. l 2 3 4 5 6 7

Families are so different that family com-
munication techniques which can be taught
are often irrelevant to most family situa-
tions. 1 2 3 u 5 6 7

 

71

MICHIGAN STATE UNIVERSITY

 

College of Comnmication Arts East Lansing - Michigan - #882M
Department of Communication

On Wednesday, May 22 and Thursday, May 23, 1974, the Department of
Communication will sponsor a lecture by Professor Donald P. Cushman on
Family Communication. The lecture will be given both nights at 7:00 P.M.
in Room 102 South Kedzie Hall. (NOTE: The room fer the lecture has been
changed.from 504 to 102 South Kedzie.) All students who are interested
are invited to attend.

We are once again asking a few minutes of your time to give us some
infOrmation regarding your reactions to this lecture. We are seeking
this infOrmation fer statistical purposes only. Your answers are entirely
confidential. No one will be identified by name. Please do not_put your
name on this questionnaire.

It is very important that you answer every question as accurately as
possible. In order to do the required statistical analyses it is necessary
that all the questions on your questionnaire be answered. Please read each
question twice befOre answering it. Upon completion of the questionnaire
please go back and check to see that you have answered all the questions to
the best of your knowledge.

Thank you,

Joseph Woelfel

Chairman

Department of Communication Lecture
Series

72

Have you previously filled out a questionnaire on your reactions to the Family
Communication Lecture?

No

 

Yes

 

Have you heard about the Family Communication Lecture from any source other than
the previous questionnaire?

No

 

Yes If Yes, briefly describe how you learned about the lecture:

 

Have you talked with anyone about the Family Communication Lecture since you
first Heard about it?

No

 

Yes If Yes, roughly how many conversations?

 

 

If Yes, briefly describe the conversations:

 

 

At this time, do you know of any reason why it will not be possible for you to
attend the Family Communication Lecture?

 

No

 

Yes If Yes, briefly describe the reason:

 

 

 

 

Cver the past few years, how often have you attended public lectures on any
subject. Please do not count regular classroom lectures.

 

Less than one per year

One or two per year

 

More than two per year

 

,73

PLEASE READ THE FOLLOWING MESSAGE, AND THEN ANSWER THE QUESTION AT THE BOTTOM
OF THE PAGE.

Both reason and research indicate that family stability is important to
the happiness and success of every person. Family stability in turn depends
upon the communication relationships which exist among the family members.

The Family Communication Lecture will provide you with a working know-
ledge of the communication principles involved in establishing and maintaining
healthy communication relationships. This should be of value to anyone who is
interested in a useful and enjoyable learning experience.

Professor Cushman's enthusiasm and competence have earned him outstanding
teacher awards from three universities, and have led to several invitations to
speak befbre family counselling groups. Over the years, many people have
benefited from his broad experience in the problems of interpersonal communi-
cation. ‘

Those of you who know Professor Cushman know that his lectures are any-
thing but boring! We are sure that you will find him entertaining, stimulat-
ing and challenging. His lecture will stress the practical side of interper-
sonal communication. You will find much that is useful in what he has to say.

Professor Cushman has developed a new approach to interpersonal communica-
tion which has important implications for family communication. Scientific
findings based on this approach can help you develop effective and enjoyable
relationships with others in your family.

Scientific research has rapidly advanced our knowledge of family communi—
cation, as researchers have accumulated a body of practical infermation about
how to improve interpersonal relationships in the family.

Studies have found that training can improve your awareness of your own
and others' communication behavior. Even a brief lecture can help you to
understand: 1) your own "style", and how this affects and is affected by
other family members; 2) how relationships develop, and how the relationship
depends upon the people in it; and 3) how conflicts arise at the relation-
ship level, and what can be done to cope with them.

In sum, there will be very few opportunities in your college career to
learn a set of principles which are relevant every day of your life. If you
are interested in the principles involved in establishing and maintaining
healthy communication relationships, and if you.want to have a useful and
enjoyable learning experience, then be sure to attend Professor Cushman's

lecture.

'What do you feel is the most important point made in the above message?

 

74

Professor Cushman of the MSU Department of Communication intends to give a pub-
lic lecture on the topic "Family Communication". Will you attend that lecture?

 

 

Even
Definitely Chance Definitely
Not Either Way Yes
'1 2 3 u 5 6 7

Attending a lecture by Professor Cushman of the MSU Department of Communication
on the topic "Family Communication" would be...

Very Quite Slightly Neutral Slightly Quite Very

punishing : : : : : : : : rewarding
l 2 3_—— E 5 6 7

 

 

 

How certain are you of this judgement?

 

 

 

 

 

 

 

 

Very
Certain Uncertain
l 4'2 " BEE“ “”"h 5 6 7
good : : : : : : : : bad
1 2 3 u 5 6 7
How certain are you of this judgement?
Very
Certain Uncertain
i’l 2 3 u 5 6 7
meleasant : : : : : : : : pleasant
l 2 3 u 5 6 7
"ow certain are you of this judgement?
Very

Certain Uncertain

75

The next question concerns whether any moral obligations which you personally

 

 

feel toward yourself or others will affect your deciSion whether to attend a
lecture on family communication.

Do you think that attending the lecture is something you ought to do or some-
thing you should not do?

 

 

Obligation No Obligation Strong
Not To Either Obligation
Attend Way To Attend

l 2 3 H S 6 7

How certain are you of this judgement?

Very
Certain Uncertain

 

76

The next few questions ask how other people would react if you discussed with
them whether you should attend a public lecture on the topic of Family Commu-
nication.

Regardless of your own personal views, would each of the following kinds of
people feel—you had a moral obligation to attend such a lecture? On the aver-
age, would each group think that this is something you ought to do, or something
you should not do?

 

 

 

 

 

 

 

Obligation No Obligation Strong
Not To Either Obligation 'r“
Attend Way To Attend
a) YOUR BEST
FRIENDS : : : : : .
WOULD SAY l 2 3 1+ 5 6 7
b) YOUR BEST
PROFESSORS : . : : . . . . A
WOULD SAY 1 2 3 u 5 6 7 :‘t
0) YOUR
FAMILY : : : . . .
WOULD SAY l 2 3 1+ S 6 7

Below is a list of five factors which peOple often take into account in making
decisions. Please rank these factors according to their importance from in
deciding whether to attend a lecture on Family Communication. Please rank the
five factors in order of importance-4U" for most important, "2" for next most
important, and so forth.

 

What would be pleasant for me
What would be rewarding for me
What I feel I should do

What my best friends would say I should do

 

What my best professors would say I should do
What my family would say I should do
'31-? certain are you of these rankings?

Very
Certain Uncertain

 

l 2 3 LL 5 6 7

77

Now consider each of the five factors by7itself.

 

How important to you will each

of these factors be in deciding whether to attend a lecture on Family Communica-

 

tion?

 

a) What would be pleasant for me:

Very

 

 

 

 

 

 

 

 

Important - : : :
l 2 3 u 7
How certain are you of this judgement?
Very
Certain
l 2 3 4 7
b) What would-be rewarding for me:
Very
Important : : :
l 2 3 u 7
How certain are you of this judgement?
Very
Certain
l 2 3 u 7
c) What I feel I should do:
Very
Important : : :
l 2 3 4 7
How certain are you of this judgement?
Very
Certain
l 2 3 4 7
3) What my best friends would say I should do:
Very
Important : : :
l 2 3 u 7
EC»! certain are you of this judgement?
Very
Certain
l 2 3 4 7

 

very

. Unimportant

: Uncertain

Very

:mmmfiwt

: Uncertain

Very

: Unimport ant

: Uncertain

Very

: Unimport ant

: Uncertain

. 3.4 «v ~13. L1'..-&§'.‘Brr~‘a-.~an.i

 

78

e) What my best professors would say I should do:

 

Very Very
hmmtan : : : : : : : :Ihhmmtam
l 2 3 u S 6 7

 

How certain are you of this judgement?

Very
Certain : : : : : : z : Uncertain

 

 

f) What my family would say I should do:

Very Very
Import ant : : : : : : : : Unimportant 1
l 2 3 4 5 6 . 7 *

 

How certain are you of this judgement?

Very .
Certain : : : : : : : : Uncertain

 

79

The next few questions concern how mudh you feel you know about certain things.
In answering, consider any infermation you.have gain ed from any_source, includ-
ing personal experience, from talking to other people, or from media like books,
newspapers or television. How much do you feel you know about each of the £61-
lowing?

 

 

 

 

Know Do Not Know Know
Nothing Know Fairly Extrenely
At All Very Much Well Well
a) FAMILY . : : : : : :
COMMUNICATION l 2 3 H 5 6 7
-b) PUBLIC . : : : . . :
‘ LECTURES 1 2 3 1+ 5 5 g 7
c) MSU DEPART- : : : . . .
MENT OF COM- 1 2 3 1+ 5 6 7
MUNICATION
d) PROFESSOR : : : : : : : :
CUSHMAN 1 2 3 4 5 6 7
(Exposure)

Have you ever attended a lecture or seminar about the specific tOpic of family

connumication?
Yes No

Eiave you ever taken a course in which family communication was a major tOpic of
discussion?
Yes No

 

Family Communication is. . .

8O

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Neutral
Punishing : : Rewarding
1 2 3 14 7
Good : __ , Bad
I 2 3 1+ 7
Unpleasant : _ : Pleasant
l 2 3 1+ 7
The MSU Department of Communication is. ..
Neutral
Punishing : : Rewarding
l 2 3 1+ 7
Good : . Bad
1 2 3 1+ 7
Unpleasant : . Pleasant
l 2 3 u 7
Public lectures are...
Neutral
Punishing : f : Rewarding
l 2 3 1+ 7
Good : : Bad
1 2 3 1+ 7
Unpleasant : I : Pleasant
l 2 3 1+ 7
Professor Cushman is. . .
Neutral
unishing : : Rewarding
l 2 3 u 7
Good : : Bad
1 2 3 u 7
Pleasant : : Unpleasant
l 2 3 u 7

81

(Demographics 6 Miscellaneous)
Please answer the following questions about yourself:

1) Is Communication your Major field? Yes

 

 

No
2) What is your class standing? Freshman __ Senior
SoPhomore __ Graduate Student
Junior __ Special

3) What is your six? __ Male
Female
4) What is your marital status? __ Single, never married
Single, formerly married
Engaged
Married
5) Do you have any children? __ Yes If yes, how many? __

No

 

6) How close do you feel to your family?

Average

Very Close : : : : : : . __ : Very Distant

l 2 3 4 5 6 7

'7) What is your date of birth?

 

 

Month hDay Year

 

82

The following are some reasons that professors and others have given for 3315 in-
structing people in techniques of family communication. How much merit do you
find in each of these statements as objections to classes about family communi—
cation techniques?

 

The statement has

No Some Very Much
Merit Merit Merit
After learning family communication tech—
niques, people become overly selfeconscious
about their communication with their
families. _ l 2 3 u 5 6 7

 

 

After learning family communication tech-
niques, family members often disturb pat-
terns of communication inside the family,
creating {:31 communication problerrs. l 2 3 1+ 5 6 7

 

Other family members often resent attempts
by college students to change family commu-
nication patterns according to what they

learn at school. 1 2 3 Lt 5 6 7

 

People who have learned family communica-
tion techniques develop rigid and unreal-
istic ideas about the ways their families
should communicate. l 2 3 u 5 6 7

 

Families are so different that family com-
munication techniques which can be taught

are often irrelevant to most family situa-
tions. 1 2 3 H 5 6 7

 

83

PLEASE READ THE FOLLOWING MESSAGE, AND THEN ANSWER THE QUESTION AT THE BOTTOM
OF THE PAGE.

Both reason and research indicate that family stability is an inportant
responsibility of every person. Family stability in turn depends upon the
communication relationships which exist among the family members.

The Family Communication Lecture will provide you with a working knot:-
ledge of the conmunication principles involved in establishing and maintaining
healthy comunication relationships. This should be of value to anyone who
ha a sense of responsibility for the well-being of his or her family.

Professor Cushman's enthusiasm and competence have earned him outstanding
teacher awards from three universities, and have led to several inVitations to
speak before family counseling groups. Over the years, many peeple have bene-
fited from his broad experience in the problems of interpersonal communication.

Professor Cushman feels that it is important to disseminate knowledge
about family communication, because he believes that the family is an institu~
tion which is basic, both to society and to the human needs of each of us. His

lecture will stress the ways in which good communication in the family is every-

one 's responsibility.

Professor Cushman has develOped a new approach to interpersonal comunica-
tion which has important implications for family communication. Scientific
findings based on this approach can help you to better perform the duties as-
sociated with your role in the family.

Scientific research has rapidly advanced our knowledge of family communi—
cation, as researchers have realized the importance of learning more about in—
terpersonal relationships in the family.

Studies have found that training can improve your awareness of your own
and other's communication behavior. Even a brief lecture can help you to
understand: 1) your own "style", and how this affects and is affected by oth-
er family members; 2) how relationships develop, and how the relationship de-
pends upon the peOple in it; and 3) how conflicts arise at the relationship
level, and what can be dore to cepe with them.

In sum, there will be very few Opportunities in your college career to
learn a set of principles which are relevant every day of your life. If you
are interested in the principles involved in establishing and maintaining
healthy coumunication relationships, and if you have a sense of responsibility
for the well-being of your family, then be sure to attend Professor Cushman's

0| 0 O O O O O O O O O O O O O O O O O O O C O O O O O O O O O O O O O O O O .

ell Hi

\VE..;,. IT“ . .‘J

 

 

84

PLEASE READ THE FOLLOWING MESSAGE, AND THEN ANSWER THE QUESTION AT THE BOTTOM
OF THE PAGE.

Both reason and research indicate that family stability is an iuportant
desire of every family. Family stability in turn depends upon the connumica-
tion relationships which exist among the family menbers.

The Family Communication Lecture will provide you with a working knowledge
of the communication principles involved in establishing and maintaining healthy
communication relationships. This should be of value to anyone who wants to
provide the kind of intelligent interpersonal leadership which peeple expect of f”
a student trained in communication. '

Professor Cushman's enthusiasm and competence have earned him outstanding 3
teacher awards from three universities, and have led to several invitations to :1
speak before family counseling grows. Over the years, nanny people have bene- 1
fited from his broad experience in the problems of interpersonal communication. 3

 

Professor Cushman has found that most families today are eager to learn
more about communication. Lecturers and courses on interpersonal communication
are always well-attended. Several books on the subject have made the best-
seller lists. Training centers have grown up around the country. Professor
Cushman's lecture will stress the kinds of information that have been found to
be of interest to most families.

Professor Cushman has developed a new approach to interpersonal connumica-
tion which has important implications for family communication. Scientific
findings based on the approach can help you contribute to the stability your
family wants.

Scientific research has rapidly advanced our knowledge of family communi-
cation, as researchers have responded to the increased public demand for infor-
mation about interpersonal relationships in the family.

Studies have found that training can improve your awareness of your own
and others' comunication behavior. Even a brief lecture can help you to
understand: 1) your own "style", and how this affects and is affected by
other family members; 2) how relationships develop, and how the relationship
depends won the peeple in it; and 3) how conflicts arise at the relation-
ship level, and what can be done to cope with them.

In sum, there will be very few opportunities in your college career to
learn a set of principles which are relevant every day of your life. If you
are interested in the principles involved in establishing and maintaining
healthy communication relationships, and if you want to provide the kind of
intelligent interpersonal leadership which peeple expect of a student trained
in cmunwication, then be sure to attend Professor Cushman's lecture.

85
Ausmt 12, 1971:

On fine next page of this questionnaire is a message which ha been need
in past conmunication research in other ooumunication classes. The message
was used to persuade students to attend a lecture on family comunications
given by Professor Donald Cushman. What we seek from you is an estimate of
how well this message was designed.

The message was designed to give students information about the lecture
which would lead them to go for a variety of reasons. Among other muons,
we thought that students might attend the lecture because they thought

it would be pleasant to attend

it would be rewarding to attend.

they ougnt to (or had a duty to) attend.
their families would expect them to attend.
their friends would expect them to attend.

fineir best professors would expect them to attend.

As you are reading the message, think about these reasons for attending
the lecture. What we want to know is how much emphasis the message you are
reading puts on each of these reasons. Note - these reasons themselves may
not be stated in the message, but information in fine message may sgpport or
be relevant to these reasons. 0r there may be no information relevant to
some offthe reasons.

 

Read the message through carefully, once only. Do not turn back to look
at the reasons listed above, and do not try to count the pieces of infomatim
supporting each point. What we want is your general impression of the rela-
tive emphasis placed on each reason in this message.

Please answer all the questions below - if you skip any one questian, the
whole questionnaire is invalid.

 

 

86

PLEASE READ THE FOLLOWING MESSAGE, AND THEN ANSWER THE QUESTION AT THE BOTTOM
OF THE PAGE.

Both reason and research indicate that family stability is important to
the happiness and success of every person. Family stability in turn depends
upon the communication relationships which exist among the family menters.

The Family Communication Lecture will provide you with a working know-
ledge of the communication principles involved in establishing and maintaining
healthy communication relationships. This should be of valm to anyone who is for
interested in a useful and enjoyable learning experience. A

Professor Cushman's enthusiasm and competence have earned him outstanding
teacher awards from three universities, and have led to several invitations to
Speak before family counseling groups. Over the years, many people have bene—
fited from his broad experience in the problem of interpersonal communication.

_ We G‘JAIWNL ——.-u-_ a .

Those of you who know Professor .Cushman know that his lectures are any-
thing but boring! We are sure that you will find him entertaining, stimulating 5--
and challenging. His lecture will stress the practical side of interpersonal
communication. You will find much that is useful in what he has to say.

 

Professor Cushman has developed a new approach to interpersonal comics-
tion which has important implications for family communication. Scientific
findings based on this approach can help you develop effective and enjoyable
relationships with others in your family.

Scientific research has rapidly advanced our knowledge of family communi-
cation, as researchers have accumulated a body of practical information about
how to improve interpersonal relationships in the family.

Studies have found that training can improve your awareness of your own
and others' communication behavior. Even a brief lecture can help you to
understand: 1) your own "style", and how this affects and is affected by oth-
er family members; 2) how relationships develop, and how the relationship de-
pends upon the peOple in it; and 3) how conflicts arise at the relationship
level, and what can be done to cope with them.

In sum, there will be very few opportunities in your college career to
learn a set of principles which are relevant every day of your life. If you
are interested in the principles involved in establishing and maintaining
healthy communication relationships, and if you want to have a useful and en—
fioyable learning experience, then be sure to attend Professor Cnshman's lec-
cure.

87

PLEASE no nor TURN BACK TO m: nmssass vans Msvsants THESE ths'rrows.

Listed below are the six reasons, mentioned on the first page of instructions,
for attending the family communication lecture. In the bi s proVided, please
rank-order the reasons as suggested by the message - puts 'beside the rea-
son most strongly suggested or supported, a '2' beside the next most strongly
suggested reason, etc.

 

J ‘z‘.’
Remember -'we want to see how much information in the messagg_is relevant to
each reason, not how important or persuasive y__ think each reason is. But
don' t forget that the message may suggest a reason to you without actually
stating that reason.

Reasons:

1. Attending the lecture would be pleasant.

2. Attending the lecture would be rewarding.

3. I ought to (have a duty to) attend the lecture.

 

u. My friends would expect me to attend.

Illllg
‘2’:

5. My family would expect me to attend.
6. My best professors would expect me to attend.

 

7. Have you ever attended a lecture on family communication?

Yes No

 

 

8. Have you ever taken a course on family communication?

Yes No

 

 

9. Have you ever taken a course in which family communicatlou.lll fur-I11!
discussed in more than three lectures or class meetings?

Yes No

 

 

10. Have you ever seen the message you just read, or a similar me about
the same lecture, before today?

Yes No

 

88

Now’we would like 2335 opinions about reasons for attending a.family coauumication
lecture. Please rank-order the following reasons, according to how Wt 31
think they are.

Reasons: Rank:

11. Attending the lecture would be pleasant.

l2. Attending the lecture would be rewarding.

13. I ought to (have a duty to) attend the lecture.
in. My friends would expect me to attend.

15. My family would expect me to attend.

 

16. 'My'best professors would expect me to attend.

 

17. How important do you think it is to learn about effective family communica-
tion in college?

'_____fVery important
Important

Neither important nor unimportant
Unimportant

Very unimportant.

Finally, we would like you to estimate the amount of information in the message
which ggggested, supported, or was relevant to the six reasons listed above.
Please try to compare the different reasons when you answer - if very much in-
formation was relevant to one reason, the message couldn't contain very much
information suggesting other reasons.

How much information suggested each of the following reasons?
18. Attending the lecture would be pleasant.

All the information in the message

A very large amount of information
A fairly large amount of information
A moderate amount of information

HH

A fairly small amount of information
A very small amount of information

.None of the informa ion

 

19.

20.

21.

89

Attending the lecture would be '35.

All the information in the message
A very large amount of it

A fairly large amount

A moderate amount

A fairly small amount
A very small amount
None of the information

I ought to (or have a gg£y_to) attend.

All the information in the message
A very large amount of it

A fairly large amount

A moderate amount

A fairly small amount

A very small amount

None of the information

 

My friends would expect me to attend.

All the information in the message
A very large amount of it

A fairly large amount

A moderate amount

A fairly small amount

A very small amount

None of the information

HHII

 

My family would expect me to attend.

All the information in the message
A very large amount of it
A fairly large amount

A moderate amount

___~_ A fairly small amount

A very small amount
None of the information

Tu 1.9.

-o-MKZ'. m.

.LI-Lv: .n

 

W ffr , r4 .

23.

2%.

90
Hy beat professors would expect me to attend.

uAll the information in the message
A very large amount of it

A fairly large amount

A moderate amount

A fairly small amount

A very ems}! amount

None of the information

 

How mudh of the information in the message was irrelevant to the six.roesons
listed above?

All the information

A very large amount of it

A fairly large amount

A moderate amount

A fairly small amount

A very anvil trarnt

HHIH

A f’ ..'. .’_.-'_“,‘,....‘3A
“-213 Q). ~ .ee --...‘.-:-L.:lui-b 5&‘en

 

Ajzen
1971

Ajzen and Fishbein
1969

Ajzen and Fishbein
1973

Dulany
1961

Dulany
1962

Dulany
1963

BIBLIOGRAPHY

91

Ajzen, Isek, "Attitudinal versus
Normative Messages: An Investi-
gation of the Differential Effects
of Persuasive Communications on
Behavior," Sociometry, 34:263-80.

’4‘ a.

 

and M. Fishbein, "The
of Behavioral Inten-
Choice Situation,"
Journal of Experimental Social
Psychology, 5:400-416; reprinted
in Attitudes and Behavior, ed a~+
Kerry Thomas, Penguin Books,

London, 1971, pp. 251-270.

Ajzen, I.,
Prediction
tions in a

 

 

 

Ajzen, Icek, and Martin Fishbein,
"Attitudinal and Normative Vari-
ables as Predictors of Specific
Behaviors," Journal of Person-
ality_and Social Psychology,
27:41-57.

 

 

Dulany, Don E., "Hypotheses and
and Habits in Verbal Operant
Conditioning'," Journal of
Abnormal and Social Psychology,
63:251-263.

 

 

. "The Place of Hypo-
theses and Intentions: An
Analysis of Verbal Control in
Berbal Conditioning," in C.
Eriksen, ed, Behavior and
Awareness, Durham, Duke Univ.
Press, pp. 102-129.

 

 

. "Does Partial Rein-
forcement Dissociate Verbal
Rules and the Behavior They
Might Be Presumed to Control?"
Journal of Verbal Learning and
Verbal Behavior, 2:361-372.

 

 

10.

ll.

12.

13.

14.

Dulany
1968

Fishbein
1967

Harris

1962

Mortensen and Sereno
1973

Schwartz and Tessler
1973

Seibold
1974

Werts and Linn
1970

Woelfel
1973

92

. "Awareness, Rules, and
Propositional Control: A Con-
frontation with SR Behavior
Theory," in D. Horton and T. Dixon
(eds.) Verbal Behavior and SR
Behavior Theory, Prentice-Hall.

 

 

Fishbein, Martin. "Attitude and
the Prediction of Behavior," in
M. Fishbein (ed.) Readings in

Attitude Theory and Measurement, as
New York: Wiley, pp. ‘“

 

 

Harris, Chester W. (ed), Problems
in Measuring Change, Univ. of
Wisconsin Press, Madison.

 

Mortensen, C. David, and Kenneth
Sereno (eds) Advances in Communi-
cation Research, New York, Harper ”r"
& Row.

 

 

 

Schwartz, Shalom H., and Richard
C. Tessler, "Test of a Model for
Reducing Measured Attitude-Behav-
ior Discrepancies," Journal of
Personality and Social Psychology,
24:225-236.

 

 

Seibold, David R. "The Attitude-
Verbal Report-Overt Behavior
Relationship in Communication
Research: A Critique and Theoretic
Reformulation," Paper presented

at the Annual Convention of the
Association for Education in
Journalism, Aug. 18-20, 1974.

Werts, Charles E., and Robert L.
Linn, "A General Linear Model

for Studying Growth," Psychological
Bulletin, 73:17-22.

Woelfel, Joseph. Sociology and
Science.

HICHIGQN STQTE UNIV. LIBRRRIES

31293006491009