\
_ “——

PURPOSEFUL AMBIGUITY AS
A PERSUASIVE MESSAGE STRATEGY

Thesis for the Degree of Ph. D.
MICHIGAN STATE UNIVERSITY
GARY B. WILSON
1971

|1lHlHl|HHIIHHHHHHHllllltllfll“\lllmtllllml

3 1293 10604 042

 

This is to certify that the

thesis entitled

PURPOSEFUL AMBIGUITY AS A PERSUASIVE
MESSAGE STRATEGY

presented by

Gary B. Wilson

has been accepted towards fulfillment
of the requirements for

 

 

 

 

 

 
   

   

 

Ph . D . degree in Conununication
l
V A
.. \
Datei’flé
0-7639
__ 1...___
I
_._J " amomc er 7“ i

 

 

“DAB 8: SONS
BUUK BMW" M

LIBRARY emozns
”musing. mall“;

      

ABSTRACT

PURPOSEFUL AMBIGUITY AS A PERSUASIVE
MESSAGE STRATEGY

BY

Gary B. Wilson

This study investigated the relationship between
the level of purposeful ambiguity and: (l) assimilation
of perceived source position; (2) source credibility;

(3) attitude change.

In this study purposeful ambiguity was defined
as the use of high levels of abstraction with connotatively
positive words.

Prior research indicated that receivers interpret
ambiguous messages to support prior expectations or
desires. It was hypothesized that with relatively high
levels of purposeful ambiguity the message would be
assimilated toward the receiver's own position. It was
also predicted that high purposeful ambiguity would lead
to higher ratings of source credibility. Finally, it
was predicted that receivers would show more favorable

attitude change with a message high in purposeful ambiguity.

Gary B. Wilson

Message elements were manipulated to achieve
three levels of abstraction. The subjects first indi—
cated their attitude toward open housing. They then
read one message and completed the posttest question-
naire. Subjects indicated the perceived position of
the source toward open housing, rated the credibility
of the source, rated the ambiguity of the message, and
indicated their own posttest attitude toward open
housing.

The results did not support the hypotheses. In
the analysis of the data it was found that the moderate
level of purposeful ambiguity had a greater tendency to
support the predicted differences than did the high
purposeful ambiguity message.

Two problems were evident. First, it was felt
that the abstraction index was an inadequate instrument
for this type of manipulation. While it indicated great
differences in the abstraction levels of the messages,
the respondents in a pretest found only small but con-
sistent differences. Ratings obtained in the main study
indicated no differences. A second problem was that to
adequately test the hypotheses a heterogeneous sample
was necessary. The final sample was quite uniformly

positive toward Open housing.

PURPOSEFUL AMBIGUITY AS A PERSUASIVE

MESSAGE STRATEGY

BY

LLLL ,
Gary Bu Wilson

A THESIS

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY
Department of Communication

1971

Accepted by the faculty of the Department of
Communication, Michigan State University, in partial ful-
fillment of the requirements for the Doctor of Philosophy
Degree.

Xfli/mfl 6? Wfi/

Director of Thesis

Guidance Committee:fififégaagédllglfiZQZézggy' , Chairman

-_ d AQIAAA,’ . /
‘ Q

 

ACKNOWLEDGMENTS

I would like to express my sincerest thanks and
gratitude to Professor Gerald R. Miller who served as my
Academic Advisor as well as Director of this thesis.
Throughout my graduate program he served as advisor,
counselor, and friend. Without his support I would not
have survived the many crises of my graduate career. I
can only hope that my future work reflects the quality
of scholarship and dedication that he exemplifies and
attempts to instill in his students.

My appreciation also to Professors Eugene
Jacobson, David Ralph, and Lawrence Sarbaugh. Their
advice and counsel in the preparation of this thesis
was invaluable. I also profitted greatly from my
exposure to them both as teachers and friends through-
out my time at Michigan State University.

Lucy Tilton typed the many preliminary drafts
of this thesis in addition to her regular duties as
secretary in the Speech Department at the University
of Connecticut. Without her cheerful willingness to

give extra time and effort the completion of the thesis

ii

would surely have been greatly delayed. My sincerest
thanks, Lucy.

Finally, to my wife, thanks and congratulations.
In reality the degree is at least half hers. Without
her encouragement and faith I would never have completed
my program. And to my two sons my thanks for just
being there when I needed reminding that the world was

a fun place to be.

iii

TABLE OF CONTENTS

Chapter
I. THE PROBLEM . . . . . . . . .

Definition of Purposeful Ambiguity.
Literature and Hypotheses. . . .

I I 0 METHOD . O O O O O O O O O 0

Topic . . . . . . . . . .
Message Development. . . . . .
main Study 0 O O O O I O I 0

Subjects. . . . .
Study Design . . .
General Procedures .
Measuring Instruments

III. RESULTS . . . . . . . . . .

Pretest of Message Manipulation. .
Pretest Equivalency of Experimental
Groups . . . . . . . .
Perceived Similarity of Source's and
Subject's Position . . . . .
Ambiguity and Source Credibility .
Ambiguity and Attitude Change . .
Rated Ambiguity of the Messages. .

IV. DISCUSSION . . . . . . . . .

Range of Initial Scores on Attitude
Toward Open Housing . . . . .

Hypothesis 1 . . . . . . .
Hypothesis 2, 3, and 4 . . . .
Hypothesis 5 . . . . . . .

Implications for Further Research .

iv

Page

24
24
27

27
27

30
33
33
34
34
39
43
43

45

Chapter Page

BIBLIOGRAPHY . . . . . . . . . . . . 57
APPENDICES
Appendix
A. Abstraction Index . . . . . . . . 60
B. Experimental Messages. . . . . . . 62
Message I . . . . . . . . . . 62
Message II. . . . . . . . . . 64
Message III . . . . . . . . . 65
C. Measuring Instruments. . . . . . . 66

12.
13.
14.

LIST OF TABLES

Anova of Subjects' Initial Position on
Open Housing . . . . . . . . . .

Perceived Similarity Between Source and
Subject Attitude Toward Open Housing . .

Correlation Between S's Pretest Score and
Perceived Position of the Source . . .

Perceived Similarity Between Source's and
Subject's Position on Open Housing. . .

Perceived Position of the Source on Open
Housing. . . . . . . . . . . .

Perceived Qualification of the Message
Source . . . . . . . . . . . .

Analysis of Variance of Perceived Qualifi-
cation of the Source . . . . . . .

Perceived Safety of the Message Source . .
Anova of Perceived Safety of Message Source
Perceived Dynamism of the Message Source .

Anova of Perceived Dynamism of the Message
Source . . . . . . . . . . . .

Attitude Change Scores . . . . . . .
Anova of Rated Ambiguity of the Message. .

Subjects' Pretest Attitudes and Perceived
Position of the Source. . . . . . .

vi

Page

35

36

37

38

39

40

40
41
41

42

43

44

50

CHAPTER I

THE PROBLEM

We communicate to influence--to affect with
intent. In analyzing communication, in trying
to improve our own communication ability, the
first question we need ask is, what did the
communicator intend to have happen as a result
of his message? What was he trying to accomplish,
in terms of influencing his environment? As a
result of his communication, what did he want
people to believe, to be able to do, to say? In
psychological terms, what response was he trying
to obtain? (Berlo, 1960, p. 12)

Considerable research has indicated that highly
credible sources are more effective in influencing
receivers than are sources with lower credibility.
McGuire (1969) has provided an adequate review of find-
ings concerning source credibility. The results of previ—
ous studies indicate that any source who hopes to be an
effective communicator should seek to establish high
credibility with his receivers. Even so, little research
has examined the effects of message variables on subse-
quent perceptions of source credibility. One common
technique of studying the effects of source credibility
has been to use a standard message and to vary the

attributes of the source to whom the message is attributed.

A second common technique has been to use the names of

l

persons known to differ in initial credibility as sources
of a common message. These techniques have been useful
in establishing the importance of source credibility in
communication.

In many communication situations the introduction
of the source does not afford the opportunity for immedi-
ate establishment of very high credibility. Few com-
municators are nuclear physicists, Presidents, famous
political figures, or national experts in particular
fields of concern. If these attributes are lacking,
communicators must establish high credibility by means of
the message itself.

Using the terminology of Miller (1966) few
speakers are perceived to possess such high exogenous
sources of credibility that they need not concern them-
selves with the endogenous sources of credibility.

Miller defines the endogenous sources of credibility as
variables associated with the communication act itself.
Exogenous sources of credibility are attributes that the
source brings with him to the communication situation
(Miller, 1966, p. 36). Examples of exogenous variables
cited by Miller are education, occupation, physical
attraction, and moral characteristics. Endogenous vari—
ables are such things as use of grammar, word selection,
and fluency of the speaker. The review of source credi-

bility research by McGuire (1969) indicates that studies

have focused upon the exogenous sources of credi-
bility.

Establishing credibility endogenously can most
easily be done by presenting material that supports the
prior beliefs of the receiver (see, e.g., Byrne, 1961).
That technique should work well given either a single
receiver or a homogeneous group of receivers whose
beliefs and attitudes are known by the source. If the
source does not have sufficient prior knowledge about
his receivers, it is difficult to use this message
strategy effectively. If the source is faced with
receivers who do not have homogeneous beliefs or atti-
tudes, the strategy is not available to him in the
first place.

The present research examined a potentially
useful message strategy for situations in which either
the source has little information about the attributes
of the audience or he is aware that the receivers have
heterogeneous attitudes toward the message issue. The

message strategy studied is labeled purposeful ambiguity.

 

Definition of Purposeful Ambiguity

 

For a symbol or group of symbols to be labeled
purposefully ambiguous, they must meet two general cri-
teria: First, there must be disagreement on the denota—
tive meanings of the symbols used. Coupled with this

is the requirement that individual receivers must have

denotative meanings for the symbols used. The first
requirement is the criterion for ambiguity. The second
requirement provides the means of separating ambiguity
from a more general vagueness of symbols. Given vagueness,
the receiver is not certain he knows the referents of the
symbols; given ambiguity, the receiver feels certain he
knows the referent intended by the source. Ambiguity

can only be discovered by comparing referents across a
group of receivers.

In order to be used purposefully in obtaining a
desired response, symbols must meet a second general
criterion. If symbols are to be purposefully ambiguous,
they must have high connotative uniformity among re-
ceivers. To employ purposeful ambiguity, the source
must not only be aware that there will be denotative
disagreement among receivers, he must also be certain
that the symbols used have a high degree of connotative
uniformity among receivers.

If denotative and connotative meaning are viewed
as dichotomies--either agreement or disagreement between
individuals--the requirements for a statement to be
purposefully ambiguous are clearer. This paradigm is
adapted from Osgood (1961, p. 102). The four possible
situations are:

A. Denotative agreement----Connotative agreement

B. Denotative agreement----Connotative disagreement

C. Denotative disagreement--Connotative agreement
D. Denotative disagreement--Connotative disagreement

When source and receiver agree on both denotative
and connotative level (A above), the source has his best
chance of achieving his communicative goals. There is
minimum ambiguity in the situation. Both source and
receiver agree on the referents of the symbols used, and
both have similar feelings toward those symbols. An
example of situation A occurs when two people are dis-
cussing where to meet for dinner. Person 8 suggests a
local steakhouse where he has enjoyed eating in the
past. Person R recognizes the name of the steakhouse
and has also been favorably impressed by the meals he
has had there. In this situation, with agreement on
both the denotative and connotative levels, it is quite
likely that the suggestion of person S will result in
acceptance on the part of person R.

Situation B (above) suggests that two communi-
cation participants agree on the denotative level but
have differing connotative reactions to the referents
of the symbols used, or to those symbols. If, in the
example cited above, person R recognized the name of
the suggested steakhouse but had received poor service
in the past his connotative reaction to the name of the
steakhouse would vary from that of person S. There is

agreement on the referent (the steakhouse) but the two

individuals have differing connotative reactions. In
this situation there is less likelihood of person S
achieving his goal, i.e., eating at his favorite steak-
house.

Situation D probably represents a total loss so
far as meaningful communication is concerned. Here the
communication participants disagree on both the refer-
ents of the symbols used and on their connotative
reactions to the symbols or their referents.

In situation C, there are two possible reasons
for denotative disagreement. The first involves the
use of different labels for the same object. A second
possibility involves the use of the same label (e.g.,
democracy), with differing referents for the label and
with similar feelings for the "object" (both agree that
democracy is "good"). It is this last situation that
allows for the use of what has here been termed purpose-

ful ambiguity.

Literature andeypotheses
The general notions of balance theory suggest
that purposeful ambiguity could be a useful message
strategy for gaining message acceptance or improving
source credibility. Cognitive consistency models assume
that the individual desires to maintain consistency
among relevant cognitions, that imbalance or inconsis-

tency leads to psychological discomfort, and that

psychological discomfort in turn motivates the individual
to regain or restore consistency. Kelman and Baron
(1968) suggest that inconsistency reduction can take
one of two general forms: First, "the individual may
avoid the implication of the inconsistency by perceiving
or interpreting the discrepant element in such a way
that it no longer appears to be inconsistent with the
potentially challenged element" (pp. 670-671). Second,
the individual may confront the inconsistency and change
either the attitude or the behavior involved (p. 671).
In terms of the use of purposeful ambiguity,
it is likely that the first of the two techniques of
inconsistency reduction would be used. If an individual
encounters a message concerning a tepic of interest,
he will tend to read the message and form a cognitive
element from it. If the message offers many connota-
tive anchors, the connotative anchors will influence the
placement of the message in the cognitive structure of
the individual. A message using positively valued
abstract words should allow that process to occur with

relative consistency. The process could be pictured thus:

 

S R

+
Figure 1 Figure 2

Figures 1 and 2 are adapted from Abelson's (1967) dis-
cussion of transcedence (pp. 353). Figure 1 represents
the inferred state of the system immediately upon the
receiver's receipt of the message. The link between

the source (S) and the connotation of the message (Me)
is positive because the source produced the message.

The link between the source and the denotative meaning
of the message (Md) is positive for the same reason.

The link between the denotative meaning of the message
and the connotative meaning of the message is positive
because of the identity of the symbols used to convey
the two forms of meaning. The link between the receiver
(R) and the connotative meaning of the message is
inferred to be positive because of the positively valued
abstractions used in the message. The receiver needs

to adjust the structure to supply the links between him-
self and the denotative meaning of the message and
between himself and the source of the message.

The relationship between the receiver's denota-
tive and connotative meanings of the message should lead
to a solution that is similar to transcendence as dis-
cussed by Abelson (1967). The receiver should be unable
to psychologically tolerate the inconsistency between
valuing something positively and recognizing that it has
little meaning. He should tend to assume a meaning that

he views as positive and thereby balance the structure.

By viewing the message as a whole, encompassing both

the denotative and the connotative meaning, he forms a
positively valued single unit from what was previously
a positively valued portion and an ambivalent portion.
This solution implies that if the receiver were then
asked to supply the denotative meanings for the message,
he would not define the terms in evaluatively neutral
referents but rather would seek to define them with
referents that were positively valued by him. Finally,
with the message now positively valued, the link between
the source and the receiver should also be established
as positive to achieve balance.

Skinner (1957) has also emphasized the idea that
ambiguity could be useful. He states that ambiguity
" . . . should increase the chances for a successful
match between the reader and the literary work." He
further asserts that while ambiguity should make the
work more universally acceptable, it should also make
the work less likely to be any particular reader's
faVOrite book (p. 275).

In 1962, James Roever authored a paper entitled,
"Understanding Misunderstanding: Toward a Theory of
Purposeful Ambiguity." In that paper Roever forecast
great utility for the strategy of purposeful ambiguity.
He states that "through the use of purposeful ambiguity,

we can satisfy various needs of various individuals

10

when they are in the same audience" (Roever, 1962, p. 4).
Roever asserts that purposeful ambiguity may be estab-
lished through the use of "high levels of abstraction
and situations whereLwe have 223 differentiated between
similarities and differences" (Roever, 1962, p. 5). He
suggests that the choices of meaning of the communication
be left entirely up to the audience.

At a recent communication symposium, S. N.
Eisenstadt cited a problem that was solved through the
use of ambiguity. He described the solution reached
when a symbol selection problem threatened to prevent
the completion of the Israeli declaration of independence.
The orthodox religious group insisted that the term "God"
should be used in the document while the nonorthodox
group insisted that it should not be written into the
declaration. The solution reached was the use of the
purposefully ambiguous Biblical expression "the Rock of
Israel." For the orthodox group, it meant God, for the
nonorthodox, it meant the denstiny of Israel or the iden-
tity of the country (Thayer, 1967, p. 475). Thus, a con-
flict was clearly defined, and for a time it appeared
that neither side would allow the other to prevail. But
through the use of compromise wording the two groups
found language symbols they could both agree were "good."
They disagreed on what the symbols denoted, but both felt

they had accomplished their goals concerning the inclusion

11

or exclusion of the word "God" from the declaration of
independence. The intentional use of an ambiguous ex-
pression solved a communication problem that seemingly could
not be solved through more concise use of symbols. As
George Gerbner has stated: "It is a political necessity
to engage in calculated ambiguity in the use of symbols
so that different constituent groups may derive different
types of gratification from the symbols . . . " (Thayer
1967, p. 477).

Fearing (1953) discusses the effects of ambiguous
content upon receivers. He defines ambiguity in terms
of the variety of possible structurizations possible by
the interpreters and states that ambiguity "is concerned
with the properties of communications content which make
it susceptible to varient structurizations by interpret-
ers" (Fearing, 1953, p. 82). Fearing further asserts
that the "important variables are in the content (includ-
ing context). These include structural simplicity or
complexity, amount of detail, etc. . . . The familiarity
with the symbols used in particular content, and the
degree to which they have common significations for com-
municators and interpreters are, of course, fundamentally
important variables" (p. 83). The preceding statement
fits well in the paradigm adapted from Osgood and pre-

sented previously.

12

Fearing also discusses involvement with the
tOpic in terms of receiver set, which is defined in
terms of the receiver's having "specific and persisting
goal integrations, strong value orientations and stereo-
types, specific prior experience in or involvement with
particular content . . . " (Fearing, 1953, p. 83). Fear-
ing foresees interaction between receiver set and ambiguity
of the content. (Specifically, he predicts that with
highly ambiguous content and strong receiver set, the
receiver will give structure to the content--structure
that will be in the direction of the receiver's set:>
With high ambiguity and low receiver set, the receiver
should seek greater structure or simply be indifferent
to the message. With low ambiguity and high receiver
set, Fearing predicts two possible outcomes: If the
material supports the beliefs of the receiver, it will
be accepted; if the material conflicts with the beliefs
of the receiver, inconsistency or imbalance will result.
Such conflict will be resolved by evasion, rejection of
the materials, or by leaving the field entirely. With
low ambiguity and low receiver set, the receiver may
accept the content but no conflict will result. Conflict
will be precluded by the receiver's low involvement
(Fearing, 1953, p. 85).

Fearing's paper stresses the differential effects

that the receiver's involvement with the topic may have

13

on his reception of the message. However, he makes
little attempt to Operationalize the concept of ambi-
guity. The definition cited above (p. 12) suggests that
the researcher might pretest the material by using
some open-ended technique or some predetermined category
system to which the receiver of the content reacts. If,
however, the subject has a strong set, Fearing suggests,
it would prevent variations in the interpretation of the
content. Comparisons across subjects might rectify this
problem, if subjects have various orientations previous
to being exposed to the content. These problems are not
adequately discussed by Fearing.

Based on the earlier work of Flesch, Paul J. Gillie
(1957) developed a simplified index of the abstraction
level of a message. The index was used in a study by
Haskins (1960), who found that the proportion of readers
who rated a magazine article as excellent was correlated
.80 with the Abstraction Index score of that article.
He invokes amount of effort involved in reading the
article to explain his findings. "Presumably, more satis-
faction is derived from the completion of a tough task
than an easy one. Equating abstractness with toughness,
we expect more satisfaction in the reading of an abstract

item than in a concrete item" (Haskins, 1960, p. 104).

14

The same results would be predicted from the
theoretic framework presented earlier in this chapter.
According to the rationale presented by Roever, abstract
messages should allow readers to infer their own mean-
ings for the abstract words. According to balance
theories, these meanings should support the receiver's
own beliefs. The entire process leads to the prediction
that abstract messages should be rated more highly than
less abstract messages--the results obtained by Haskins.

Bousfield (1961) reported in a study in which
Turkish words were paired with their purported English
meanings. The subjects learned the words as pairs.
After allowing one week for forgetting, subjects were
given the Turkish word and asked to recall the purported
translation. They also rated the Turkish word on the
good-bad semantic differential scale. Bousfield reports
that:

. . . it was found that the subjects tended to
retain the connotative meaningful responses
assumed to have been acquired by the Turkish
words when they were unable to recall the sup-
posed translations. The present explanation of
this type of recall is that, even though the sub—
jects were unable to recall the supposed trans-
lations, a sufficient number of acquired implicitly
produced meaningful responses remained to enable
them to give appropriate ratings (Bousfield, 1961,
While the analogy is presently tenuous, it seems

possible that the preceding outcome may also occur even

when there is no language change. A precise definition

15

of the word democracy_in the English language is not

 

simple. When the existing institutions in this country
must also be included in the definition, it becomes

even more complex and has few referents for the average
schoolboy who is forced to learn the book definition.
However, in this culture the connotative "definition" of
the word democracy is "very good." Perhaps for abstract
words, the results found by Bousfield are directly com-
parable; that is, the person may forget the exact deno-
tative meaning of the word but retain the connotative
meaningfulness.

Sherif, Sherif, and Nebergall (1965) report a
study completed during the 1960 election campaign. Their
study is of special interest here because it demonstrated
7 the displacement of a communication that was reported
to be completely neutral or unidentifiable in terms of
the political party it supported. Moreover, it was
similar to the purposefully ambiguous communication of
concern in this study.

Sherif, Sherif, and Nebergall found that the
unidentifiable communication was assimilated toward
the receiver's position, i.e., each receiver predicted
that the author would vote for the candidate favored
by the receiver (Sherif, Sherif, and Nebergall, 1965,

“p. 161). The other communications were all tied to

support for one of the major political parties and were

16

therefore less susceptible to assimilation by those of
opposing views. Finally, there was a tendency for those
rated as less highly involved to assimilate the state-
ment that mildly favored the opposing political party
(Sherif, Sherif, and Nebergall, 1965, pp. 158-159).

In contrast, statements that were strongly in
favor of one party or the other showed no tendency for
assimilation. Both supporters of the Republicans and
the Democrats rated the extreme statements in a similar
manner. The extremely Democratic statement was rated
accurately by both the Republicans and the Democrats.
Thus, the very clear or unambiguous statements were not
susceptible to varying interpretations by various re-
ceivers, instead, these statements were acceptable only
to those holding similar views.

An early study in psychological distortion of
judgments was conducted by Campbell, Hunt, and Lewis
(1957). Psychology students judged the degree of
imbalance of bogus psychotics on the basis of various
messages supposedly composed by those individuals. In
general, the messages designed to portray very unbalanced
individuals were correctly interpreted. However, when
the message was near the center of the stimulus range
(very unbalanced to very stable), there were distortions.
Specifically, the researchers found a tendency for assimi-

lation in the direction of the expectations of the person

17

doing the judging. Those who expected the author of

the message to be highly disturbed tended to judge the
author more disturbed than those who expected the author
of the message to be only mildly disturbed. Campbell,
Hunt, and Lewis conclude that when the judgment task
results in ambiguity or confusion, there will be a
tendency for the judge to assimilate the judgments in
the direction of his own eXpectations (Campbell, Hunt,
and Lewis, 1957, p. 355).

Dillehay (1965) used a single message that was
moderately in favor of adding flouride to water. The
experimental groups were composed of nurses and other
women from the community. Subject attitudes ran from
highly favorable to mildly negative toward the use of
flouride. Dillehay found that all groups showed a
tendency to assimilate the message; i.e., all groups
tended to perceive the message as being closer to their
own view than its placement had previously indicated.
Since the message was not included in the report, it
is not possible, at this time, to judge its ambiguity.
However, if the message was only moderately in favor of
fluoridation, it could conceivably be considered a
slightly ambiguous message.

A study by Kelman and Eagley (1965) investigated
the effects of source credibility upon perception of the

message and the source's position on the issue. Given

18

a positive source, there was a tendency toward assimi-
lation--subjects perceived the message as favoring their
own views and placed the source's attitude in a position
similar to their own. Given a negative source, there
was no apparent tendency toward assimilation, nor was
there a tendency toward contrast. Kelman and Eagley
concluded:

Displacement (assimilation) and attitude change

can be viewed as alternative ways of achieving

the congruity between source and message that

is so important when source orientation prevails.

One can assume that displacement is the most

likely mechanism to be used when the message is

highly ambiguous and thus easily lends itself to

different interpretations (p. 76).

The preceding discussion would lead to the following
hypothesis:

Hla. Given a message high in purposeful ambiguity
and a receiver group with heterogeneous attitudes toward
the topic, the receivers will rate the source's position
significantly more similar to their own position than
they will with a message low in purposeful ambiguity.

Given that the subjects will rate the source's
position more similar to their own if the message is
high in purposful ambiguity than if it is low in pur-
poseful ambiguity, it would follow that the variability
of the subjects' ratings of the source's position would
also be effected. With high variability in the subjects'
initial ratings of their own position, the act of rating
the source's position as similar to their own would

lead to high variability in the rating of the source's

position. With low purposeful ambiguity the subjects'

19

initial ratings of their own position should have rela-
tively less effect. Therefore, the following effect is
hypothesized:

Hlb. Given a message high in purposeful ambiguity
and a receiver group with heterogeneous attitudes toward
the topic, the receivers will show significantly greater
variability in their ratings of the source's position
than they will with a message low in purposeful ambi-
guity.

Manis (1961a) studied the effects of ambiguity
and receiver position on message interpretation. The
topic used was fraternities, and the receivers' views
toward fraternities varied from positive to negative.
Ambiguity was introduced into the messages by deleting
every other word. Three messages were used: One
favoring fraternities, one neutral, and one Opposing
fraternities. These three messages were counter-
balanced to form eighteen treatment groups.

Ambiguity, as manipulated by Manis, served to
increase the uncertainty of the message position and
resulted in the ambiguous messages regressing toward
the neutral point on the rating scales. The important
dependent measure, as far as the present research is
concerned, was the subjects' impression of the position
of the source of the message on fraternities. The
author's conclusions were:

1. When responding to messages that they essen-

tially agreed with, subjects tended to dis-

place the communicators toward their own
position.

20

2. When responding to messages that deviated
moderately from their own stands, there was

a consistent tendency for subjects to dis-

place messages toward the point on the scale

that they themselves favored; this trend was
shown by the pro and the anti groups in
response to the neutral messages, and by the
neutral group in response to both the pro

and the anti-fraternity messages.

3. When the subjects responded to statements that
were definitely opposed to their own stands,

no consistent displacement trends were

obtained . . . (Manis, 1961a, pp. 80-81).

The second of these three conclusions is of
greatest interest here. A neutral message, rated by
itself, showed a trend toward assimilation by both
those favoring and those opposed to fraternities. In
each case, the readers of the message felt that the
author of the statement was basically in their camp.
Those neutral toward fraternities felt that each author
was more neutral than did the other two groups. It is
the measurement of the author's perceived stand that
allows the type of ambiguity of concern in this study
to best be detected. A neutral message best typifies
the strategy of purposeful ambiguity for it offers the
reader few clues as to the "true" position of the author
of the statement.

In a replication with variation, Manis (1961b)
eliminated the variable of ambiguity and used instead
high and low credible source manipulations. Again, the

dependent variable of concern was the receiver's esti-

mation of the source's attitude toward fraternities.

21

Manis found that with a low credible sourse, there was
no consistent trend toward either assimilation or con-
trast by any of the treatment groups. With a high
credible source, the results were quite different:
The subjects tended to show a general assimilation effect.
All treatment groups demonstrated a trend to displace
the estimation of the source's attitude in the direction
of the receiver's attitude toward fraternities (Manis,
1961b, p. 84). Thus, there seemed to be a tendency on
the part of the receivers to wish that the high credible
source would display attitudes similar to their own.
Finally, Byrne (1961) reported a study which

demonstrated that similarity of attitudes leads to
greater acceptability of the other person. This sug-
gests that if the source of a message is perceived as
supporting views similar to one's own, then the source
will be rated higher in perceived credibility than
would an unknown source. Using this study in combination
with those previously reported, it would seem that
proper use of ambiguity, built around the use of conno—
tatively positive words, would be a viable means of
enhancing the source's credibility. Thus the following
hypotheses are proposed:

H2a. Given a message high in purposeful ambiguity
and a receiver group with heterogeneous attitudes toward
the topic, the receivers will show significantly less

variability in their ratings of the source's qualification
than they will with a message low in purposeful ambiguity.

22

H2b. Given a message high in purposeful ambiguity
and a receiver group with heterogeneous attitudes toward
the tOpic, the receivers will rate the source signifi-
cantly higher in qualification than they will with a
message low in purposeful ambiguity.

H3a. Given a message high in purposeful ambiguity
and a receiver group with heterogeneous attitudes toward
the topic, the receivers will show significantly less
variability in their ratings of the source's safety than
they will with a message low in purposeful ambiguity.

H3b. Given a message high in purposeful ambiguity
and a receiver group with heterogeneous attitudes toward
the topic, the receivers will rate the source signifi—
cantly higher in safety than they will with a message
low in purposeful ambiguity.

None of the research reviewed deals directly with
the receiver's perception of the source's dynamism and
how it might be affected by the use of purposeful ambi-
guity. Berlo, Lemert, and Mertz (1966) suggest that
the rating of the dynamism of the source, though statis-
tically independent "may not be psychologically inde-
pendent of the other two factors" (p. 21). Intuitively,
it would seem that if the receiver perceives a purpose-
ful ambiguous message to emanate from a safe and quali-
fied source, he would also attribute some amount of
vigor to that source. Though evidence is lacking at
the present time, the following hypotheses would seem
in order:

H4a. Given a message high in purposeful ambiguity
and a receiver group with heterogeneous attitudes toward
the topic, the receivers will show significantly less
variability in their ratings of the source's dynamism

than they will with a message low in purposeful ambi-
guity.

23

H4b. Given a message high in purposeful ambiguity
and a receiver group with heterogeneous attitudes toward
the tOpic, the receivers will rate the source signifi-
cantly higher in dynamism than they will with a message
low in purposeful ambiguity.

As indicated in the opening paragraph of this
chapter, there is ample evidence that a message from a
high credibility source leads to greater attitude change
than the same message from a low credible source. Because
it is felt that a message high in purposeful ambiguity
will lead to the receiver attributing greater credibility
to the source of the message than will a message of low
purposeful ambiguity, the message high in purposeful
ambiguity should lead to greater attitude change in the
direction advocated in the message. Formally stated:

H5. Given a receiver group with heterogeneous atti-
tudes toward the topic, a message high in purposeful
ambiguity will lead to significantly greater attitude

change in the direction advocated in that message than
will a message low in purposeful ambiguity.

CHAPTER II

METHOD

221.2

Four topics were used in a pilot study with
adults and students in the Lansing, Michigan area. The
four tOpics were: Open Housing, Rioting in the Cities,
Withdrawal from Vietnam, and Negotiations in Vietnam.
On the basis of the results of that pretest of the topics,
Open Housing was selected as the topic for the main study.
It was felt that it offered both a split in attitudes

and high interest value for adults.

Message Development

 

The original message was written by the experi-
menter to represent a position favorable to stronger
Open Housing laws. From the first message two additional
messages were deve10ped by manipulating the abstraction
level according to the indices developed by Gillie

(1957) and tested by Haskins (1960).* The primary

 

*
A copy of the index can be seen in Appendix A.

24

25

manipulations involved the use or nonuse of the article
the, the verb form tg'b§_or its derivatives, and nouns
with specified endings (e.g., -tion, -ship, -y).

The final sentence in the messages offers an
example of the manipulation of the article. The use of
the finite article the lowers the abstraction index
score. The final sentence in the least abstract message
was: "All of the citizens must have equal access to all
of the products of our society." The most abstract
version reads: "We must offer our citizens access to the
products of our society."

The use of finite verb forms lowers the abstrac-
tion index score while the use of nouns which end in -ttg§
raises the score. An example of that manipulation can
be found in the following sentences: "Prejudice toward
blacks is being handed down like some birthright";
"Prejudices are handed down like some birthright"; "Preju-
dices can be passed down from generation to generation
unaltered." The sentences are in order of increasing
abstraction.*

In each case an attempt was made to keep as much
commonality as possible while manipulating the sentence
elements required to change the abstraction index score.

The completed messages were ranked according to

perceived level of ambiguity by five adult judges. The

 

*
The complete messages will be found in Appendix B.

26

criterion of acceptability was that all five judges must
rank the messages in the same order and the order must
be identical with the abstraction index ranking.

The final test of the acceptability of the mes-
sages consisted of submitting them to ten undergraduate
and ten graduate students at the University of Connecti-
cut. Before reading the three messages, the students
read the following instructions from the front of the
experimental booklet:

Please read all three messages carefully. You
will find that they all deal with open housing.
However, there are differences in the manner in
which they express the author's position on the
issue. You are being asked to judge the clarity
with which they express the author's View.

Please indicate which message most explicitly
states the author's position and which message is
most ambiguous on that position. You are not to
judge which message you like best or least. Please
judge the messages only on their relative ambiguity
concerning the author's position.

At the bottom of each message you will find the
labels "most precise" and "most ambiguous." When
you decide which message most precisely expresses
the author's position on open housing circle the
label "most precise." When you decide which message
is most ambiguous concerning the author's position
on open housing circle the lable "most ambiguous."

The subjects were then given time to read and
rate the messages. When all the subjects were finished,
time was allowed for questions. They were thanked for

their cooperation and dismissed. The criterion for

acceptability was 80% ranking the message as predicted.

27

Messages meeting the criterion were considered adequate

for use in the main study.

Main Study

 

Subjects

S's were 77 adults recruited from a large work
group in a Hartford, Connecticut, insurance company and
50 students recruited from basic courses at the Univer—
sity of Connecticut. Of these, 66 adults and 46 students
returned usable data. The remaining S's failed to com-
plete the questionnaire.

The adult S's completed the materials during
free time in their work day. In all cases the materials
were returned before lunch on the day they were dis-
tributed. The distribution of the materials was handled
by a friend of the experimenter who worked at that com-
pany.

The 50 students participated during a regularly
scheduled class period. The classes were told that the
last 20 minutes of the period would be devoted to parti-
cipation in a communication study and volunteers were

requested.

 

Study Design
Three conditions were employed in this study.
They were: High, Moderate, and Low Purposeful Ambi-

guity. The experimental materials were arranged so that

28

every third subject in an existing group would be
assigned tO the Low Ambiguity condition, every third g
to the Moderate Ambiguity condition, and the same for

the High Ambiguity condition. This was done tO insure
that any extraneous variables operating in the intact
groups would be distributed among the three experimental
conditions. The S's were asked tO fill out some prelimi—
nary personal data On the second page Of the experimental
package. After they had completed that portion Of the
experiment they were asked tO read the experimental mes-
sage and complete the evaluation sheets following the
message. Included in the evaluation was a scale which
asked the S's tO indicate how closely the views in the
message coincided with their own views on Open housing
and a scale assessing their post-communication attitude

toward Open housing.

General Procedures

When the instructor Of a class was contacted he
was told the general nature Of the study; i.e., a study
Of political communication, and was asked if it would be
possible to use some time during a class period for volun-
tary participation in this study. The three instructors
contacted all agreed tO cooperate.

When the adult experimental aide was contacted
he was told only that it was a study concerned with

learning more about how people react tO and evaluate

29

political messages such as those they might encounter in
their newspapers. He was then asked if he would attempt
to collect the data from those in the area where he
worked. Upon agreeing, he was given instructions on the
administration Of the materials.

The experimental package was designed tO be self-

administering tO help insure uniformity.*

At the experimental session S's were given the

following information:

I am interested in learning more about how people
react tO and evaluate messages. What I am asking you
to dO is tO read a communication and evaluate it as
you normally would if it appeared in your own news-
paper.

The aide was instructed tO alter the instructions

as little as possible. He was instructed to state:

An instructor in the Speech Department at the
University Of Connecticut is interested in learning
more about how peOple react tO and evaluate messages.
He has asked me tO see if some Of you would help him
with his study.

If they agreed, as all did, he then continued:

Please read the instructions, and complete all
the scales. Read the message as you would if you
saw such a message in your own neWSpaper. Return
it tO me when you finish.

At this point the experimental materials were

distributed and the S's were asked tO read the cover sheet.

At this point they were reminded that participation was

voluntary and that if they wished they would be excused

 

*
The complete final instrument will be found in
Appendix C.

30

now. NO S's refused at this point. The S's were then
asked to turn to the second page Of the booklet which
contained instructions concerning the completion Of the
scales. They were given time tO read the instructions
and any questions were answered.

When it appeared that all were satisfied that
they understood the instructions they were asked to turn
to page three and proceed until they had completed the
entire questionnaire. When all S's had completed the
questionnaire, the materials were collected, any questions
were answered and the class dismissed.

The same procedure was used with the adult S's
with the exception that the instructions were given tO
small groups Of §'s who worked in close proximity tO each

other.

MeasuringpInstruments

 

For measuring both attitudes toward Open housing
and perceptions Of the source's attitude toward Open
housing, six seven-step bipolar scales were used. The
scales were: Positive-Negative, Fair-Unfair, Good-Bad,
Valuable-Worthless, Honest-Dishonest, and Wise—Foolish.
The direction was alternated tO reduce the effects Of
response sets. The six scales were scored by assigning
a value Of seven tO the response indicating the most
favorable attitude toward Open housing and a value Of one

to the response indicating the least favorable attitude.

31

Each S's score was based on a summation Of his responses
over the six scales. A total score Of 42 thus represents
a maximally favorable attitude toward Open housing and

a score Of six a minimally favorable attitude.

TO assess the perceived credibility Of the
source, scales representing the three dimensions Of
source credibility found by Berlo, Lemert, and Mertz
(1966) were used. TO assess the safety dimension Of
credibility five bipolar scales were used. The scales
were: Safe-Unsafe, Just-Unjust, Kind-Cruel, Friendly—
Unfriendly, and Honest-Dishonest. TO assess the qualifi-
cation dimension the scales used were: Trained-Untrained,
Experienced-Inexperienced, Skilled-Unskilled, Qualified-
Unqualified, and Informed-Uninformed. TO assess the
dynamism dimension the scales used were: Aggressive-Meek,
Emphatic-Hesitant, Bold-Timid, Active-Passive, and
Energetic-Tired. All scales were seven steps and the
direction Of polarity was rotated tO minimize the effects
Of response sets. A score Of seven was assigned to the
response indicating greatest safety, qualification, or
dynamism and a score Of one to the reSponse indicating
the least safety, qualification, or dynamism. For each
dimension the subject's ratings were summed over the
five scales for that dimension. Thus, a score Of 35 was
maximum for each dimension and a score Of five was

minimum.

32

Perceived similarity between the views Of the g
and the views Of the source Of the message was measured
in two ways. The S was asked tO give his own views on
Open housing as a pretest measure and asked tO indicate
his perceptions Of the source's attitude on the post-
test measures. The correlation between those two measures
serves as one indication Of perceived similarity Of views.
In addition, the following single measure was used:
Please indicate how similar the source's views on Open
housing are to your own: Exactly Like Mine, Very Much
Like Mine, Quite a Bit like Mine, Somewhat Like Mine,

Not at All Like Mine. A value Of five was assigned tO
the response indicating least similarity Of views.

The final instrument used was a measure Of the
g's perceived ambiguity Of the message. The scales used
were: Clear—Unclear, Precise-Vague, and Unambiguous—
Ambiguous. These were seven step scales and the polarity
was rotated tO minimize the effects Of response sets. A
value Of seven was assigned tO the response indicating
greatest clarity and a value Of one tO the response
indicating the least clarity. The S's score was Obtained
by summing across the three scales.

A complete copy Of the measuring instruments used

will be found in Appendix C.

CHAPTER III

RESULTS

Pretest Of Message Manipulation

The messages were first submitted tO five members
Of the Speech Department faculty. These judges were
asked tO rank the messages in order Of the preciseness
with which they felt the message expressed the author's
position on the issue Of Open housing. All five judges
ranked the messages in the order predicted by the
abstraction index. The messages were then submitted tO 20
students in the Speech Department at the University Of
Connecticut. Of those students, 10 were undergraduate
and 10 were graduate students. They also were asked tO
rank the messages according tO the preciseness with which
they felt the messages expressed the authors' positions
on Open housing. Sixteen Of the 20 S's ranked the mes—
sages in the exact order predicted by the abstraction

index and the faculty rankings (p < .001).

33

34

Pretest Equivalency Of
Experimental Groups

 

TO check on the effectiveness Of the method of
assigning S's to treatment groups an analysis Of variance
was run on pretest attitude toward Open housing. It
was felt desirable to check the comparability Of the
college students with those S's who were nO longer asso-
ciated with the college campus. This was partially
because the work group was located very near an area Of
the city that has experienced riots for four consecutive
summers. The results Of that analysis are found in
Table l and indicate that the groups do not differ
significantly on pretest attitudes toward Open housing.

Perceived Similarity Of Source's
and Subjectrs Position

 

 

The first hypothesis predicted that a message
high in purposeful ambiguity would lead tO subjects
rating the source's position on Open housing as more
similar to the subject's own position than with a mes-
sage low in purposeful ambiguity. The instruments in
this study provide three methods Of measuring the per-
ceived degree Of similarity. First, one can look at
the difference scores between the S's pretest attitude
toward Open housing and the perceived position Of the
source toward Open housing. Second, one can correlate

the two measures just mentioned. Finally, one can use

35

 

 

 

 

 

 

Table l. Anova Of Subjects' Initial Position on Open
HouSIng.
Message I Message II Message III
Score
C* NC** C NC C NC
37-42 4 5 7 6 5 5
31-36 7 7 3 6 3 5
25-30 5 4 3 9 6 7
19-24 0 4 3 3 1 3
13—18 0 1 0 0 0 1
6-12 0 O 0 0 0 0
x 33.38 30.90 33.69 31.54 33.54 31.14
sd 4.38 7.41 6.60 5.65 5.78 7.17
*C = College Subjects.
**NC = Noncollege Subjects.
3232?.ch SS df MS F P
Between Groups 132.92 1 132.92 3.16 n.s.
Between Messages 5.51 2 2.76 0.06 n.s.
Group X Messages 1.09 2 0.54 0.01 n.s.
Error 503.26 107 42.09
642.78 112

 

36

the S's rating Of the degree Of agreement with the
source's position.

Table 2 shows the results Of a test Of the differ-
ence between the S's pretest position and the perceived
position Of the source Of the message. As the table
indicates, the predicted difference between the high and
the low purposeful ambiguity conditions is small and non-
significant (t = 0.92, p > .05). The larger difference
between the low and the moderate purposeful ambiguity also
was nonsignificant (t = 1.46, p > .05).

Table 2. Perceived Similarity Between Source and Subject
Attitude Toward Open Housing.

 

 

Message I Message II Message III
N 37 40 36
D 5.86 3.75 4.36
6.17 6.51 7.69

t—I vs. III = 0.92 n.s.
= 1.46 n.s.

 

Using correlational methods tO investigate the
effects Of increasing purposeful ambiguity lead to the
results in Table 3. As can be seen, the correlation for
the low ambiguity condition was greater than the corre—
lation for the high ambiguity, though not significantly.
If the hypothesized assimilation had occurred tO a greater

degree there should be a significantly higher correlation

37

in the Message III condition than in the Message I con-
dition. The Opposite trend was found. Message II showed
the greatest tendency for assimilation. Even this was
not significantly different than the correlation found
with Message I. The largest difference (Message II vs.
Message III) also falls short Of significance (Z = -l.75
p i .08 2 tailed).

Table 3. Correlation Between S's PreteSt Score and
Perceived Position Of the Source.

 

 

 

Message I Message II Message III
N 37 40 36
X -5- 31.97 32.40 31.97
Score
Y S°ur°e 37.84 36.15 36.33
Score
* *1):
r g x 0.40 0.56 0.21
Source
*
p < .05.
**
p < .01.

A third measure Of perceived similarity was the
degree Of agreement with the source's position. As the
data in Table 4 indicate, there was very little difference

in the Obtained results.

38

Table 4. Perceived Similarity Between Source's and
Subject's Position on Open Housing.

 

 

Message I Message II Message III
N 37 40 36
i“ 2.73 2.88 2.78
s.d. 0.96 0.96 1.02

t-I vs. III = 0.21 n.s.

 

As should be expected, the three methods Of
assessing the perceived similarity Of views toward Open
housing between source and receiver agree. The data
Obtained dO not support hypothesis la. There is no evi-
dence that increased purposeful ambiguity leads tO in-
creased perception Of similarity Of views toward Open
housing between the source and the receiver Of the communi-
cation.

The second part Of Hypothesis 1 stated that there
would be greater variance in the ratings Of the sources'
positions with the more ambiguous message. The variance
ratio between the most and the least ambiguous message
resulted in a nonsignificant difference (Table 5).

As Table 5 indicates, the greatest variance was
found in the ratings Of the middle message. The differ-
ence between Message I and Message II was significant

while the difference between Message II and Message III

39

is Opposite to the direction predicted. It must be con-

cluded that the data do not support Hypothesis lb.

Table 5. Perceived Position Of the Source on Open

 

 

Housing.
Message I Message II Message III
N 37 40 36
2 37.84 36.15 36.33
s.d. 4.59 7.46 5.34

Variance Ratio

I vs. III--l.35 df-36/37 n.s.
I vs. II--2.64 df—40/36 p < .05.

 

Ambiguity and Source Credibility

The second hypothesis stated that the variability
Of the ratings Of qualification would be greater for the
least ambiguous message than for the most ambiguous mes—
sage. A test Of the variance ratio indicates that the
difference was nonsignificant (Table 6).

The second hypothesis also stated that with a
message high in purposeful ambiguity the source would
be rated significantly higher on the qualification
dimension Of credibility than with a message low in purpose-
ful ambiguity. The data shown in Table 7 indicate that
the hypothesis was not supported. Virtually nO differ-
ence was found in the ratings Of qualification (F = 0.57,

p > .05).

40

Table 6. Perceived Qualification Of the Message Source.

 

 

Message I Message II Message III
N 37 40 36
X 24.08 25.15 24.00
s.d. 5.88 4.53 5.44

Variance Ratio

I vs. III--l.16 df-37/36 n.s.
I vs. II--l.68 df—37/40 n.s.

 

Table 7. Analysis Of Variance Of Perceived Qualifica-
tion Of the Source.

 

Source Of SS df MS F

 

Variance p
Messages 31.90 2 15.95 0.57 n.s.
Error 3083.86 110 38.04

Total 3115.75 112

 

The first portion Of Hypothesis 3 predicted that
the variance for the ratings Of the safety dimension
would be greater for the least ambiguous message and
smallest for the most ambiguous message. As Table 8
shows, the variance ratio was very small and nonsignifi-
cant.

The third hypothesis also predicted that the

message high in purposeful ambiguity would lead to a

41

Table 8. Perceived Safety Of the Message Source.

 

 

Message I Message II Message III
N 37 40 36
3? 25.40 26.18 24.92
s.d. 4.83 6.71 5.40

Variance Ratio

I vs. III--l.25 df-36/37 n.s.
I vs. II—-1.93 df-40/37 p < .10 (2 tailed)

 

higher rating Of source credibility on the safety dimension.
Table 9 contains the results concerning the safety
dimension Of credibility. As with Hypothesis 2 the

results were nonsignificant. As with Hypothesis 2 the
subjects rated the safety dimension Of credibility higher
for the middle message than for the most ambiguous or

the least ambiguous message (Table 9). The difference

in ratings was, however, tOO small tO be statistically

reliable.

Table 9. Anova Of Perceived Safety Of Message Source.

 

Source Of

 

Variance SS df MS F p
Messages 30.75 2 15.38 0.47 n.s.
Error 3616.44 110 32.88

Total 3647.19 112

 

42

The fourth hypothesis stated that the variance
in ratings Of dynamism for the most ambiguous message
would be less than the variance found in the ratings
associated with the least ambiguous message. The test
Of those variances indicate no significant differences
(Table 10). The only significant difference was between
the variance on message II and message III. That differ-

ence is the direction Opposite tO the prediction.

Table 10. Perceived Dynamism Of the Message Source.

 

 

Message I Message II Message III
N 37 40 36
3? 27.08 27.35 26.06
s.d. 5.45 4.17 6.10

Variance Ratio

I vs. III--1.25 df-36/37 n.s.
II vs. III--2.14 df—36/40 p < .10 > .02
(2 tailed)

 

Hypothesis 4 also predicted that the subjects
receiving the most ambiguous message would attribute the
highest ratings Of dynamism tO the source. As Table 11
indicates, this was not supported. NO significant differ-
ences were found in the subjects' ratings Of dynamism
when compared across the three messages. The very low F
(0.62) would seem tO indicate lack Of any stable differences

in perceived dynamism across the three treatment groups.

43

Table 11. Anova Of Perceived Dynamism Of the Message
Source.

 

Source Of

 

Variance SS df MS F p
Messages 34.70 2 17.35 0.62 n.s.
Error 3049.74 110 27.72

Total 3084.44 112

 

Ambigtity and Attitude Change

The fifth hypothesis predicted that the group
reading the most ambiguous message would show greater
attitude change than the group reading the message lowest
in ambiguity. Table 12 shows the attitude change scores
for all three experimental groups. As the table indicates,
the group receiving the most ambiguous message did not
show significantly more attitude change (F = 0.44, Table
12). It must be concluded that the data fail to support
Hypothesis 5. In all treatment groups the changes are
too small tO conclude that the message did indeed lead tO

attitude change as measured in this study.

Rated Ambiguity Of the Messages
As stated in Chapter I in order for a message to
be purposefully ambiguous the receiver must attribute
denotative meaning to the statements. As a check on the
receiver perception Of the ambiguity Of the messages

they were asked tO rate the ambiguity Of the message they

44

 

 

 

 

 

Table 12. Attitude Change Scores.
Message I Message II Message III
N 37 48 36
5 0.65 0.48 1.17
s.d. 2.07 3.28 4.89
3°“Fce 0f 85 df MS F p
ariance
Messages 11.39 2 5.69 0.44 n.s.
Error 1412.63 110 12.84
Total 1424.02 112
received. The results Of that rating indicate that the

subjects perceived the messages as approximately equally

precise (Table 13).

Table 13.

Anova Of Rated Ambiguity Of the Message.

 

Source Of

 

Variance SS df MS F p
Messages 18.69 2 9.34 0.58 n.s.
Error 1760.18 110 16.00

Total 1778.87 112

 

CHAPTER IV
DISCUSSION

Range Of Initial Scores on Attitude
Toward Open Housing

 

 

The experimenter attempted tO gain access tO
a number Of groups in the local area which were per-
ceived tO represent a range Of views toward Open housing.
Being a largely rural area, there were only a few organi-
zations with memberships large enough to permit adequate
distribution Of subjects tO treatment groups. The
primary group selected was the volunteer fire depart-
ments. After contacting eight such groups it was possible
to gain access tO only two groups during a regular meet-
ing time. The other groups flatly refused admittance to
the experimenter when they learned the tOpic Of the
message. Attempts to explain that the tOpic was Of
secondary importance and the primary task was tO study
how people read and evaluate messages had nO effect.
The two groups that gave permission tO use regular
meeting time were located in the immediate university

area. At the first meeting the experimenter was stalled

45

46

until very late after the group learned about the task.
In the end it was asked that instructions be given and
the experimenter leave the meeting. The experimental
materials were left with the group along with the request
that the materials be left at the fire station (the usual
place Of meeting for the group) at the close Of the meet-
ing and they would be picked up in the morning. One
questionnaire was completed. Comments received by the
experimenter indicated that the group felt that the
message was tOO political and probably was, in fact,
related to the current political canvassing in the area.
The local groups are in frequent contact with each other
and it became evident that the second group learned Of
the happenings at the first meeting and shared many Of
the same views therefore that contact was cancelled.

The other noncollege group worked at a large
insurance company. Though they were nO longer on campus
they were college educated and most were recent college
graduates--within the last five years. They were not
as different from the college sample as had been hoped.
Table 1 (Chapter III) shows the distribution Of scores by
group and message condition.

An additional factor possibly causing refusals
could have been the frequency Of black rioting in the
nearest large city. In each Of the last four summers

major riots have occurred.

47

Whatever the cause for refusals, it seems that
the remaining cooperative pOpulation may have been tOO
much in favor Of Open housing tO provide an adequate

test Of the theory prOposed.

Hypothesis 1

 

The first hypothesis stated that the message
highest in purposeful ambiguity would cause subjects tO
rate the perceived position Of the source as more simi-
lar to their own than would the message lowest in pur-
poseful ambiguity. The data dO not support the
hypothesis. There are nO significant differences in
perceived position Of the source across the three treat-
ment groups.

Hypothesis 1 was based on the rationale that
greater abstraction in the message would Offer subjects
fewer anchors for placement Of the message and therefore
allow them tO provide their own interpretations for the
symbols used. The data reported in Chapter III indi-
cates that the subjects found sufficient anchors to
allow them to place the position Of the source at a
relatively firm pro-Open housing position. Furthermore,
ratings Of message clarity showed nO significant differ-
ences across the three messages, deSpite the fact that
the pretest with a different group Of subjects clearly

indicated perceived differences.

48

In retrospect, it appears that pretest data
should have been Obtained tO index not only the ranking
but also to assess the perceived magnitude Of differences
as well. It is possible that the differences between
messages were Of the order Of just noticeable differences
and not sufficient tO be reliably indexed when the
messages were seen in isolation. If this conjecture is
correct, this problem plagues all the hypotheses investi—
gated.

Another possible source Of difficulty could have
been the Abstraction index score itself. As can be
seen in Appendix A, the index score is quite easily
manipulable if one so desires. It would seem quite
possible that the method employed in the construction
Of the three messages rendered the Abstraction scores
invalid. The robustness Of the instrument tO intentional
manipulation has not been carefully investigated. How-
ever, On the basis Of the data Obtained in this study
relative to the variations in the messages, it appears
that the scores bore little relation tO receiver per-
ceptions Of the abstraction level. It would seem that
even when viewing written material in isolation,
receivers should perceive and report some differences
when receiving material as divergent in abstractness

as Reader's Digest and a college level philosophy text

 

49

(the approximate variation between the least and the

most abstract message).

Hypotheses 2, 3, and 4

 

The second, third, and fourth hypotheses dealt
with the effects Of purposeful ambiguity on the three
dimensions Of source credibility isolated by Berlo,
Lemert, and Mertz (1966). It was predicted that the
perceived qualification, safety, and dynamism dimensions
Of source credibility would be higher when the receiver
was exposed to the highest level Of purposeful ambiguity.
Because support was not found for the three related
hypotheses, they will be discussed together.

These hypotheses were based on the assumption
that ambiguity would allow assimilation Of the source's
position tO one similar tO the subject's own. It was
reasoned that this should be a very likely occurrence
if the terms used held positive connotations for the
receivers. As Table 5 in Chapter III indicates, there
were no significant differences in the perceived position
Of the source across the messages. This coupled with
the similarity Of the means Of the experimental groups
before exposure would almost preclude support for the
hypotheses. Most subjects were in favor Of Open housing
and they also perceived the source to be in favor Of

Open housing, therefore, little assimilation could occur.

50

Table 14. Subjects' Pretest Attitudes and Perceived
Position Of the Source.

 

 

Message I Message II Message III
Subject *
Attitude 31.97 32.40 31.97
Source
Attitude 37.84 36.15 36.33

 

*
Possible Range = 6-42, neutral = 24.

Hypothesis 5

The fifth hypothesis predicted that subjects
receiving the most abstract message would show the
greatest attitude change in the direction advocated.
As Table 12 in Chapter III indicates, the hypothesis
was not supported. The initial differences between
the subjects' attitudes and the perceived position Of
the source allowed little room for attitude change.
The overall mean for subject attitude toward Open housing
was about 32 on a 42 point scale. The overall perceived
attitude Of the source was about 36.6. These scores
allowed little room for measured attitude change within
the range that would normally be expected (between pre-
message score and perceived position advocated in the
message). However, even with the narrow range allowed,
the attitude change was only about one scale unit or

22% or that which could be considered tO have been

51

advocated. The data seem to indicate that abstraction
is probably not the strategy best suited tO generating
attitude change toward the topic. However, assuming
that a totally unknown source has neutral credibility
the data in Tables 6, 8, and 10 indicate that the middle
message generated slightly more credibility than did the
other twO messages. Again, in this case, the order is
not as predicted in the rationale. It would seem that
careful investigation Of the differences in the three
messages on factors not included in the indices used
here is needed. The next section will incorporate some

suggested alternatives.

Implications for Further Research

 

In several cases the middle message Of the three
used showed more tendency toward the acceptance pattern
predicted than did the most abstract message. This
would seem tO indicate that perhaps there were variables
Operating in the messages Of which the experimenter was
unaware. Another possible explanation is the existence
of a curvilinear relationship which was beyond the sensi-
tivity Of the present measuring instruments. Either
possibility suggests that more development in the area
Of measurement is needed before more definitive results
should be expected.

Any further research must first investigate the

problem Of indexing the abstraction level Of the message

52

and the relationship Of that tO denotative variations
in meanings across populations. While it appears conf
sistent that more abstract words should elicit more
variations in meanings across subjects the present
methodology allows that tO remain in assumption. One
possible solution to that problem would be tO request
the subject to summarize the message in a limited number
Of words and compare the subjects' statements across
messages. Greater variation in subject encoded messages
would indicate greater ambiguity in the stimulus message.
This basic technique was attempted in the pilot study
previously mentioned and the results were not satisfactory.
The results indicated two problems: A portion Of the
subjects merely restated the message using the words
already provided in the message; a second group Of
subjects evaluated the message in terms Of their own
preferences. These two groups constituted a substantial
proportion Of the subjects responding and rendered the
resulting material Of little value. If the subjects
were limited tO a fraction Of the number Of words in
the original message the problem might less severe,
e.g., one-tenth the length Of the stimulus message or in
the present case 20 words.

A second possible approach tO indexing the
ambiguity Of the messages is suggested by information

theory. Information theory states that information

53

reduces the number Of alternatives available. It would
be possible to have the subjects react to a list Of
alternative behaviors related to the message topic. The
changing probabilities Of selection Of alternatives
could be used as an index of the information in the
message. The message generating the most consistent
election Of alternative behaviors across subjects would
be classified as the most precise or least ambiguous.

A related approach would use a list Of bipolar
adjectives related tO the concepts discussed in the
message. Assuming a normal distribution Of choices given
no information, the deviation from a normal distribution
would serve as an index Of the amount of information in
the message. Of the two methodologies suggested as the
second alternative, the latter would be the more generally
applicable. The scales could remain relatively constant
and only the concepts to be rated would have to change.
This would provide at least some generalizability to the
measure Of information in the message.

Another possible method Of indexing the ambi-
guity Of the messages would be the use Of physiological
measures. According to Berlyne (1960) a more complex
stimulus leads to longer orienting responses (pp. 99-102).
He further reports that the presentation Of stimuli
cause adaptive responses translatable to GSR and Alpha

wave measurement (pp. 88-92). Alpha waves are suppressed

54

during periods of arousal. It would seem that a message
lacking in firm referents for the subject would lead tO
mental activity designed tO clarify the message. That
is, the subject should attempt tO reconstruct the stimu-
lus tO fit his past experiences. In the end this would
involve finding or supplying referents for the message.
During the process Of resolving the problem caused by
the lack Of referents the subject should be in a state
Of arousal. By monitoring the period Of suppression Of
the alpha waves or GSR increase an index Of the relative
ambiguity Of the stimulus message could be developed.
By exposing the subject tO a number Of brief messages
(one or two sentences) and comparing the arousal periods
recorded one should be able to develop an index Of the
relative ambiguity Of the messages.

TO summarize, it is felt that the present method
Of indexing the ambiguity Of the messages is unsatis-
factory. First, the robustness Of the index tO intentional
manipulation is unknown. Also, the index was designed tO
be used with samples Of material from longer total
messages. The fact that only about 200 words were
available for the entire message may have affected the
reliability Of the instrument. For those reasons alter-
native methods are suggested. Two are based on an infor-
mation theory type Of measure Of ambiguity. The first

involves having the respondent encode a message a fraction

55

the length Of the original attempting to express the
central content Of the original message. The second
consists Of having the subjects respond to bipolar
scales or a series Of positions that are apprOpriate tO
the tOpic Of the message. The third involves physiologi-
cal measures. It is felt that any Of these techniques
are at least equal in validity to the measure actually
used in terms Of assessing the ambiguity Of the message.
An advantage Of these alternative techniques would be
that they allow manipulation Of message elements in a
new attempt tO locate the portions Of the message that
lead to increased ratings Of ambiguity.

A second change in the methodology would be
tighter control Of the subjects' initial attitudes.
In the pretest Of topics for this study a split was found
in expressed attitude toward Open housing. However, when
the final study was run it was found that the distribution
Of attitudes was strongly skewed in favor Of Open housing.
Time may have been a factor but also the self selection
Of participating subjects probably was a greater factor.
The refusal Of some groups Of subjects tO complete the
experimental task appeared to be at least partially based
upon attitudes toward Open housing. Future investigations
should be designed to insure that attitude is more tightly
controlled. It would be helpful tO separate the pretest

on attitude and the message evaluation sessions. In that

S6

manner the experimenter can adjust the tOpic to control
the attitude distribution. If all attitudes are located
at one end Of the continuum the researcher cannot be
certain that the resulting ratings Of communicator position
is not the result Of his own perception Of the message
before the experiment was run. That is, ambiguity should
allow assimilation Of the message toward the attitude
position Of the receiver. TO be certain that this has
occurred it is necessary tO show that the message was
actually assimilated in the direction Of the receiver
based on pretest positions that occupy very different

positions Of the issue in question.

BIBLIOGRAPHY

A
J

BIBLIOGRAPHY

Abelson, R. P. Modes Of resolution Of belief dilemmas.
In: Fishbein, M. Readings in Attitude Theoty»
and Measurement. New York: John Wiley and
Sons, Inc.,I1967.

 

Berlo, D. K. The Process Of Communication. New York:
Holt, Rinehart, and Winston, 1960.

 

Berlo, D. K., Lemert, J. B., and Mertz, R. J. Dimensions
for evaluating the acceptability Of message
sources. Research Monograph, M. S. 0., Depart-
ment Of Communication, 1966.

Berlyne, D. E. Conflict, Arousal, and Curiosity.
New York: McGraw-HiII, 1960.

Bousfield, W. A. The problem Of meaning in verbal
learning. In: COfer, C. N. Verbal Learning
and Verbal Behavior. New York: McGraw-Hill,
1961.

 

 

Byrne, D. Interpersonal attraction and attitude
similarity. J. Abnorm. Soc. Psychol., 1961,
22" 713-7150

 

Campbell, D. T., Hunt, W. A., and Lewis, N. A. The
effects Of assimilation and contrast in judg-
ments Of clinical materials. Am. J. Psychol.,
1957, lg, 347-360.

~-Dillehay, R. Judgmental processes in response to a

persuasive communication. J. Pers. SOC. Psychol.,I99h47
J ‘ .3311
DOdd, S. C., and Gerbrick, T. R. Word scales for

degrees Of Opinion. Language and Speech, 1960,

2' 18-310

Fearing, F. Toward a psychological theory Of human m _
communication. J. Pers., 1953, 22, 71—88. ‘;\ \ (3~

57

58

‘I“‘ Gillie, P. J. A simplified formula for measuring
abstraction in writing. J. Applied Psychol., u?\,3fijn
1957, 5;, 214-217.

“r— Haskins, J. B. Validation Of the abstraction index as
a tOOl for content--effects analysis. J. Applied
PSYCh010 ( 1960' fl’ N0. 2' 1.02-1060

 

Kelman, H. C. and Baron, R. M. Determinants Of modes
Of resolving inconsistency dilemmas: A functional
analysis. In: Abelson, R. P., Aronson, B.,
McGuire, W. J., Newcomb, T. M., Rosenberg, M. J.,
and Tannenbaum, P. H. Theorigs Of Cognitive
Consistency:_ A SourcebOOk. Chicago: Rand
McNally and CO., 1968.

 

Kelman, H. C. and Eagley, A. H. Attitude toward the
communicator, perception Of communication content,
and attitude change. J. Pers. SOC. Psychol.,
1965, l, 63-78.

Kerlinger, F. N. Foundattpns Of Behavioral Research.
New York: Holt, Rinehart, and Winston, 1966.

‘* Manis, M. The interpretation Of Opinion statements as
a function Of message ambiguity and recipient
attitude. J. Abnorm. Soc. Psychol., 1961,
‘gg, 76-81.

. Interpretation Of Opinion statements as a
function Of recipient attitude and source
prestige. J. Abnorm. Soc. Psychol., 1961,
2;, 82-86.

McGuire, W. J. The nature Of attitudes and attitude
change. In: Lindzey, G. and Aronson, E.
The Handbook Of Social PsychOlOgy, VOl. III,
2nd. editiOn, Reading, Mass.: Addison-Wesley,
1969.

Miller, G. R. Speech Communication: A Behavioral
Approach. IndIanapOlis: Bobbs-MerrIll, 1966.

 

 

Osgood, C. E. Comments on Professor Bousfield's paper.
In: COfer, C. N. Verbal Learning and Verbal
Behavior. New York: McGraw-Hill, 1961.

 

 

Osgood, C. E., Suci, G. J., and Tannenbaum, P. H. The
Mgasurement Of Meaning. Urbana: University Of
IlIinoIs Press, 1957.

 

59

Roever, J. E. Understanding misunderstanding: Toward
a theory Of purposeful ambiguity. A paper
presented at the Speech Association Of America
Convention, 1962.

Sherif, C., Sherif, M., and Nebergal, R. Attitude and
Attitude Changt. Philadelphia: Saunders, 1965.

 

 

Skinner, B. F. Verbal Behavior. New York: Appleton-
Century-Crofts, 1957.

 

Thayer, L. Communication Theorytgnd Research. Spring-
field, Ill.: Charles C. Thomas, 1967.

 

APPENDICES

APPENDIX A

ABSTRACTION INDEX

APPENDIX A

ABSTRACTION INDEX

1. Count the number Of finite verbs per 200
words. Count all verbs Of any tense which are in the
first, second, or third person and which have subjects,
either eXpressed or understood. DO not count nonfinite
verb forms or verbals. In verb forms with auxiliary
words, count the auxiliary rather than the main verb.

DO not count any form Of the verb "to be" (1&1 was,

 

 

 

Egg, were, will be, have been, etc.) when used only as
a COpula tO link the subject with a predicate complement.
2. Count the number Of definite articles and
their nouns per 200 words. Count both the article tng
and the noun it modifies, but only if that noun is a
single word not otherwise modified, either by an inter-
vening adjective or by a clause or phrase following the
noun. DO not count ttg_when modifying adjectives or

noun-adjectives, as in the best, the Irish.

 

 

3. Count the number Of nouns Of abstraction per

200 words. Count all nouns ending in the suffixes -ness,

60

61

-_t£p, -d2m, -ngg, -tpn, and -y, including the plurals

Of such nouns. Count nouns ending in -y even when it

is the end Of a longer suffix like -;ty_or -plpgy_but not
when it is used as a diminutive.

4. Add the numbers found in Steps 1 and 2 and
add 36 to this sum.

5. Multiply the number found in Step 3 by 2.

6. From the total found in Step 4, subtract the
result Of Step 5. The result Of this subtraction is the
abstraction score. Scores should be interpreted as
follows: 0-18, very abstract; 19-30, abstract; 31-42,
fairly abstract; 43-54, standard; 55-66, fairly concrete;

67-78, concrete; 79-90, very concrete.

APPENDIX B

EXPERIMENTAL MESSAGES

APPENDIX B

EXPERIMENTAL MESSAGES

Message I

 

I would like to take this Opportunity tO discuss
one Of the difficult issues facing us tOday. That issue
is: how vigorously shall we enforce Open housing laws
as they currently exist.

Laws today exist which make it illegal to refuse
to sell or rent a house or an apartment tO a person
because Of his race or religion. Yet there are few
places where blacks and whites live side by side. The
neighborhoods are segregated, therefore, the schools also
are segregated. Separation Of the races in the neighbor-
hOOds and in the schools leaves few chances for blacks
and whites tO get to know each other socially. Lack Of
social contact between the races causes a major problem
tOday.

Prejudice toward blacks is being handed down like
some birthright. We cannot prevent this unless the
present leaders take positive action tO get the races

together. Only in that way will blacks and whites get tO

62

63

know one another better. This can happen only if the
white leaders see that Open housing is vigorously
enforced.

We can no longer afford tO wall blacks and other
minorities into the centers Of our cities. All Of the
citizens must have equal access tO all Of the products

Of our society.

64

Message II

I would like tO take this Opportunity tO discuss
one Of the difficult issues facing us tOday. That issue
is: how vigorously shall we enforce Open housing laws
as they currently exist.

We currently have laws which make discrimination
in the sale or rental Of housing illegal. Yet when we
look at the housing patterns in America we see that
segregation is clearly evident. Segregation in housing
leads tO segregation in schools and community life. This
in turn leads tO the perpetuation Of prejudice toward
minority groups. The majority population has little
Opportunity tO gain first hand knowledge Of other groups.

Prejudices are handed down like some birthright.
This creates a cycle which can be broken only if community
leaders take steps tO break down the wall Of ignorance
between the various groups in the community. Knowledge
must replace the myths and the legends. This can come
about only if everyone lives by the rules Of our society.
This will occur only if reasonable housing is made
available tO all our citizens.

This nation can no longer afford to build ghettoes
in our central cities. We must give all Of our citizens

access tO all the products Of our society.

65

Message III

I would like tO take this Opportunity to discuss
one Of the difficult issues facing us tOday. That issue
is: how vigorously shall we enforce Open housing laws as
they currently exist.

We currently have laws making discrimination in
the sale or rental Of housing illegal. Yet when we lOOk
at housing patterns in almost any community in America
we see segregation. Segregation in the community leads
to segregation in our schools and social institutions.
This in turn leads tO perpetuation Of prejudice toward
the segregated groups because the majority is afforded
no Opportunity tO gain first hand knowledge Of minorities
through informal social contact.

Prejudices are passed down from generation to
generation unaltered. It is time that responsible
leaders in the community should seek action tO break down
the wall Of ignorance that separates the various racial
and ethnic groups in our midst. This can occur only if
everyone abides by the rules Of our society. It can
occur only if everyone has access tO adequate housing.

We can nO longer afford tO house huge minority
populations in our central cities. We must give our

citizens access to the products Of our society.

APPENDIX C

MEASURING INSTRUMENTS

APPENDIX C

MEASURING INSTRUMENTS

Inside this booklet you will find a number Of
scales and one message. We are interested in learning
more about how peOple react tO and evaluate political
messages such as those that might appear in their local
newspaper. This is not a test Of any type. It is a
request for your personal reaction to a message.

You will notice that the first information
requested is about you. This information is necessary
so that we can know more about the people who are
evaluating the messages. Please notice that at nO time
are you requested tO put your name or any other identifying
marks on the booklet. There will be nO way the information
collected here today can be associated with you or this
organization. SO please fill in all the requested infor-
mation as honestly and as candidly as you can.

Please read the message as you might any such

political message appearing in your local paper. Then

66

67

complete the rest Of the booklet based upon your reaction
tO the message.

Your cooperation in this study is sincerely
appreciated and will aid us in better understanding

communication among peOple.

Gary B. Wilson
Speech Department
University Of Connecticut

68

INSTRUCTIONS FOR THE COMPLETION OF THE RATING SCALES
Inside this booklet you will find rating scales
which lOOk like this:
GOOd ___ : ___ : ___.3 ___ : ____: ___ : ___ Bad

Consider the Spaces tO represent the following degrees

Of Opinion:

X ____. ___ . ___ : ___ : ___ : : Y
E Q s N s T "E—
x u l e 1 u x
t i i u i i t
r t g t g t r
e e h r h e e
m y a t m
e l l l e
l y y 1
Y Y

You will be asked to rate someone or something on that
scale. For example, suppose you were asked tO rate
Senator DOdd on the scale. A check (/ ) in the center
space would indicate that you felt neutral about Senator
DOdd.

Good ___ : ___ : ___ :__£“ : ___ : ___ : ___ Bad
As you place your check closer tO the ends Of the scale
it indicates stronger feelings. If you felt Senator DOdd

were extremely gOOd you would place your check in the

space closest tO "Good."

Good Z : : : . : : Bad
If you felt he were only slightly gOOd you would place
your check in the Space nearest the center or neutral

space toward "Good."

GOOd : : / : : : : Bad

69

DO not place any checks between the spaces:
This Not This
Good ____: _£__: ___ : ___ ( ___ : ___ : ____Bad
Please make one check mark on each and every scale--dO

not skip any.. Work fairly quickly, but not carelessly,

it is your first impression that is important.

70
Age Sex

 

MARITAL STATUS:

 

PLACE OF RESIDENCE:

Own or are buying present home
Rent present home

Rent apartment

Live with parents or relative
Live in college housing

Other (Please specify)

 

 

POLITICAL PARTY PREFERENCE:
Democrat
Republican

Other (Please specify)

 

PLEASE INDICATE THE STRENGTH OF PREFERENCE FOR YOUR
POLITICAL PARTY:

Very strong
Strong
Moderate

Indifferent

 

PLEASE RATE YOUR POLITICAL BELIEFS:

Liberal __ : : : : : : __ Conservative

PLEASE INDICATE YOUR VIEWS TOWARD OPEN HOUSING

Good __ : __ : __ : __’: __’: __': __ Bad
Unfair ___: __ : __': __ : ___: __ : __ Fair
Valuable ___: __.‘ __ : __ : __ : __ : __ Worthless
Honest __ : __ :‘__ .‘__ : __ : __.: __ Dishonest
Foolish __ : __ : __ °.__ :‘__ : __.: __ Wise
Positive : : : : : : Negative

71

PLEASE INDICATE HOW YOU BELIEVE THE SOURCE FEELS
TOWARD OPEN HOUSING:

 

Wise_:_:_:_:_:_:_FOOlish
Negative__:__:_: _: _: _: __Positive
Bad__:_:_:_:__:__:__Good

Fair__:_:_:_:___:_:_Unfair
Worthless __ : __ : _ : _ : _ : _ : __Valuable
Dishonest_:__:_:_:_:__:_Honest

PLEASE RATE THE SOURCE OF THE MESSAGE:
Safe___:_:_:_:_:_:__Unsafe
Unjust__:___:__:_:_:__:_Just

Crue1_:_:_:_:_:_:_Kind
Friendly _ : _ : _ : _ : _ : _ : _Unfriendly
Dishonest__:_:_:__:_:_:__Honest

Trained_ : _ : __ : _ : _ : _ : _Untrained
Inexperienced __ : _ : _ : _ : __ : __ : __Experienced
Skilled__ : _: __ : _: _:___: __Unskilled
Unqualified___ : __ : _ : __ : _ : __ : ___Qualified
Informed _ : __ : __ : _ : _ : __ : ___Uninformed
Aggressive_:_:_:__:_:_:__Meek
Emphatic_:_:_:__:_:_: _Hesitant
Timid__:___:___:_:_:__:___Bold
Active__:_:_:_:_:_:__Passive
Tired : : : : : : Energetic

72

PLEASE INDICATE HOW SIMILAR THE SOURCE'S VIEWS ON OPEN
HOUSING ARE TO YOUR OWN:

Exactly like mine
Very much like mine
Quite a bit like mine
Somewhat like mine

Not at all like mine

 

PLEASE INDICATE HOW CLEAR YOU FELT THE MESSAGE WAS:

Clear__:_:__:__:_:_:_Unclear
Ambiguous_: ___:_:_:_: _: _Unambiguous
Precise __.: : : : - : __ Vague

PLEASE INDICATE YOUR VIEWS TOWARD OPEN HOUSING:

Good_:_:__:__:__:_:_Bad
Unfair_:_:_:_:_:_:__Fair
Valuable __ : __ : __ :‘__ : __ : __’: __ Worthless
Dishonest_:__:_:_:_:_:___Honest
Foolish_:_:_:__:_:_:_Wise
Positive : : : : : : Negative

HICHIGRN STATE UNIV. LIBRARIES
INllll‘lllllllllflllllllllimltHIMIIIHWI‘IIHHIIWHI
31293106040425