Close
About
FAQ
Home
Collections
Login
USC Login
Register
0
Selected
Invert selection
Deselect all
Deselect all
Click here to refresh results
Click here to refresh results
USC
/
Digital Library
/
University of Southern California Dissertations and Theses
/
An Experimental Study Of The Effect Of Listener Feedback On Speaker Attitude
(USC Thesis Other)
An Experimental Study Of The Effect Of Listener Feedback On Speaker Attitude
PDF
Download
Share
Open document
Flip pages
Contact Us
Contact Us
Copy asset link
Request this asset
Transcript (if available)
Content
This dissertation has been
microfilmed exactly as received 68— 1689
MATTOX, Paul Richard, 1933-
AN EXPERIMENTAL STUDY OP THE EFFECT OF
LISTENER FEEDBACK ON SPEAKER ATTITUDE.
U niversity of Southern California, Ph.D., 1967
Speech
University Microfilms, Inc., Ann Arbor, Michigan
AN EXPERIMENTAL STUDY OF THE EFFECT OF
LISTENER FEEDBACK ON SPEAKER ATTITUDE
by
Paul Richard Mattox
A Dissertation Presented to the
FACULTY OF THE GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(Speech)
September 1967
UNIVERSITY O F S O U T H E R N C A LIFO RNIA
T H E G R A D U A T E S C H O O L
U N IV E R S IT Y PA R K
L O S A N G E L E S , C A L IF O R N IA 9 0 0 0 7
This dissertation, written by
.......................... P au l__R ichard M a t t o x ........................
under the direction of hi£.....Dissertation Com
mittee, and approved by all its members, has
been presented to and accepted by the Graduate
School, in partial fulfillment of requirements
for the degree of
D O C T O R OF P H I L O S O P H Y
...
Date S e p te m b e r ., 19 6.7.........................
DISSBRT^^^N ^
Chairman
TABLE OF CONTENTS
Page
List of Tables......................................... v
List of Figures....................................... vi
CHAPTER
I. INTRODUCTION ................................... 1
The Problem................................... 1
Definitions of Terms ........................ 2
Attitude ................................... 2
Negative feedback .......................... 3
Positive feedback .......................... 3
Negative position .......................... 3
Affirmative position ...................... 4
Limits of the Study.......................... 4
Significance of the Study .................... 5
Organization of the Remainder
of the S t u d y .............................. 10
II. REVIEW OF LITERATURE.......................... il
Introduction ................................. 11
Survey of Related Literature ............... 13
Cognitive Dissonance.......... ............. 30
Attitude Measurement Instruments ........... 35
CHAPTER Page I
i Summary ............................... 4 6 j
| III. DESIGN OF THE STUDY............................. 47
Preview................................. 47
1 Speakers...................................... 48
Listener...................................... 49
Materials...................................... 50
Facilities ........................ ..... 56
Procedures................................... 57
Training the listener .................... 57
Determining pre-experimental attitudes . . 61
Scheduling the speakers .................... 63
Maintaining the schedule .................. 64
Introducing the speaker and providing
overt feedback........................... 65
Statistical Procedures .................... 67
IV. PRESENTATION AND ANALYSIS OF DATA............. 71
Presentation of D a t a ........................ 71
Analysis of D a t a ............................. 71
V. SUMMARY, CONCLUSIONS, AND IMPLICATIONS .... 77
Summary ...................................... 77
Conclusions................................... 78
Implications for Future Research ............ 79
: BIBLIOGRAPHY............................................ 84
APPENDIXES............................................... 91
iii
CHAPTER Page j
Appendix A. Listener's Personal Reaction j
and Observations 92 j
I
Appendix B.
Photograph of Listener ............... 98
Preliminary Faculty Survey
Memoxundum............................. 100
Initial Opinion Survey Form ......... 102
Faculty Notification of Speaker
Selection Memorandum ............... 104
iv
LIST OF TABLES
! TABLE
1. The Woodward Ballot Responses of 66 Speakers
after Exposure to Overt Listener Feedback
Page
72
v
LIST OF FIGURES
FIGURE Page
1. Modified Woodward Ballot ........................ 53
2. Speaker Room.................... 58
vi
CHAPTER I
INTRODUCTION
The central concern of this study was to investi
gate the effect of listener feedback upon the attitudes of
a speaker toward his topic. If a person is speaking in
support of one of his strongly held beliefs, and if his |
1
listener overtly responds during that speech, will this j
listener feedback affect the speaker's belief? Will feed- j
|
back strengthen the speaker's conviction or weaken it?
Will the effect be different when the observable listener
feedback is positive (favorable) as compared with negative
(unfavorable) toward the speaker's expressed ideas?
The Problem
The problem of this study was co test the following
null hypotheses:
1. That no significant number of speakers shift
their initial attitudes toward their topic
after exposure to overt listener feedback.
2. That no significant number of speakers reverse
their initial attitudes, i.e., from affirmative,
to negative or vice-versa, after exposure to
1
overt listener feedback. ;
3. That there is no significant difference in ini
tial attitude changes after exposure to overt
j
!
listener feedback, between speakers who ini
tially favor a highly controversial proposition
and those who oppose it.
4. That among those speakers who change their ini- j
j
tial attitudes after exposure to overt listener j
j
feedback, there is no significant difference j
between the number whose attitude is reinforced I
i
f
and the number whose attitude is weakened.
5. That there is no significant difference between
i
the number of shifts of opinion after exposure j
to positive listener feedback as compared with
negative feedback.
Definitions of Terms
The following definitions of terms were used in
this study:
Attitude.— "An orientation toward or away from some
object, concept, or situation; a readiness to respond in a
predetermined manner to the object, concept, or situa
tion. "Attitude is more than a statement that something
is good or bad, desirable or undesirable, for judgment is
■^Ernest R. Hilgard, Introduction to Psychology (2d
ed.; New York: Harcourt, Brace & Co., 1957), p. 570.
; 3
accompanied by an affective response,"3 This definition is
also supported by Ruch who defines attitudes as "an emo
tionalized system of ideas which predisposes us to act in a ;
I certain way under certain conditions."3 The term "belief" |
! j
!is closely associated with "attitude^ j
j
I I
; Negative feedback.— Visual and auditory responses
i
! made while listening to a speech to indicate disagreement
with the position advocated by the speaker.
Posifive feedback.— Visual and auditory responses
made while listening to a speech to indicate agreement with
the position advocated by the speaker.
Negative position.--That group of subjects who were
initially opposed to the topic used in the study.3
3Ibid., p. 163.
3Floyd L. Ruch, Psychology and Life (5th ed.; Chi
cago: Scott, Foresman & Co. , 1958) , p. 305'.
4"Belief" is now more commonly equated with the
term "prejudice" as a sub-part of attitude. Hilgard has
defined "prejudice" as an "attitude that is firmly fixed,
not open to free discussion, and resistant to change" (p.
589). For further development of the relationship of "be
lief/prejudice," see Ruch, pp. 364, 602, and Hilgard, pp.
519-20. For a description of the historical development of:
the term "attitude" and its interpretations see Gardner
Lindzy (ed.) Handbook of Social Psychology (Cambridge: Ad-;
dison-Westley Publishing Company, Inc., 1954), I, pp. 43-
46 .
3Note that the term "negative" is used twice, once
to mean a type of feedback the subjects were overtly ex
posed to; second, to indicate the attitude or position
taken by the subjects in reference to the topic. In the
! Affirmative position.--That group of subjects who
1 .
iwere initially in agreement with the topic used in the
I
| study.
Limits of the Study
This study was limited by the facts that all sub
jects were undergraduate male students, chosen from begin
ning speech classes at one college, limited to five minutes j
iof speaking, and to one common topic.
; Additionally, the study utilized a highly contro- j
Iversial, emotional topic (race riots), deliberately chosen |
j
in preference to a more "academic" controversial subject |
: I
such as student parking problems, optimum class length time:
; * i
i
I
periods, etc., where opinions would probably not be so
deep-seated.
The study was limited to a one-to-one relationship
between the speaker and a listener; to the use of a male
Caucasian listener; and to the use of male Caucasian fresh
men and sophomore college students as speakers.
The study specifically was not concerned with such
items as delivery, gestures, choice of evidence, plans of
organization, or overt speaker adaptation behavior.
I course of the study the word "no" is used synonymously in
place of "negative" to indicate attitude wherever practica
ble to minimize the possibility of confusion in terminol
ogy.
j Signif i cance of the' S tudy j
This study was thought to be significant, first,
because of several possible implications for the teaching
iof speech. If the study showed substantial influence upon
i i
ia student speaker's attitudes, resulting from overt feed- j
‘ i
'back from a peer group listener, then the speech teacher I
t
should be aware of this influence and perhaps adjust to it j
|in his classroom procedures. Certainly the teacher should j
. i
not concentrate so intensely upon the effectiveness of the ;
!speaker in molding listeners’ attitudes as to overlook the
, i
Ifact that-meanwhile the audience behaviors are molding the |
speaker's attitudes. The speech teacher has an ethical I
responsibility in relation to everything that happens in
his class which affects the personal, civic, and moral at
titudes of his students. Furthermore, there was the possi- :
bility that the study's findings might suggest classroom
units or projects that would make the teaching of "listen
ing" more specific than is usually suggested in textbooks.
Finally, there might be practical implications regarding
the overt feedback from the teacher himself as he listens
to his students speak. A teacher's listening habits, such
as smiling or frowning, might be sufficiently important to
justify special research.
The study was thought to be significant, second,
because of its theoretical implications for the study of
human communication. A conceptual framework for such
6 |
I
researches was suggested by Dickens: i
Each man is physiologically an island. He experi
ences & myriad of sensations, perceptions, thoughts,
and feelings; he can communicate with himself (in vari
ous ways) about his experiences. He may assume that
other people have experiences similar to his but how
can he know? How can he bridge the physical and physi
ological gap that separates him from other humans? How|
do the thoughts and feelings of one person become inter-?
mingled with those of another? How does the "intra" ;
become "inter?" That is the central question in human ;
communication research.^ j
Dickens wentT~oh to compare and contrast intra- and inter-
;
individual communication as areas for research. The proc
ess of intra-individual communication can be directly ob-
i
served— but only by introspection of one person. The only
portion of inter-individual communication susceptible to j
|
direct verification by independent observers is the spoken |
(or written or pictorial, etc.) portion. Research on signs'
and symbols of several people, introduces the study of re
lationships, such as that of sequence. Another important
relationship is that of agreement-disagreement.7 It was
from this last suggestion that the present study originated;
Thus the significance of the study appeared to be
sustained becai.se it might contribute to knowledge concern-
!
ing inter-individual relationships between intra-individuali
processes.
^Milton Dickens, class lecture notes, University of
Southern California, Los Angeles, California, February 16,
1965.
7Ibid.
; In designing the study the experimenter included
|
janother set of variables that held promise of filling a
;theoretical research gap. An overwhelming majority of rhe-
i
Itorical and other communication researches has for centu-
iries focused on the effectiveness of speaker behaviors as
measured by audience responses. The focus of this study
:was upon the relatively neglected area of the effectiveness ;
of certain audience behaviors as measured by speaker
; i
‘responses. '
■ Another relatively neglected research area was also j
■ i
chosen. Most researches in speech and communication have !
! |
|dealt with verbal, vocal, and pictorial signs and symbols, j
e.g., tape recordings, manuscripts, motion pictures, etc. j
In this study the emphasis was upon facial expressions and
head gestures. These signs and symbols were considered to
be the ones most commonly used by listeners responding to a
speech. Furthermore the speakers experienced these stimuli
directly from the listener (rather than by use of pictures,
signal cards, or the like) because this direct experience
was considered to be more realistic.
Therefore, the significance of the study was sup
ported by both practical and theoretical implications. It
seemed reasonable to hope that a step might be taken toward j
filling several relatively neglected research areas. Addi-j
tional support for the significance of the study is given
in the following chapter.
j Earlier studies or oral communication were mainly
i
jconcerned with various types of speaker behaviors; the re-
i
|sponses of listeners were observed only as a measurement of
I speakers1 effectiveness. In recent years researchers have
;shown increasing interest in the nature and effects of lis
tener "feedback," a term and concept that was stressed by
.Shannon and Weaver in their information theory** and Wiener
in his cybernetics theory.** Wiener's concept of feedback
;included an analogy between servo-mechanisms in electronic
|and mechanical systems as compared with human systems,
I
j
li.e., the feedback serves the function of "correcting" the
|sender's behaviors. A broader concept of feedback was im-
;plied by Ruch who defined feedback as the "interchange of
;information on the part of human beings in a communication
i or problem-solving situation. " 1** A more detailed examina
tion of the concept of feedback is discussed in Chapter II.
This correction/exchange function may describe the
feedback at the machine level, but the writer doubted that
the concept was broad enough to explain the possible addi
tional functions of feedback at the human level. The pre
ceding statement was put to test in the present study.
^Claude Shannon and Warren Weaver, The Mathemati
cal Theory of Communication (Champaign-Urbana! University
of Illinois Press, 1949).
9Norbert Wiener, The Human Use of Human Beings
(New York: Doubleday and Co., Inc., 1954) .
l^Ruch, op. clt., p. 595,
I Some specific contributions to the field of speech j
I this study could provide were thought to be as follows:
i
; |
1. The study provides an indication of the effects,!
if any, of listener feedback on modifying
| speaker attitudes. It is basic to the field to
i
| know the ways in which speakers as well as lis-
I
teners react to feedback, i.e., is the communi-
i
cation situation truly circular in nature, and j
I
if so, to what extent? !
2. The study provides a resource for teachers of
1 j
speech to better enable them to evaluate the j
j performance of their students. j
3. The study provides to the writer's knowledge
the first source of empirical evidence specifi-
•cally concerned with listener feedback effect
on fact of, as well as direction of, speaker
attitude shift. _ _ .
4. This study provides further insight into how
speaker opinion is formed and offers additional :
criteria for the consideration of the research
er in speech.
5. The study serves as the next logical "stepping
stone" leading to further research in a hereto
fore relatively uninvestigated area of oral
communication.
! 10 (
Organization of' the' Remainder of the Study
| The remainder of the dissertation is organized into j
j
■four chapters, the bibliography, and two appendices. Chap
iter II deals with a review of previous literature. Chapter
;III is concerned with the design of the study. Chapter IV i
I presents and interprets the data. Chapter V includes sum-
jmary, conclusions, and implications. A bibliography of
i
:sources used in this study follows Chapter V. I
Appendix A is composed of the written observations j
iand reactions of the listener. Appendix B contains a pho- j
! j
;tograph of the listener, an initial opinion survey form, |
‘ i
: i
land faculty memoranda. I
CHAPTER II
REVIEW OF LITERATURE
Introduction
Doctoral research in speech began during the
1920's. Most of the early speech researches, graduate and
i
jpost-graduate, were historical or critical (non-quantita-
i
Itive) but during subsequent years the percentage of empiri-
;cal and experimental (quantitative) studies steadily
i
i
I increased.! From the beginning another characteristic of
1
I
;speech research has been the strong emphasis upon analyses
of speakers or speeches in terms of their effectiveness.
Sometimes effectiveness has been assumed and its causes
sought; sometimes effectiveness has been estimated or meas
ured in terms of audience response. Meanwhile, students in
psychology were exploring various aspects of human communi
cation which were either wholly or partially concerned with
listener response in relation to speaker stimulus.
1-Franklin H. Knower, "Graduate Theses: An Index of
Graduate Work in Speech," Speech Monographs, II (October,
1935), 1-51; "Graduate Theses," Speech Monographs, XXXII
(August, 1965), 336-385; Waldo W. Braden, The Communicative
Arts and Sciences of Speech, ed. Keith Brooks (Columbus,
Ohio: Charles E. Merrill Books, Inc., 1967), pp. 91-93.
11
I 12 |
| In recent years, however, speech researchers have
|joined with the psychologists in exploring the active role
!
|of the listener (or audience) during oral communication;
!this interest is reflected in the present study. Addition
ally, the relationship existent between intra-inter-indi-
Ividual communication was considered to be of special im- j
! j
jportance. The focus was upon the effects of listener be- j
havior as measured by a comparison of pre- and post-experi-I
;mental speaker attitudes. Therefore, the literature was
jreviewed with special interest in quantitative analysis of j
I :
' I
ilistener feedback and when applicable to intra-inter-indi-
j |
:vidual relationships. Researches most relevant to the
;present study are discussed in the remaining sections of j
this chapter.
The following indexes and bibliographies were con
sulted; ~
1. Graduate Theses: An Index of Graduate Work in '
Speech, 1902-1965, Franklin Knower (ed.), Speech
Monographs, II-XXXII (1935-1965).
2. Abstracts of Theses in the Field of Speech, 1946-
1965, Clyde W. Dow (ed.T, Speech Monographs, XIII-
XXXII (1946-1965).
3. Doctoral Dissertations in Speech: Work in Prog
ress, 1963-1965, J. Jeffery Auer (ed.), Speech Mon
ographs, XXX-XXXII (1963-1965).
4. Doctoral Dissertations Accepted by American Uni
versities , 1933-1955.
5. The Index of American Doctoral Dissertations,
1957-1965.
6. Dissertation Abstracts: An Index to American Doc-
13 j
toral Dissertations, 1957-1965. !
7. The Index to the Quarterly Journal of Speech,
1915-1965.
8. The Index to Speech Monographs, 1934-1965.
9. The Index to The Speech Teacher, 1952-1965.
10. The Index to Western Speech, 1936-1965.
11. The Index to The Southern Speech Journal, 1935-
1965.
12. Psychological Abstracts, 1940-1965.
Survey of Related Literature
In recent years numerous experimental studies have
been devoted to the oral communication situation. Some of
this effort has been directed toward an examination of
feedback, its effect upon overt behavior, and attitude
change. Studies representative of these areas are consid
ered in the following pages.
However, where the effect of feedback has been con
sidered in relation to attitude changes, in the majority of
the studies the dependent variable has been the audience
attitude rather than the speaker attitude. It was hypothe
sized, then, that the speaker's attitude could be changed
by audience reaction. Such studies, however, did not ap
pear to be numerous enough to fully develop this hypothe
sis. Some of these studies are discussed later in this
chapter.
Feedback has generally been operationally defined
|as a modifier of overt speaker behavior during the speaking ;
jprocess, i.e., as an indication to the speaker to modify
|his gestures, vocal variety, or content of the speech while
I speaking. Typical of those using this concept of the func
tion of feedback were Lewis and Nichols, who said that:
The feedback which the persuader receives from his
listeners helps him effect a new analysis of his audi
ence as he proceeds from point to point in the process
of the talk. If the attitude of the audience overt be
havior is perceived correctly, the speaker can then
make the adjustments in the presentation which will
best advance the purpose of the speech.2
i Berio also considered feedback to be primarily use
ful for modifying the overt behavior of the speaker during
ithe speaking process. j
For example, you are giving a talk . . . At one point j
in your talk you tell a joke. The audience is supposed j
to laugh. If they laugh, it tells you that you were
successful. It tells you to keep going, your messages
are having an effect.2
It should be noted, however, that while the current |
study was not concerned with modifying speaker behavior
while speaking, the use of overt audience reactions (feed
back) was necessary. As Clevenger said:
The overt feedback of an audience is used for in
ferring process responses to communication. Facial
expressions and the like are used to infer whether they
are interested, find the message agreeable, or under
stand what is being said. The basis for such inferences
2Thomas R. Lewis and Ralph G. Nichols, Speaking and
Listening (Dubuque: William C. Brown Company, 1965), p.
: 163.
2David Berio, The Process of Communication (New
York: Holt, Rinehart & Winston, 1960), p. 113.
I is the assumption that there exists (at least in our
culture) fairly uniform connections between internal
! states and certain behavioral manifestations. For ex-
j ample . . . an auditor who scowls and draws his lips in
| a thin line is thought to be angry; one who shakes his
| head, disagrees or fails to understand . . .
As indicated in^Chapter I in this study, feedback
i is now generally considered to be more complex than the
servomechanism self-correction approach first associated
Iwith it.
One of the most intensive efforts utilizing the
i
|preceding feedback concept was that of Powers, Clark, and
i
1 R .
jMcFarland. Their two-part theoretical consideration of
:feedback essentially conceived it to be basically a self-
'correcting process. They extended this function of feed
back to include six levels of human behavior, beginning
with simple muscle reflexes and extending through communi
cation (fifth level) to the construction of "systems" such
as governments and symphonies.® Perhaps the best summary
of their concept of feedback's function is in relation to
problem solving where they noted that "... [feedback]
merely alters the properties of a system, thus changing the
^Theodore Clevenger, Jr., Audience Analysis (Indi
anapolis: 'Bobbs-Merrill Co., Inc., 1966), pp. 62-63.
^W. T. Powers, R. K. Clark, and R. L. McFarland, "A
General Feedback Theory of Human Behavior," Perceptual and
Motor Skills, XI (February, December, 1960), 71-88, 309-
323.
®Ibid., p. 316 .
transformations applied to existing information."7 I
Powers, Clark, and McFarland's application was to j
this writer's knowledge, the most extensive effort as of
! !
iits date to expand and adapt the servomechanism function of
;feedback to all levels of human behavior. Their study pro
vided excellent background information on one concept of
feedback function.
A variation of those studies concerned with the use j
;of feedback as a factor influencing overt speaker behavior j
jwas that of Amato and Ostermeier who used colored cards to j
I provide positive, neutral, or negative feedback.^ m this
I experiment the audience held up red cards to indicate disa
greement (negative feedback), green cards for no opinion
I(neutral feedback), and white cards for agreement (positive
feedback). The audience had been previously instructed to
provide specific types of feedback. The results of the
study indicated that negative feedback, i.e., a preponder
ance of red cards, caused a significant deterioration of
the speaker's delivery.
In the current study, feedback was defined as overt
visual and auditory responses indicating agreement or disa
greement with the speaker while listening to a speech.
^Ibid., p. 317.
^Philip P. Amato and Terry H. Ostermeier, "The Ef
fect of Audience Feedback on the Beginning Public Speaker,"
The Speech Teacher, XVI (January, 1967), 56-60.
17
This use of feedback as an overt manifestation agreed with
the definitions previously cited. As indicated previously,
however, the focus was upon determining whether these overt
behaviors defined as feedback effected changes in the atti
tudes of the speakers rather than an attempt to evaluate
feedback's effect as a force causing the speaker to adapt
his speech to the audience while speaking.
Typical of the literature dealing with feedback and
attitude was that of Lana.^ In this study, Group "A" was
presented with persuasive messages in a pro- con- order and
Group "B" with con- pro-. This experimenter's significance
in respect to the current study was the reported high cor
relation between interest and attitude change, the implica
tion being that attitude change and interest increase pro
portionally. While Lana's study and the writer's both
dealt with attitude change, the former was not concerned
with feedback influence on the speakers' attitudes.
One common exploration of feedback has been in the
area of information transmission and learning. Leavitt and
Mueller, for example, investigated the relationship of
feedback to the accuracy of descriptive oral
^Robert E. Lana, "Interest Media and Order Effects
in Persuasive Communications," American Journal of Psychol
ogy, LVI (1963), 9-13.
18
communications.10 Their question was:. "How is the trans
mission of information from 'A' to 'B' influenced by the
return of information from 'B' to Their methodol
ogy provided three degrees of feedback: (1) zero feedback
where the speaker was out of sight of the Ss; (2) limited
feedback where the communicator was in view of the Ss and
answered only "yes" or "no" to questions asked by the Ss;
(3) free feedback where the speaker was also in view of the
Ss and answered in detail any questions asked. In all
three groups geometric patterns were described by the
speaker; then the Ss attempted to draw pictures of these
designs from the oral descriptions. Leavitt and Mueller
found that significantly greater accuracy in reproducing
the geometric patterns was obtained with the free feedback
group.
Related findings were uncovered by Zajonc who found
that an individual's reaction time improved when the indi
vidual was made aware of his failures (and theoretically
exposed to feedback) and that the individual's reaction
time improved again when knowledge (feedback) about his as
well as the other participants' performance abilities was
lOHarold J. Leavitt and Ronald A. H. Mueller, "Some;
Effects of Feedback on Communication," Human Relations, IV
(1951), 401-410.
H-Ibid. , p. 401.
! 19 !
jmade freely available to the entire group.-**2
i The previous two studies utilized feedback primar
ily as a manipulated variable and are representative of one
! concept of feedback function. They provided greater in-
:sight into the concept of feedback which broadened the
I !
foundations upon which the current study was based. |
Feedback's function has also been examined in stud- ;
j
ies involving the combined influence of group discussion and
jindividual speeches on.attitude in oral communication situ- !
iations. Lewin, for example, demonstrated that group dis
cussion following a lecture produced attitude changes among ;
'all those involved.H
Similar results were obtained by Coch and French,
who found that resistance to new concepts was significantly
decreased and attitude change brought about by participa
tion in group discussion.14 Thus it would appear that dis
cussion in groups affects speaker/participant attitudes.
Whether this concept holds true in individual-to-individual
communications has yet to be fully demonstrated.
l2Robert B. Zajonc, "The Effects of Feedback and
Probability of Group Success on Individual and Group Per
formance , Huma^^ XV (May, 1962), 149-161.
-*-3jCurt Lewin, "Group Decision and Social Change,"
Readings in Social Psychology, eds. Guy E. Swanson, Theo-
dore Newcomb and Eugene Hartley (New York: McGraw-Hill
Book Co.,1952), pp. 457-474.
l^Lester Coch and John R. P. French, "Overcoming
Resistance to Change," Human Relations, I (1948) , 512-532.
More closely related to this researcher's tech
niques and procedures was that of Miller, et al.^5 This
study indicated that carefully manipulated experimenter
Iresponses could influence overt speaker behavior. Miller
ifound that by saying "good" every twenty seconds during the
;speech of a student confederate and then omitting this re- i
Enforcement in speeches of students who heard and then fol- ;
: !
lowed the confederate produced disruptive patterns of ver- j
jbality in those Ss who had the positive reinforcement with- j
jheld. Apparently just the withholding of praise by the j
| [
experimenter in the presence of peers produced anxiety and j
: i
^resulting disruptive performance. A similar disruptive j
1 ' j
effect on verbal proficiency (behavior) was also encoun
tered by Vlandis, who alternated rewards, i.e., comments of
"good," with punishment, i.e., comments of "no," during the
course of the subject's speech.it would be valuable to
know if the instructors' verbal responses in the foregoing
experiments influenced the feedback from the student audi
ences. Additionally, the question might be raised as to
whether disruptions of speakers' delivery were brought about
by changes in their, attitudes toward their topics.
l^Gerald R. Miller, et al., "The Effect of Differ
ential Reward on Speech Patterns," Speech Monographs,
;XXVIII (March, 1961), 9-15.
^6John W. Vlandis, "Variations in the Verbal Be- !
havior of a Speaker as a Function of Varied Reinforced
Conditions," Speech Monographs, XXXI (June, 1964), 116-119.
| One interesting study investigating the influence
j of refutative feedback upon attitude change was that of
|Thistlethwaite, Kemenetzky, and Schmidt.^ The main focus
i
:of their study was to determine whether implicit or explic-
:it refutation was more effective in reducing intial atti-
• ' f
tudes. Initial attitudes were determined through essays
I written by the Ss one month before the experiment. A dif- !
; |
ferent topic was used for each of the two test groups in- ;
; i
Ivolved. Messages were then recorded and played back to the;
i I
jSs using explicit refutation for one group and implicit ■
refutation for the other. As the Ss listened to the re- j
corded messages, and indicated their responses by pushing
one of two buttons which recorded either a "like" or "dis
like" reaction. The results indicated that explicit refu
tation produced more "dislike" responses than implicit
refutation. Since the Ss recorded only "like" or "dislike"
reactions during the communications, it is possible that
the actual effect of feedback may have been minimized
through the restriction of limiting them to only two
responses.
Two experiments by Scott tended to produce related
l^Donald L. Thistlethwaite, Joseph Kemenetzky, and
Hans Schmidt, "Factors Influencing Attitude Change through
Refutative Communications," Speech Monographs, XXIII
(March, 1956), 14-25.
I attitude shift findings.He found that a previously im- i
|planted psychological set in the speaker which categorized
|the potential audiences as being either hostile or favor-
jable to the topic produced limited attitude changes in the
|speaker facing the friendly audience. In the second study
Scott "dramatized" the reward or punishment of the speaker
at the end of each message by publicly announcing the sup
posed results of the audience's change in conviction. This ;
:procedure itself produced an attitude change in the !
i
! !
;speakers who were thus "rewarded" with "success" in persua- |
! I
jsion. In both studies, however, some subjects were re
quired to defend topics they disagreed with. Additionally,
iwhile the experimenter "made known" the supposed audience
1
!position in advance of the delivery of the speech or at the j
conclusion, no effort was made to manipulate actual audi
ence reactions during the delivery of the speeches. In
fact, the feedback the speaker received from the experi
menter may have contradicted that from the audiences. In
summary, although attitudes shifted toward the advocated
position in the subjects who anticipated facing "favorable"
audiences, and in those speakers who "won," there was no
attitude shift in those who faced "hostile" audiences or
■^William A. Scott, "Attitude Change by Response
Reinforcement: Replication and Extension," Sociometry,
XXII (19 59), 328-335; and "Attitude Change Through Reward
of Verbal Behavior," Journal of Abnormal and Social Psy
chology , LX (1957), 72-75.
j"lost." It was thus difficult to determine to what extent
speaker's conviction shifts were due to uncontrolled audi-
I i
! I
jence feedback, group pressure of having "won" or "lost," or '
I ' I
|as Miller's study suggested, a psychological set produced
by anticipation of a favorable audience reception.
Hovland, Janis, and Kelley, in their studies of j
; I
;self-reinforcement, attempted to measure speaker attitude
'.shifts caused by the influence of group pressure or "public j
commitment."19 They felt that "the act of communication
with others should increase the speaker's own convic-
t
i i
;tions."2° Or, in other words, once a speaker takes a pub- ;
ilic stand on an issue, advocacy, especially if in an im- |
j i
promptu situation, should increase the Ss conviction toward j
;the position supported. Hovland forced speakers to impro
vise new supporting arguments during a speaking situation.
No effort was made, however, to consistently vary the
causes of the improvisations. Attitude change was deter
mined by the number and frequency of personal references to
himself made by the subject while he was speaking. Thus,
in effect, what was actually measured was a type of behav
ior during the communication process rather than attitude
changes produced by feedback reactions.
l^Carl I. Hovland, Irving L. Janis, and Harold H. j
Kelley, Communication and Persuasion (New Haven: Yale Uni- :
versity Press, 1953), p. 236.
20Ibid.
! 24 |
: I
Similar results were obtained by Verplanck who used |
|feedback to strengthen existing attitudes in his Ss.21
jVerplanck attempted to determine whether positive or nega
tive feedback was more effective in reinforcing predeter
mined initial attitudes. The study was related to the cur-
jrent one in that a one-person-to-one-person communication
isituation was utilized. Feedback was provided by active
! j
; participation in the communicative situation through agree- ;
;ment with, paraphrasing of what the S said, and by remaining
;silent. Thus a multi-variable inter-personal reaction of j
refutation and support was used in contrast to the more
! strictly controlled person-to-person overt feedback provided
:in this writer's experiment. The number of personal opin- j
ion statements made by the Ss was used as the technique for
determining the degree of strengthening of initial atti
tude. (Rather than a pre- post- shift of opinion ballot
technique.)
Verplanck concluded that positive feedback, i.e.,
paraphrasing and silence, produced a greater strengthening
of initial attitude. Verplanck, however, was concerned
primarily with strengthening initially held attitudes
rather than attempting to determine attitude shifts per se.
His findings while contributing to attitude research left
21william S. Verplanck, "The Control of the Content!
of Conversations: Reinforcement of Statements of Opinion,"
Journal of Abnormal and Social Psychology, LI (November,
1955), 568-676.
unanswered the essential question of this writer's study, i
■i.e., the fact of shift and the direction of shift due to
i
j
jfeedback's effect.
I
A study related to Verplanck's was that of Levonian
who used motion pictures with a central theme drawn from
the predetermined attitudes of the Ss.^2 Levonian was !
I primarily concerned with strengthening of initial atti
tudes. His methodology involved the use of one film spe-
i
icifically tailored to reinforce the initial attitude of one
| |
jgroup; and a second non-tailored film, i.e., a random j
:selection of attitude items was used for the second group. '
■The films provided the feedback. As in Verplanck's study,
’initial attitudes were significantly strengthened by the
: i
tailored film as opposed to the non-tailored film. How
ever, there is some doubt that Levonian used feedback as it
has generally been defined, i.e., mutual interaction while
communication is taking place. In other words, since the
source of the feedback, i.e., the films, could not react to
the audience, the communication situation primarily was
more unidirectional than circular in nature.
Kelman used feedback to try to determine the factors
22Edward Levonian, "Opinion Change as Mediated by
Audience Tailored Film," Audio-Visual Communication Re-
view, XI (1963), 104-113.
which cause conformity.^3 Essentially what concerned him
jwas what would cause a person who superficially appeared to
!
|conform, to change his attitude to actually agree with what
|he previously had only paid "lip service" to. The study
jinvolved the use of seventh-grade students and their atti
tudes toward concepts displayed in selected comic books.
Some of the concepts concerned were expediency, attitudes
toward stereotyped heroes and villains, etc. Three groups
|were used: (1) a control group with no reward comments by |
the experimenter; (2) a group that was led to believe that |
Ithey would have to compete for a limited number of rewards; |
; !
land (3) a group that was led to believe that rewards would !
;be easily obtained. The rewards were to be based upon j
agreement with attitude comments expressed by the research
er. Initial attitudes were determined by having the Ss
write an essay prior to the experiment. Feedback was in
the form of the comments of the researcher to the group
based upon their initial attitudes.
Kelman found that the limited reward group produced I
the greatest amount of conformity. He hypothesized that
greater incentive due to competition for rewards was re
sponsible for these results. Feedback in this study served
as a technique for testing other variables, i.e., compe-
23|jerbert Kelman, "Attitude Change as a Function of
Response Restrictions," Human Relations, VI (August, 1953),
185-215.
jtition, rewards, and their influence on conformity, rather i
|than as a variable under test itself. i
i
| One of the most interesting studies related to
|feedback and speaker attitude change was that of Ring and
i i
iKelley.24 They found that "high praise for the correct j
answer" and mild censure for the "wrong one" produced no
significant shift of opinion in their subjects. However,
;they also varied their "plausibility standards," i.e.,
: criteria for evaluation which were perceptible to the sub- j
! i
ijects, for punishment and reward during the course of the j
i ' j
iexperiment. The effect of "apparently" inconsistent behav-
jior on the part of the experimenters, i.e., varying the re-
i
|inforcement criteria may have influenced the subjects' at
titudes. This procedure plus the possible effect of the
experimenter's "authority image" could have introduced ad
ditional independent variables which might have signifi
cantly influenced their findings.
The final study considered was that of Jones.2®
This was the thesis most closely related to determining
24Kenneth Ring and Harold H. Kelley, "A Comparison
of Argumentation and Reduction as Modes of Influence,"
Journal of Abnormal Psychology, LXVI (1963), 95-102.
25Dorwin Cartwright and Alvin Zander (eds.), Group
Dynamics (Evanston: Row, Peterson and Co., 1953), pp~ 48-
51; 585-612.
26stanley E. Jones, "Attitude Changes of Public j
Speakers during the Investigative and Expressive Stages of
Advocacy" (unpublished Doctoral dissertation, Northwestern
University, 1964).
^speaker attitude shift before/ during, and after the deliv- '
■ery of a persuasive speech. Basically, Jones attempted to
j I
i i
jevaluate the effects of several possible contributors to
I
!speaker attitude change, or "self persuasion" as he termed
I
;it. His study was divided into two main areas of inquiry:
| (1) the "investigative" or speech preparation stage, and
i(2) the "expressive" or delivery stage.27 The methodology i
used indicated that two separate experiments were engaged
in, extending over two academic quarters. Additionally,
iJones constructed his own type of forced choice test. All
| ;
subjects were pre-tested to determine their attitudes to-
i i
Iward a topic assigned by the experimenter. A different
topic was used in each experiment. Each subject was j
assigned to the side of the topic opposite the one indi
cated in his pre-test. The next step in methodology was to
assign each S forty-five cards containing material both pro
and con pertaining to the particular topic. After distrib
ution of the cards, the subjects were allowed forty-five
minutes to select six cards for use in preparation of a
speech supporting the side of the issue they were assigned.
Following the selection of these cards all participants
were tested to determine if an attitude shift had taken
place. This test completed the experiment for the first
group.
27ibid., pp. 6-7.
I 2 9 !
i The group in Jones’ second experiment, which took
Iplace during the following quarter, utilized the same pro-
i
!
jcedure as group one, including the post-investigative test,
i
I
jbut was then allowed an additional thirty-five minutes to
prepare a ten minute speech based upon the cards selected.
I |
iEach subject was told that the speech would be heard pub- |
|licly, but that for various reasons they were to tape re- I
j
cord the speech in private. Upon completion of the tape |
irecording, each subject was again tested for retention of
! i
jprevious attitudes and any additional attitude shifts. In j
both groups careful effort was made to avoid the "contami
nating effect" of listener feedback.2^ Jones felt that it j
was impossible to accurately or consistently control "the |
possible effects of feedback."29
Essentially Jones' findings were that a significant
attitude shift in favor of the advocated position took
place after the investigative stage in all subjects. In
those students who were allowed to deliver their speeches,
no significant additional positive changes in attitude over
those obtained in the investigative stage were found.
There appeared to be a nonsignificant conviction decrease
after the expressive state. The conclusion was that the
investigative stage produced a greater shift in speaker
attitude than did the expressive stage. Although Jones did ;
28Ibid., p. 96.
29lbid., p. 67.
! 30 !
not include feedback as part of his methodology, it was
ifelt that his findings provided an especially valuable con
tribution to attitude shift during the intra stage of human (
: communication as well as an especially useful foundation j
for the current study. [
i
Cognitive Dissonance
Of special interest relative to the theoretical
iorientation of this study, was Festinger's theory of cogni- i
jtive d i s s o n a n c e .30 Festinger defined cognitions as "the j
i
! !
|things a person knows about himself, about his behavior and
ihis surroundings."31 He hypothesized that when a person is
'confronted with inconsistent cognitions he will experience !
! i
; t
^psychological discomfort which will motivate him to active- ]
ly try to avoid conditions and information that produced
and/or increased the unpleasant feelings, i.e., the person
will try to reduce dissonance and achieve consonance.
The above theory was relevant to the intra- inter
concept (described in preceding chapter) of Dickens and his
students in the following ways. Bowman found that one
basic and identifiable relationship between the remarks by
several people during a group discussion was that of
3C>Leon Festinger, A Theory of Cognitive Dissonance
(Evanston, Illinois: Row Peterson & Co., 1957).
3'1’ ibid. , p. 9 .
! 31 I
i
32
sequence. ^
i Fear reported highly significant accuracy when Ss
!■
•'were asked to distinguish between typewritten excerpts of
I !
: t
I individual speeches as compared with excerpts containing j
I i
'consecutive remarks by several persons (all other vari- j
sables were held constant).33 The Ss were also asked to
;indicate the clues by which they identified the individual
:versus multi-individual excerpts. Fear discovered that the j
imost commonly reported clue was the agreement-disagreement
Relationship among the ideas verbalized in the typescripts.
Krueger found that undergraduate speakers succeeded
|remarkably well in identifying audience attitudes of agree-
i
i
jment, disagreement, or indecision during the progress of a
:five-minute persuasive speech before an undergraduate speech
class.34
The above three studies appeared to demonstrate
that samples of inter-individual communication can be
32predrick B. Bowman, "An Experimental Study to
Determine the Amount of Inter-active Thinking Displayed in
a Problem-solving Group Discussion" (unpublished Doctoral
dissertation, The University of Southern California, Los
Angeles, California, 1959).
33Arthur J. Fear, "An Experimental Study of the
Ability of Lay Judges to Distinguish between Typescripts of
Individualized Development and Group and Discussion Idea
Development" (unpublished Doctoral dissertation, The Univer
sity of Southern California, Los Angeles, California,
1966).
34David K. Krueger, (unpublished Doctoral disserta
tion in process, The University of Southern California, Los:
Angeles, California, 1967).
32 |
analyzed, identified, and characterized by having groups of j
[independent observers look for relationships among the ;
I
jideas in those samples. Among the most important of these
{relationships was that of agreement-disagreement. This
{relationship was recognizably different in samples of what
;Fear called "individual idea development" when compared I
with "group idea development." ;
Festinger's theory is restricted to intra-individ
ual behaviors, i.e., when one person is in conflict within '
himself he is motivated to resolve this dissonance and seek |
; . !
consonance. It was hoped that the present study might j
:throw some additional light upon this theory because the
intra-individual attitudes of the speaker were to be meas- ;
ured both before and after receiving positive or negative
feedback from the listener.
However, the present experiment was focused upon
the inter-individual agreement and disagreement (not just
the intra-). Thus the study might explore a quite differ
ent type of "dissonance" and "consonance." Theoretically
it seemed likely that each of two individuals might be com
pletely consonant within himself yet at the same time be in
complete disagreement with the other person. The motiva
tional effects of inter-individual disagreement might pro
vide an interesting contrast to those of intra-individual
dissonance.
In addition to the above theoretical considerations,
cognitive dissonance theory was' related to this researcher’
study through Festinger’s prediction that one common way of
reducing dissonance is by avoiding exposure to counter
arguments, e.g., negative feedback. Therefore, dissonant
material should induce greater changes in attitudes than
consonant material. Specifically, negative listener feed
back should produce a greater shift of opinion than posi
tive feedback.
An exhaustive examination of cognitive-dissonance
theory and application was reported by Brehm and Cohen.35
Their text considered more than 125 studies by many re
searchers who applied Festinger’s theory to a variety of
research areas including some specifically directed toward
the oral communication situation.
Chapters VII and XIV are of special concern to
researchers in oral communication theory because of their
focus on methodology and communication.
One effort to apply cognitive dissonance theory to
the oral communication situation was Mentzer’s who found
that the establishment of a pro or con audience set in the
speaker influenced the communicator's attitude.36 His
35jack W. Brehm and Arthur R. Cohen, Explorations
In Cognitive Dissonance (New York: John Wiley and Sons,
1962).
■^Ray t . Mentzer, "Cognitive Dissonance and Atti
tude Change in a Communicator: The Effects of Audience
Image and Communication Discrepancy” (unpublished Doctoral
thesis, Purdue University, Purdue, 1962).
3 4 :
findings indicated an attitude shift toward the communi
cator's initial attitude when speaking to the "pro" audi
ence. But no shift in either direction when speaking to
the "con" audience. Perhaps the lack of inclusion of man
ipulated audience feedback may have contributed to these !
findings.
Slightly different results were obtained by Gibson,
whose findings suggested that:
Persons in states of cognitive dissonance ap
proached significantly greater changes in attitudes to
ward the direction of the communication than did sub
jects in states of cognitive consonance.37
In this study Gibson measured audience attitudes after they
heard messages judged (by Gibson) to produce cognitive dis
sonance or cognitive consonance. However, the judgment of
but one person seemed to be a questionable procedure for
establishing the validity of these stimulus materials; per
haps a pilot study using several independent judges would
have strengthened the reader's confidence that dissonant
and consonant materials were what they purported to be.
Additionally, controlled audience feedback per se, as in
other studies of cognitive dissonance, was not considered
in determining the results of the study.
Essentially, Festinger's theory is of interest
37James W. Gibson, "Direct and Indirect Attitude
Scale Measurements of Positive and Negative Communications"
(unpublished Doctoral thesis, Ohio State University, Colum
bus, Ohio, 1962), pp. 82-83.
35 j
because of its interpretation of intra- inter-individual
oral communication theory as well as its possible applica-
tory use of cognitive dissonance (positive and negative
feedback) on attitude shift.
The sometimes conflicting findings of studies deal
ing with applications of Festinger's theory to oral communi
cation situations was closely examined by Cronkhite, who
probably best summed up the current status of the merits of
the cognitive dissonance concept when he said:
. . . while recent analysis and experimentation has
neither proven or disproven the theory of cognitive
dissonance, it may serve to focus future research upon
questions more basic to the theory.-^®
Attitude Measurement Instruments
The accuracy of any experimental study is dependent
upon the suitability of the measuring instruments chosen.
In selecting an instrument for use in the present study, the
following procedures were used.
First, an effort was made to find an adequate defin
ition of attitude to aid in the selection of the measuring
device. McNemar commented that "The common element of most
definitions of social attitudes is . . . that an attitude
is a readiness or a tendency to act or react in a certain
SSQary Lynn Cronkhite, "Toward a Real Test of Dis
sonance Theory," The Quarterly Journal of Speech, LII
(April, 1966), 178.
36 i
manner."^ This definition was found to be consistent with j
those offered by other sources previously cited and pro- i
vided a working base to proceed from.
The next step was to determine criteria necessary ;
for a suitable instrument applicable to this study. The j
three factors considered most important were validity, re- ;
f
liability, and unidimensionality.^0 Definitions of the
first two terms apply to all measuring devices and require
no further explanation. The third criterion as applied
specifically to this study means the measurement of a
single variable as opposed to instruments which sometimes
measure more than one variable.
Attention was directed toward those attitude-deter
mining instruments which appeared to best fit the above
criteria. The fix'st to be considered was the semantic dif
ferential advocated by Osgood.While this device has
been used in measuring some characteristics of an individ
ual's attitude and personality structure, its primary pur
pose has been to assess similarities among the meaning of
■^Quinn McNemar, "Opinion-attitude Methodology,"
Psychological Bulletin, LXIII (July, 1950), 365-373.
^Milton Dickens and Lee Edward Travis, "The Experi
mental Approach," An Introduction to Graduate Study in
Speech and Theatre, ed. Clyde W. Dow (East Lansing: Michi
gan State University Press, 1961), pp. 206-207.
^C. E. Osgood, C. J. Suci, and P. H. Tannenbaum,
The Measurement of Meaning (Urbana, Illinois: University
of Illinois Press, 1957).
words.This usage plus the doubtful ability to success
fully adapt it to the writer’s study resulted in its
rejection.^ ~
The second attitude determination device considered
was the indirect method technique. In discussing this
method of measuring attitudes, Wechsler and Bernberg com
mented that its prime asset was "the concealment of the
intent of the measure from the subject, thus permitting him
to comment without becoming personally involved," and thus
bias his responses.^
However, this obvious asset was outweighed by two
disadvantages. These liabilities, according to Sanford,
were: first, that the indirect type instrument's value was
severely handicapped by the time involved in application;
and second, that its susceptibility to unpredictable vari
ables existent in the communication situation intensified
"its tendency toward subjectivity in interpretation."^
A third factor that might have been considered was
the complex nature of the test which would probably require
^0. J. Harvey (ed.), Motivation and Social Inter
action (New York: The Ronald Press Co., 1963), p. 270.
^Berlo, op. cit. , pp. 246-248.
^^Irving R. Wechsler and Raymond E. Bernberg, "In
direct Methods of Attitude Measurement," International
Journal of Opinion and Attitude Research, IV (1950) , 211.
^Filmore h . Sanford, "The Use of a Projective De
vice in Attitude Surveying," Public Opinion Quarterly, XIV
(1950), 697-698.
an extensive knowledge of applied and theoretical psychol
ogy to assure reliability in interpretation.46
Therefore, while the indirect test had many merits,
especially in regard to concealing the purpose of the study,;
matters of practicality and questionable suitability were
the main deterrents to its adoption for use in this study.
The third instrument considered was the Thurstone i
Type Attitude Scale.47 However, the length of time re- ,
quired during administration (where application time was
significant in the study's methodology) plus its difficulty
in scoring tended to discourage its use.48 Additionally,
the Thurstone scale is more suited to measuring strength of
attitude shift, rather than fact of shift per se, and di
rection of shift.49
The specific type of test chosen was the Woodward
Ballot,50 This instrument was the subject of some contro
versy extending through the 1930's involving three main
figures, William A. D. Millson, Ernest R. Henrikson, and
4 6Robert S. Woodworth and Donald Marrquis, Psychol
ogy (New York: Henry Holt and Co., 1953), 108-111.
4 7l . L. Thurstone and B. J. Chave, The Measurement
of Attitude (Chicago: University of Chicago Press, 1929),
p. 96.
48Lindzy, op. cit., pp. 341-369.
49Gibson, op. cit., p. 90.
50h . S. Woodward, "Measurement and Analysis of
Audience Opinion," The Quarterly Journal of Speech, XIV
(1928), 94-111.
: 39
Alan H. Monroe. 53-
Woodward put forth his ballot originally as a means
of measuring shifts of attitude in an audience exposed to a
idebate. In a sense, it was to be a measure of the "effec
tiveness" of the speakers. Much of the resulting contro
versy centered around this concept of measuring speaker
"effectiveness." One of Woodward's concerns was the con
struction of an instrument which would allow an audience to
"judge" a debate in lieu of a "professional" critic judge
c 9
who might be partial to one team. ^ The implied rationale
for this wish was to make debates and debaters more moti
vated (and hopefully knowledgeable) regarding the topics
■^William A. D. Millson, "Experimental Work in
Audience Reaction," The Quarterly Journal of Speech, XVIII
(February, 1932), 13-30; "Audience Reaction to Symposium,"
The Quarterly Journal of Speech, XXI (February, 1935) , 45-
53; "Measurement of Speech Values," The Quarterly Journal
of Speech, XXII (December, 1936), 544-553; "Problems in
Measuring Audience Reaction," The Quarterly Journal of
Speech, XVIII (November, 1932), 621-637; "A Review of Re
search in Audience Reaction: Part I: Experimental Research
with Non-student Audiences," The Quarterly Journal of
Speech, XXIV (October, 1938), 4 64-483; "A Review of Research
in Audience Reaction: Part II: Misconceptions of the
Function of the Woodward Ballot and Formulae," The Quarter
ly Journal of Speech, XXII (December, 1938), 655-672; E. H.
Henrikson, "The Audience Reaction Ballot: An Evaluation,"
The Quarterly Journal of Speech, XXIV (February, 1938) , 54-
61; A. H. Monroe, "The Statistical Reliability and Validity
of the Shift of Opinion Ballot," The Quarterly Journal of
Speech, XIII (December, 1937), 577-585; "Testing Speech
Performance," Bulletin of the National Association of Sec
ondary School Principals, XXIX (November, 1945) , 156-164.
52^oodward, op. cit., pp. 98-99.
being debated. !
j
Woodward had used a "weighting" system of assigning!
I
arbitrary numerical values of plus one, zero, and minus one I
to shifts of opinion toward the affirmative, remaining
!
neutral, and toward the negative. In a series of articles,!
Millson reported the use of Woodward's ballot as an instru-;
ment to determine the "effectiveness" of discussion versus j
i
debate and the "effectiveness" of various types of deliv
ery. He also attempted to increase the accuracy of the
interpretation of the raw data through development of a
more complex formula for treating the d a t a . 54
Millson was also strongly concerned with making
classroom speeches more like a "real life" situation.
Probably Millson tended to use the ballot beyond what Wood
ward had originally intended. These extensions formed the
nucleus for the controversy which later developed.
The main critic of both the original Woodward Bal
lot and Millson's applications and formula of it was Hen-
rikson. His primary attack was directed at Millson's spe
cific applications of the ballot, the use of the formula
devised by Millson and the latter's efforts to equate the
classroom's speaking situation with "real life" speaking
^Millson, "Audience Reaction to Symposium," loc.
cit.
54jyiillson, "Measurement of Speech Values," loc.
cit.
41 ;
situations .'55 I
For example, in his criticism of the ballot, per se,i
t
'Henrikson did not directly attack validity, in fact by pro- j
posing a modification he appears to have accepted its basic j
validity. He conceded, however, that "no ballot can hope
to record fine changes in audience opinion attitude in the
way that a fine scientific instrument can. . . .56
Additionally, he criticized the scoring of the "un
decided" category in both the pre- and post-speaking por
tions which would tend to be a criticism of reliability and
related to his criticism of Millson1s formula. He also felt
that the ballot should have a provision for a "weakening of
opinion," and proposed such a modification.57 His proposed
inclusion of a "weakening of opinion" would appear to in
crease the complexity of the ballot and its scoring. Addi
tionally, the detailed instructions given to the Ss per
taining to marking the initial opinion ballot with its
additional items could have served to draw attention to an
experimenter's desired response and therefore probably did
not answer his criticism of Millson's focusing attention
upon the purpose of the ballot.^8
The third major participant in the Woodward Ballot
55Henrikson, loc. cit.
56ibid., p. 49. 57Ibid., p. 51.
58Henrikson, op. cit., pp. 50, 53-55.
controversy was Alan H. Monroe, who undertook to establish
both the accuracy of the ballot, per se, and Millson's
formula for interpreting the raw data derived from it.
Monroe conducted his own experiments utilizing the
Woodward Ballot and the "Millson Formula" for interpreting
the raw data obtained. He also attempted to prove the ac
curacy of the "Millson Formula" by mathematical analysis.
He concluded that both the ballot and the Millson formula
were accurate tools for use in the measurement of attitude
shift.
Millson ended the controversy in a two-part arti
cle. 59 Essentially these two articles were "rebuttals" to
Henrikson's criticism in the form of restatement and exten
sion of the former's arguments. Millson cited Monroe's
studies as support for the validity of both the Woodward
Ballot, per se, and the "Millson Formula" for treating the
raw data derived from the ballot. The bulk of the first
article was devoted to defining the use of the Millson
Formula in measuring speaker "effectiveness," and second,
Millson criticized Henrikson's "Montana Ballot" (in reality
a more complex form of the Woodward Ballot).
The second of Millson's articles was concerned with
specific application of the Woodward Ballot, interpretation
and application of the Millson Formula, including scoring
^Millson, "Reviews," loc. cit.
the shift of the undecided category (on the initial opinion
portion), and use of the Woodward Ballot to stimulate "real
life" conditions in the classroom. The main underlying
^concern appeared to be measuring the "effectiveness" of
speaking via the Woodward Ballot.
Millson conceded that part of his earlier formulae
had been inaccurate, but asserted that Henrikson had missed
the concept underlying the formula as a whole, that the
errors were not significant, and had already been corrected
An area of controversy quite significant to the cur
rent study was the discussion and defense of the lack of a
specific "weakening" of attitude response on the "standard"
Woodward Ballot. As Millson pointed out, there is such a
provision inherent in the final portion of the ballot.
Second, the inclusion of two additional categories, i.e.,
less affirmative and less negative, would probably add un
necessary complexity to the ballot. He further questioned
the accuracy of non-psychologically trained Ss being capa
ble of making the "finer distinctions" in attitude shifts
required by the more complex instrument.®®
Briefly summarizing the controversy, Millson at
times attempted to measure more than one variant, e.g., au
dience opinion shift and modes of delivery simultaneously,
plus devising a more complex formula for interpreting the
®®Millson, "Review, Part I," op. cit., pp. 476-477.
44 i
|
ballot's raw data. Henrikson legitimately criticized both I
these areas. Monroe defended Millson1s use of the formula
and the value of the ballot itself. j
In the current study the application of the ballot
differed in some respects from those areas involved in the
controversy. First, neither Woodward's nor Millson1s in
terval ranking formula, based upon arbitrarily assigned S
numerical values, was used. Second, the objections regard
ing the interpretation of the "undecided" category were
probably diminished by first securing definite convictions
of the Ss prior to speaking without revealing the purpose
of the study, and thus eliminating the controversy over
scoring the "undecided" response on the initial ballot.
Additionally, the Woodward Ballot appeared to lend
itself more readily to interpretation via an ordinal scale
reflecting "frequencies" than to an interval scale reflect
ing "distances" so use of chi square interpretation would
seem to be the most appropriate technique. The ballot was
used unidimensionally to measure attitude shift in a speak
er after exposure to controlled overt single listener feed
back. Thus the inability to control feedback from large
audiences and their potential group interaction was elimi
nated through the use of one highly trained listener.
Direct analysis regarding the validity of the ballot, per
se, by either Millson or Henrikson was not clearly sup
ported or denied. Their concern tended to be directed at
45 ;
J
specific situation applications of the ballot and the vari- j
!
ous formulae used in interpreting the raw data. In regard
i
to formulae used for treating this type of raw data, Guil- !
ford pointed out that treatment formulae and test instru- j
c-, I
ment validity are not synonymous.DX It is possible there- j
fore that support for validity of the Woodward Ballot could
be provided by Monroe's comparison of Woodward Ballot find- !
2 ^
ings to Thurstone-type findings which correlated at .75. *
Further discussion of the validity of the Woodward Ballot
and its suitability for interpretation via chi square are
discussed in Chapter III.
In summary, the Woodward Ballot appeared to be the
most suitable instrument for this particular study. Its
pre-test provision permitted initial opinions to be ob
tained without disclosing the purpose of the study. Its
post-test portion was uncomplicated and easy to administer.
A comparison of pre- and post-speaking responses provided
clear data for examining fact of opinion shift as well as
direction of shift; the two principal factors this study
sought to examine.
P. Guilford, Fundamental Statistics in Psy
chology and Education (3d ed.; New York; McGraw-Hill Book
Company, Inc., 1956), pp. 461-483.
^Monroe, "The Statistical Reliability and Validity
of the Shift of Opinion Ballot," loc. cit., p. 584.
46 j
i
Summary !
This chapter has examined contemporary speech and
psychological literature related to audience feedback ef- j
feet on listener attitudes. Specific attention was devoted
to the theory of Cognitive Dissonance because of the place |
that concept has assumed in oral communication theory and
its relationship to intra- inter-personal communication j
concepts. The chapter concluded with an examination of
representative attitude measuring instruments including the
rationale for the one selected.
Additionally, the literature surveyed in this chap
ter provided valuable background information for the in
vestigator in formulating and designing the current study.
Moreover, the material examined did indicate that of the
numerous related studies that have been made involving at
titude change there has been no specific effort to examine
the effect of peer-to-peer listener feedback on the fact of
speaker attitude shift, per se. Thus it appeared that a
research gap existed in a vital area of communication
theory.
CHAPTER III
DESIGN OF THE STUDY
Preview
The study was designed to allow two groups of ap
proximately thirty students each to present short persua
sive speeches to a single listener. Each speaker had pre
viously committed himself to a "yes" or "no" answer to the
topic--"Was police brutality a significant cause of the
Watts riot?" This previous commitment determined the divi
sion of the groups.
A listener had been previously rehearsed to appear
to either overtly agree or disagree with the speakers as
they spoke. Half of each group received positive feedback,
while the other half received negative feedback, i.e., the
listener appeared to either agree or disagree with them.
Immediately after giving the speech the Woodward
Shift-of-Opinion Ballot was administered to determine if
any changes in convictions had taken place after exposure
to the overt feedback. The raw data thus obtained were
then analyzed statistically.
47
48 j
Speakers !
The speakers chosen were Caucasian male students,
primarily freshmen or sophomores, between the ages of :
eighteen and twenty-two. The total of sixty-six who were
used were chosen from the beginning speech classes at San
Diego State College. Approximately forty-five sections of
Beginning Speech were offered during the spring semester. !
Average enrollment was twenty students per section, with an ;
approximately equal division between male and female stu
dents. Thus a total potential of 450 male students was
available to draw upon.
S in c e th e t o p i c chosen d e a l t w ith a h i g h l y c o n t r o
v e r s i a l a s p e c t o f _r a c i a l r e l a t i o n s , th e l i s t e n e r and th e
sp e a k e r s s e l e c t e d were C aucasian to a v o id th e p o s s i b i l i t y
o f i n j e c t i n g an u n c o n t r o lle d v a r i a b l e in th e form o f r a c i a l
r e a c t i o n s . For exam ple, i f a n on-C aucasian r e c e i v e d nega
t i v e feed b ack from th e C aucasian l i s t e n e r , any r e s u l t i n g
s h i f t s o f a t t i t u d e m ight be th e r e s u l t o f p e r c e iv e d r a c i a l
p r e j u d ic e r a th e r than th e e f f e c t o f n e g a t iv e fe e d b a c k , per
s e .
Similarly, since the listener was a male, females
were excluded to eliminate the possibility of feedback be
ing perceived as an indication of sexual attraction or dis
like. In brief, the writer attempted to eliminate as many
potential uncontrolled variables as possible.
Detailed examination of the 450 potential subjects
49 ;
narrowed the actual suitability of the group considerably.
Factors such as race, age, availability on the nights the
speeches were scheduled, limited the actual number selected j
to approximately 120. Although this was more subjects than ;
needed, all 120 were contacted to allow for "no-shows" and I
any unforeseen complications.
Provisions were made for a "pseudo-listener" to be |
available as well as tentative scheduling of a third night. 1
The "pseudo-listener" would hear any "excess" speakers and
the additional night provided opportunity for greater
flexibility should scheduling adjustments be necessary.
The final choice of approximately sixty subjects
was based upon two criteria: (1) the generally accepted
size of thirty in each group as suitable for accurate
statistical treatment;1 and (2) the wish to complete the
speeches in two nights indicated that approximately thirty
subjects was the maximum that could be efficiently processed
in each evening.
Thus the final selection of approximately sixty sub
jects (thirty in each group) was based upon suitable statis
tical criteria and the time factors involved.
Listener
The listener was nineteen years old, a freshman at
^George A. Ferguson, Statistical Analysis in Psy
chology and Education (New York: McGraw-Hill Book Co.,
Inc., 1959), pp. 175-76.
50
the University of San Diego, College for Men. He appeared
9
to be the "typical" freshman or sophomore student. He had
participated in two years of high school drama and one year
of high school debate. He was currently enrolled in the
ibeginning speech course at the University of San Diego. In
class he had demonstrated above average ability in all
phases of the course. A check was made to determine if he
was acquainted with any of the subjects. He was not.
In addition to his dramatic and public speaking
experience, the listener was also chosen because of his
general physical appearance, i.e., the "typical" freshman
student, and his ability to play the role of an interested
peer-group opponent or partisan. Preliminary rehearsals
with the researcher indicated a high degree of skill in
this area. His ability and training is discussed later in
this chapter under Procedures.
Materials
The principal materials used consisted of one
questionnaire, a nonsignificantly modified Woodward Ballot,
a concealed tape recorder, and two faculty "reminders^
2A photograph of the listener taken one month prior
to the experiment is included in Appendix B, p. 98.
•^The modifications made upon the Woodward Ballot
were not significant as far as reliability and validity are
I concerned and may have increased these two items by remov
ing one of the main points of the controversy concerning
the ballot, i.e., scoring the "undecided" category in the
first or initial opinion portion and then comparing any
As will be discussed later under Procedures, pre
testing determined the exact wording of the questionnaire.
The questionnaire administered to the subjects offered only
a "yes" and "no" response. A few students refused to an
swer on only a "yes" or "no" basis. So in fact the ques
tionnaire allowed an "undecided" choice for those who were
not strongly committed to the topic. The importance of
genuine conviction will be discussed later. Thus, the
timeliness of the topic, plus its emotional nature, and the
pre-test indicated that the majority of the students con
tacted were committed to an honest "yes" or "no" answer on
this forced-choice basis.
The questionnaire included provisions for the sub
ject's name, age, sex, race, year in college, and beginning
speech class hours and instructor, therefore making a
quick, accurate selection of the subjects suitable for the
resulting shifts to scoring on the second portion of the
ballot. Two modifications were made as follows: (1) elim
ination of the "undecided" response on the initial opinion
part of the ballot. The rationale for this change and its
non-significance has been previously discussed; (2) elimi
nation of a provision of the Ss explaining the "rationale"
for their responses. Woodward noted that while provision
for inclusion of these "reasons" were provided, "no effort
is made to secure its use by the audience" (Woodward, op.
cit., p. 95).
The original five items on the second part of the
ballot were retained verbatim and in their original order.
The non-significance of these modifications was probably
best expressed by Woodward when he commented: "The ballot
serves its fundamental purpose if it bears two X's regis
tering the opinion of the voter" (ibid., p. 96).
study possible.
The Woodward Ballot was used for determining shift
in attitude of the subjects after the speech was given.
This ballot contained a place for the student's name, the
topic of the speech, typed in capital letters across the
top of the paqe, and directly under this, the instructions.
There were five possible choices which required only a
checkmark in the appropriate box to indicate response.
These responses ranged from a strong shift toward the af
firmative of the question, to undecided, and then to a
strong shift toward the negative side of the question.
The form used is indicated in Figure 1.
The accuracy of the Woodward Ballot when used ac
cording to directions is as reliable as any other attitude
change measurement device.^ One of the prime considera
tions -necessary in this study was that the subjects have a
strong, sincere attitude supporting the position they advo
cated. Such a genuine commitment was necessary if the
researcher was to measure the effect of overt feedback. A
lack of spontaneous, definite belief would allow extraneous
factors to exert undue influence on the subject's opinion,
after it was initially indicated and before exposure to
overt feedback. In other words, a genuine, strong opinion
was unlikely to change due to minor "everyday" happenings.
^Woodward, loc. cit.
WAS POLICE BRUTALITY A SIGNIFICANT
CAUSE OF THE WATTS RIOT?
Check the answer below which most nearly indicates
your feeling about the above topic of your speech:
I believe more strongly in the affirmative of
the question than I did.
I believe in the negative of the question.
I am undecided.
I believe in the affirmative of the question.
I believe more strongly in the negative of the
question than I did.
Figure 1.— Modified Woodward Ballot
54 :
The sincerity of conviction was demonstrated by the j
reactions of some of the Ss. While no subject interrupted
his speech to comment on the listener's apparent lack of
objectivity, three subjects tried to question him at the j
conclusion of their speech, and one speaker who received j
negative reinforcement "invited the listener outside" to
discuss the matter further. The listener declined. Anoth- ;
er student, who received positive feedback, told the writer
that "That guy (the listener) was just great. He sat
there, paid attention, and agreed with everything I said."
The strong, spontaneous reactions of the Ss also
tended to lend support to the validity of the Woodward Bal
lot. The Ss were strongly concerned with their attitude
being "fairly" received by the listener. Their sincerity
of conviction plus their concern for a "fair" hearing ac
curately indicated their attitudes and, therefore, their
responses on the ballot would most probably be "attitude
position responses," i.e., validity.
The Woodward Ballot's prime value in this study was
that it could determine initial subjects' attitude and
later measure any opinion shifts which occurred without
letting these subjects know the basic nature of the study
or biasing their convictions. Such a criterion was essen
tial if the effect of exposure to listener feedback was to
be accurately recorded. In short, if the subjects did not
know in advance what was to be measured, there was no
55 :
i
i
chance of the measuring instrument influencing their con- |
victions. Additionally, the Woodward Ballot's simplicity
in administration and interpretation added to its suitabil- j
ity for this study.
|
The researcher, in selecting the Woodward Ballot, j
was aware that measurements of strength of attitude change
would be difficult to obtain and/or probably inconclusive I
in nature. However, since the primary purpose of the study
was concerned with the fact and the direction of change,
this possible limitation was not considered significant.
A speech criticism form was used by the listener to
stimulate interest in the speakers and to provide an "au
thentic" atmosphere. Giffin and Warner found that college
debaters had more confidence in the decisions of judges who
took notes while listening than in those critics who did
not.This writer felt that the inclusion of listener note-
taking would provide a worthwhile incentive for the speak
ers .
The tape recorder was intended primarily as a means
of verifying the consistency of the verbal feedback of the
listener. After concealment it was turned on prior to the
entrance of the first speaker, and it remained on constant
ly during all the speeches each night. It was turned off
^Kim Giffin and Donald Warner, "A Study of the In
fluence of Note-taking by Tournament Judges on Debaters'
Attitudes," The Gavel, XLV (March, 1963), 42-43, 56.
56
only after the last speaker left the room.
Facilities
Four rooms were utilized in the study. The room
where the speeches were given was originally part of an
experimental radio broadcasting system. Because of its
original purpose, it was partially soundproof, equipped
with one-way glass windows on two sides, and contained a
microphone hookup and wiring. The room was approximately
twenty-seven feet long by fifteen feet wide, and was
equipped with desks, a lectern, and chairs, such as are
usually found in a speech classroom. A window opening to
the outside and utilizing Venetian blinds was in the wall
opposite the door. The window and the blinds were kept
closed. A small microphone had been suspended from the
ceiling for several years. The microphone head itself was
approximately one inch in diameter. It reached to about
seven feet above the floor.
The desks and lectern were arranged so that the
microphone was above and between the speaker and the lis
tener. In practice, this position worked well in all re
spects; apparently no subjects noticed its presence.
A tape recorder was concealed on the floor behind a
small cabinet in the right front corner of the room, out of
sight of the speaker. A diagram is included, drawn to
scale, to aid in visualizing the arrangement of the speak-
ing room (Figure 2) .
The writer's office was chosen as the most suitable
location for the completion of the Woodward Ballot, as it
was almost directly across the hall from the "speaker room"
and provided the proper "academic atmosphere" for answering
questionnaires.
Procedures
Training the listener
As indicated previously, the purpose of the listen
er was to provide consistent, unmistakable feedback (to the
speaker), clearly indicating agreement or disagreement
since, as Clevenger noted, certain overt actions are gener
ally accepted as indications of inner feelings.^ Thus em
phasis in training the listener was upon the development of
simple overt actions which would remove any possibility of
doubt on the part of the speaker as to whether the listener
was, or was not, in agreement with him. The consistency
and unmistakability of listener feedback was essential if
uncontrolled variables were to be held to a minimum.
The selection of a single listener was also in part
based upon the necessity for providing a consistent, unmis
takable form of feedback. One listener could provide such
feedback whereas feedback from several listeners would be
^Clevenger, loc. cit.
Scale - 1" = 6'
27'
12'
Iff-
18'
A1
One-way Windows
E Lecturn
A2
Two-way Window
F Door
B Microphone
G Desk
C Concealed Tape Recorder G Desk
D Listener
Figure 2.— Speaker room
much more difficult to coordinate. Additionally, as Berio
noted: "Clearly person-to-person communication permits
maximum f e e d b a c k . Since there had been no previous stud
ies specifically devoted to determining the effect of overt
listener feedback on fact of, as well as direction of
speaker attitude change, the logical starting place, to
this writer, seemed to be with one listener. Then, should
it be proved feasible to train one listener to provide con
sistent, unambiguous feedback, future research could look
toward two or more listeners.
The listener was rehearsed prior to hearing the
speeches to insure his consistent feedback responses.
These training sessions lasted two hours per afternoon for
three days. The listener was provided first with a brief
description of the procedures and nature of the study.
However, the exact purpose was omitted. Since the speeches
were to be five minutes maximum in length, it was decided
that approximately five visual feedback indications were to
be provided by him. Additionally, if possible, an equal
number of vocal feedback indications were to be combined
with the visual ones. It was clearly understood that these
latter feedback indicators, while loud enough to be heard
by the speaker, were not to be so loud as to interruptor
override the speaker's voice. The listener was further
Berio, op. cit., p. 114.
60 :
i
cautioned to avoid direct verbal communication with the sub~|
jects. He was to verbalize noise phonemes only, not words. ;
He was to appear to be muttering to himself and vocalizing j
"uh-huh" or "Huh-uh," dependent upon the type of feedback ■
being provided.
The first two sessions were devoted to developing
the vocal volume level suitable for the above criteria.
Rehearsal took place in the room later used for the experi
ment to aid in obtaining optimum listener efficiency.
Since the speakers would not all use the same amplitude
level, the researcher varied the force of his voice to give
the listener practice in adapting to changing levels of
loudness. By the end of the second practice session, the
listener demonstrated to the researcher's satisfaction the
ability to adequately and consistently reproduce the de
sired verbal response technique.
The third practice session was devoted to develop
ing believable facial reactions and coordinating them with
the vocal feedback. In practice the development of the
visual reinforcement reactions provided the least diffi
culty in reproducing, since they were not dependent on the
"force" of the speaker's actions, as was the audible feed
back. Coordination of both feedback elements likewise was
readily mastered by the listener. At the- completion of the
third session his performance was deemed by the researcher
to be- highly consistent and convincing.
61 ;
Determining pre-experimerital
attitudes :
It was decided to use a topic that would evoke a j
j
definite, spontaneous, honest response from the subjects.
I
Since the first part of the Woodward Ballot provides a |
place for an agreement or disagreement answer, a topic
which would lend itself to a "yes" or "no" only answer on
the part of the subjects was highly desirable as previously j
indicated. Such a positive commitment would simplify as
well as make more accurate the utilization and interpre
tation of raw data obtained from the Woodward Ballot.
The topic had to be timely and of a highly emotion
al nature if the desired genuine attitudes of the subjects
were to be obtained. Memories of the Watts riot at this
time were still vivid and the riot was a frequent topic in
the researcher's beginning speech classes. Students fre
quently expressed strong convictions regarding the riot.
Because of these factors it was decided to investigate the
feasibility of using some aspect of this topic.
A pre-test to aid in determining the suitability
of this topic area was administered to a fundamentals
speech class at Grossmont College. The first choice of
phraseology was arbitrarily decided upon to read, "Was
police brutality the major cause of the Watts riot?" On a
sample three-choice ballot, all males answered "no," six
females answered "no," and three females answered "yes."
62 ;
I
Two female students were absent. No subjects answered "un- j
decided." From the overwhelming negative response of the
males it was concluded that this phraseology was unsuita- j
ble. The wording was then modified to read, "Was police j
j
brutality a significant cause of the Watts riot?" and re
submitted to the class. The results were: four males an
swered "no," four males answered "undecided," four males
answered "yes;" five females answered "no," five females,
"yes," no females answered "undecided." One male and one
female were absent when the second questionnaire was admin
istered. These results indicated a sufficient dispersion
of responses among the males to make the question and its
phraseology acceptable for the study.
These students were told only that the question
naires had to do with a "research" comparison to answers
given at another college. They were asked to indicate sex
only on both questionnaires, and to omit their names. This
was done to minimize any effort on the part of the students
to apply the "right" response, i.e., the one they might
anticipate the instructor as desiring or agreeing with.
Subsequent informal class discussion indicated that pres
ences of strong, sincere convictions existed in the students;.
Thus it was felt that the criterion of having a
topic which would provide the spontaneously honest "yes" or
"no" response necessary for use with the Woodward Ballot
6 3 j
was met. This assumption later proved to be correct. When !
the revised questionnaire was submitted to the subjects at
San Diego State College (without explaining the true pur
pose of the questionnaire) less than a dozen declined to
]
answer on a "yes" or "no" basis.
Scheduling the speakers j
With the cooperation of several faculty members at
San Diego State College, the initial opinion questionnaire
was distributed to beginning speech classes. As previously :
indicated, approximately 120 subjects were found who fitted :
the required criteria. In order to minimize independent
variables, which might influence the subjects' opinions,
efforts were made to have the speeches delivered as soon as :
possible after the administration of the questionnaire. In
practice this was approximately two weeks. Fortunately, no
new incidents of police brutality or race rioting occurred
during this interim.
The experiment was held on two consecutive evenings
and lasted approximately three hours each night (from 6:45
p.m. to 9:45 p.m.).
In assigning the time and date of the subject's
speech, two factors were considered. First, each group of
subjects, i.e., those who answered "yes" and those who an
swered "no," were divided into two sub-groups. One of each
sub-group was scheduled to speak both evenings. On Wednes-
i 64
day, one "no" sub-group was scheduled to speak first. On
Thursday, the second "no" sub-group spoke last. The oppo-
i
site positioning and procedure was used for the two "yes"
sub-groups. This process enabled the listener to supply
the same response, i.e., positive or negative, all evening.
Maintaining the schedule
The subjects for each evening were divided into
three time blocks of one hour each. Their time and date of
speaking had been told them previously by their class in
structor. Each block of students assembled in the assigned
classroom where names were checked against the schedule
list by one of the researcher's assistants. As soon as a
student was "checked in," the researcher took him to the
speaker room to deliver his speech.
After concluding his speech, the subject was es
corted to the researcher's office across the hall and once
inside asked to be seated at the desk and to complete the
Woodward Ballot. If he was hesitant or had questions about
his reactions, he was told to check the box which most ac
curately reflected his feelings at that moment. Upon com
pleting the ballot, the assistant would then lead him to
the outer door of the building and thank him for his coop
eration. Any questions by the subject were parried with a
comment to "Ask Mr. Mattox next week when the project is
over." The speakers were not allowed to return to the
6 5 i
i
building during the remainder of the experimental period.
Introducing the speaker and
providing overt feedback j
Just before entering the "speaker room" the re
searcher would say, "We have a student in here who will lis-j
i
ten to your speech. Because of the large number of speakers:
we have to listen to tonight, please do not have any con
versation with him. The time limit, as you know, is three
to five minutes, and he will have time cards showing how
much time you have left. Please leave just as soon as you
finish your speech."
Upon entering the speaker room the following pro
cedure was always used. The researcher would say, "This is
(the student's name). He is going to speak on police bru
tality and the Watts riot." The listener would then, im
mediately, make his initial biasing comment. If he were to
react positively he would always say, "Good, I agree with
that," or if he were to react negatively, he would always
say, "Well, okay, but I can't agree with that." A playback
of the tape recorder verified the listener's consistency of
response using this dialogue. In only three cases did a
subject question the fact that the listener had expressed
an opinion on the merits of the speaker's position before
that position was actually disclosed. In these instances
the listener did not comment directly on this observation,
but instead requested the subjects to begin their speech
because of the shortage of time. Immediately after the
listener made his opening comment, the researcher would
leave the room. There was no further conversation between
the researcher and the subject.
At the end of the first thirty seconds (approxi
mately) of the speech, the listener began his feedback re
actions. If his response was negative (as on the first
night), he would shake his head from side to side, as re
hearsed, and mutter softly, "huh-uh." Simultaneously, he
would also grimace in apparent disapproval. He made the
verbal feedback reactions on an almost inaudible level,
i
without interrupting the speech but within the threshold of
hearing of the speaker as planned. Naturally, there was
some variation in the timing of the feedback responses due
to minor variations in each speech. However, the listener1
verbal reactions were consistent enough (as verified by the
tape recording) to insure the consistency of his feedback
reaction. Throughout the speeches the listener appeared to
be rating the speaker via the speech rating form.
Verification of the consistency of the listener's
visual feedback responses was made by observing through the
one-way windows in the speaker room.
A trial run, including testing of mechanical equip
ment and rehearsal of assistants and listener, was made
prior to the actual experiment, thus aiding in the effi
cient execution of the procedures involved.
St at istical Procedures
The chi square test for goodness of fit of an ob
served to a theoretical frequency distribution, was the
statistical procedure used to test the five hypotheses com
prising the problem of this study.
Chi square was chosen because it deals with fre
quency data, and because it is appropriate for use with
data at the level of an ordinal or ranking scale. The
Woodward Ballot used in this study yielded only ordinal
data. That is to say, the subject recorded his initial at
titude toward the proposition as either affirmative or neg
ative; with reference to that initial professed belief. On
the final ballot the subject could record one of three
choices: (1) strengthening of initial belief, (2} no
change, or (3) weakening of initial belief. This "stronger
than" characteristic of the ballot data described an ordi
nal scale. For certain purposes, such as measuring "effec
tiveness" of debaters in audience-decision intercollegiate
debates, Woodward Ballot scoring might be legitimately
elaborated by assigning arbitrary numerical values to the
shifts. This translates frequencies into scores and may
give a spurious sense of precision. However, this scoring
procedure does not produce data that meet the requirements
of an interval scale because the distance between adjoining
ballot categories is unknown. For example, a shift from
affirmative to affirmative more strongly is usually
68 j
i
assigned an arbitrary distance-value of plus one, and a ,
shift from affirmative to undecided is given an equivalent
value of minus one; the equivalence of these shifts is open !
j
to doubt, to say the least. Thus in the present study a
conservative decision was made--the data were considered to !
be on an ordinal scale. This decision precluded the possi- ;
bility of using parametric statistics, such as analysis of j
variance or Millson-type weighted responses. For example,
Siegal notes that for an analysis of variance application
(or a Millson arbitrarily weighted interval type formula)
to yield non-spurious finds, five criteria must be met:
1. The observations must be independent. That is, the
selection of any one case from the population for in
clusion in the sample must not bias the chances of any
other case for inclusion, and the score which is as
signed to any case must not bias the score which is
assigned to any other case.
2. The observations must be drawn from normally dis
tributed populations.
3. These populations must have the same variance (or,
in special cases, they must have a known ratio of
variances).
4. The variables involved must have been measured in
at least an interval scale, so that it is possible to
use the operations of arithmetic (adding, dividing,
finding means, etc.) on the scores.
In the case of the analysis of variance (the F
test), another condition is added to those already
given:
5. The means of these normal and homoscedastic popula- ;
tions must be linear combinations of effects due to
columns and/or rows. That is, the effects must be
69 j
i
additive.8 <
The most important of these criteria is, in regard
to the Woodward Ballot, Number Four, that "the variables I
involved must have been measured in at least an interval j
|
scale so that it is possible to use the operations of !
arithmetic (adding, dividing, finding means, etc.) on the
scales."9
As previously indicated, the Woodward Ballot does
not provide such an interval scale. Guilford additionally
notes that in regard to computations of interval scales,
"we must think in terms of exact limits.Since it is
impossible to provide exact limits (distances) from the
Woodward Ballot except by assigning arbitrary values, such
interval-type scales are not readily applicable to raw data
obtained from the Woodward Ballot. Siegal strongly empha
sized the probability of spurious findings when attempting
to apply interval-type (Millson) treatment methods to in
struments which yield ordinal data (such as the Woodward
Ballot) when he said:
At the risk of being excessively repetitious, the
writer wishes to emphasize here that parametric statis
tical tests, which use means and standard deviations
(i.e., which require the operations of arithmetic on
^Sidney Siegal, Nonparametric Statistics for the
Behavioral Sciences (New York: McGraw-Hill Book Co., Inc.,
1956), p. 19.
9Ibid. lOGuilford, op. cit., p. 37.
the original scores) ought not to be used with data in j
an ordinal scale.H j
The Woodward Ballot yields ordinal data and thus
nonparametric methods of treatment such as the chi square
would therefore be, as indicated by Siegal and Guilford,
the statistical technique most appropriate for use with
this instrument.
Hsiegal, op. cit., p. 26.
CHAPTER IV
PRESENTATION AND ANALYSIS OF DATA
Presentation of Data
The data for all sixty-six subjects are presented
in Table I. This table subdivides the subjects into four
groups :
1. Affirmative speakers who received positive feedback.
2. Affirmative speakers who received negative feedback.
3. Negative speakers who received positive feedback.
4. Negative speakers who received negative feedback.
Table I shows how each speaker marked his Woodward
Ballot after exposure to overt listener feedback. This
shift-of-opinion balloting technique was described in the
preceding chapter. The five alternatives offered on the
ballot are identified in abbreviated form at the bottom of
the table.
Analysis of Data
The chi square technique was used for the statis
tical analysis of the data. The reasons for choosing this
technique instead of some parametric treatment, such as
analysis of variance, were discussed in the preceding
71
TABLE 1
THE WOODWARD BALLOT RESPONSES OF 66 SPEAKERS
AFTER EXPOSURE TO OVERT LISTENER FEEDBACK
Speakers
Positive
Feedback
Negative
Feedback
Affirmative A. 4 A. 4
Affirmative B. 1 B. 1
Affirmative C. 0 C. 2
Affirmative D. 10 D. 8
Affirmative E. _0 E. 0
15 _ 15 30
Negative A. 0 A. 0
-
Negative B. 8 B. 9
Negative C. 0 C. 2
Negative D. 2 D. 1
Negative E. 10 E. _4
20 16 36
Total 66
A. Affirmative more strongly
B. Negative
C. Undecided
D. Affirmative
E. Negative more strongly
chapter.
As described in Chapter I, the research problem in
this study comprised five null hypotheses. Chi square was
applied to each hypothesis in order to determine the sig
nificance of the difference between the actual or observed
frequencies and the predicted or theoretical frequencies.
In testing the first two hypotheses the most
logical theoretical expectation seemed to be that none of
the subjects would shift his opinion. The reasoning behind
this choice was as follows. If nothing had been done but
request the subjects to express their attitudes toward the
proposition (police brutality and the Watts riot) and then
repeat the request a week later, the most logical expecta
tion would be that the number of attitude shifts would fall
well within the boundaries of chance— unless something
relevant to the proposition happened during the interim
between votes. As previously mentioned, in the present
study no new incidents of police brutality or rioting were
reported through the news media between the time of the
initial ballot and the final one. Two things that did
occur in all instances were that (1) each subject prepared
his speech and (2) delivered it to an overtly responsive
listener. It was not feasible in this study to measure
possible effects of speech preparation; therefore, it was
assumed that the effects of this variable would be about
the same for all subjects.
74
In testing the last three hypotheses a different
theoretical frequency was indicated. The most logical
expectation in all three hypotheses seemed to be a fifty-
fifty split, i.e., affirmative speakers are just as likely
to be influenced (or not influenced) by feedback as the
negative speakers.
Results of chi square analysis of each hypothesis
were as follows:
Hypothesis 1; That no significant number of
speakers shift their initial attitudes toward their topic
after exposure to overt listener feedback.
Table 1 shows that thirty-one of the sixty-six
speakers shifted their attitudes in one way or another.
Chi square analysis showed that this number of shifts can
not be attributed to chance; the level of significance was
.01. The null hypothesis was rejected.
Hypothesis 2: That no significant number of
speakers reverse their initial attitudes, i.e., from af
firmative to negative or vice versa, after exposure to
overt listener feedback.
Only five subjects reversed sides on the question,
i.e., from affirmative to negative or vice versa. Chi
square showed that this number could have occurred by
chance. The null hypothesis could not be rejected.
Hypothesis 3: That there is no significant differ-
ence in initial attitude changes after exposure to overt
listener feedback, between speakers who initially favor a
highly controversial proposition and those who oppose it.
Table 1 shows the following:
Affirmative Speakers Negative Speakers
No change 18 No change 17
Changed 12 Changed 19
Chi square showed that the above differences could be due
to chance. The null hypothesis could not be rejected.
Hypothesis 4: That among those speakers who change
their initial attitudes after exposure to overt listener
feedback, there is no significant difference between the
number whose attitude is reinforced and the number whose
attitude is weakened.
In testing this hypothesis the N became thirty-one
(those who shifted attitudes) rather than sixty-six. The
focus here was on the direction of shifts. Table 1 shows
that twenty-two of the shifts were in the direction of
strengthening the initial belief in comparison with nine
shifts in the direction of weakening the original belief.
Chi square showed that this difference was beyond chance at
the .05 level. The hypothesis was rejected.
Hypothesis 5: That there is no significant differ
ence between the number of shifts of opinion after exposure
76
to positive listener feedback as compared with negative
feedback.
Again, N was thirty-one. Of these, seventeen had
been exposed to positive feedback and fourteen to negative
feedback. The difference was not significant. The hypoth
esis could not be rejected.
CHAPTER V
SUMMARY, CONCLUSIONS, AND IMPLICATIONS
Summary
The purpose of this study was to investigate the
effect of overt listener feedback on speaker attitude
changes.
Sixty-six subjects gave short persuasive speeches
on an assigned controversial proposition— "Was police bru
tality a significant cause of the Watts riots?" Approxi
mately half of the subjects had previously answered "yes"
to the above question and the remainder had previously an
swered "no." One specially trained "listener" constituted
the audience for each of the subjects. The listener had
previously been rehearsed to appear to either agree (posi
tive feedback) or disagree (negative feedback) with the
subjects immediately before and during their speeches.
Half of the "yes" group and half of the "no" group received
positive feedback and the remainder received negative feed
back. A shift-of-opinion ballot was used so as to permit
measurements of the possible effects of the two kinds of
feedback.
77
7 8 !
I
j
Five hypotheses were proposed:
1. That no significant number of speakers shift
their initial attitudes toward their topic
after exposure to overt listener feedback.
I
2. That no significant number of speakers reverse
their initial attitudes, i.e., from affirmative .
]
to negative or vice-versa, after exposure to j
overt listener feedback.
3. That there is no significant difference in ini
tial attitude changes after exposure to overt
listener feedback, between speakers who ini
tially favor a highly controversial proposition
and those who oppose it.
4. That among those speakers who change their ini
tial attitudes after exposure to overt listener
feedback, there is no significant difference
between the number whose attitude is reinforced
and the number whose attitude is weakened.
5. That there is no significant difference between
the number of shifts of opinion after exposure
to positive listener feedback as compared with
negative feedback.
Conclusions
The five hypotheses comprising the problem of this
study were presented in Chapter I and were repeated
79 :
verbatim in Chapters IV and V (immediately above). To
avoid additional repetition the conclusions regarding these |
hypotheses, numbered the same as previously, are given \
below in briefest form:
1. Rejected
2. Not rejected
3. Not rejected
4. Rejected
5. Not rejected
Implications for Future Research
The findings in the present study seemed encourag
ing enough to justify the suggestion that further research
to verify, modify, extend, or refine the results of this
study might be profitably undertaken. Several possibili
ties occurred to the writer, as described below.
In the beginning the writer had hoped to carry the
analysis of his data a step further. He wondered if posi
tive feedback might not usually tend to encourage the
speaker into a stronger belief in his existing attitude;
and likewise, if negative feedback might not discourage the
speaker into a weaker belief. Insufficient relevant data
prevented reaching conclusions regarding the above ques
tions. The diminishing size of the groups occurred as
follows:
Positive 14
— More so 22
N = 31 Negative 8
(who shifted)
Positive 3
Less so 9
Negative 6
Taking into account the character of the data, the size of
the groups as shown in the right-hand column above did not
seem adequate for statistical treatment. By inspection
only these data suggest interesting possibilities.
As indicated in Chapter II, there were no other
studies which closely parallel this writer's investigation.
Further, contradictory findings in generally related stud
ies were not conclusive in either supporting or rejecting
the writer's hypotheses.
There were, however, several studies which offered
interesting implications. Among these were studies con
cerned with Festinger's theory of cognitive dissonance.
Applied to the present experiment, Festinger's concept ap
peared to predict a strong change in speaker attitude that
would reduce dissonance and at the same time achieve con
sonance. Specifically, those speakers who faced negative
feedback should have shifted their position toward the
center and away from the experiences (negative feedback)
which produced the dissonance. A non-statistical observa-
81 :
tion of the number of subjects who faced negative feedback I
and shifted away from their initial attitudes appears to
support his theory. However, a substantial minority of j
subjects who also received negative feedback apparently did j
i
not try to avoid dissonance and indicated a strengthening ;
of initial attitude. Therefore, in the absence of a suffi
cient number of responses to permit statistical interpre
tation, Festinger's theory can neither be accepted nor
denied on the basis of this study's findings. It should
also be noted that more than one researcher in cognitive
dissonance theory has uncovered conflicting findings.
Other studies which appeared to offer stimulating
implications to the findings uncovered in the researcher's
study were those of Hovland, Scott, and Jones. Hovland's
concept of "public commitment" being a factor in "forcing"
a speaker to adhere to his original position would seem to
agree with this writer's findings that both types of feed
back produced a strengthening of convictions. Apparently,
as in the writer's hypothesis, once the speaker felt himself
to be "publicly committed" to an issue, his attitude toward
that position increased regardless of the type of feedback
he was exposed to. His hypothesis that "the act of communi
cating with others should increase the speaker's own com
mitments" seemed to be borne out. However, why this theory
appears to operate is beyond the scope of the current
study. Without additional data this writer felt that
82
further observations and/or attempted correlations to Hov-
land's theory would be premature.
An interesting comparison can also be made with
Scott's study, which appears to be supported by the find
ings uncovered here. His results were in part similar to
this researcher's, i.e., a more pronounced shift toward
strengthening of initial convictions than toward weakening
them. However, as noted in Chapter II, Scott "informed"
his subjects that the audience was either in favor of or
opposed to the speaker's topic before or after the actual
speaking situation and made no effort to control feedback
during the speeches.
More closely related to this researcher's findings
were those of Jones, who found that the act of constructing
a speech supporting a given attitude tended to increase
speaker conviction. Jones' findings also appeared to be in
agreement with the writer's research. However, again it
must be remembered that Jones' study specifically excluded
speaker feedback. His findings that an attitude shift in
favor of the speaker advocate position after preparation of
a persuasive communication does agree to an extent with the
findings of the researcher's study. Why this effect took
place is also considered beyond the range of the current
study. Again it should be noted that Jones made a con
certed effort to avoid what he called the "contaminating
effect" of listener feedback.
83
To the writer of the current study, one of the more
|interesting implications involves the use of two trained
listeners, who might simultaneously provide positive or
|negative feedback, or a combination of both types. Such a
procedure involving two listeners could possibly indicate
the practicality of using even larger groups and thus pro
vide a stimulating challenge for future researchers.
An additional implication might be the use of an
"academic” controversial topic which did not involve such
deeply held convictions as the Watts riot, a topic such as
optimum class size, revision of student government, etc.
The findings derived from a topic like one of the above
could provide an interesting comparison for the findings
uncovered in the present study.
A final implication is the possibility of the de
velopment of some other attitude scaling instrument than the
Woodward Ballot which would permit the measuring of an addi
tional dimension of opinion change— namely, the strength of
a shift in terms of distance along an attitude scale.
I
B I B L I O G R A P H Y
84
B I B L IO G R A P H Y
Books
Berio, David K. The Process of Communication. New York:
Holt, Rinehart and Winston, 1960.
Braden, Waldo W. The Communicative Arts and Sciences of
Speech. Columbus, Ohio: Charles E. Merrill Books,
Inc., 1963.
Brehm, Jack W., and Cohen, Arthur. Explorations in Cogni-
tive Dissonance. New York: John Wiley and Sons,
1962.
Cartwright, Dorwin, and Zander, Alvin (eds.). Group Dynam
ics . Evanston: Row, Peterson and Co., 1953.
Clevenger, Theodore, Jr. Audience Analysis. Indianapolis:
Bobbs-Merrill Co., Inc., 1966.
Dickens, Milton, and Travis, Lee Edward. The Experimental
Approach. In An Introduction to Graduate Study in
Speech and Theatre, ed. Clyde W. Dow. East Lan
sing: Michigan State University Press, 19 61.
Ferguson, George A. Statistical Analysis in Psychology and
Education. New York: McGraw-Hill Book Co., Inc.,
1959.
Festinger, Leon. A Theory of Cognitive Dissonance. Evans
ton: Row, Peterson and Co., 1957.
Guilford, J. P. Fundamental Statistics in Psychology and
Education. 3rd ed. New York: McGraw-Hill Book
Co., Inc., 1956.
Harvey, O. J. (ed.). Motivation and Social Interaction.
New York: The Ronald Press Co., 1963.
Hilgard, Ernest R. Introduction to Psychology. 2d ed.
ed. New York: Harcourt, Brace and Co., 1957.
85
86
Hovland, Carl L., Janis, Irving L., and Kelley, Harold H. |
Communication and Persuasion. New Haven: Yale
University Press, 19 53. 1
Lewin, Kurt. Group Decision and Social Change. In In j
Readings in Social Psychology. Edited by Guy E.
Swanson, Theodore Newcomb, and Eugene Hartley. New j
York: McGraw-Hill Book Co., Inc., 1952. j
|
Lewis, Thomas R., and Nichols, Ralph G. Speaking and Lis
tening. Dubuque: William C. Brown Company, 1965. j
Lindzy, Gardner (ed.). Handbook of Social Psychology. 2 ;
vols. Cambridge: Addison-Westley Publishing Co.,
Inc., 1954. ;
Osgood, C. E., Suci, G. J. , and Tannenbaum, P.. H. The
Measurement of Meaning. Urbana: University of
Illinois Press, 1957.
Ruch, Floyd L. Psychology and Life. 5th ed. Chicago:
Scott, Foresman and Co., 1958.
Shannon, Claude, and Weaver, Warren. The Mathematical The
ory of Communication. Champaign-Urbana: University
of Illinois Press, 1949.
Siegal, Sidney. Nonparametric Statistics for the Behavioral
Sciences. New York: McGraw-Hill Book Co., Inc.,
1956.
Thurstone, L. L., and Chave, E. J. The Measurement of Atti
tude. Chicago: University of Chicago Press, 1929.
Wiener, Norbert. The Human Use of Human Beings. New York:
Doubleday and Co., Inc., 1954.
Woodworth, Robert S., and Marrquis, Donald. Psychology.
New York: Henry Holt and Co., 1953.
Periodicals
Amato, Philip P., and Ostermeier, Terry. "The Effect of
Audience Feedback on the Beginning Public Speaker,"
The Speech Teacher, XVI (January, 1967) , 56-60.
Coch, Lester, and French, John R. P. "Overcoming Resist
ance to Change," Human Relations, I (1948), 512-
532.
87 ;
i
Cronkhite, Gary Lynn. "Toward a Real Test of Dissonance
Theory," The Quarterly Journal of Speech, LII ;
(April, 1966), 176-182.
i
Giffin, Kim, and Warner, Donald. "A Study of the Influence I
of Note Taking by Tournament Judges on Debaters' i
Attitudes," The Gavel, XLV (March, 1963), 42-43, 56.;
Henrikson, E. H. "The Audience Reaction Ballot: An Evalu- j
ation," The Quarterly Journal of Speech, XXIV (Feb- j
ruary, 1938), 54-61. ;
Kelman, Herbert. "Attitude Change as a Function of Re
sponse Restrictions," Human Relations, VI (1953),
185-215.
Knower, Franklin H. "Graduate Theses: An Index of Gradu
ate Work in Speech," Speech Monographs, II (Octo
ber, 1935), 1-51.
________ . "Graduate Theses: An Index of Graduate Work in
Speech," Speech Monographs, XXXII (August, 1965),
336-385.
Lana, Robert A. "Interest Media and Order Effects in Per
suasive Communications," American Journal of Psy
chology , LVI (1963), 9-13.
Leavitt, Harold J., and Mueller, Ronald A. H. "Some Ef
fects of Feedback on Communication," Human Rela
tions, IV (1951), 401-410.
Levonian, Edward. "Opinion Change as Mediated by Audience
Tailored Film," Audio-Visual Communications Review,
XI (1963), 104-113.
McNemar, Quinn. "Opinion-Attitude Methodology," Psycholog
ical Bulletin, LXIII (July, 1950), 365-373.
Miller, Gerald R., et al. "The Effect of Differential Re
ward on Speech Patterns," Speech Monographs, XXVIII.
(March, 1961), 9-15.
Millson, William A. D. "Experimental Work in Audience Re
action," The Quarterly Journal of Speech, XVIII
(February* 1932), 94-111.
"Problems in Measuring Audience Reaction," The
Quarterly Journal of Speech, XVIII (November,
1932) , 621-637.
88
Millson, William A. D. "Audience Reaction to Symposium,"
The Quarterly Journal of Speech, XXI (February,
1935), 45-53.
________ . "Measurement of Speech Values," The Quarterly
Journal of Speech, XXII (December, 1936), 544-553.
________ . "A Review of Research in Audience Reaction:
Part I: Experimental Research with Non-student Au
diences," The Quarterly Journal of Speech, XXIV
(October, 1938), 464-483.
"A Review of Research in Audience Reaction:
Part II: Misconceptions of the Function of the
Woodward Ballot and Formulae," The Quarterly Journal
of Speech, XXII (December, 193877 655-672.
Monroe, Alan H. "The Statistical Reliability and Validity
of the Shift of Opinion Ballot," The Quarterly
Journal of Speech, XXIII (December^ 1937), 577-585.
________ . "Testing Speech Performance," Bulletin of the
National Association of Secondary School Principals,
XXIX (November, 1945), 156-164.
Powers, W. T., Clark, R. K., and McFarland, R. L. "A Gene
ral Feedback Theory of Human Behavior," Perceptual
and Motor Skills, XI (1960) , 71-88.
Ring, Kenneth, and Kelley, Harold H. "A Comparison of
Argument and Reduction as Modes of Influence,"
Journal of Abnormal Psychology, LXVI (1963), 92-
102. —
Sanford, Filmore H. "The Use of a Projective Device in
Attitude Surveying," Public Opinion Quarterly, XIV
(Winter, 1950), 690-700.
Scott, William A. "Attitude Change by Response Reinforce
ment: Replication and Extension," Sociometry, XXII
(1959) , 328-335.
________ . "Attitude Change through Reward of Verbal Behav
ior," Journal of Abnormal and Social Psychology, LX
(1957), 72-75.
Thistlethwaite, Donald L., Kemenetzky, Joseph, and Schmidt,
Hans. "Factors Influencing Attitude Change through
Refutative Communications," Speech Monographs,
XXIII (March, 1956), 14-25.
89
Verplanck, William S. "The Control of the Content of Con
versations: Reinforcement of Statements of Opin
ion," Journal of Abno'rmal and Social Psychology, LI
(1955), 71-88.
Vlandis, John W. "Variations in the Verbal Behavior of a
Speaker as a Function of Varied Reinforced Condi
tions , " Speech Monographs, XXXI (June, 1964), 116-
119.
Wechsler, Irving R., and Bernberg, Raymond E. "Indirect
Methods of Attitude Measurement," International
Journal of Opinion and Attitude Research, IV
(1950), 201-215.
Woodward, H. S. "Measurement and Analysis of Audience
Opinion," The Quarterly Journal of Speech, XIV
(1928), 94-111.
Zajonc, Robert B. "The Effects of Feedback and Probability
on Group Success on Individual and Group Perform
ance," Human Relations, XV (1962), 149-161.
Unpublished Materials
Bowman, Fredrick B. "An Experimental Study to Determine the
Amount of Inter-active Thinking Displayed in a Prob
lem-solving Group Discussion.1 1 Unpublished Doctoral
dissertation, The University of Southern California,
Los Angeles, 1957.
Dickens, Dr. Milton. Class lecture. University of Southern
California, Los Angeles, February 16, 19 65.
Fear, Arthur J. "An Experimental Study of the Ability of
Lay Judges to Distinguish between Typescripts of
Individualized Development and Group Discussion Idea
Development." Unpublished Doctoral dissertation,
The University of Southern California, Los Angeles,
California, 1966.
Gibson, James W. "Direct and Indirect Attitude Scale Meas
urements of Positive and Negative Communications."
Unpublished Doctoral dissertation, Ohio State Uni
versity, Columbus, Ohio, 19 62.
Jones, Stanley E. "Attitude Changes of Public Speakers dur
ing the Investigative and Expressive Stages of
Advocacy." Unpublished Doctoral dissertation,
Northwestern University, Evanston, 19 64.
90
;Krueger, David H. Unpublished Doctoral dissertation in
! process. The University of Southern California,
! Los Angeles, California, 19 67.
Mentzer, Ray T., Jr. "Cognitive Dissonance and Attitude
Change in a Communicator: The Effects of Audience
Image and Communication Discrepancy." Unpublished
Doctoral dissertation, Purdue University, Purdue,
1962.
APPEND IXES
91
APPENDIX A
LISTENER'S PERSONAL REACTIONS
AND OBSERVATIONS
In listening to some sixty speeches there were sev
eral things which caught my attention. On the first night
we began with the negative approach in which I would begin
by stating outright that I would be willing to listen to
the speaker but that I was dead set against his ideas be
fore he even began. During the course of the speech I
would shake my head as if in disagreement or agreement
every twenty-five to thirty-five seconds and then shake my
head and utter a noise at the same time. During the course
of the evening I naturally tried to be as consistent as
possible and keep the same timing for each individual, but
this proved to be difficult. Sometimes he wouldn't present
an opening and instead of interjecting a negative thought
at a point where it would be senseless I merely waited un
til the first chance to disagree.
With the negative speakers I found that there were
basically two classifications which the speakers fell
under. The first general classification included those
speakers who, when seeing that I was definitely against
their ideas, immediately fought back and tried to convince
93
94 ;
me that their argument was sound. They couldn't conceive
of anyone thinking different than them. They used comments |
!
such as "I think anyone can see" or "any logical person can }
i
tell that ..." with the underlined portions obvious ref- ;
erences to myself. When I shook my head the speakers would !
often become indignant and even mad as if my disagreement
were a personal attack. A few made caustic remarks and one I
even told me that he wanted to speak to me afterwards. One
ingenious speaker even worked the phrase "with more manners
than most judges" into his speech. One gentleman couldn't
believe that I was disagreeing with him and he kept asking
"why are you doing that?" "Why are you saying that?"
The final characteristic of this classification was
that I noticed that this type of speaker would become even
more determined after each negative thought emitted by me.
Perhaps it was my imagination, but I felt that their voices
rose in tempo and ferocity after each negation. Some rose
to such a pitch that they would stutter in their haste to
convince me that I was wrong.
The second classification that I would place the
negative speakers under would be those who threw in the
towel or just gave up. They wouldn't even try to convince
me that their idea was sound.
The first characteristic of this classification was
those who gave up right at my opening statement. They were
despondent at the beginning of the speech and usually grew
95
progressively worse as their speech was presented. At the
end of their speech they sounded like they didn't believe
in their ideas themselves.
The second characteristic included those who seem
ingly unbothered by my opening statement became as bad as
the first group after a few negative statements and ended
their speeches in much the same manner.
The final impression that I received from this
classification was that they seemed apologetic and every
time I disagreed with them I almost expected some of them
to agree with me. Also, they generally left the room as
soon as possible and it seems as though some of them even
cut their speeches short so they could leave earlier.
After the negative approach we switched to the pos
itive one, in which all of the circumstances were the same
except everything I did was now positive instead of nega
tive .
All of the speakers in this group reacted in about
the same manner. If they started out weakly or unsure of
themselves, they became bolstered by my actions and seemed
to radiate confidence. If they started strongly, they be
came encouraged by my comments and ended even stronger.
After most of the speakers saw that I agreed with
them, they would make a statement and then look for and ex
pect a positive reaction from me. If they received it,
they would nod themselves and proceed.
One very typical example of this group would be one
young man who came in and stated that he thought the Com
munists were the cause of the Watts riots. He started un
certainly and he seemed nervous, but after my first posi
tive comment he became an entirely different speaker. He
became poised and confident. Instead of being nervous, he
was suddenly relaxed. He became so engrossed in his sub
ject that I had to bring it to his attention that he had
run out of time. Truly the fact that I was positive with
this individual made it a completely different speech.
A P P E N D I X B
97
98
PHOTOGRAPH OF LISTENER
PRELIMINARY FACULTY SURVEY MEMORANDUM
99
March 25, 1966
Dear College Professor:
In regard to our conversation last week about administering
your questionnaire let me again briefly state what I would
like you to do. Give the questionnaire to each male student
in your Speech 3 classes. Make sure each student fills in
all the necessary information. Inform them that their an
swers on the questionnaire are in no way connected with the
class, grades, etc. Pick up the questionnaires from the
male students as soon as they have answered them. Please do
not leave them in the hands of the students for any reason.
If the students wish to know anything about the question
naire use your discretion and tell them that it pertains to
a survey being done by another faculty member. Please re
turn the marked questionnaires to my box in the speech arts
office. Your co-operation with this questionnaire is
greatly appreciated.
Paul R. Mattox
100
INITIAL OPINION SURVEY FORM
NAME
AGE ______________________________________
sex ______________________________________;
RACE ______________________________________
CIRCLE YOUR YEAR IN COLLEGE F S J S
SPEECH 3 INSTRUCTOR ______________________
HOUR AND DAY OF SPEECH 3 ________________
WAS POLICE BRUTALITY A SIGNIFICANT CAUSE OF THE WATTS RIOT?
CHECK THE ANSWER THAT INDICATES OR MOST NEARLY INDICATES
YOUR OPINION.
YES
102
FACULTY NOTIFICATION OF SPEAKER SELECTION MEMORANDUM
.Dear Colleague:
The following male students have been selected from your
class to deliver a five minute speech supporting the answer
they indicated on the preliminary questionnaire.
■ The time limits on the speeches are four to six minutes.
!The students are to report to Room ________ at _____________
April ________.
It is strongly recommended that you make this assignment in
the form of a requirement otherwise there would be no in
centive for the students to prepare and deliver a speech at
the prescribed time and hour. Let me stress that several
years of research are involved in this project and even one
or two "failures to show" can jeopardize the entire study.
Any rewards or compensations you wish to give the students
would be an additional incentive but should not be in lieu
of the requirement criteria.
When the study is over I will be pleased to discuss its
purposes and answer any questions that you or your students
may have.
LET ME STRESS AGAIN THE IMPORTANCE OF THE STUDY AND THE
NECESSITY FOR THE STUDENTS TO BE THERE ON TIME AND GIVE
THEIR SPEECHES.
The students and their answers are listed below. (In case
they forgot.)
:Very truly yours,
Paul R. Mattox
104
Linked assets
University of Southern California Dissertations and Theses
Conceptually similar
PDF
An Experimental Study Of An Application Of Game Theory To The Selection Of Arguments By College Debaters
PDF
An Experimental Comparison Of Spoken Communication Developed Individually and Interindividually
PDF
An Experimental Study Of The Accuracy Of Experienced And Inexperienced Speakers In Identifying Audience Behavior As Indicative Of Feelings Of Agreement, Indecision, Or Disagreement
PDF
A Historical Study Of The Speechmaking At The Abilene Christian College Lectureship, 1918-1961
PDF
An Experimental Study Of The Retention And Comprehension Of Poetry Resulting From Silent Reading And From Oral Interpretation
PDF
An Experimental Study Of Effects Of Interest And Authority Upon Understanding Of Broadcast Information
PDF
An Experimental Application Of 'Cloze' Procedure As A Diagnostic Test Of Listening Comprehension Among Foreign Students
PDF
An Empirical Study Of The Behavioral Characteristics Of Sincere And Insincere Speakers
PDF
An Experimental Study Of The Influence Of Subliminal Cue Words On Audience Responses To A Filmed Speaker'S Sincerity, Effectiveness, And Subject Matter
PDF
A Historical And Critical Study Of The Public Address Of James Harvey 'Cyclone' Davis (1853-1940) Of Texas
PDF
An Empirical Study Of Classroom Teaching Of Ethics In Beginning College Public Speaking Courses
PDF
An Experimental Study Of The Ability Of Lay Judges To Distinguish Betweentypescripts Of Individual Idea Development And Group Idea Development
PDF
Rhetoricians On Language And Meaning: An Ordinary Language Philosophy Critique
PDF
An Experimental Study Of Several Methods Of Teaching Basic College Speechcourses With Emphasis On Conservation Of Teachers' Time And Varying Class Size
PDF
A Historical Study Of Speech Education At The University Of Southern California (1880 Through 1950)
PDF
An Intensive Rhetorical Analysis Of Selected Speeches Of Robert Maynard Hutchins: 1940-1955
PDF
A Critical Study Of Frank C. Baxter'S 'Shakespeare On Tv'
PDF
A Descriptive-Analytical Study Of Ethical Standards In Contemporary American Public Address
PDF
An Experimental Study Of Factors Influencing The Perception Of Auditory Flutter
PDF
Christian Science and the rhetoric of argumentative synthesis
Asset Metadata
Creator
Mattox, Paul Richard
(author)
Core Title
An Experimental Study Of The Effect Of Listener Feedback On Speaker Attitude
Degree
Doctor of Philosophy
Degree Program
Speech
Publisher
University of Southern California
(original),
University of Southern California. Libraries
(digital)
Tag
OAI-PMH Harvest,Speech Communication
Language
English
Contributor
Digitized by ProQuest
(provenance)
Advisor
Dickens, Milton (
committee chair
), Borgers, Edward W. (
committee member
), McBath, James H. (
committee member
)
Permanent Link (DOI)
https://doi.org/10.25549/usctheses-c18-591516
Unique identifier
UC11360066
Identifier
6801689.pdf (filename),usctheses-c18-591516 (legacy record id)
Legacy Identifier
6801689.pdf
Dmrecord
591516
Document Type
Dissertation
Rights
Mattox, Paul Richard
Type
texts
Source
University of Southern California
(contributing entity),
University of Southern California Dissertations and Theses
(collection)
Access Conditions
The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the au...
Repository Name
University of Southern California Digital Library
Repository Location
USC Digital Library, University of Southern California, University Park Campus, Los Angeles, California 90089, USA