The effects of androgyny and message expectations on resistance to persuasive communication

MISSING IMAGE

Material Information

Title:
The effects of androgyny and message expectations on resistance to persuasive communication
Physical Description:
xi, 101 leaves : ; 28 cm.
Language:
English
Creator:
Montgomery, Charles Ledford, 1946-
Publication Date:

Subjects

Subjects / Keywords:
Persuasion (Psychology)   ( lcsh )
Sex role   ( lcsh )
Genre:
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )

Notes

Thesis:
Thesis--University of Florida.
Bibliography:
Includes bibliographical references (leaves 94-99).
Statement of Responsibility:
by Charles L. Montgomery.
General Note:
Typescript.
General Note:
Vita.

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 000063613
notis - AAG8812
oclc - 04214249
System ID:
AA00003511:00001

Full Text













THE EFFECTS OF ANDROGYNY AND MESSAGE EXPECTATIONS
ON RESISTANCE TO PERSUASIVE COMMUNICATION










By

CHARLES L. MONTGOMERY


A DISSERTATION PRESENTED TO THE GRADUATE COUNCIL OF
THE UNIVERSITY OF FLORIDA
IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE
DEGREE OF DOCTOR OF PHILOSOPHY







UNIVERSITY OF FLORIDA


1977
















ACKNOWLEDGEMENTS


To say thank you to those people who are responsible

for my completing this last act of faith in the ritual of a

doctoral program is redundant for most and unnecessary for

the rest. But, somehow, not to include their names in this

document would be inappropriate at best and arrogant at

worst.

So here they are.

In my time at Florida I have inadvertently, uninten-

tionally, accidentally, mindlessly, unwittingly, and by sheer

chance caused no small measure of anger, ire, and turmoil

with my manner of acting and reacting to people and situa-

tions. So let it be perfectly clear to all that there is no

ulterior motive for or hidden meaning in the order in which

I talk about those who have been so important to me over

these last three years. They are simply listed in the order

of their appearance in my life.

TO...

...My parents: who at the same time know me the best

and the least...Roger and Joanne: who enabled me to be

extravagant in my times of need...Barbara: who watched my

complexities grow with the understanding and compassion of

a lifetime...Angela: for whom "Dr. Daddy" says it all...









Michael: who gave me the insight into a better way and who

had the ultimate patience to see it through...Judee: for

whom being androgynous means the best of both worlds...

Bill: who showed me what it means to relax and whose inner

strength I have grown to admire...Miller: a friend who

knows more than most but too often keeps it to himself...

Tom: who balances the best of New York and Evanston...

Marshall: without whom the Odd Couple would be just another

television show and who may never surprise us some day...

Don, Don, Mike, Gill, and Bill: the boys for whom exciting-

unexciting takes on more meaning than even the KIS can

measure...Marv: for whom the word scholar was invented...

Joe, Rick, and Ralph: who made my old job look easy and

whose friendship helped me to keep my sanity...Sharon: who

gave me strength when I needed it most and whose dreams

should always come true...Bob: who made me glad I was

left-handed...Ron: for whom a broken promise is less

important than an ambition fulfilled...Jenny: whose blue

eyes gave me the strength for the final effort...

...THANKS.

Without all of these people this project would have

been impossible. Without some of them I would have been

done a year earlier, but I would have been ten years more

foolish.


iii
















TABLE OF CONTENTS

CHAPTER Page

ACKNOWLEDGEMENTS . ii

LIST OF TABLES .. . vi

LIST OF FIGURES . v

ABSTRACT .................

I INTRODUCTION. ... . 1

The Problem. . 1
The Resistance to Persuasion Paradigm. 2
Sex of the Receiver in the Resistance
Paradigm. . 9
Sex-Roles in American Society 13
Androgyny . 16
Sex-Role and Resistance to Persuasion. 18
Summary of Rationale . .. 24
Hypotheses . 28

II METHODS AND PROCEDURES. . 30

Overview . .. 30
Preparation of Experimental Materials. 31
Dependent Measure. . .. 32
Independent Measures. . .33
Manipulation Checks. . .. 35
Experimental Procedures. . 36
Experimental Design. . 38

III RESULTS . . 42

Viability of Experimental Messages 42
Control Group Analysis .. 44
Threat . .. 44
Expectancy. . .. 45
Credibility . 45
Anticipatory Attitude Change Measures. 46
Manipulation Checks for Experimental
Groups. ... . .. 48













Threat . .
Expectancy .
Reliability . .
Homogeneity of Variance. .
Tests of Hypotheses .
Experimental and Control Group
Comparisons . .
Supplemental Credibility Analysis.

DISCUSSION . .

Competing Explanatory Principles .
Research Implcations .
Summary . .


Page

. 48
* 50
* 52
* 52
. 53

. 57
. 57

. 60

. 70
. 73
. 75


APPENDICES


BEM SEX ROLE INVENTORY .

SITUATION INDUCTION . .

SOURCE AND EXPECTANCY INDUCTIONS. .

EXPERIMENTAL MESSAGES . .

DEPENDENT MEASURE ... .

CREDIBILITY, THREAT AND EXPECTANCY
MEASURES . .

DISTRIBUTION OF DIFFERENCE SCORES ON
ATTITUDE MEASURES FOR EXPERIMENTAL
GROUPS . .

REFERENCES . .

BIOGRAPHICAL SKETCH . .


. 78

. 81

. 83

. 86

. 89



. 91


. 93

. 94

* .100
















LIST OF TABLES

Table Page

1 Means, standard deviations, and t-tests for
difference scores comparing the no message
control group with the message only control
groups. . ... .. .42

2 Means for sex of receiver and message type
manipulation check. . ... 43

3 Means for threat scores of control groups 45

4 Means for expectancy scores of control groups 46

5 Means for extroversion ratings for control
groups. . ... .. .47

6 Means, standard deviations, and t-tests for
difference scores comparing the no message
control group with the anticipatory attitude
change control groups .. . 48

7 Means for threat scores in experimental
conditions. .. . . 49

8 Means for expectancy scores for receiver sex-
type X expectancy of message interaction in
experimental conditions . ... .50

9 Means for expectancy scores for source sex-
type X expectancy of message interaction in
experimental conditions . ... 51

10 Results of Cronbach's alpha test for internal
reliability of dependent measures and
manipulation checks . 53

11 Mean difference scores and analysis of
variance summary for difference scores of
experimental groups . 55









Table Page

12 Dunnett's t-test analysis comparing no
message control group difference score
means with those of the experimental groups 58

13 Means for sex-type of source X message
expectancy interaction for experimental
subjects' sociability scores. . 59


vii

















LIST OF FIGURES

Figure Page

1 The design of the investigation ... 40

2 Descriptions of control and experimental
groups. . . .. 41


viii









Abstract of Dissertation Presented to the
Graduate Council of the University of Florida
in Partial Fulfillment for the Requirements for the
Degree of Doctor of Philosophy



THE EFFECTS OF ANDROGYNY AND MESSAGE EXPECTATIONS
ON RESISTANCE TO PERSUASIVE COMMUNICATION

By

Charles L. Montgomery

August 1977

Chairman: Michael Burgoon
Major Department: Speech

A review of the literature suggests several key factors

in the resistance to persuasion process. Receivers must be

made aware of the existence of an impending persuasive mes-

sage. Defenses against the impact of the coming message

must be available. Receivers must be motivated to use avail-

able defenses. Receivers must be given the opportunity to

build defenses against an attack message. Finally, when

motivated to defend themselves, receivers often produce

counterarguments which they use to combat the appeals in the

attack message.

The literature also reveals that while attitude change

researchers have paid close attention to receiver variables,

those who have studied resistance to persuasion have not

been so careful. More specifically, few resistance studies

have considered the sex of the subjects in their theoretic

formulations, and no resistance to persuasion researchers

have investigated the effects of sex-role orientations on

individuals' abilities to resist persuasive discourse. Since

ix









sex-role stereotypes are pervasive and commonly held in this

society, this study uses them as a basis for operationally

defining the basic principles in the resistance to persuasion

paradigm. It is argued that differing levels of threat and

motivation will affect receivers' productions of counter-

arguments.

Two hypotheses are tested. First, there will be an

interaction between expectancy and the sex-typing of the re-

ceiver and source of a persuasive message such that:

(A) Given traditional receivers, unexpected messages from

traditional sources will elicit more resistance to persuasion

than expected messages while expected messages from non-

traditional sources will elicit more resistance to persuasion

than unexpected messages. (B) Given nontraditional receivers,

expected messages from traditional sources will elicit more

resistance to persuasion than unexpected messages while

unexpected messages from nontraditional sources will elicit

more resistance to persuasion than expected messages.

(C) Given traditional sources, expected messages will

elicit more resistance to persuasion from nontraditional

receivers than traditional while unexpected messages will

elicit more resistance to persuasion from traditional re-

ceivers than nontraditional. (D) Given nontraditional

sources, unexpected messages will elicit more resistance to

persuasion from nontraditional receivers than traditional.

Second, given traditional receivers, males will demonstrate

more resistance to persuasion than females.









The experiment is a 2 X 2 X 2 X 2 design. The factors

are: (1) sex, (2) receivers' sex-role: traditionally sex-

typed versus nontraditionally sex-typed, (3) sources' sex-

role: traditionally sex-typed versus nontraditionally

sex-typed, and (4) sex-role related message expectations:

expected message versus unexpected message. The dependent

variable is resistance to persuasion and is measured on a

four-item, seven interval semantic differential-type instru-

ment. Receivers' sex-roles are measured with the BEM Sex

Role Inventory. Manipulation checks include threat scores,

expectancy scores, anticipatory attitude change measures, and

credibility measures. These measures have sufficiently high

internal reliabilities. Proper control comparisons are made.

The procedure involves inducing source characteristics, sug-

gesting the existence of an impending persuasive message, and

creating expectations concerning the content of the message.

The first hypothesis is confirmed; the second hypothesis

is not. A significant three way interaction indicates that

receivers' sex-roles, sources' sex-roles, and sex-role re-

lated message expectations combine to be important mediators

of resistance to persuasion. The exaggeration effect of

positive versus negative violations of expectancies is also

found. The results of the threat and expectancy scores in-

dicate support for the deductive model developed in the theo-

retic rationale. The results of the other manipulation

checks reveal that credibility and anticipatory attitude

change can be ruled out as competing explanations for the

results.
















CHAPTER I
INTRODUCTION


The Problem


A significant amount of research has focused on iso-

lating factors which influence individuals' abilities to

resist persuasive attacks. Much of this research has been

aimed at establishing the comparative merits of differing

theoretical perspectives within the resistance to persuasion

paradigm. This concentration on general theoretical issues

has caused most resistance to persuasion researchers to

treat the receivers of a persuasive communication as an

isomorphic, homogeneous group. Thus, very little research

effort has been made to determine how variations in people

exposed to persuasive communication affect their abilities

to withstand persuasive appeals.

In contrast, researchers who have focused on persuasion

rather than resistance to persuasion have given careful

attention to the receiver variables that affect persuasibil-

ity. In this area considerable attention has been given to

the sex of the receiver as an important mediator of the

success of persuasive attempts. More recently, the sex-role

(cultural and societal expectations linked to sex) of the

receivers and sources of persuasive communication have been





2


studied. However, sex and sex-role have not been given the

same consideration by resistance to persuasion investiga-

tors. The purpose of this investigation is to determine how

the sex-roles of receivers and sources of persuasive com-

munication influence the ability of receivers to resist

persuasive attacks.


The Resistance to Persuasion Paradigm


Certainly common factors affect individuals' suscepti-

bility to persuasive communication and their ability to

withstand attempts to influence. While researchers in both

paradigms are often concerned with similar principles,

creating attitude change and inducing resistance to persua-

sion are not best viewed as opposite extremes of the same

process (Ulman & Bodaken, 1975). Lack of demonstrated

attitude change in any one study should not be equated with

resistance to persuasion. The resistance to persuasion

paradigm requires an active process of inducing people to

resist subsequent persuasive discourse. Before resistance

to persuasion can be elicited, then, the receivers must be

made aware of the probability of an impending attack.

There are several different theoretical formulations

concerning the cognitive operations which produce resistance

to persuasion (c.f. McGuire, 1964; Papageorgis, 1968;

Tannenbaum, 1966). However, it is generally accepted that

before notification of a coming attack can produce resistance,

that notification must motivate the receiver to resist the








attack. Research has demonstrated that merely having knowl-

edge of the existence of an attitudinally discrepant message

is not usually sufficient to produce resistance to that

message (Cooper & Jones, 1970; Papageorgis, 1968). In

addition, the simple presence of a warning is also not

sufficient to produce resistance to an impending attack

(McGuire & Papageorgis, 1962; Tannenbaum, 1967). Again,

these findings indicate that individuals must somehow be

motivated to resist persuasion.

However, motivation to defend is also not, in and of

itself, sufficient to produce resistance. In addition to

motivation, receivers must also be given time in which re-

sistance can be built (McGuire, 1964; Tannenbaum, 1966).

Apsler and Sears (1968) suggest that warning individuals of

an impending attack gives those'individuals time to delib-

erate and to form responses which are consistent with their

prior cognitions.

A common explanatory principle advanced in the re-

sistance to persuasion literature is that in order for

notification or warning of a coming attack to elicit re-

sistance, that notification must in some way be threatening

to the receiver (McGuire, 1973; McGuire & Papageorgis, 1961).

Roberts and Maccoby (1973) suggest that threat may change

the disposition of a receiver such that it directly induces

resistance to persuasion. In addition, threat may motivate

the receiver to prepare and practice defenses against the

impending attack.









Papageorgis (1968) suggests that there are basically

two types of warnings. One type simply consists of inform-

ing the receiver of the topic of a coming attack and the

direction of the arguments in that message. The second type

consists of informing the receivers of the "persuasive

contexts" of the coming communication; that is, giving the

receiver more specific information about the nature of the

communication and the communicator. Papageorgis contends

that there are three kinds of persuasive contexts: neutral,

positive, and negative. Neutral contexts generally have no

effect on communication acceptance. Positive contexts tend

to enhance attitude change by establishing favorable expec-

tations of the coming message, and negative contexts tend to

inhibit attitude change by creating basically negative

expectations concerning the impending attack message.

Freedman and Sears (1965) tested the hypothesis that

there could be three methods by which receivers, when warned

that a threatening communication is coming, could become

resistant to that communication. Receivers could derrogate

the source of the communication, pay less attention to the

message, or activate belief defense mechanisms. These belief

defense mechanisms, according to Freedman and Sears, could

consist of rehearsing, recalling, or constructing arguments

that would weigh against impending attack messages. This

"counterargumentation" would reduce the receivers' suscepti-

bility by giving them more confidence in their initial

positions and more information with which to combat competing









arguments. The results of the Freedman and Sears experi-

ment demonstrated no effects for either source derrogation

or decreased attentiveness. However, they did find a trend

toward the direction of activation of belief defense mech-

anisms. Freedman and Sears indicate that their experimental

procedures may not have provided enough time for their

subjects to develop defenses and that given sufficient time,

the subjects in their study may well have generated

significantly more counterarguments when warned of a coming

attack.

Festinger and Maccoby (1964) and McGuire (1964) suggest

that persons become more resistant to persuasive communica-

tion when they are motivated to carry on a "silent dialogue"

with the communicator. This counterarguing serves as a

belief defense mechanism and lowers the effectiveness of a

persuasive message.

Brock (1967) found that the more discrepant receivers

perceived an anticipated message to be, the more they were

motivated to counterargue. Brock also found a trend to

produce more counterarguments when receivers perceived an

intent to persuade in the communication. Finally, Brock's

data supported his prediction that giving receivers a sample

counterargument enhanced their ability to counterargue and

increased their ability to resist a persuasive message.

Brock found a strong correlation between the production of

counterarguments and the ability to resist a persuasive

attack. Brown's (1971) findings lend additional support to









Brock's analysis. Brown found that threat stimulates the

production of counterarguments in individuals and that there

is a significant negative correlation between levels of

threat and attitude change.

In two experiments, Osterhouse and Brock (1970) gen-

erally supported the counterargumentation formulation for

eliciting resistance to persuasion. They studied the rela-

tionships among distraction, perceived intent of a communi-

cator to persuade, the amount of threat in a communication,

and the amount of counterargument production. Osterhouse and

Brock found that subjects who were distracted during the

presentation of a persuasive message were inhibited from

producing counterarguments and thus were more easily

persuaded than those subjects who were not distracted during

the communication. They also found an interaction between

perceived threat of the communication and the intent of the

source to persuade such that when exposed to a highly

threatening message, more counterarguments were produced by

subjects who were highly cognizant of persuasive attempts

than by those subjects who perceived a low level of intended

persuasion on the part of the communicator. While Osterhouse

and Brock did not find a significant relationship between

the threatening nature of a communication and the amount of

counterargumentation that communication would produce, they

blamed this finding on an impure threat manipulation that

may have been too complex for .some of their subjects to

understand.









Roberts and Maccoby (1973) indicate that three general

factors affect the production of counterarguments: the

availability of defenses, the opportunity to utilize those

defenses, and some motivation to use those defenses.

Osterhouse and Brock indicate that the maximal condition for

the production of counterarguments exists when messages

"advocate action having negative consequences for the recipi-

ent or [contradict] a strongly held opinion in an area

having important implications for the individual" (p. 335).

Taken together, the above findings indicate that counter-

argumentation is an important method by which individuals

protect themselves from threatening persuasive communication.

These results are not inconsistent with the formula-

tions of McGuire's (1964) inoculation theory. McGuire

suggests that by giving persons'weakened doses of a coming

persuasive attack, those persons can be made more resistant

to that attack. He argues that these weakened versions of

the persuasive message motivate persons to build defenses

against that attack and thus reduce its effectiveness.

Roberts and Maccoby (1973) note that two assumptions are

fundamental to McGuire's conceptualization of the resistance

to persuasion process: (1) resistance to persuasion is a

function of practicing the defense of a belief, and (2) prac-

tice is a function of motivation to defend. Essentially,

McGuire contends that when threatened, persons will be

motivated to create defenses against an impending attack in

a manner which will allow them to counter the arguments in









that attack. Thus, they will become immune to the effects

of that persuasive attempt.

It should be noted that all researchers do not adhere

to the counterarguing conceptualization of the resistance to

persuasion process. Miller and Barron (1973) suggest that

there are major weaknesses in the counterarguing explanations

of resistance to persuasion. They argue that there are via-

ble alternatives for the counterarguing predictions which do

not depend upon counterargumentation, e.g. Tannenbaum's (1967)

Congruity Theory, and that there are serious methodological

problems involved in measuring counterarguing. While there

are certainly alternative explanations for many of the

counterarguing formulations, Miller and Barron do not at-

tempt to provide evidence or reasoning to suggest that one

formulation is more sound, more accurate, or more precise

than another. They simply indicate that alternate explana-

tions to the counterarguing explanation of resistance to

persuasion exist without any attempt to defend them. They

offer no higher order explanations that would subsume the

results of research supporting the counterarguing representa-

tions, nor do they suggest any method by which one explana-

tion could be tested against its alternatives. Moreover,

Miller and Barron's methodological criticisms, while largely

correct, in no way disturb the theoretical ground upon which

the counterarguing conceptualizations are built. In fact,

in their conclusion, Miller and Barron indicate that the

counterargumentation theoretical approach should not be









abandoned. Rather, these formulations need more refinement

and better measurement techniques.

While the counterargumentation explanation of the re-

sistance phenomenon is not universally accepted, the approach

is theoretically sound and has received significant empiri-

cal support. It is very likely that when receivers become

aware of an impending, attitudinally discrepant persuasive

attack, they will be motivated to build defenses against that

attack message. These defenses will take the form of counter-

argument production which will enable receivers to cogni-

tively refute the arguments in the message and thus become

more resistant to its appeal.


Sex of the Receiver in the Resistance Paradigm


While the effects of threat, motivation and warning on

individuals' ability to withstand persuasive attacks have

received a considerable amount of attention, these efforts

have produced "few generalized findings" (Hollander, 1974,

p. 272). Hollander notes that different experimental

designs and differences in the combination and operationali-

zation of variables have reduced the possibility of an

important synthesizing breakthrough in the study of

resistance to persuasion.

A major factor that has limited cross experimental

generalization and inhibited advanced theory building is the

great discrepancy among the samples that resistance re-

searchers have studied. These differences are very









significant when the sex composition of the samples is

considered. Hollander (1974) criticizes the McGuire and

Millman (1965) study for its use of an imbalanced ratio of

female to male subjects. Eighty-eight per cent of the indi-

viduals in McGuire and Millman's investigation were female,

and Hollander argues that this unrepresentative sample

probably biased their results. McGuire and Millman concluded

that forewarning alone had little impact on the reception of

a persuasive message. However, when Hollander replicated

their study with a more representative distribution of males

and females, he found that there were sex differences that

affected subjects' resistance to persuasion. He found that

females were able to resist persuasive attempts better than

males with no warning while males were more resistant than

females in the warning condition.

Dean, Austin and Watts (1971) note that the results of

Apsler and Sears' (1968) study might have been biased be-

cause they used only females in their sample. Dean, Austin

and Watts partially replicated the Apsler and Sears study

and found an interaction such that forewarning inhibited

change for males while making females more susceptible to

change. Apsler and Sears reported simply that only warning

facilitated attitude change. Thus, in two instances analy-

sis of the sex differences among the receivers in the samples

studied added further specification to the results of

earlier experiments in resistance to persuasion.








The analysis of Hollander (1974) and Dean, Austin and

Watts (1971) takes on added importance when the sex of the

subjects in a cross section of resistance studies is ex-

amined. Several studies fail to even mention the sex of

the subjects they studied (Brock, 1967; Freedman & Sears,

1965; Hardyck, 1966; Manis & Blake, 1963; McGuire, 1962;

McGuire & Papageorgis, 1961; Papageorgis, 1970). Festinger

and Maccoby (1964) used an all male sample. Three studies

utilized all females in their research (Apsler & Sears,

1968; Cooper & Jones, 1970; Deaux, 1968). Two investiga-

tions employed both males and females but the authors failed

to report the proportion of each (Kiesler & Kiesler, 1964;

Dinner, Lewkowitz & Cooper, 1972). Several investigations

studied samples that included significantly more females

than males (Manis, 1965; McGuire, 1961, McGuire & Millman,

1965). Anderson's (1967) sample consisted of significantly

more men than women. Only Dean, Austin and Watts (1971),

Hollander (1974) and Papageorgis (1968) reported using ap-

proximately equal numbers of males and females. It is

readily apparent that most investigations in the resistance

to persuasion paradigm have failed to adequately consider

the sex of the receiver of a persuasive message in their

analyses.

Generally, randomization of sex across experimental

conditions would preclude criticism of these investigators'

failure to consider this variable in their formulations.

However, the above review clearly demonstrates that most of









the investigations of resistance to persuasion have failed

to either randomly select or assign subjects in the samples

of the population to which they wish to generalize their

results. Thus, the generalizability of much of the previous

research in resistance to persuasion may be questioned. The

advancement of theory in the resistance to persuasion

paradigm partially depends upon the ability of researchers

to compare and contrast the results of a variety of experi-

ments. The lack of continuity among the subjects studied by

investigators in this area makes such comparisons difficult

at best and probably significantly inhibits the development

of theory in this major area of research.

More importantly, the sex-role of the receiver has

been given even less attention in the resistance to persua-

sion literature. In fact, no published study in this review

considered the sex-role of the receiver of a persuasive

communication in the resistance to persuasion paradigm. In

the persuasion literature, Montgomery and Burgoon (1977)

predicted and found an interaction between sex and sex-

typing such that traditional males were significantly less

susceptible to persuasive appeals than either traditional

females or nontraditional individuals. While Montgomery

and Burgoon did not use a warning approach, their data do

indicate that the sex-role of the receiver is an important

mediating variable in determining individuals' ability to

withstand a persuasive attack .and that it is dangerous to








base predictions concerning persuasibility solely on the

basis of the sex of the receiver.


Sex-Roles in American Society


Despite the recent emphasis on equalizing the places of

men and women in this society, most individuals still view

males and females as having different roles in our social

structure (Beuf, 1974; Ellis & Bentler, 1973; Garske, 1975;

Halas, 1974). Garske notes that most research indicates

that sex-role stereotypes are "pervasive, clearly delin-

eated, and stable" (p. 32). Other research has demonstrated

that among college students, sex-role stereotypes are clearly

defined and held in common agreement by men and women

(Rosenkrantz, Bee, Vogel, Broverman & Broverman, 1968).

While there are some individual 'differences, Spence,

Helmreich and Stapp (1975) conclude that college students'

perceptions of the behaviors of traditionally oriented males

and females accurately represent the behaviors of those

individuals in this society. Finally, O'Leary (1974) notes

that people expect males to exhibit masculine behavior and

females to demonstrate feminine behavior. She concludes

that "norms governing the approved masculine or feminine

image are clearly defined and consensually endorsed" (p. 813).

In short, members of this society tend to have strong,

lasting, and socially consistent expectations of the

differing sex-roles attributed to males and females.









While the concept of sex-role stereotyping has been

studied by many investigators, few researchers have taken

care to explain what definition of role they are using or

have bothered to define exactly what they mean by the concept

of a "sex-role stereotype." For example, none of the studies

reviewed thus far have specifically delineated what is meant

by a sex-role. Instead, the researchers seem to depend upon

the reader's intuitive judgment of what constitutes a sex-

role stereotype. Kagan (1964) defines a sex role standard

as the "learned association between selected attributes,

behaviors, and attitudes, on the one hand, and the concepts

of males and female, on the other" (p. 138). He suggests

that persons' sex-roles are the sum of their perceptions of

culturally accepted characteristics for men and women.

Strictly speaking, the term role may be inappropriate

when discerning among individuals' behaviors based upon

their sex. To most role theorists, these kinds of behavior-

ally oriented determinations are usually labeled positions

not roles. Thomas and Biddle (1966) define position as a

"collectively recognized category of persons for whom the

basis of such differentiation is their common attribute,

their common behavior, or the common reactions of others

towards them" (p. 29). Role, on the other hand, is typically

defined as those behaviors linked to a person's occupying a

certain position (Thomas & Biddle, 1966; Sarbin & Allen,

1969; Shaw & Costanzo, 1970). Thus roles may be defined as








those sets of behaviors persons exhibit because of the

positions they occupy in this society.

Researchers investigating sex-roles are most often con-

cerned with one of two types of behaviors: (1) actions of

persons based upon their being either male or female and

(2) the actions of those persons who expect others to behave

to certain patterns based upon the target person's being a

member of a class of people delineated by sex. Both of

these concerns clearly fit into the category of exploring

sex-related positions and not sex-roles. However, since

there is no consensus among role theorists on exactly what

constitutes a role (Burris, 1976), and since the term

"sex-role" consistently appears in the literature, it is

probably not inappropriate to continue to use this classifi-

cation as long as it is made clear what definition the term

implies.

Shaw (1976) points out that role may take on three

meanings: expected role, perceived role, and enacted role.

An expected role consists of a set of behaviors which in-

dividuals anticipate others exhibiting. A perceived role

can be defined as the behaviors actors believe they should

exhibit. Finally, an enacted role consists of those be-

haviors which persons actually demonstrate. Shaw notes that

there is usually considerable agreement between perceived

and expected roles among individuals. At the same time,

enacted roles tend to depend upon perceived role and the

person's behavioral tendencies.








An overview of the sex-role research indicates that

most researchers are chiefly concerned with how individuals

view the actions of others in terms of their sex linked be-

haviors. Thus, most sex-role research seems to concentrate

on what Shaw (1976) identifies as expected roles. In addi-

tion, research involving sex related communication variables

most often deals with how persons perceive the actions of

others and how they expect others to act. Accordingly, this

line of research may also be termed as studying expected

sex-roles.


Androgyny


Originally, research involving sex differences relied

on attributing behavioral differences only to one's being

either male or female. However; explanations for phenomena

based solely on this dichotomous differentiation often proved

to be inadequate (Bem, 1974; Montgomery & Burgoon, 1976).

More recently, researchers have been concerned not simply

with sex but also with the concept of masculinity/femininity.

Bem (1974), for example, places individuals into one of

five categories based upon self-rating of femininity/masculin-

ity. According to Bem, individuals of both sexes may be

classified as being masculine, near-masculine, androgynous,

near-feminine or feminine. Masculine persons are those who

typically exhibit traits and behaviors traditionally associ-

ated with maleness, and feminine individuals are those who

typically exhibit traits and behaviors associated with being








female. The "near" categories include those persons who are

not completely masculine or feminine but who are approaching

those positions. The key concept in Bem's work, and her

most important contribution to the field, is that of

androgyny. Instead of viewing persons along a continuum of

feminine through masculine, Bem's formulations depend upon

the concept that individuals may have both masculine and

feminine characteristics in their behaviors. These an-

drogynous people are able to exhibit either masculine or

feminine behaviors (or both) depending upon the social situa-

tion. According to Bem (1974), androgynous individuals may

be "both masculine and feminine, both assertive and yielding,

both instrumental and expressive--depending on the situation-

al appropriateness of these various behaviors" (p. 155).

People may also be classified as either traditionally

sex-typed or nontraditionally sex-typed based upon both sex

and masculinity/femininity. In this categorization, mascu-

line males and feminine females are considered traditionally

sex-typed. On the other hand, androgynous individuals,

masculine females, and feminine males are considered non-

traditionally sex-typed. Several sex-role researchers have

expressed positions which are essentially similar to Bem's

conceptualizations of the traditional and nontraditional sex-

roles in this society (Carlson,1971; Constantinople, 1973;

Heilbrun, 1976; Spence, Helmreich & Stapp, 1975).

The foregoing review of the role and sex-role literature

may be summarized in the following manner. Traditional








sex-roles are filled by persons whose behaviors are associ-

ated with actions typically expected of males and females:

feminine females and masculine males. Nontraditional sex-

roles are filled by those individuals whose behaviors other

persons would not normally associate with a particular sex:

feminine males, masculine females, and androgynous persons.

In addition, the roles generally expected of traditionally

oriented persons are clearly delineated and consensually

accepted by most members of this society.


Sex-Role and Resistance to Persuasion


An important link between sex-role research and the

resistance to persuasion literature lies in the connections

between role expectations and violations of those expecta-

tions. It has already been demonstrated that sex-role

expectations permeate our society. Thus, it is likely that

violations of these expectations have major ramifications.

In the persuasion literature, Brooks (1970) indicates

that receivers generally have a set of expectancies con-

cerning what behaviors they anticipate sources to exhibit.

The result of sources violating these expectations may be

that receivers exaggerate the behaviors actually being ex-

hibited. For example, if a source is initially perceived as

being negative, and that source demonstrates more positive

behaviors than expected, the receivers will tend to over-

estimate how positive the unanticipated behaviors are.

McPeek and Edwards (1975) found that positive violations








increase persuasibility only for dissimilar sources and sug-

gested that observer source and situational factors may be

important mediating factors to the exaggeration effect.

Kohn and Snook (1976) demonstrated that positive violations

of audience's expectations concerning the nature of a per-

suasive message led to attitude change in the direction of

the position advocated in the message. The reverse is also

true. If receivers anticipate a communication to be favor-

able to a position those receivers hold, and the communica-

tion actually advocates an opposite stance, they will

probably exaggerate their evaluation of that communication

in a negative manner.

Burgoon and Chase's (1973) data can be interpreted as

indicating that when the source of a persuasive message

violates receivers' expectations about the nature of the

message in a positive manner, the result is increased atti-

tude change in favor of the position being advocated by the

source. Burgoon, Jones and Stewart (1974) extend this

reasoning and argue that when a source violates receivers'

anticipations in a negative manner, those receivers will

become more resistant to the position being advocated by

that source. In fact, Burgoon, Jones and Stewart indicate

that this situation may lead to attitude change away from

the position supported by the source. Clearly, it can be

argued that receivers may have definite expectations concern-

ing people's communication behaviors and that violations of

these expectations may have either beneficial or deliterious








effects depending on whether or not the violations are per-

ceived in a positive or negative manner by the receivers.

To date, no persuasion or resistance to persuasion

studies have investigated the effects of sex-role viola-

tions on attitude change and resistance to attitude change.

However, there is some evidence in the sex-role literature

to indicate what the relationships between sex-role expecta-

tions and persuasibility may be. Bem (1975) suggests that

sex-role stereotyping restricts an individual's behavioral

choices such that traditionally sex-typed persons strive to

identify the actions of others with a traditionally sex-typed

bias. That is, traditionally oriented people tend to expect

others to behave in a traditional manner. Accordingly, they

could be expected to anticipate that other communicators

would exhibit typically sex-typed behaviors. Moreover, these

expectations would probably be heightened if the receivers

were given information about the sex-role of the communica-

tor. Thus, it is likely that traditionally sex-typed

receivers would expect traditional sources to exhibit more

traditional communication behaviors than nontraditionally

sex-typed sources.

Some research indicates how people's sex-roles influ-

ence what behaviors they expect from others and how they

react to sex-role violations. Ellis and Bentler (1973)

found that more liberal and less status seeking males and

more liberal and masculine females both tend to oppose

traditional sex-role standards of behavior. Taynor and Deaux









(1975) discovered that traditionally sex-typed persons

least expect a man to perform in a feminine role and a woman

to perform in a masculine role. Those studies provide some

evidence to suggest that sex-typing determines how both

traditional and nontraditional persons evaluate the sex-role

behavior of others.

The specific consequences of persons violating tradi-

tional sex-role standards has received limited attention.

However, a series of experiments completed by Bem and Lenney

(1976) provides some indication as to what the consequences

of sex-role violations may be. She notes that tradition-

ally sex-typed persons react negatively to inappropriate or

out-of-role behavior. In one experiment it was found that

traditionally oriented subjects actually avoided cross-sex-

typed tasks even if these tasks were more rewarding than

traditionally oriented tasks. Both masculine men and

feminine women actively avoided inappropriately sex-typed

tasks. In another experiment, Bem found that traditionally

sex-typed persons feel very uncomfortable when forced to

perform a nontraditional task.

In contrast, Bem and Lenney (1976) reported that an-

drogynous persons have less trouble adapting to nontraditional

roles and have no feelings of discomfort while performing what

are typically considered inappropriately sex-typed tasks.

Heilbrun (1976) found that androgynous individuals,

while still holding expectations of sex-typed behaviors

similar to their sex-typed counterparts, were better








adjusted and more capable of integrating nontraditional

behaviors. In summary, there is evidence to suggest that

there are considerable differences in how traditional and

nontraditional persons react to unexpected sex-typed be-

havior. Traditionally oriented persons are made to feel

uncomfortable and react negatively to nontraditionally sex-

typed behavior while nontraditional persons are generally

able to adapt to such unanticipated behavior.

Linking the above findings to the resistance to persua-

sion literature, it is probably correct to assume that

traditional persons who observe nontraditional behavior from

sources of communication may be threatened by that behavior.

Moreover, this effect may be magnified when traditionally

sex-typed persons see traditionally sex-typed sources exhibit

inappropriate sex-typed behavior. On the other hand, among

nontraditional persons, out-of-role communication behavior

will not be as threatening. These receivers are better able

to cope with nontraditional behaviors than their sex-typed

counterparts. Thus, the sex-typing of receivers, the sex-

typing of sources, and the sex-role related behavior of

sources may all be important mediators of the receivers'

ability to resist persuasive communication.

While the foregoing summary of the sex-role and resist-

ance to persuasion literature forms the basis for predictions

concerning resistance to persuasion grounded in sex-typing

of receivers and sources and expectations of communication

behaviors, there is additional evidence to suggest that there









are differences in abilities to resist influence among

traditional receivers. The persuasion literature indicates

that females in general may be more susceptible to persua-

sive attempts than men (Burgoon, Jones & Stewart, 1974) and,

more importantly, that feminine females are more susceptible

to persuasive influences than masculine males (Montgomery &

Burgoon, 1977).

Bem (1975) found that traditional females were less

aggressive, more conforming, and demonstrated lower levels

of self-esteem than their masculine male counterparts. In

addition, Spence, Helmreich and Stapp (1975) found that

feminine females had significantly lower levels of self-

esteem than did masculine males. Goldberg's (1975) data

support Bem's conclusion that traditionally oriented females

tend to conform more than traditional males.

Again, these personality differences may be linked to

the resistance to persuasion literature by way of threat and

motivation. Aggressive, nonconforming males will probably

be more involved with an issue than less aggressive, con-

forming females. This difference in involvement will

directly affect the amount of threat an impending attack

message will cause such that traditional males may be more

threatened by belief discrepant messages than traditional

females. Thus, the traditional males will be more motivated

to create defensive mechanisms than traditional females.

This conclusion is supported by the results of Hollander's

(1974) and Dean, Austin and Watts' (1971) studies which









indicate warning inhibits change in men but enhances change

among women.


Summary of Rationale


The preceding review of literature indicates that there

are three factors which mediate sex-typed receivers' ability

to resist a persuasive message: (1) the amount of threat

and motivation to counterargue, (2) the communicators' ful-

filling of receivers' expectations of behavior, and (3) the

persuasive context of the communication situation. In

addition, it has been demonstrated that individuals in this

society have definite expectations concerning the behaviors

of sources of communication, and these expectations are

pervasive and deeply held. Moreover, it has been suggested

that there are differences in how traditional and nontradi-

tional receivers react to out-of-role behavior by sources of

persuasive messages.

In summary, this review of literature suggests the

following lines of argument. Traditionally sex-typed indi-

viduals who view traditional communication behavior from

traditional sources of persuasive communication will likely

be very amenable to change. These receivers' expectations

will be fulfilled, and they will not be threatened by the

communication. Thus, a generally positive persuasive context

will be generated which will result in change in the direc-

tion of the position being advocated in the message. In

contrast, when traditional receivers view nontraditional








communication behaviors from a traditional source, that out-

of-role behavior will constitute a negative violation of

those receivers' expectations. These receivers will be

threatened by such atypical behavior and motivated to pro-

duce counterarguments against the position being advocated by

the source. Hence, a negative persuasive context will be

generated, and those traditional receivers will become

resistant to the persuasive message.

Traditional receivers who are subjected to persuasive

attempts by nontraditional sources advocating a nontradi-

tional message will be less swayed than when presented with

a traditionally oriented message from a nontraditional

source. In the latter case, the traditional receiver will

perceive the traditional behavior of the nontraditional com-

municator as a positive violation of expectations. This

view will create a very positive persuasive context and will

not be threatening to those traditional receivers. Thus,

they should be susceptible to opinion change in the direction

of the position being advocated by the source.

A similar line of argumentation should apply to non-

traditionally sex-typed receivers. The traditional source

advocating a traditional position should carry very little

weight with these receivers. Nontraditional persons are

bothered by such traditional stands and will be motivated to

produce counterarguments against the position being advocated

by the traditional source giving a traditional message.

However, when the traditional source exhibits out-of-role








communication behavior, the nontraditional receiver will

view that behavior as a positive violation of expectancies

and be significantly less resistant to this source's per-

suasive appeal than when the traditional source advocates a

traditional position. In addition, when the nontraditional

receiver views a nontraditional source exhibiting nontradi-

tional behaviors, that receiver should be receptive to the

message. However, when nontraditional receivers view a

nontraditional source advocating a traditional position,

they will react negatively to this negative violation of

expectancies. These receivers should be motivated to produce

counterarguments against the position being advocated by the

source and thus be more resistant to the persuasive appeal

than when presented with nontraditional sources giving

nontraditional positions.

Focusing on the source instead of the receiver, the

same basic explanatory principles may be used to predict

differences in how traditional and nontraditional receivers

will accept messages from traditional and nontraditional

sources. It has been demonstrated that nontraditional re-

ceivers view traditional communication behavior from

traditional sources in a negative manner. However, tradi-

tional receivers perceive the same communication behavior

positively since it corresponds with their expectations of

how traditional people should behave. Thus, when viewing a

traditional source exhibiting.traditional communication








behaviors, the nontraditional receiver should be more

resistant to persuasion than the traditional receiver.

When the traditional source exhibits nontraditional

communication behaviors, traditional receivers should be

more resistant to change than nontraditional receivers. In

this instance, the violation of expectancies is negative for

the traditional receiver and positive for the nontraditional

receiver. At the same time, traditional receivers should be

more threatened and more motivated to produce counter-

arguments than the nontraditional receiver. Thus, the

negative violation of expectancies combined with the

threatening nature of the unexpected behavior will result in

traditional receivers being more resistant to unanticipated

persuasive appeals from traditionally sex-typed communicators.

Nontraditional receivers will be amenable to change since

the out-of-role behavior constitutes a positive violation of

their expectancies and does not promote counterarguing.

When a nontraditional source advocates an unexpected

position, traditional receivers should perceive that be-

havior as a very positive violation of their expectancies and

be more amenable to change than nontraditional receivers

viewing the same behavior. Nontraditional receivers will be

threatened by such behavior and be more motivated to create

counterarguments against a message which constitutes a

negative violation of the nontraditional receiver's

expectancies.









The literature reviewed previously provides no basis

for predicting differences in how traditional and nontradi-

tional receivers will react to nontraditional sources

exhibiting expected communication behaviors.

Finally, within the traditional groups of receivers, the

more conforming nature of traditional women combined with

their low levels of self-esteem and aggressiveness should

make them less threatened by persuasive attempts and thus

less motivated to produce counterarguments than male tradi-

tional receivers. This difference in the production of

counterarguments should make males more resistant to a

persuasive communication than females.


Hypotheses


The foregoing review of literature can be formally

summarized in the following two research hypotheses:

HI: There will be an interaction between expectancy
and the sex-typing of the receiver and source
of a persuasive message such that:

A. Given traditional receiver, unexpected
messages from traditional sources will
elicit more resistance to persuasion than
expected messages while expected messages
from nontraditional sources will elicit
more resistance to persuasion than
unexpected messages.

B. Given nontraditional receiver, expected
messages from traditional sources will
elicit more resistance to persuasion than
unexpected messages while unexpected mes-
sages from nontraditional sources will
elicit more resistance to persuasion than
expected messages.





29


C. Given traditional sources, expected mes-
sages will elicit more resistance to
persuasion from nontraditional receivers than
traditional receivers while unexpected mes-
sages will elicit more resistance to per-
suasion from traditional receivers than
nontraditional receivers.

D. Given nontraditional sources, unexpected
messages will elicit more resistance to
persuasion from nontraditional receivers
than traditional receivers.

H2: Given traditional receivers, males will demonstrate
more resistance to persuasion than females.
















CHAPTER II
METHODS AND PROCEDURES


Overview


The experimental subjects in this investigation were

given two packets of experimental materials. The first

packet contained situation, source, and expectancy induc-

tions which specified the background of the experimental

situation, the sex-typing of the source, and the direction

of the arguments in an expected persuasive message. Ap-

proximately 30 minutes after reading the first packet, the

subjects were given a second set of materials. The second

packet contained a persuasive message and posttest attitude,

threat, expectancy, and credibility measures.

The experimental subjects (N=162) were members of

speech classes at the University of Florida and were ran-

domly assigned to the experimental conditions such that no

class constituted a single cell in this investigation. The

subjects were told that a colleague of their instructor

wanted their opinions of the choices in the experimental

situation.

The dependent variable was a difference score computed

by comparing responses on two four-item semantic differen-

tial scales. The independent variables were sex-typing of








the receivers, sex-typing of the source, and the expec-

tancy of the position supported by the source. Receivers'

sex-typing was determined on the basis of androgyny scores.

Sex-typing of the source and the message expectancy were

manipulated.

Subjects in control conditions (N=188) completed at-

titude, anticipatory attitude, threat, expectancy and

credibility measures.


Preparation of Experimental Materials


The experimental situation in this investigation in-

volved making a decision about where children in a finan-

cially troubled orphanage should be taken on a field trip.

The subjects were told that these children had not taken a

field trip in five years and that there were two options of

approximately equal educational value. The first choice was

to take the orphans to the local college's homecoming foot-

ball game. The second alternative was to take the children

to the town's annual arts and crafts show. (See Appendix B.)

This topic was chosen because the subjects were not expected

to have prejudicial initial opinions of where the children

should be taken. Moreover, people should have definite

sex-role related expectations of which trip the different

sources would favor the orphans taking, e.g., a traditional

male should be most inclined to advocate the football game

over the arts and crafts show in the minds of most subjects.









Two messages were constructed for this investigation:

one arguing for taking the orphans to the football game and

the other arguing for taking them to the arts and crafts

show. The messages were highly similar with the only dif-

ferences being minor editorial changes made to keep each

message consistent with its respective topic. (See

Appendix D.)

The messages contained four straightforward arguments

which were chosen for their applicability to both experi-

mental choices. In addition, the language in the messages

was judged by a panel of speech teachers to be moderate in

intensity. Moderate messages were created to avoid the

relative ineffectiveness of low intense language and the

boomerang effects of messages consisting of highly intense

language (Bowers, 1964; Burgoon'& Miller, 1971).


Dependent Measure


The dependent measure in this study was a difference

score that reflected the subjects' evaluations of taking

both the football and arts and crafts trips. Since not all

subjects received the same persuasive message, it was

necessary to have a method to directly compare the attitudes

of these groups. By subtracting the evaluative score for

the trip the message opposed from the score of the trip the

message supported, such a direct comparison among groups

was made possible. For example, if a group received the

message supporting taking the orphans to the football game,








that group's difference score was computed by subtracting

the evaluation of the arts and crafts trip from the evalua-

tion of the football trip. On the other hand, if a group

received the message favoring the arts and crafts show, that

group's difference score was computed by subtracting their

evaluation of the football trip from their evaluation of

the arts and crafts trip.

The subjects' evaluations of the trips were recorded on

identical four-item, seven interval semantic differential-

type scales (Osgood, Suci & Tannenbaum, 1957). The four

items were good-bad, foolish-wise, pleasant-unpleasant, and

worthless-valuable. Each item was scored on a one to seven

basis with a seven reflecting the most positive evaluation

of the trip. The scores for the four items were summed for

each scale. The maximum evaluation for one trip was 28 while

the minimum evaluation was 4. Accordingly, the maximum

favorable difference score (comparing the evaluations of

both trips) was 24, and the maximum unfavorable difference

score was -24. (See Appendix E.)


Independent Measures


Three independent variables were examined in this in-

vestigation: receiver sex-typing, source sex-typing, and

message expectancy.

The Bem Sex Role Inventory (BSRI) was used to classify

the receivers of the experimental manipulations as either

traditionally sex-typed or nontraditionally sex-typed. The








BSRI is a self report instrument which measures masculinity/

femininity based upon reactions to 60 personality charac-

teristics. (See Appendix A.) Bem (1974) indicates that the

BSRI has high internal consistency (r=.86), test-retest

reliability (r=.93), discriminant validity, and convergent

validity when compared to other measures of masculinity/

femininity.

The subjects were divided into traditionally sex-typed

and nontraditionally sex-typed groups on the basis of both

androgyny scores and sex. The groups were split at the

medians on androgyny such that males scoring below the median

(masculine and approaching masculine) and females scoring

above the median (feminine and approaching feminine) were

classified as traditional individuals. Males scoring above

the median (feminine and approaching feminine) and females

scoring below the median (masculine and approaching masculine)

were classied as nontraditional individuals. The above

classification system meets Bem and Korula's (1974) re-

quirements for research using the BSRI.

The messages in this study were attributed to two

sources: one traditionally sex-typed and one nontradition-

ally sex-typed. Both sources were males. The traditional

male was identified as William Harper, a member of the track

team, a physical education major, a construction worker, and

one who enjoys outdoor activities during his spare time. The

nontraditional male was identified as William Harper, a

male nurse, a member of the student artisans guild, and one








who works with small children and enjoys cooking in his

spare time. (See Appendix C.)

It is traditionally accepted in this society that men

should be sports minded and outdoor enthusiasts. However,

men are not generally associated with being nurses, cooking,

and working with children. Thus, the above attributions

seem to correspond with the commonly held views of tradi-

tional and nontraditional male sex-roles in this society.

The expectancy manipulation was created by indicating

in the source induction which positions the sources were

going to argue in an imminent persuasive message. The

expected condition included the traditional source arguing

for taking the orphans to the football game and the non-

traditional source supporting taking the children to the

arts and crafts festival. The unexpected condition included

the traditional source favoring taking the children to the

arts and crafts festival and the nontraditional source

arguing for taking the children to the football game. (See

Appendix C.)


Manipulation Checks


The two scales were constructed to measure the sub-

jects' reactions to the threat and expectancy manipulations

in the present investigation. Both scales were five-item,

seven interval semantic differential-type scales taken from

items suggested by Osgood, Suci, and Tannenbaum (1957).

The items on the threat scale included safe-dangerous,








threatening-not threatening, risky-not risky, intimidating-

unintimidating, and harmful-not harmful. The items on the

expectancy scale included predictable-unpredictable,

unexpected-expected, surprising-not surprising, unlikely-

likely, and anticipated-unanticipated. (See Appendix F.)

The viability of the expectancy manipulations was

tested in a pilot study conducted at Clemson University.

Subjects' (N=120) reactions to the expectancy inductions

verified that traditional men were expected to argue for

taking the orphans to the football game and not expected to

favor the arts and crafts show (p<.05). At the same time,

nontraditional men were expected to favor the arts and

crafts show over the trip to the football game (p<.09).

The results of the pilot study seemed to indicate that the

variables isolated as independent factors in this investiga-

tion were essentially equivalent to the descriptions of

those factors presented in the theoretic rationale.


Experimental Procedures


Each subject in the experimental groups received two

packets of written materials during this investigation. The

first packet contained the situation induction, the source/

expectancy induction, and the BSRI. The second packet con-

tained an experimental message, the attitude scale, and the

threat, expectancy, and credibility measures.

At the beginning of a class period, classroom instruc-

tors distributed the first packet to the subjects and asked









them to read through the materials. Since the source/

expectancy manipulations indicated that a persuasive mes-

sage should have been in the packet and that message was

missing, the instructors were told to answer inquiries con-

cerning the absence of the message by suggesting that a

mistake may have been made when the packets were assembled.

In addition, the instructors were asked to tell the subjects

that the mistake might or might not be remedied. The sug-

gestion that there may or may not be an attack message

forthcoming reflects the findings of Burgoon et al. (1976)

that a moderate probability of attack elicits the most

resistance to persuasion. They found that persons are un-

motivated to defend themselves against attacks that are

unlikely to occur or attacks that are too overpowering. In

the first instance persons are unconcerned about a highly

improbable message. In the second case, they give up in the

face of a strong threat.

Approximately 30 minutes after the first administration

of experimental materials, an experimenter arrived at the

classroom with the second packets. Half of these experi-

menters were male; half were female. The experimenter ex-

plained that indeed a mistake had been made when the original

packets had been collated and that the second packet contained

the remainder of the material for the project. The subjects

were then asked to read the message and to complete the

scales which followed the message.









The first packets were randomly distributed among the

subjects such that each experimental condition was rep-

resented in each class, and no intact class constituted any

cell in the design. Because the information in the second

packet had to match the information in the first, the second

packets were distributed in the same order as the first.


Experimental Design


To insure the persuasiveness of the experimental mes-

sages, one control group was given the situation induction

and the football message and a second control group was

given the situation induction and the arts and crafts mes-

sage. Difference scores from these two groups were compared

with a control group that received only the situation

induction.

To check for anticipatory attitude change (change caused

by simply learning the direction of an impending attack mes-

sage), four control groups were given the situation induc-

tion and one of the four different source-expectancy

inductions. Difference scores from these four groups were

compared with those from the no message, situation induction

only group.

The subjects' opinions of the credibility of the source

of the message were measured on a fifteen-item, seven in-

terval semantic differential-type scale (Osgood, Suci &

Tannenbaum, 1957). Three items representing each of the

five dimensions of credibility isolated by McCroskey and









his colleagues (e.g. McCroskey, Holdridge & Toomb, 1974) were

included on the credibility scale. The dimensions of charac-

ter, extroversion, composure, competence, and sociability

were measured.

An additional four control groups were given the

situation induction, one of the source/expectancy induc-

tions, a matching persuasive message, and the threat,

expectancy, and credibility measures. These groups served

as a control for the threat and expectancy explanations

advanced and as a check for credibility differences among

the message sources.

The basic design of this investigation is outlined in

Figure 1. The explanation of the eight experimental and

eleven control groups are listed in Figure 2.





40



Threat
Source/ Expectancy
Situation Expectancy Attack Attitude Credibility
Group Induction Induction BSRI Message Measures Measures

C1 X X
C2 X X X X

C3 X X X X

C4 X X X X

Cg X X X X

C6 X X X



C9 x x x
C0 X X X

Cg X X X
Cl x x x



E X X X X X





E3 X X X X X X
E2 X X X X X X

E5 X X X X X X

E6 X X X X X X


E7 X X X X X X

E X X X X X X


Figure 1. The design of the investigation








Group

Cl

C2


C3


Description

Attitude measure only

Traditional source/expected message: track star
favoring football trip

Traditional source/unexpected message: track
star favoring arts trip


C4 Nontraditional source/expected message: nurse
favoring arts trip

C5 Nontraditional source/unexpected message: nurse
favoring football trip

C6 Traditional source/expected message

C7 Traditional source/unexpected message

C8 Nontraditional source/expected message

C9 Nontraditional source/unexpected message

C10 Attitude measure: football message

C Attitude measure: arts message


E Traditional receiver/traditional source: expected
message

E2 Traditional receiver/traditional source: unexpected
message

E3 Traditional receiver/nontraditional source:
expected message

E Traditional receiver/nontraditional source:
unexpected message

E5 Nontraditional receiver/traditional source:
expected message

E6 Nontraditional receiver/traditional source:
unexpected message

E7 Nontraditional receiver/nontraditional source:
expected message

E Nontraditional receiver/nontraditional source:
unexpected message

Figure 2. Descriptions of control and experimental groups
















CHAPTER III
RESULTS


Viability of Experimental Messages


To insure that the messages used in this investigation

were successful in creating attitude change, difference

scores from the no message control group were compared with

scores from the two message only control groups--one receiv-

ing the arts and crafts message and the other receiving the

football message. The results of this analysis (Table 1)

indicated that both experimental messages were successful in

creating significant attitude change in the direction that

each message advocated.



Table 1. Means, standard deviations, and t-tests for
difference scores comparing the no message control
group with the message only control groups.


Group Mean SD t P

No Message .86* 3.94

Football Message 3.71 5.58 2.95 <.05

Arts and Crafts Message 5.94 8.29 2.50 <.05

t =1.68
t.05, df=40, 6

*Since difference scores were used, the mean for the no mes-
sage group was positive when compared with the football mes-
sage and negative when compared with the arts and crafts
message.









Further analysis indicated that there was no signifi-

cant difference between the persuasiveness of the two

experimental messages (t
the messages used in this investigation were successful in

creating attitude change in the direction of the position

each message supported.

In addition to the general persuasiveness checks, an

analysis was performed to determine whether there were any

sex differences that mediated the effects of the experimental

messages. A 2X2 analysis of variance with sex and message

type (arts or football) as the two factors indicated no

significant main effect for sex or message type. However,

there was a significant sex X message-type interaction,

F (1, 160) = 11.2, p<.05. (See Table 2.) A second analysis

was performed to determine if sex-typing of the receivers

affected their reception of the experimental messages. This

analysis revealed no significant main effect for sex (p>.05)

or message type (p>.05) and no significant interaction

(p>.05).


Table 2. Means for sex of receiver and message type manipu-
lation check.


Male Receivers Female Receivers

Football Message 6.13 2.77

Arts and Crafts Message 1.94 8.29








Taken together these results indicated that sex of the

receiver, but not sex-typing of the receiver, affected the

reception of the persuasive messages in this investigation.

With the receivers' sex confounding the interpretation of

the results of the present investigation, this variable was

added as a fourth factor when the first hypothesis was

tested.


Control Group Analysis


Four control groups received the four combinations of

the source/expectancy manipulations: traditional source

with expected message, traditional source with unexpected

message, nontraditional source with expected message, and

nontraditional source with unexpected message. Seven 2X2

analyses of variance were performed with sex-typing and

expectancy of message as the two factors. The seven de-

pendent measures were threat, expectancy, and the five

dimensions of source credibility.

Threat

The analysis of variance for threat produced two main

effects: (1) Unexpected messages were perceived as more

threatening than expected messages, F (1, 61) = 11.7,

p<.05. (2) Messages from traditional sources were more

threatening than messages from nontraditional sources,

F (1, 61) = 11.9, p<.05. While there was no significant

interaction (p>.05), an examination of the cell means

(Table 3) revealed that the two significant main effects








could best be attributed to the highly threatening nature

of the unexpected message from a traditional source.


Table 3. Means for threat scores of control groups.



Group Mean*

Traditional Source: Expected Message 24.92

Traditional Source: Unexpected Message 17.44

Nontraditional Source: Expected Message 26.88

Nontraditional Source: Unexpected Message 24.68


*Scores range from 5 (high threat) to 35 (low threat)



Expectancy

The analysis of variance for expectancy produced two

main effects: (1) Messages from traditional sources were

less expected than messages from nontraditional sources,

F (1, 61) = 8.8, p<.05. (2) Receivers anticipated the ex-

pected message more than the unexpected message, F (1, 61) =

38.6, p<.05. The cell means (Table 4) reveal that dif-

ferences can be attributed to the traditional source/

unexpected message cell. However, there was no significant

interaction (p>.05).

Credibility

The analyses of variance for the sociability, composure,

and competence dimensions of credibility produced no signi-

ficant main effects or interactions. The analysis for the

character dimension revealed that nontraditional sources









Table 4. Means for expectancy scores of control groups.


Group Mean*

Traditional Source: Expected Message 24.54

Traditional Source: Unexpected Message 16.06

Nontraditional Source: Expected Message 24.33

Nontraditional Source: Unexpected Message 20.63


*Scores range from 5 (unexpected) to 35 (expected)



were perceived as having higher levels of character than

traditional sources, F (1, 61) = 3.8, p<.05. The nontradi-

tional source had a mean character rating of 15.59 while the

traditional source had a mean character rating of 14.22.

The analysis of variance for the extroversion dimen-

sion produced a main effect for expectancy. However, this

effect was overridden by a significant interaction between

the sex-typing of the source and the expectancy of the mes-

sage, F (1, 61) = 5.7, p<.05. An examination of the indi-

vidual cell means again indicates that the Traditional

Source: Unexpected Message conditions produced the signi-

ficant effects (Table 5).


Anticipatory Attitude Change Measures


To provide a check for the subjects' changing attitudes

solely on the basis of learning that a persuasive attack was

imminent, four control groups were given the situation









Table 5. Means for extroversion ratings for control groups.


Group Mean*

Traditional Source: Expected Message 17.62

Traditional Source: Unexpected Message 13.94

Nontraditional Source: Expected Message 15.61

Nontraditional Source: Unexpected Message 15.25


*Scores range from 3 (introverted) to 21 (extroverted)



induction, the source/expectancy induction, and an attitude

measure. Difference scores from each of these four groups

were compared with the scores from the no message control

group to determine if simply learning about the coming per-

suasive message produced any differences in attitudes among

the groups. Results of this analysis (Table 6) revealed that

there were no significant differences between the no mes-

sage group and the four anticipatory change control groups.

In addition, a 2X2 analysis of variance was performed

on the difference scores with sex-typing of the source and

expectancy of the message as the two factors. This analysis

produced no significant main effects and the interaction

was not significant. Anticipatory attitude change cannot

be presumed to be a factor in this investigation.








Table 6. Means, standard deviations, and t-tests for dif-
ference scores comparing the no message control
group with the anticipatory attitude change
control groups.



Group Mean SD t P

No Message .86 3.94

Traditional Source:
Expected Message 1.64 2.54 .71 NSD

Traditional Source:
Unexpected Message 1.29 5.80 .27 NSD

Nontraditional Source:
Expected Message 1.94 3.43 .86 NSD

Nontraditional Source:
Unexpected Message 1.88 5.29 .68 NSD


t =1.68
-.05, df=40,


Manipulation Checks for Experimental Groups


Since differing levels of threat and expectancy provide

a basis for the underlying rationale of this investigation,

it was necessary to determine if the experimental manipula-

tions were successful in creating varying levels of these

two presumed antecedents to resistance to persuasion.

Threat

A 2 X 2 X 2 X 2 analysis of variance was computed on

the threat scores from the experimental conditions. The

four factors were sex-type of the receiver, sex-type of the

source, sex of the receiver, and message expectancy. Results








of this analysis revealed a significant interaction among

receiver sex-type, source sex-type, and message expectancy,

F (1, 146) = 3.97, p<.05. An inspection of the individual

cell means indicated that they followed the general pattern

predicted for attitude scores in the first hypothesis of

this investigation (Table 7).



Table 7. Means for threat scores in experimental conditions.



ConditionI Mean*

TR:TS:EM 23.79
TR:TS:UM 19.94
TR:NS:EM 22.06
TR:NS:UM 24.85
NR:TS:EM 21.68
NR:TS:UM 24.18
NR:NS:EM 22.00
NR:NS:UM 22.50

*Scores range from 5 (high threat) to 35 (low threat

TR = Traditional Receiver; NR = Nontraditional Receiver;
TS = Traditional Source; NS = Nontraditional Source;
EM = Expected Message; UM = Unexpected Message




In addition, there was a significant interaction among

sex of the receiver, sex-type of the source, and expectancy

of the message, F (1, 146) = 6.39, p<.05. Two three way

interactions do not override each other. Moreover, an

examination of the means for this interaction revealed no

discernible pattern. These results indicate that the sub-

jects in the experimental conditions did perceive differing









levels of threat that essentially followed the pattern pre-

dicted by the theoretic rationale leading to the hypotheses

advanced.

Expectancy

A second 2 X 2 X 2 X 2 analysis of variance was com-

puted on expectancy scores in the eight experimental con-

ditions. The four factors were receivers' sex, receivers'

sex-type, source's sex-type, and expectancy of message.

The results produced two significant interactions. An

interaction was found between sex-typing of the receiver and

message expectancy, F (1, 147) = 4.15, p<.05. Inspection of

the individual cell means (Table 8) revealed that most of

the variation was due to the traditional receivers' not

anticipating the unexpected message. A similar pattern was

found in the control group expectancy results.



Table 8. Means for expectancy scores for receiver sex-type
X expectancy of message interaction in experimental
conditions.


Condition Mean*

Traditional Receiver: Expected Message 22.91

Traditional Receiver: Unexpected Message 20.55

Nontraditional Receiver: Expected Message 21.88

Nontraditional Receiver: Unexpected Message 21.79


*Scores range from 5 (unexpected) to 35 (expected)









Another significant interaction was found between the

sex-typing of the source and the expectancy of the message,

F (1, 147) = 8.6, p<.05. An analysis of the cell means

(Table 9) indicated that the greatest difference was found

in the experimental subjects' perceptions of the traditional

source's communication behavior. Subjects anticipated the

expected message more than the unexpected message with both

sources. However, the difference was four times greater

with the traditional source than with the nontraditional

source.



Table 9. Means for expectancy scores for source sex-type X
expectancy of message interaction in experimental
conditions.


Condition Mean*

Traditional Source: Expected Message 22.69

Traditional Source: Unexpected Message 21.04

Nontraditional Source: Expected Message 21.87

Nontraditional Source: Unexpected Message 21.53


*Scores range from 5 (unexpected) to 35 (expected)





Taken together, these results indicate that traditional

subjects anticipated expected behavior more than nontradi-

tional subjects, and all of the subjects anticipated expected

behavior more from traditional sources than nontraditional

sources.








Reliability


To determine the internal consistency of the attitude

measures, the threat and expectancy scales, and the scores

for the five dimensions of credibility, Cronbach's alpha

(Cronbach, 1951) was computed for these measures. The

results of these computations are summarized in Table 10.

These results generally indicate high levels of reliability

for the dependent measures and the manipulations checks

utilized in this investigation with the exception of the

expectancy scale. The low reliability (a = .40) for this

scale could place severe limitations on the interpretation

of any results based on the expectancy scores alone. The

composure and extroversion scores also had low reliabili-

ties. Since the credibility measures were mainly a check

against competing explanations for the results of this study,

the lack of high reliability for the composure and extro-

version dimensions is of less importance to the theoretical

paradigm presently being investigated.


Homogeneity of Variance


A basic assumption of the analysis of variance tech-

nique is that the variances of the groups being analyzed

must be homogeneous. The F Max test (Winer, 1971) was

computed for the experimental group as a test of homogeneity

of variance. The results indicated that the assumption of

homogeneity of variance was not violated in this investiga-

tion and that analysis of variance was the proper method for

hypothesis testing.








Table 10. Results of Cronbach's alpha test for internal
reliability of dependent measures and
manipulation checks.


Variable a

Threat .90
Expectancy .40
Football Message Attitude Measure .88
Arts and Crafts Message Attitude Measure .89
Sociability .76
Composure .51
Competence .76
Character .72
Extroversion .59


Tests of Hypotheses


The first hypothesis of this study was tested by a

2 X 2 X 2 X 2 analysis of variance. The four factors were

receivers' sex, receivers' sex-type, sources' sex-type, and

expectancy of message. The dependent measure was a dif-

ference score which measured subjects' overall evaluation of

both choices in the experimental situation. The significance

level employed for all analyses was .05. The first

hypothesis stated:

There will be an interaction between expectancy and the sex-
typing of the receiver and source of a persuasive message
such that:

A. Given traditional receivers, unexpected messages from
traditional sources will elicit more resistance to
persuasion than expected messages while expected mes-
sages from nontraditional sources will elicit more
resistance to persuasion than unexpected messages.

B. Given nontraditional receivers, expected messages from
traditional sources will elicit more resistance to per-
suasion than unexpected messages while unexpected









messages from nontraditional sources will elicit more
resistance to persuasion than expected messages.

C. Given traditional sources, expected messages will
elicit more resistance to persuasion from nontradi-
tional receivers than traditional receivers while
unexpected messages will elicit more resistance to
persuasion from traditional receivers than nontradi-
tional receivers.

D. Given nontraditional sources, unexpected messages will
elicit more resistance to persuasion from nontradi-
tional receivers than traditional receivers.

Table 11 presents the results of the analysis of vari-

ance computations. The significant receiver sex-type, source

sex-type, and message expectancy interaction confirms the

first hypothesis. The pattern of cell means exactly conforms

to the complex pattern predicted in Hypothesis 1.

Specific planned comparisons of individual cells pro-

duced two significant differences. In the traditional

source-expected message condition, nontraditional receivers

were more resistant to persuasion than traditional receivers,

t(47) = 2.1, p<.05. In the traditional source/nontraditional

receiver condition, the expected message elicited more

resistance to persuasion than the unexpected message,

t(45) = 2.2, p<.05. None of the other planned comparisons

was significant.

A very large error term seemed to mitigate against the

planned comparisons being significant. Appendix G gives the

means, standard deviations, high scores, and low scores for

the experimental groups. As indicated by the data, there is

extreme variation in the experimental subjects' responses

on the attitude measures. This large within error did not















0

U
Ca
u



a4
Ic

a)



o
4
.-4




0














U
rd



















0
>1




(n

















(0)
-4

cU)





S(0
C04-I
(9




H
Q)


0







,--4
O C





0)-
C H


r),

0)0)

0, U)










4-)
0U0)
a) j'-
04 m u)
X (n 4J
(1) Un 4
r.a a)








Wa) rd

( Ma 0
,-

)U) 0
04 (

Q) rJO
Fm n


ou

* II
u-oz
koZ













* II






umZ


















>
- I I
n Z




















0
O,
.-4
(r










5-4
u


m)mm)mm
ZZZZZZZ


N -4
,-4 o ,-4 T
r e1< ,ii-


HHNC\
Nr- N


NM r-i


oo -in ,r





Ln-N


Hn () oo


QQQQQmLnQl

ZZZZ *Z .Z
v v


O0H

0N 1co
Hce0\JH


in r- oo co
co rN co a)

o m
r-i






omor
0 0m


-1 -1 r-l -4r4 -4-r-l H H H H H -4D Hr--q-l
H H3


c r-
coH
* II
< Z












CN N













mN
* II

























Sin
u
* II














0
z

*d



O
-H(



rJ
0


r- CN k.0


r\i r in

1-4


I .-
'--- Q

0) -
-.04 (>.
> >i





S- 4 En
XXE-44



- U)X


U o in oj
S-

*rA Ar( U rd m u


0000;ena


CDr-4 M -4r
-4H(


mCn

l00
me


x

U u Q


XXXXXXxXX


c1m
mN
ro


-CO N t CmN
00 (N (\) (N
* II II
rZ rmZ









prevent the interaction term from reaching the prescribed

level of significance, but it did seem to preclude finding

significant individual cell comparisons.

A sex of the receiver X sex-type of source X expec-

tancy of message interaction also proved to be significant

in the four way analysis of variance, F (1, 146) = 10.5,

p<.05. Since males and females differentially evaluated the

two experimental messages, this interaction is not surpris-

ing. However, it does not affect the confirmation of

Hypothesis 1.

In short, the significant receivers' sex-type X sources'

sex-type X expectancy of message interaction combined with

the pattern of means provide strong support for Hypothesis 1.

A second hypothesis of this investigation stated:

Given traditional receivers, males will demonstrate more
resistance to persuasion than females.

The four cells involving traditional receivers were

collapsed into a simple one way analysis of variance to

test this hypothesis. The results of this analysis revealed

no significant sex differences among traditional receivers

(F<1). The mean difference score for traditional males was

6.2 while the mean difference score for traditional females

was 5.5. Therefore, no support for Hypothesis 2 can be

claimed in the present investigation.








Experimental and Control Group Comparisons


Dunnett's t-test was used to compare the experimental

groups' mean with that of the no message control group.

Two comparisons were significant with groups that received

the football message. The traditional receiver/traditional

source/expected message mean differed significantly from

the no message control group, t(6/124) = 2.64, p<.05. In

addition the traditional receiver/nontraditional source/

unexpected message mean was significantly different from

the no message control mean, t(6/124) = 3.25, p<.05. While

several of the other means in the experimental condition

approached being different from the no message control, none

were statistically different. Table 12 presents the results

of the Dunnett's analysis.

While all but one of the experimental group means appear

to be different from the no message control group in magni-

tude, the large variation in individual's responses to the

attitude measures (See Appendix G) reported in the previous

section prevented those differences from reaching statistical

significance.


Supplemental Credibility Analysis


Analyses were performed on the five dimensions of

credibility to determine whether the experimental manipula-

tion caused any differences in the experimental subjects'

perceptions of the sources' credibility. A 2 X 2 X 2 X 2

analysis of variance was computed for each of the five









Table 12. Dunnett's t-test analysis comparing no message
control group difference score means with those
of the experimental groups.



Group Mean Dunnett's t P

No Message .86
TR:TS:EM 5.91 2.64 .05
TR:TS:UM 3.44 .95 NSD
TR:NS:EM 6.40 2.01 NSD
TR:NS:UM 7.81 3.25 .05
NR:TS:EM .36 .48 NSD
NR:TS:UM 6.27 2.17 NSD
NR:NS:EM 4.82 1.47 NSD
NR:NS:UM 3.36 1.76 NSD
t.95, df = 6, 120 = 2.26

TR = Traditional Receiver; NR = Nontraditional Receiver
TS = Traditional Source; NS = Nontraditional Source
EM = Expected Message; UM = Unexpected Message




credibility dimensions. The factors were those used in the

test of Hypothesis 1: sex of receiver, sex-type of receivers,

sex-type of source, and message expectancy. There were no

significant main effects or interactions for the dimensions

of character, extroversion, composure, or competence. How-

ever, there was a significant interaction on the sociability

dimension between sex-type of source and expectancy message,

F(l, 147) = 16.8, p<.05. An inspection of the cell means

(Table 13) indicates that traditional sources exhibiting

unexpected communication behavior were thought to be less

sociable than the other source-message combinations. There








Table 13. Means for sex-type of source X message
expectancy interaction for experimental subjects'
sociability scores.


Group Mean

Traditional Source: Expected Message 16.55
Traditional Source: Unexpected Message 13.95
Nontraditional Source: Expected Message 15.28
Nontraditional Source: Unexpected Message 16.44




were no other significant main effects or interactions on


the sociability dimension.
















CHAPTER IV
DISCUSSION


The confirmation of the first hypothesis is very en-

couraging. The significant interaction term and the pattern

of means in the experimental groups indicate that receivers'

sex-typing, sources' sex-typing, and sex-role related mes-

sage expectations combine to be important mediators of

resistance to persuasion. Moreover, the results of the

threat, expectancy, anticipatory attitude change, and credi-

bility measures suggest strong support for the underlying

explanatory principles advanced in the theoretic rationale.

Threat, motivation, opportunity, and expectancy all appear

to be important contributors to the resistance to persuasion

phenomenon when operationalized in the form of receiver,

source, and message sex-role stereotypes.

In order to completely interpret the results of this

investigation, it is necessary to probe multiple differences

among the cells in the experimental groups. While the pat-

tern of means corresponds with the pattern predicted in the

first hypothesis, the large within error term prevented all

but two of the post hoc individual cell comparisons from

reaching statistical significance. However, the correct

pattern of means and the significance of the interaction

term justify further exploration of the differences among









individual cells. Since the reliability of the expectancy

measures was low, any interpretation of the results of this

investigation based solely on those scores would be tenta-

tive at best. However, with the means of the difference

scores and the threat measures corresponding to the pattern

predicted in the theoretic rationale, the expectancy re-

sults become a third check on the accuracy of the underlying

explanations advanced earlier. Thus, while the results of

the expectancy scores must be viewed with some reservation,

they are not completely meaningless to the interpretation of

the results.

Specifically, the data indicate that in the conditions

where the attack message corresponded with expected sex-role

communication behaviors, the subjects who were most threat-

ened by the message were most resistant to its appeal. This,

of course, confirms one of the basic tenets of McGuire's

inoculation explanation of the resistance to persuasion

phenomenon. Nontraditional receivers were significantly more

resistant than traditional receivers when exposed to a

message from a traditional source arguing for a sex-role

stereotyped position--taking the orphans to the football

game. The nontraditional receivers in this condition had

by far the lowest mean difference score and the second

highest level of threat in the eight experimental cells.

By contrast, the traditional receivers in this condition

were significantly less resistant to the message and were

not as threatened by the message. These results clearly








support the Bem (1976) and Ellis and Bentler (1973) findings

that nontraditionally sex-typed individuals are opposed and

react negatively to traditionally sex-role stereotyped be-

haviors. In terms of the theoretic rationale, nontraditional

persons are apparently more motivated than traditional re-

ceivers to produce counterarguments against traditional

communication behaviors which they feel are threatening to

them.

This finding becomes more important when the mean dif-

ference scores for the nontraditional source/expected mes-

sage (arguing for the arts and crafts show) are examined.

Although there was no basis in the literature for making

specific predictions about the behaviors of these groups,

they intuitively might be expected to exhibit similar ten-

dencies as the traditional source/expected message groups.

They did not. In this condition, traditional receivers were

less resistant to the message than nontraditional receivers

while the threat scores for the two cells were nearly

identical. It appears that nontraditional persons are

bothered by traditionally expected communication behaviors

but the converse seems not to be true. Traditionally

oriented receivers generally did not react negatively to

expected nontraditional sex-role communication behaviors.

Evidently, when persons do not adopt or have given up tradi-

tional sex-role orientations, they become predisposed to

react negatively to those who still demonstrate those tradi-

tional behaviors. This in turn causes them to be threatened








by messages which portray the traditional positions and

causes the nontraditional receiver to produce defenses in the

form of counterarguments against the position advocated by

the traditional source arguing for a traditionally accepted

position.

In short, in the expected message conditions the only

persons who proved to be resistant to the persuasive appeals

were the nontraditional receivers who were exposed to a

traditional source who argued for a traditionally sex-role

stereotyped position.

In the conditions in which the sources argued for posi-

tions that were not anticipated by the receivers, the role

of positive and negative violations of expectancies seems

quite clear. Brooks (1970), Burgoon and Chase (1973), and

Burgoon, Jones, and Stewart (1974) all suggest that viola-

tions of expectations may affect the manner in which re-

ceivers accept or reject persuasive messages. Positive

violations tend to enhance the persuasive effect of a mes-

sage while negative violations tend to inhibit change. The

difference scores of the experimental groups in the present

study seem to indicate that a similar process occurs in the

resistance to persuasion paradigm.

In both cells where the message could be considered a

positive violation of the receivers' expectations, little

resistance to the persuasive appeal was demonstrated. In

contrast, in the two cells in which the message could be

considered a negative violation of the receivers'









expectations, more resistance to the message was shown.

While the lack of significance in the post hoc comparisons

hinders more exact interpretation, the cell means correspond

perfectly to the pattern supported by the rationale de-

veloped in Chapter I.

Specifically, traditional receivers exposed to a non-

traditional source supporting the unexpected position of

taking children to the football game and nontraditional

receivers exposed to a traditional source supporting the

unexpected position of taking the orphans to the arts and

crafts show showed very little resistance to the messages'

appeals. On the other hand, the traditional receivers who

were exposed to the unanticipated situation of having a

traditional source argue for taking the children to the arts

and crafts show and the nontraditional receivers who were

exposed to the unexpected situation of having a nontradi-

tional source argue for taking the children to the football

game were much more resistant to the messages' persuasive

appeals.

Examination of the threat scores for these groups sup-

ports the results of the mean difference scores. In both

cells where the receivers were exposed to a positive viola-

tion of expectancies, the messages were not perceived as

threatening. In the two cases where the message could be

interpreted as negative violations of the receivers' ex-

pectancies, the messages were.considered much more threaten-

ing. Again, it can be argued that when the messages were









threatening to the receivers, they were motivated to build

defenses in the form of counterarguments which left them

resistant to the appeals contained in the messages. When

the messages were not threatening to the receivers, there

was no motivation to counterargue, and the receivers were

much less resistant to the effects of the message.

Examination of the expectancy scores adds further speci-

fication as to how sex-typing affects receivers' abilities

to resist persuasive attacks. Taynor and Deaux (1975) and

Heilbrun (1976) suggest that nontraditionally sex-typed

individuals are better able to integrate nontraditional be-

havior than traditionally sex-typed individuals. The results

of the expectancy measures in the current investigation tend

to support this view. Generally, in both the control and

experimental groups, traditional receivers were more sur-

prised than nontraditional receivers by the unexpected be-

haviors of both traditional and nontraditional sources.

Moreover, in the experimental groups while both traditional

and nontraditional receivers were surprised by the unantici-

pated behaviors of the two sources, the unexpected message

from the traditional source was most surprising to both

groups of receivers. Thus, traditional receivers and sources

seemed to create more surprise in unanticipated communica-

tion behaviors than nontraditional receivers or sources.

The key to understanding how these results affect the

receivers' abilities to resist persuasive attacks is not

whether the behaviors were absolutely unexpected by the









receivers. More important is whether the behaviors represent

positions that the receivers were predisposed to expect.

Clearly, nontraditional individuals were not totally sur-

prised by the unexpected, out-of-role behaviors of the

sources. At the same time, the threat scores indicate that

the nontraditional receivers reacted to these messages in a

manner essentially similar to the traditional receivers.

Negative violations of expectancies were threatening to both

groups of receivers while positive violations of expecta-

tions were not as threatening to both groups of receivers.

Moreover, the mean difference scores indicate that both

traditional and nontraditional receivers reacted the same to

the persuasive appeals in the messages: positive violations

inhibited resistance while negative violations created

resistance.

Thus, the expectancy violations do not seem to have to

be measured in absolute terms in order for those violations

to affect receivers' abilities to resist persuasive appeals.

It appears as if the violations only need to represent

violations of commonly held social anticipations of be-

havior for the resistance to persuasion process to be

affected.

Brooks (1970), Burgoon and Chase (1973), and Burgoon,

Jones, and Stewart (1974) point out that violations of re-

ceivers' expectations tend to exaggerate the effect that a

persuasive message has when compared to its persuasiveness

under more normal circumstances. That is, individuals react








more favorably than usual to positive violations of expec-

tations and more negatively than would otherwise be anti-

cipated to messages that are perceived as negative violations

of expectations. The results of the present study suggest

that this exaggeration effect also occurs in the resistance

to persuasion paradigm. In three of the four cases where

the effectiveness of a message which represented a violation

of the receivers' expectations could be measured against

the effectiveness of the same message from a source which

did not violate the receivers' expectations, the exaggera-

tion phenomenon was found.

Specifically, traditional receivers were less resistant

to the message from the nontraditional source arguing for

taking the children to the football game (unexpected positive

violation) than they were to the traditional arguing for the

same position (expected and no violation). On the other

hand, traditional receivers were more resistant to the

traditional source arguing for taking the orphans to the arts

and crafts show (unexpected negative violation) than they

were to the same message from the nontraditional source

(expected and no violation). Finally, the nontraditional

receivers were less resistant to the traditional source

arguing for the arts and crafts show (unexpected positive

violation) than they were to the nontraditional source taking

giving the same message (expected and no violation). The

threat scores for these conditions correspond exactly to the

above pattern. Positive violations were perceived as less








threatening than the messages under expected conditions while

negative violations were seen as more threatening than when

the same messages were presented under expected conditions.

The only exception to the exaggeration pattern was

found when the nontraditional receivers were exposed to the

football message from both the traditional (expected and no

violation) and the nontraditional (unexpected negative viola-

tion) source. As has been pointed out earlier, the non-

traditional receivers had extremely negative reactions to

stereotypic behaviors of the traditional source. The threat

scores demonstrate that the nontraditional receivers were

more threatened by the expected behavior of the traditional

source than the unanticipated behavior of the nontraditional

source even though the latter probably constituted a negative

violation of the nontraditional receivers' expectations.

The reaction seems to have overridden the exaggeration

effect of the expectancy violation for these receivers.

In short, with the exception of the nontraditional re-

ceivers' very negative sentiments against traditional sex-

role behaviors, messages which violated receivers' expecta-

tions in this study tended to exaggerate the effect that

the message would have under normal circumstances. Positive

violations caused less resistance than would be normally

anticipated and negative violations created more resistance

than would normally be expected.

The second hypothesis in this investigation received no

empirical support. Traditional females were not less









resistant to persuasion than their male counterparts. The

rationale supporting Hypothesis 2 depended largely on per-

sonality differences between traditional men and traditional

women. Traditional males are generally more aggressive,

less conforming, and have higher levels of self esteem than

do traditionally oriented females. It was argued that men

would be more likely to produce counterarguments against an

impending attack message because they would be more involved

with their own positions on the issue and therefore more

threatened by a message which would challenge that point of

view. Females, on the other hand, would be less likely to

produce counterarguments against an impending attack since

they would be less threatened by influence attempts because

of low levels of aggression and self-esteem and high levels

of conformity. The data simply do not support this

conclusion.

Maccoby and Jacklin (1974) and Rosenfeld and Christie

(1974) have both pointed out that most recent investigations

into sex differences and persuasibility have shown that

there are no systematic differences which make females

generally more susceptible to persuasive appeals than males.

Montgomery and Burgoon (1977) suggest that gender in and of

itself may not cause differences in persuasibility among

males and females. They suggest that sex-role variations

may be more significant predictors of differential attitude

change. The reasoning presented in support of Hypothesis 2

suggested that when considering only traditional males and








traditional females and not including androgynous or undif-

ferentiated individuals, differences in abilities to resist

persuasive attempts may become more apparent. They did not.

Furthermore, nothing in the data gives any indication as to

why the second hypothesis was not confirmed. There is no

reason found in the present investigation to suggest that

traditional males are better able to resist persuasive ap-

peals than traditional females.


Competing Explanatory Principles


The underlying assumption of the present investigation

is that when receivers are threatened by an impending per-

suasive attack, they will be motivated to build defenses

against that attack. These defenses take the form of counter-

arguments with which receivers may negate the effects of the

arguments in the attack message. Thus, the more threatened

individuals are by an impending attack the more they will be

motivated to produce counterarguments against that message

and the more resistant they will become to the message's

appeal.

While this research contained no direct measure of the

subjects' counterarguing behavior, a major weakness in re-

sistance to persuasion research according to Miller and

Barron (1973), it did provide conditions optimal for the

production of counterarguments. Roberts and Maccoby (1973)

suggest three factors which influence the production of

counterarguments: motivation to use defenses, opportunity









to use defenses, and the availability of defenses. Through

the manipulation of sex-role related message expectations,

differing levels of threat were created in the experimental

procedures. It is logical to assume that these differing

threat levels in turn produced different levels of motiva-

tion among the receivers which led to differences in their

counterarguing behaviors and finally to differences in the

receivers' abilities to resist the persuasive messages used

in this study. The threat scores and means of the experi-

mental groups support this reasoning.

Ample opportunity to produce counterarguments was also

provided in the experimental procedures. Subjects were given

approximately 30 minutes to think about the attack message

from their first warning of its existence to its actual

presentation. This is ample time for the receivers to pro-

duce counterarguments. The proof of the availability of

counterarguments rests on several factors. First, subjects

were motivated to defend and given the opportunity to build

defenses. These combinations led to significantly different

abilities to resist the persuasive attack among the eight

cells in the experimental design. Second, the two major

competing explanations for the resistance phenomenon were

not a factor in this investigation.

Miller and Barron (1974) suggest that it is important

to consider the effects of anticipatory attitude change when

assessing the viability of the counterarguing explanation of

the resistance to persuasion process. To this end, four









control groups were given attitude measures immediately

after they were warned of the impending persuasive attacks.

None of the groups demonstrated any amount of attitude change

that was different from the no message control group. More-

over, there were absolutely no differences among the four

control groups in the amount of attitude change created by

simply warning the subjects that an attack message was com-

ing. Clearly, differences in the subjects' abilities to

resist the persuasive message in this study were not the

result of differing levels of anticipatory attitude change.

A second major competing explanation for the resistance

to persuasion process is that subjects differentially react

to the sources of the attack messages. That is, subjects

will change their opinions of the credibility of the sources

of persuasive messages and these changes in credibility

ratings account for the differing levels of resistance. The

results of the credibility measures indicate absolutely no

differences in how the subjects in the experimental condi-

tions reacted to the sources on the competence, composure,

character, and extroversion dimensions. The single signi-

ficant credibility measure indicated that traditional

sources who advocated unexpected positions were thought to

be less sociable by all receivers. This finding does not

correspond with the differences in attitude or threat scores

for the experimental group and thus cannot be considered a

confounding element in this study. Clearly, differences in

source credibility ratings did not cause the differences









in the receivers' abilities to resist the persuasive attacks

in this investigation.

In short, the maximal conditions for the productions of

counterarguments were present in this investigation and the

two major competing explanations of anticipatory attitude

change and source derrogation were not factors. Thus, the

theoretic rationale advanced earlier remains the best

explanation for the results of this study.


Research Implications


The results of this study have several important impli-

cations for resistance to persuasion researchers. First,

although the basic assumptions of McGruire's (1964) inocula-

tion theory are essentially similar to the assumptions of

this study, his concentration of refutational or supportive

pretreatment strategies is not a sufficient explanation for

the results of this study. These results could not be pre-

dicted from McGuire's formulations. Differing message

related expectancies as well as different levels of threat

may be more basic elements of pretreatment strategies that

are found in McGuire's inoculation concept.

Second, the sex-role of the subjects studied by re-

sistance to persuasion researchers must be considered. As

was pointed out earlier, few researchers considered even the

sex of the subjects during resistance to persuasion experi-

ments and none took into account the sex-roles of their

subjects. The results of this study indicate that receivers'









sex-roles, sources' sex-roles, and sex role related message

expectancies combine to be important mediators of the re-

sistance to persuasion process. Resistance researchers

must take care to insure that subjects' sex-role orienta-

tions are randomly distributed among their samples.

Several additional areas of research are suggested by

the results of the current investigation. First, the

procedures in this study need to be duplicated using both

male and female sources. Pragmatic considerations mandated

using only a male source during this study, but the results

cannot be generalized to any large population until the

experiment is repeated using both male and females as

sources.

Second, a direct measure of the subjects' counterarguing

behaviors would provide researchers with a more defensible

conceptualization of the counterarguing process. Miller and

Barron (1974) suggest several methods with which counter-

arguing behaviors can be measured. Any of these avenues

would add to our knowledge of the resistance process.

Third, the expectancy violations predictions need to be

tested with other than sex-role violations. It is important

to insure that the effects of positive and negative viola-

tions are not limited to the method in which they were

operationalized in the current investigation. Both natural

and induced expectation should be studied.

Fourth, Burgoon et al. (1976) noted that the severity

of the outcome of persuasive message significantly affects









the abilities of individuals to resist that message's

appeals. An investigation incorporating differing severi-

ties of outcomes within the framework of this study may

provide some useful information on how expectancies and

severity of outcomes interact to influence the resistance

to persuasion process.


Summary


Hypothesis 1 predicted that receivers' sex-roles,

sources' sex-roles and sex-role related message expectations

would be significant mediators of resistance to persuasion.

It was argued that differing levels of threat and motiva-

tion would produce differing amounts of counterarguing which

would in turn produce differences in the experimental

subjects' abilities to resist the persuasive messages in

this study. A complex pattern of means conformed exactly to

the predictions in Hypothesis 1 and were supported by the

results of both the threat and expectancies measures.

Anticipatory attitude change and source derrogation were

tested and ruled out as competing explanations for the

results of this study.

Hypothesis 2 predicted that traditional males would be

more resistant to persuasive appeals than traditional females

but was not supported by the data in this investigation.

The results were discussed in terms of their applica-

bility to McGuire's inoculation theory and the importance of





76


resistance researchers taking into consideration the sex

role of the subjects in the sample they study.

Several extensions of the present study and several new

lines of research were suggested.



































APPENDIX A
BEM SEX ROLE INVENTORY









I.D. NUMBER


INSTRUCTIONS: A number of personality characteristics are listed

on the following page. Please use those characteristics to de-

scribe yourself. That is, please indicate, in a scale from 1 to

7, how true of you these various personality characteristics are.

Please be careful to mark all of the items.


Example

would


1: Consider the personality characteristic SLY. You

mark a 1 if you are NEVER OR ALMOST NEVER sly.

mark a 2 if you are USUALLY NOT sly.

mark a 3 if you are SOMETIMES BUT INFREQUENTLY sly.

mark a 4 if you are OCCASIONALLY sly.

mark a 5 if you are OFTEN sly.

mark a 6 if you are USUALLY sly.

mark a 7 if you are ALWAYS OR ALMOST ALWAYS sly.


Example 2: The following is an illustration of four markings

and their meanings.


--you are sometimes but infrequently sly.

--you are never or almost never malicious.

--you are always or almost always irresponsible.

--you are often carefree.


THANK YOU FOR YOUR HELP.


Sly 3

Malicious 1

Irresponsible 7

Carefree 5


SEX AGE











03

>41





U)E-
O E4

















U)







z


U
0 n
E-4














0
O E-










>4
S
02

IE-1






















O H
U
z-










0












> >
z4 E
GE-1



>i H
i-l


(0

Uc)
ro

wQ 0u
14H 4
a) 0) u r0
O 0 *H .C !
l Cl 4-J U a)
4.> ,-4 -,A U) -,-i 0) ro r
0 Q) C ::1 43 > -1C
4 > 0 C w 0 0 -M -l O
> -H l q M I 0 4) ) aE -rC
J I-I --I o t0 U) ) a- u r-H l 0 40
C UC l -rH 4 M ( H- C! (0 4- 4- : H -H aQ C
S-I H a r 4- O > Cl) 0) Ul 4- 4-J r--l Q
( 10 rl T ) H r-H 4-4 Ul) r- 04 -H E tT >i 04 a) 4J -H 4J >
-4 -4 C "r-H -- ) 4-) -* 3 ( 0a) C Cl) E > C ) C :
-'i u i t y c u a o c o m r o o fo e a o









0)




.0 a) -
41 T) 1 ro In

.( () 4

) C 4v1 a)
41 -- $ C :
*-H 0 .0 ) ) 4-l
.C 0 4- -*, 4- *- 0
U I 4- -4 Cl UO 0
H- -H 0 r -H -H LO a)
S4- a) 4-J c a U 0 44 (a)

4Q 4, J 0 Qa) .4- 4 i C 4-> 0 ) -. W ) -H rd n -I r- -
(C >i Od 0 r-l 4 4 H C-4 a) ) Mn 0) I 4 -r4 a) r 1 1 )
*Hc -r Q4 '- MH a ) 4- a) 04( U 44 (.) -I U -H 4-) fl U
H-i Cl 0 Cl n cl y i -A OI U -S r H r r- 0) a) E L-i En
q) (0 Qi 44 aH a) (C r j Q l 0o w a tC l 0 0 ml
) u3 ^ U)Q H, a o U) ) W4-i uWo Q 44 U M









4-4O ,-

a ) 4- a
-HI O o-He OO


.Q 0 C a)
4- 3 0 O) 0 r-
c 4-i 0 4-) a) Cl)
(Cl H (ml H4 H C-' l
-* 0 C a) .Q 4
- t3 H C: U 000 > l 04 OH- Q)
) C H r- ) : -H -4 -A -H C- -Hq r.
4 -H-I '4O a) *H 4- -i C 4-i a) o(T '0 -4 -Hq
'0 4- C -r >1 Q U a) U 4J C 4J >i C r- a)( 0) C
4-H Qa) a)' ) En -4 Q) Cd 01 4Ji 040 ( C -4-
- a) H-i a4 ) O >i r. .1 4-4 Q) LO fl 04 -4 >i1 a P E
a) -H l) a) 0: o O X 0 o 4- l- X tn C- H lj 4-) 0 C 0 a)1
In Q U S H Ul E U) 5
4 P4UE, U4r I I I



































APPENDIX B
SITUATION INDUCTION








The children in a financially troubled orphanage have

been given a chance to take a field trip for the first time

in five years. The director of the orphanage has to make a

decision between two possible trips the orphans may take.

One choice is to take the children to the local college's

homecoming football game. The other choice is to take the

children to the local town's annual arts and crafts festival.

The director has determined that both events will provide

the children with an equally important experience, but he

cannot decide which trip to take.


































APPENDIX C
SOURCE AND EXPECTANCY INDUCTIONS









TRADITIONAL SOURCE: EXPECTED POSITION


William Harper is a member of the track team and is

majoring in physical education. He does construction work

for the telephone company during the summer and enjoys

outdoor activities during his spare time.

For this survey you will be asked to react to Bill's

opinions concerning the field trip which support taking the

children to the homecoming football game.









TRADITIONAL SOURCE: UNEXPECTED POSITION


William Harper is a member of the track team and is

majoring in physical education. He does construction work

for the telephone company during the summer and enjoys

outdoor activities during his spare time.

For this survey you will be asked to react to Bill's

opinions concerning the field trip which support taking the

children to the arts and crafts festival.









NONTRADITIONAL SOURCE: EXPECTED POSITION


William Harper is in the nursing program and is a mem-

ber of the student artisans guild. He works with small

children at the local day care center and enjoys cooking

in his off-hours.

For this survey you will be asked to react to Bill's

opinions concerning the field trip which support taking the

children to the arts and crafts festival.









NONTRADITIONAL SOURCE: UNEXPECTED POSITION


William Harper is in the nursing program and is a mem-

ber of the student artisans guild. He works with small

children at the local day care center and enjoys cooking

in his off-hours.

For this survey you will be asked to react to Bill's

opinions concerning the field trip which support taking the

children to the homecoming football game.



































APPENDIX D
EXPERIMENTAL MESSAGES









HOMECOMING FOOTBALL GAME


William Harper


To me there is no real choice involved in deciding

where to take the orphans. Obviously, the football game

with all of its excitement and pagentry is the place to take

children who rarely are given an opportunity to leave the

confines of an institution. There is something magic about

a homecoming football game that cannot be matched by any

other experience I know. It will give the children a chance

to become a part of the outside world. The football game

will let them forget about their troubles for a few hours

and become involved in a totally new experience.

Going to the arts and crafts show won't necessarily

give the children any experience they couldn't get inside

the orphanage. The orphans can view art in books and on

television and do crafts work at the institution. Moreover,

the homecoming football game is a much more worthwhile

event than the arts and crafts festival because it will

provide the children with an experience that will help them

when they leave the orphanage and get into the real world.

I am convinced that the orphans should be taken to the

once in a lifetime experience of a homecoming football game.








ARTS AND CRAFTS FESTIVAL


William Harper


To me there is no real choice involved in deciding

where to take the orphans. Obviously, the arts and crafts

festival with all of its excitement and pagentry is the

place to take children who rarely are given an opportunity

to leave the confines of an institution. There is some-

thing magic about an arts and crafts festival that cannot

be matched by any other experience I know. It will give the

children a chance to become a part of the outside world.

The arts and crafts festival will let them forget about

their troubles for a few hours and become involved in a

totally new experience.

Going to the homecoming football game won't necessarily

give the children any experience they couldn't get inside

the orphanage. The orphans can play football at the insti-

tution and watch football games on television. Moreover,

the arts and crafts festival is a much more worthwhile event

than the homecoming football game because it will provide

the children with an experience that will help them when

they leave the orphanage and get into the real world.

I am convinced that the orphans should be taken to the

once in a lifetime experience of an arts and crafts festival.



































APPENDIX E
DEPENDENT MEASURE









Please indicate your reactions to the following statements by
placing a mark in the space that most accurately reflects your
opinions of those statements.

1. The orphans should be taken to the college's homecoming
football game.

On the scale below the closer your mark is to one of the
words, the more you feel that word indicates your attitude
towards the statement. A mark in the center scale indicates
a neutral attitude towards the statement.

good: : : : : : : :bad
foolish: : : : : : : :wise
pleasant: : : : : : : :unpleasant
worthless: : : : : : :valuable

On the following scale place a mark directly above the word
that you feel best represents your opinion of the statement.



2. re orphans s d be takn to (D t a a
3 P- D (D W o (0D P) 0 (D x 3
rt cn ti DI 0 0 0O O hI rt+
(D h F rkh CL Ct< (D (D (D
h H- i- H :1
(DcD H (D (D
H (D 3i H


2. The orphans should be taken to the town's annual arts and
crafts festival.

Use the above instructions to complete the following two
scales.

good: : : : : : :bad
foolish: : : : : : : :wise
pleasant: : : : : : : :unpleasant
worthless: : : : : : : :valuable


H rt C< Cr 1 t < CD M
H- (D W (D Pi 0 (D 0 0 OC(D x 3 H
rt hi Cln CIL Q 0 r 0 0 i 0 t ;
(D h- 1< F1 Ft h C < & C, r rDD
3 -H H Ih*i HI::
m v PJ H En
(D(D H H (D (D
H (D H


3. Please indicate how you feel about the decisions you have
made during this survey.

right: : : : : : : :wrong
confident: : : : : : : :not confident
certain: : : : : : :uncertain
negative:_:___: : : :_ :positive
sure: : : : : : : :unsure




89