1
5h

HW,HWIHHWIM

WW!

1

3&3 IIHHHWIIVIH

OZDVJO/SX

lliiiiillgi

C“ ,

LIBRARY
Michigan State
University

A
l

in *

 

 

This is to certify that the
thesis entitled

ATTRIBUTIONS 0F HOSTILITY AND DEPENDENCY RELATED TO
DIVERGENT RATINGS 0F SELF AND OTHERS

presented by

David Bennett Rosenberg

has been accepted towards fulfillment
of the requirements for

Master's degreein Clinical Psychology

%£%

Major professor /

 

 

 

Date 3/7/79

0-7639 MS U it an Ami-native Action/Equal Opportunity Institution

 

PLACE N RETURN BOX to remove thlo checkout from your record.
TO AVOID FINES return on or baton duo duo.

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

MSU lo An Afflrmdlvo Action/Equal Opportunity Institution
Walla-9.1

—.———E _

ATTRIBUTIONS OF HOSTILITY AND DEPENDENCY RELATED TO
DIVERGENT RATINGS OF SELF AND OTHERS

By
David Bennett Rosenberg

MASTER'S THESIS

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

MASTER OF ARTS

Department of Psychology
1 990

7

I?
‘4)

eosgi

ABSTRACT

ATTRIBUTIONS OF HOSTILITY AND DEPENDENCY RELATED TO
DIVERGENT RATINGS OF SELF AND OTHERS

BY

David Bennett Rosenberg

While participating in 11 small interpersonal groups, 56 undergraduates
rated own and each other's conduct on behaviorally oriented scales of
acceptance of self and others (Hurley,1989). Intrapersonal discrepancies A
(one's self-rating minus one's mean rating of peers) were separately derived for
each scale. After groups’ terminated, 47 participants described themselves and
all same—group members on Lorr and McNair's (1965) Interpersonal Behavior
Inventory (lBl).

Hypothesized linkages of each discrepancy with the lBl’s two major
unipolar factors were investigated. Self-acceptance based discrepancies
correlated significantly with these IBI factors in over 80% of all significance tests
as did 50% of the other-acceptance based discrepancies. Supporting
Sullivan's (1953) view of personal security operations and suggesting
considerable ego involvement in these ratings, persons who rated themselves
above others tended to be regarded as more Hostile, while those who rated

others above self tended to be viewed as more Dependent.

ACKNOWLEDGEMENTS

I would like to thank Dr. Frankmann and Dr. Abeles for their helpful
suggestions and comments throughout this thesis. They made themselves
easily accessible and their patience was also greatly appreciated. Most of all, I
would like to thank Dr. John Hurley. He has been an especially important
person in my life. I could not have received better guidance, tutelage, and
friendship throughout this and other projects. It is for these reasons that I have
dedicated my thesis in his honor. Finally, I would like to give a special thanks
to my wife Riva. Her love, nurturance, and companionship have helped to make

these past few years rich and fulfilling.

TABLE OF CONTENTS

Interpersonal Discrepancies
Intrapersonal Discrepancies
Operational Definitions
Hypotheses

Method

Participants

Measures and Procedures
Acceptance vs. Rejection of Self and Others (ARS & ARO)
Interpersonal Behavior Inventory (IBI)
Statistical Analyses

Results

Reliability

Present Versus Normative Samples
Hypotheses

Correlations Between All Measures
Discussion

References

Appendix A

Appendix B

Appendix C

Appendix D

(DCOQva

10
13
15
16
16
16
18
22
26
32
38
4O
41
42

LIST OF TABLES
Group Members' Mean Ratings Received from Peers on Acceptance
vs. Rejection of Self (ARS) & Others (ARO)
Results From the Analyses of Variance for All Hypotheses

Multiple Comparisons Between Intrapersonal Rating Style and
IBI-4 Means .

Correlations Between Intrapersonal Discrepancies on ARS and
ARO with IBI-4 Factors

17

19

21

24

ATTRIBUTIONS OF HOSTILITY AND DEPENDENCY RELATED TO DIVERGENT
RATINGS OF SELF AND OTHERS

”First, there is Peter as he wishes to appear to Paul, and Paul as he
wishes to appear to Peter. Then there is Peter as he really appears to
Paul-«that is, Paul's image of Peter . . . and similarly there is the

reverse situation. Further, there is Peter as he appears to himself,

and Paul as he appears to himself. Lastly, there are the bodily Peter

and the bodily Paul. Two living beings and six ghostly appearances

which mingle in many ways in the conversation between the two”

(Buber, 1957, p. 107).

l | l D' .

Psychologists have long believed that interpersonal discrepancies, or
differences between an individual’s self-perceptions and others' perceptions of
that individual, provide rich clinical data (James, 1890; Freud, 1923; Sullivan,
1953; Carson, 1969). Cline (1970) asserted that "most kinds of mental and
emotional illness involve fairly serious distortions or breakdowns in
communication as well as in interpreting and evaluating social and interactional
cues” (p.221 ). Empirical works have supported this assertion. In a study using
advanced student nurses, McGreevy (1962) found that individuals with greater
self vs. other discrepancies were viewed by clinicians as more defensive and
maladjusted than those with lower discrepancies. Other researchers have
reported similar findings (Spiegel, 1970; O'Leary and Donovan, 1974; Donovan
and O'Leary, 1976). For example, Donovan and O'Leary (1976) found that
patients who displayed markedly distorted self-perceptions ”appeared to be

significantly more anxious, obsessional, socially immature and alienated, less

insightful, and to have increased difficulties in their thinking and communication,
skills“ (p. 18) than those inpatients with less distorted self-images.

Although the study of interpersonal discrepancies has been traditionally
linked to psychopathology, recent work by Taylor and Brown (1988) has
challenged these beliefs and provided evidence that unrealistically positive
self-evaluations are associated with mental health. Most of us are confronted
with discrepancies at varying intensities throughout our daily living. Thus, a
further understanding of these and related discrepancies would appear
beneficial for the ”normal" as well as the emotionally disturbed. A program at
Michigan State University involving undergraduates who participated in small
interpersonal groups for experiential learnings (SIGEL) has yielded pertinent
data in this area. These groups channel students' attention toward a fuller
understanding of their interpersonal behavior through various methods (i.e.,
appropriate self-disclosure of feelings and thoughts, empathic listening, and
respectful confrontations) described in the assigned textbook Egan's (1976)
Interpersonal Living.

The course description stated that the aim of SIGEL groups is “to build an
atmosphere of concern and respect for each member's personhood while also
attempting to respond both constructively and honestly to each participant's
behavior within a here-and-now context.” Twice each term, SIGEL participants
rate self and each other group member on subscales representing
interpersonal behaviors two principal dimensions; Acceptance versus
Rejection of Others (ARO) and Acceptance versus Rejection of Self (ARS).
Although more commonly labeled affiliation and dominance (Wiggins, 1982),

these dimensions have been corroborated by many researchers‘. The present

 

lThe ARS & ARO labels appear to better capture the underlying psychological
processes (Adams, 1964; Hurley, 1976).

study also utilized SIGEL groups. More detailed information regarding the
structure and nature of these groups can be obtained from Hurley (1986a;
1986b) and Hurley and Rosenberg (1990).

Employing data from earlier SIGEL groups, Dillavou (1978) found that
members who rated their own behavior as significantly more favorable than it
was rated by peers, denoted Self-Overraters, scored lower in cognitive
complexity than their peers. This pattern extends into the social domain, as
Hurley and Rosenthal (1978) found that mental health professionals who
participated in a small experiential group viewed group-peers who were
extreme self-overraters as lower in interpersonal skills than those who were
viewed as extreme self-underraters. Recently, Flores (1987) confirmed this
finding in a study involving sets of 100 female and 100 male undergraduates
who participated in SIGEL groups. Hurley (1988) has also found that group
members who rated themselves more modestly were viewed more favorably by
their peers, while members with inflated self-ratings were regarded less
favorably. Earlier researchers and theorists had generally considered only the
amount of such discrepancies. The above small group studies appear to be
among the first works to consider both the direction and the magnitude of
discrepancies between one's ratings by self and others.

Social and cognitive psychologists have also contributed to the
understanding of these phenomena. (Festinger, 1957; Norman, 1969; Bem,
1972; Alperson, 1975; Shrauger & Shoeneman, 1979; Funder, 1980; Higgins,
1987; Markus & Wurf, 1987). Higgins (1987) recently presented the most
comprehensive theory of self-discrepancies to date. He stated that
discrepancies between self-state representations (or intrapsychic
representations) are related to various types of emotional vulnerabilities. A

primary assertion of this theory is that the self is represented in three basic

provinces (actual, ideal, and ought), and each province can be viewed from two
primary states or standpoints, your own and that of a significant other. The
theory holds that discrepancies between pairs of either provinces or standpoints
link to specific emotional states. For example, if there is a large discrepancy
between an individuals actual and ideal self-image, the person will experience
dejection-related emotions, while actual vs. ought discrepancies are purported
to engender agitation-related emotions. Additionally, the magnitude of these
discrepancies have been found to relate strongly to an individuals experience
of specific negative emotions and their intensities. Higgins' work has further
extended the literature by looking at how individuals' intrapsychic discrepancies
impact their effective states. However there exists another type of personality
discrepancy that researchers seem to have largely overlooked. This is the
difference between how an individual perceives and feels about her/himself in
relation to her/his perceptions and feelings regarding others. This class of
discrepancies will be referred to as intrapersonal discrepancies.

l | l D' .

Johnson (1981) observed that most of the commonly used personality
questionnaires yield more information about the respondent than about his or
her views of others. There appears to be a general agreement among
personality theoreticians and researchers that people tend to make judgements
about, describe, and generally evaluate others on traits that are central to
themselves (Lemon & Warren, 1974; Shrauger & Patterson, 1974; Markus &
Fong, 1982; Lewicki, 1983). Thus, marked differences between a person's self-
ratings on a given dimension and his/her ratings of another on the same
dimension seem likely to yield important information about the rater.

Surprisingly little attention has been paid to discrepant intrapersonal

perceptions and their relation to interpersonal behavior. The present study
seeks to elucidate some salient correlates of these discrepancies.

Intrapersonal discrepancies would appear manifested by at least two
distinct interpersonal styles. The first style, other-disparagers, is characterized
by those who rate themselves markedly more favorable than they rate their
peers. While self-effecers, or the second style, are individuals who rate their
peers as markedly more positive than they rate themselves. Although these
styles appear to be opposite extremes of relating to others, each can also be
viewed as one way of maintaining the individuals sense of personal worth.
Sullivan (1953) suggested that ”security operations” constantly monitor the
individuals perceptions in efforts to protect her/his self-esteem (p. 373).
Similarly, Markus (1980) noted that "countless studies indicate that the self
works to maintain a good image of itself” (p. 125). These views have been
empirically corroborated by Zuckerman (1979).

One way the self can maintain a positive self-view is to disparage others.
Disparagement, or having “to protect your feeling of personal worth by noting
how unworthy everyone around you is,“ represents a security operation alleged
to be a ”common phenomenon on the American scene” (Sullivan, 1953, p. 242).
He further noted:

If you have to maintain self-esteem by pulling down the standing of others,
you are extraordinarily unfortunate in a variety of ways. . . When security is
achieved that way, it strikes at the very roots of that which is essentially

human—the utterly vital role of interpersonal relations (p. 242).

Disparagement is an aggressive posture. According to Webster's (1983)

unabridged dictionary, to disparage is ". . . to match unequally . . . to lower in

esteem; to discredit . . . to speak slightingly of; to show disrespect for; to belittle"
(p. 527). It is likely that those who are disparaged perceive such acts as
unfavorable and hostile. Group members may feel especially irritated and/or
threatened by disparagers, since the latter attempt to lower the formers' feelings
of personal worth.

Individuals who favor disparaging styles have been associated with low
self—esteem and high scores on measures of prejudice (Cracker & Schwartz,
1985). These relationships have been demonstrated through the empirical
testing of downward comparison theory Millie, 1981). This theory impfies that
peOple with low self-esteem, and who also feel threatened, are most likely to
regard others unfavorably, especially outgroup members. These individuals
also maintain enhanced views of ingroup members. To test this theory,
Crocker, Thompson, McGraw, and lngerrnan (1987) conducted a two-part
experiment using students at Northwestern University. The results showed
some support for downward social comparisons, although low self-esteem
students displayed ”a generalized negativity toward all targets” (Cracker et al.,
1987, p. 911), regardless of in- or outgroup status. Thus, these findings only
partially support downward social comparison theory, and better match
Sullivan's notion of disparagement.

Self-effacement, which can be assessed by intrapersonal discrepancies,
may also be construed as a defense against low self-esteem. Although
Sullivan did not specify this style, Reyher (1981) operationalized a similar
phenomenon which he labeled self-eclipsing or an approval-seeking security
operation. Webster's (1983) unabridged dictionary defines abasement as “the
act of humbling or bringing low" (p. 2). Individuals employing a self-effacing
style probably see themselves as ”beating the other guy to the punch.‘ Markus
8: Wurf (1987) supported this notion by observing that ”people may even be self-

denigrating in the service of self-enhancement." (p. 320). Hence, strategies like
self-handicapping and self-defeating actions prior to a performance can serve
“to provide a ready-made excuse for failure" (Markus & Wurf, 1987, p. 320).

Studies that have explored self vs. other phenomenon have often been
flawed methodologically. In reviewing this literature, Wylie (1974) concluded
that many studies employed measures which have inadequate or unexplored
construct validity, solo raters, and methods which require many inferences by
the rater regarding the target person. Others have also voiced conceptual and
methodological concerns (Shrauger 8: Shoeneman, 1979). Recently, Kenrick
and Funder (1988) asserted:

research new indicates quite clearly that anyone who seeks predictive
validity from trait ratings will do better to use (a) raters who are thoroughly
familiar with the person being rated; (b) multiple behavioral observations;
(c) multiple observers; (d) dimensions that are publicly observable: and (e)

behaviors that are relevant to the dimension in question (p. 31).

This present study addressed these issues by using: (a) well-established
measures of interpersonal behavior; (b) multiple raters; (0) ratings based on
naturally occurring behavior in group sessions; (d) raters had full knowledge of
who would be rating them and on the behaviors being rated; and (e) raters who
were well-acquainted with each other.

:1 I. I D [i '|i

Intrapersonal discrepancy scores will be derived from differences between
how individuals rated self and their pooled group peers on brief scales of
Acceptance vs. Rejection of Self (ARS) and Other (ARO). These scores were

standardized (z-score) for each group to reduce the impact of groups having

atypically high or low rating norms. The three relevant rating styles derived
from these discrepancy scores will be defined as follows:

Other-disparagers are individuals who positioned self in the upper third of
the distribution of ratings separately on either ARS or ARO self-ratings, while
their mean ratings of others fell in the bottom third of either measure.

Self-effacers are defined as individuals who placed self in the bottom third
of the distribution of either ARS or ARO self-ratings, while their mean ratings of
others placed them in the relevant distribution's top third.

The least discrepant raters are all individuals who did not fit either of the
two previous definitions.

This study will examine the relationship between the above rating styles
and these rater's self and peer-based scores on a well-accepted inventory of
interpersonal behavior, the Interpersonal Behavior Inventory (IBI-4). The IBI-4
has been shown to be comprised of essentially three major factors; Affiliation,
Hostility, and Dependency (Hurley, 1989). Hypothesized linkages between the
Hostility and Dependency factors and the aforementioned intrapersonal rating
styles are listed below. Additionally, this study will also consider the role of
gender in relation to these intrapersonal discrepancies and IBI-4 factors.
WM
H1: Other-disparaging raters will be viewed by their peers as highest on the

IBI—4's Hostility factor, above both least discrepant and self-effacing raters.
H2: Self-effacers will be rated by peers as highest on the IBI-4's Dependency

factor, above other-disparagers and least discrepant raters.

Ha: Self-effacers will rate self highest on the IBI-4's Dependency factor, above

self-ratings of other-disparagers and least discrepant raters.

METHOD

E I. . I

The participants were 56 (17 male; 39 female) undergraduates at a large
midwestem university enrolled in an upper-level psychology course: Small
Interpersonal Groups for Experiential Learning (SIGEL) instructed by Professor
John Hurley for each term of the regular academic year since Fall term, 1971.
The present study employed data from all SIGEL groups for the Spring and Fall
terms of 1983. These 11 groups consisted of 3 to 7 members, usually juniors or
seniors (average age about 21 years) plus one or two leaders. Criteria for
placement into the groups were as follows: (a) students were not to be well
acquainted with any other member in their group, especially the leader(s); (b)
the student's ability to meet at the scheduled times; and (c) balancing the
male/female ratio in each group.

Nine groups were co-led and two were solo-led. Prior to leading a group,
leaders participated as SIGEL group members and then spent a subsequent
term in preparation for leadership. This latter term entailed weekly direct
observations of SIGEL groups, readings and discussions of the small group
literature, and participation in an advanced SIGEL group. Leaders (11 males; 9
females) ranged in age from 19 to about 30 with a mean age of approximately
24 years.

This study omitted members' ratings of leaders because SIGEL members
tend to rate leaders quite positively. This may be true because SIGEL leaders
are generally perceived as authority figures who are close in age to members.

Similarly, SIGEL leadership selection and training may also play a role in this

9

10

phenomena Lastly, low ratings of a leader by a member would likely be
viewed by the leader as an invitation to confront the member for rating him/her
unfavorably, and this may seem scary to members. Consequently, the leaders'
intrapersonal discrepancy scores were excluded. However, leaders' ratings of
members were retained to strengthen the data base which would have been
weakened by excluding the ratings given by leaders who are often the most
experienced and best-informed participants. All members' ratings of self and
each other were included in subsequent analyses.

Groups convened during a period of about nine weeks, meeting twice
weekly for 90-minutes, plus two 12-hour (uninterrupted) marathon sessions
usually held near the third and seventh weekend of the term, totaling roughly 50
hours of group interaction. At the end of each term, members were strongly
encouraged to anonymously rate this course on scales ranging from
exceptionally bad (1)--below average (3)--average (5)--above average (7)--to
exceptionally good (9). Returns from about 98% of all group members for the
past five years on theitem , ”Describe SlGEL's value to you, as compared with
other Department of Psychology courses” yielded mean ratings between “above
average“ and ”exceptionally good." Similar ratings were obtained when
members were asked how they would describe SIGEL to other students
unfamiliar with the SIGEL course. A very high rate of attendance at SIGEL's
small group meetings, 96.7% for the past five years, also suggests that the

course was well received by the students.

Measumndfimcedute
E | B' l' ISIIEQII IEBSEEBQI
Several prior studies (Hurley, 1976; Hurley & Rosenthal, 1978; Hurley,
' 1986a; Hurley, 1989; Small & Hurley, 1978) have established the construct

11

validity of the ARS and ARO interpersonal measures. Evidence for their
convergent and divergent validity has been demonstrated through expected
patterns of correlations with prototypical measures of dominance and affiliation
(Wiggins, 1982). For example, Gerstenhaber (1975) found that the LOV factor,
akin to affiliation, of LaForge and Suczek's (1955) Interpersonal Checklist (lCL)
linked significantly to ARO ([ = .55, n 5 .001) but not at all to ARS (r = .00).
Conversely, the ICL's DOM factor, akin to dominance, correlated .70 with ARS
(315.001) but nonsignificantly to ARO (1 = .18). In an unpublished study (Hurley,
1983) partly using data from the present sample, 47 SIGEL group members
described self and all other same-group members on Lorr and McNair's (1965)
Interpersonal Behavior Inventory (IBI) after 50-hours of group interaction. It was
found that members' mean peer-based ratings on ARS (midway, and near
group’s end) correlated positively (Time 1 = .41 & Time 2 = .63) with peer-based
ratings on lBl's five-scale Dominance factor, but linked inversely (Time 1 = -.39
& Time 2 = -.44) with lBl's four-scale lntropunitive factor. Similar ratings on
ARO, which did not link significantly to either of these factors, correlated strongly
(Time 1 a .73 & Time 1 = .74) with the lBl's six-scale Affiliation factor.

The following 10-point, bipolar, semantic differential scales assessed ARS:
Shows feelings--Hides feelings, Expressive--Guarded, Active-Passive, and
Dominant--Submissive. ARO's components were: Warm-Cold, Helps others--
Harms others, Gentle--Harsh, and Accepts others—Rejects others. All
participants completed rating booklets (See Appendix A pp. 38-39) containing
these scales twice during each group. The initial scale presented in these
booklets, Liked-Disliked, did not contribute to ARS or ARO but provided the
rater an opportunity to express strong feelings that might otherwise interfere
with his/her ability to rate accurately. Previous findings by Smith (1979) partially
supported this tactic. After Liked-Disliked, the booklet alternated ARS and

12

ARO subscales (as shown above) until all were presented (starting with Shows
feelings--Hides feelings and ending with Accepts others--Rejects others). The
scales' favorable and unfavorable anchors were irregularly ”staggered to reduce
the influence of response sets.

During SlGEL's weekly class lecture, attended by members of all groups,
the instructor thoroughly reviewed the mechanics and implications of these
ratings while also emphasizing that members' ratings had no impact on course
grades. Additionally, participants were informed that the ratings would be most
useful if they were filled-out candidly. Each group elected a data coordinator
whose duties included responsibility for distributing, collecting, and scoring the
rating booklets (these functions were closely monitored by SlGEL's instructor).
This further separated the ratings from lectures and grades.

Raters were instructed to mark the space that ”best represents your
personal impression of each member's actual behavior within all group
sessions up to now“ for each scale. The instmctions also stated that it “will be
most useful if you use the full range of possible ratings“ and that ”these ratings
will be fully shared with all group members later." The marked spaces were
translated into scores ranging from 0 to 9 for each subscale, yielding a possible
range of scores from 0 to 36 for both ARS and ARO.

The initial administration of these booklets was after each group's first
postmarathon session, following about 23 hours of group interaction. About
one week later each person received a full set of all ratings given and received
in their group for review and discussion at that group's next meeting. Following
roughly another 20 hours (or after a total of about 43 hours) of small group
interaction, the ratings were readministered and subsequently shared and

reviewed as before.

13

WWII

Circumplex models of interpersonal behavior surfaced in the literature
near the 1957 publication of Leary's seminal work, Interpersonal Diagnosis of
Personality. Building on this, and the work of others (i.e., Schutz, 1958;
Schaefer, 1959; Foa, 1961), Lorr and McNair (1963) developed the IBI in an
attempt to better assess the circular ordering of 13 categories of interpersonal
behavior. Subsequent IBI revisions (Lorr & McNair, 1965; Lorr, Bishop, &
McNair, 1965; Lorr & McNair, 1966) have added two more categories (15 total).
The lBl's original Yes-No format was also changed to a four-point scale: (1) Not
at All, (2) Occasionally, (3) Fairly Often, and (4) Quite Often.

Although the IBI was initially developed from psychotherapist's ratings of
their patients, Lorr and McNair (1965) found that relatively untrained normal
raters reproduced the lBl's circular ordering of variables and its major factor
structure. In a comprehensive and scholarly review of the literature on
interpersonal measures, Wiggins concluded that 'on both substantive and
psychometric grounds, the Interpersonal Behavior Inventory appears to be a
useful device for assessment of patient characteristics and evaluation of
therapeutic outcomes” (p. 15).

Lorr and Suziedelis (1969) sought to ascertain the underlying structure of
the lBl's specific scales and also to determine its higher-level factors. They
used three populations--nonpsychotics (clients being seen in treatment for at
least 4 months; n = 525), college undergraduates (whom they labeled normals;
‘ n = 290); and neurotics (clients in "once-a-week treatment'; n a 60). A friend or
acquaintance of at least one year filled-out an IBI-4 that described each
undergraduate. The patient's psychotherapist completed an IBI-4 for the

nonpsychotics and neurotics. The results indicated that the IBI-4's 15 scales

14

formed five unipolar factors. However, other evidence suggests fewer factors
can account for the IBI-4's underlying stnrcture (Bochner & Kaminski, 1974;
Hurley, 1989).

In a partial replication of the above study, Bochner and Kaminski (1974)
found that only three factors accounted for the variance among the IBI-4's 15
scales in a sample of 267 undergraduates at Michigan State University. They
labeled these factors: 1) Hostility--Affection, 2) Dominance, and 3)
Submissiveness. Unlike Lorr and Suziedelis (1969), Bochner and Kaminski
used only one population, nonclinical undergraduates at Michigan State
University. Furthermore, each of their subjects provided two sets of ratings; self
and an acquaintance they “liked very much.“ Unlike the works of Lorr and
Suziedelis (1969) and Bochner and Kaminski (1974), the present study
employed multiple raters (4 to 8) of each target person and obtained ratings
from naturally occurring behaviors during small group sessions.

Utilizing data including that of the present sample, Hurley (1989) factor
analyzed IBI-4 ratings of 76 SIGEL group members, again finding only three
primary factors. Labeled after the specific IBI-4 scale most central to each
factor, these were: 1) bipolar Affiliation (affiliation, nurturance, agreeableness,
sociability, minus scales inhibition, and detachment), 2) unipolar Hostility
(hostility, competitiveness, exhibition, dominance, and mistrust), and lastly, 3)
unipolar Dependency (dependency, abasiveness, submissiveness, and
deference). Hurley's factor structure was used in the present study primarily
due to sample similarities.

Participants voluntarily completed the IBI-4 for all members of their group
including self at the end of the term. In an effort to discourage ”good subject”
behavior, the participants were reminded that these IBI-4 responses would not

impact their grades. In addition to the standard IBI-4 instructions, members were

15

told to base impressions solely on their in-group observations (see Appendix B
p. 40). These ratings were completed and returned within one month of the

group's termination.

StatisticaLAnalvses

All hypotheses were tested using one-way analyses of variance (ANOVA).
Then, if the ANOVA produced a significant E (as .05), the differences between
the means were analyzed using Tukey's method of multiple comparisons. Each
hypothesis was separately tested using the data from the 23- and 43-hour
ratings for intrapersonal discrepancies on ARS and ARO. Additionally, a
summary score, which aggregated the two rating occasions, was also tested.
Thus, six one-way ANOVAs were made for each hypothesis (ARS 23 m, ; ARS
43 mu, ; summary ARS, or 23- plus 43-hour ratings; and the three parallel ARO
ratings). This served to detect differences that may have been limited to either a
specific dimension or to a single occasion.

Also examined were Pearson product-moment correlations between each
person's six intrapersonal discrepancy scores (self-rating minus the individuals
mean rating of all other group members excluding leaders) and the three salient
IBI-4 factors. All means and standard deviations for ARS and ARO
intrapersonal discrepancies are reported in Appendix C (p. 41). Statistical
significance tests used the .05 level and employed nondirectional rejection
regions for the ANOVAs and multiple comparisons. The one-tailed rejection
region was used for all correlations since direction was predicted by the

hypotheses.

 

RESULTS

B I. I ll
May (1988) has reported that peer- and self-based (23- & 45-hour)

Cronbach alpha's ranged from .84 to .93 for Acceptance versus Rejection of
Self (ARS), and from .77 to .93 for the Acceptance versus Rejection of Others
(ARO). The present data, using self-based ratings, yielded Cronbach alpha's of:
ARS = .82 and ARO = .81 after 23 hours of group interaction, and ARS = .84
and ARO = .88 after 43 hours.

This study also reflects prior works (Hurley, 1986; Hurley, 1989, Hurley &
Rosenberg, 1989) that have found these ARS and ARO measures to be
relatively independent. The present interscale correlations were .27 (self-
based) and .13 (peer-based) after 23 hours of group interaction, and rose to .44
and .36 (self and peer-based, respectively) after 43 hours. Thus, at this latter
time, each dimension accounted for only about 19% of the others' variance for

self-ratings and 13% for peer-based ratings.

WW
Table 1 shows the results of all t-tests performed between this study's

 

Insert Table 1 about here

 

sample and normative data for both the ARS and ARO measures (Flores, 1987),
and also for the IBI factors (Lorr & Suziedelis, 1969). The Flores (1987) study

16

l. 0 tl3ll'3

ABS
23-hour

43-hour

ABQ
23-hour

43-hour

EactoLSentesMean

Affiliation
Hostility

Dependency

. .
tr=-I :-

TABLE 1

z. =0 orr'::t~or.‘...:or-r =
B' I' [S IIIEBSI B Q” [EEO]
Flores 1986 Present Study
(N = 200) (N = 56)
Mean SD Mean 5.0 d t
22.2 6.5 21.7 6.3 .5 .54
25.0 5.1 24.5 5.7 .5 .67
25.8 4.1 26.2 3.9 .4 .89
27.5 3.6 27.3 4.4 .3 .48
:10: ;:| I. :go. 3 you. ': 3| e
1968 Norms Present Study
(N= 290) (N= 56)
SD Mean an d t
98.5 N/A 95.3 13.8 3.2 1.1
87.6 N/A 94.9 17.7 7.3 1.3
112.4 N/A 126.1 17.2 13.7 2.2 In: .05]

17

18

was chosen because of its large size (N = 200) and gender balance (100 men &
100 women). Mean differences between the present sample and Flores'
sample for the ARS and ARO measures did not exceed .5 on a scale which
ranged from 0 to 36. Thus, no statistically significant differences between the
samples were found.

Similarly, statistically significant differences were not found between the
sets of means for the lBl's Affiliation factor (mean difference = 3.2 with a scale
range from 58 to 232) and Hostility factor (mean difference = 7.3 with a scale
range from 47 to 188). However, a statistically significant difference was
obtained on the lBl's Dependency factor (125.05 two-tailed; mean difference =
13.7 with a scale range of 35 to 140) in which the present sample averaged
13.7 points higher than the original norms. The meaning of this difference is
unclear, and may be related to the dated norms for the IBI (the original data was
collected over 20 years ago). Additionally, the present study utilized about five
times as many raters for each participant yielding much more stable means than
the Lorr and Suziedelis (1969) study. For the purposes of this study, the

present sample appears reasonably representative of the normative data.

masthesea

The ANOVAs and multiple comparisons for each hypothesis are
summarized in Tables 2 and 3 respectively (ANOVA's are presented fully in
Appendix D, pp. 42-46).

 

Insert Table 2 about here

 

ABS
23 hour
43 hour

Summary

ABQ
23 hour

43 hour

Summary

Results from Analyses of Variance for all Hypotheses'

II II . I
E n
5.3 .008
6.8 .002
6.2 .003
.18 .82
.52 .60
.22 .81

TABLE 2

II II . 2
E D.
3.3 .05
4.8 .01
7.8 .001
.05 .96
.26 .78
.33 .72

II II . 3
E D.
3.2. .05
2.1 .13
1.1 .33
2.3 .12
1.9 .16
2.0 .14

H1: Other-disparaging raters will be viewed by their peers as highest on the
IBI-4's Hostility factor, above self-effacers and least discrepant raters.

H2: Self-effacers will be rated by peers as highest on the IBI-4's Dependency
factor, above other-disparagers and least discrepant raters.

H3: Self-effacers will rate self highest on the IBI-4's Dependency factor, above
self-ratings of other-disparagers and least discrepant raters.

' ANOVA results are fully reported in Appendix D (pp. 42-46).

19

2O

 

Insert Table 3 about here

 

The findings fully confirmed this hypothesis for ARS discrepancies, but not for
ARO based discrepancies. The ANOVAs for ARSza hour and ARS43 ho“, were
both significant (as .01), with other-disparagers rated significantly higher
(T ukey p 5 .05) than self-effacers for Hostility each time. Similarly, the summary
ARS score also yielded a significant E (p s .01 ). Other-disparaging raters were
rated highest on the Hostility factor; significantly (T ukey p 5 .05) above self-
effacers, but not significantly higher than least discrepant raters.

Individuals whose intrapersonal rating style was other-disparaging were
also rated higher than self-effacers after the 23-hour ARO ratings, but after 43-
hours, the self-effacers were rated highest. For summary ARO, other-
disparagers were viewed by their peers as highest on Hostility, followed by
least discrepants, then by self-effacers, although none of these ARO-related

differences achieved statistical significance.

"=0=|0:I0t an .m : .. r . r: .‘ .. -o:r~ .m :.1 .' :-.r

The results partially supported this hypothesis. All three ARS related ANOVA's
attained statistical significance (AR823 m," p 5 .05; ARS43 hour, 9 s .01 ; and
summary ARS, p s .01). Self-effacers, as measured by summary ARS, were
rated highest on the Dependency factor, followed in order by least discrepant

raters and by other-disparagers.

TABLE 3

tr rt: one-1.01 ==ttz=r r 1.: on. LI- \ : an =-'tr:-_11

H1: Other-disparagng (0-D) raters will be viewed by their peers as highest on the IBI-4's hostility
factor, above self-effacing (SE) and least discrepant (LD) raters.

ABS

23 hour 00 (103.6) minus S-E (86.0) 17.6 4 .6'
23 hour O—D (103.6) minus LD (94.6) 9.0 2.4
43 hour O-D (104.1) ninus S-E (84.6) 19.5 5 .2 '
43 hour 00 (104.1) minus LD (95.4) 8.7 2.3
Summary O-D (104.7) minus S-E (86.1) 8.6 4 .9 '
Summary O-D (104.7) minus LD (94.8) 9.9 2.4

H2: Self-effacers (S-E) will be rated by peers as highest on the IBI-4's Dependency factor, above
other-asparagus (OD) and least discrepant (LD) raters.

W 9.0mm W028 Iulsex
ABS

23 hour S-E (130.7) ninus O-D (118.2) 12.5 3.2
23 hour S-E (130.7) ninus LD (129.6) 1.1 <1

43 hour S-E (135.0) ninus O-D (118.7) 16.3 4 .3'
43 hour S-E (135.0) ninus LD (125.0) 10.0 2.6
Summary S-E (132.3) m'nus O-D (114.3) 18.0 5 . 1 *
Summary S-E (132.3) m’nus LD (130.9) 1.4 <1

H3: Self-effacers will rate self highest on the IBI-4's Dependency factor, above self-ratings of
other-disparagers and least discrepant raters.

Batlnoeeeasien Qonnadeenltieanl Meennitletence Iulsex
ABS

23 hour S-E (129.8) minus O-D (112.3) 17.5 3.5*
23 hour S-E (129.8) minus LD (123.6) 6.2 <1

 

1 Summary score means are not equivalent to the mean of 23-hour plus 45-hour ratings. They
represent the highest, middle, and lowest thirds of the aggregated distribution.

21

22

The difference between self-effacers and other-disparagers attained
statistical significance (p 5.05), while the difference between self-effacers and
least discrepants did not. This pattern also held for 23- and 43-hour ARS
ratings with self-effacers rated higher than other-disparagers, and above,
although not significamly so, the least discrepants. No ARO based ANOVA was
statistically significant. However, in each instance self-effacers were rated
highest, twice followed by least discrepants and other-disparagers (summary
and 23-hour ARO), and once followed by other-disparagers and least
discrepants (ARO43 hour).

.° : -:r. : .r' ._ = : rot: or t: = -' Dun-:10: an
2.. : = - . to . . r: .' .. .o: .n :4 o. :-.t . : .Onlyoneof
the three ARS related ANOVAs was significant (ARS 23 hour, :1 5.05). Summary
ARS self-effacers were rated higher than least discrepants and other-
disparagers, although this difference was not statistically significant. For AR823
hour, self-effacers were rated significantly higher (p 5 .05) than other-
disparagers. Although the pattern was the same at ARS“ how, the differences
were not statistically significant. No ARO based ANOVA was statistically
significant, although the pattern of these means fit the hypothesis. Thus,

effacers self-ratings were always higher than self-ratings for either least

discrepants or other-disparagers.

QenelatieneBetweenALMeaeuLes

Table 4 presents all correlations between ARS and ARO intrapersonal
discrepancy measures and the three IBI-4 factors. Two-thirds (44 of 66) of
Table 4's correlations achieved statistical significance and about 70% of these

(31 of 44) were positive. Gender was omitted from Table 4 because it

23

 

Insert table 4 about here

 

participated in only one statistically significant relationship. Self-rated hostility
was linked inversely (g 5 .01) to gender, indicating that men's self-perceptions
included more hostility than did womens'.

All 15 correlations involving the six ARS and ARO discrepancy measures
were positive, and over two-thirds achieved statistical significance. Each
measure's 23-hour discrepancy correlated significantly (as .001) with the
same index at 45-hours. Thus, individuals who displayed intrapersonal
discrepancies relatively early on tended to maintain these discrepancies later in
the group. 01 the four intersource and interoccasion correlations, only one
attained statistical significance. AR823 ho... discrepancies accounted for about
13% (r - .36) of the variance in ARO 23 hour discrepancy scores. At 45-hours,
ARS discrepancies accounted for only 4% of ARO discrepancy scores. The
correlation between summary scores on each measure also attained statistical
significance ([ = .29, 515.05), providing further evidence of the relative
independence of these indices. As expected, both summary scores correlated
significantly (p 5.001) with all of its own components.

Among the IBI-4 factors, eight of 15 correlations achieved statistical
significance ((12.05). Showing strong to modest intersource agreement, each
same-factor rating by self vs. peers attained statistical significance: Affiliation [ =
.71 (95 .001 one-tailed); Hostility 1 =29 (Q5 .05 one-tailed); and Dependency I
:- .48 ((25 .001 one-tailed). The two remaining significant correlations were
negative, linking peer-based Affiliation inversely with Dependency, both peer-
based ( r a -.29, n 5 .05 one-tailed) and self-based ([ = -.34, p 5 .05 one-tailed).

24

3.8.2: none 3.... 3.3.25 mono mvi. 3:863. Soda 3.: 3:865 Sofia «v.1

3.8.82 36d 8.1. 3:265 none 3.: 8.3.82 8nd 3.: . 33.2.6 8%.. 8.... .
335.82 mono ami. 3.2.0.8 mono vui. 3.2.22 none 3.... 33.23 mono «Ni.
85...: Ronni-a 13:28:12 228a 1m. 3398... 305358“. 3.2
3.. a. an . as. . an. . a. . am. 3 . 8. . mu. . a. . a. one
3. 2. . ow . 2. . R. . , 8. 8. 3. o... . .8. . 3. . E. amo
a. 9. . an 8. 2. . 2.. 9.. 2. . B. 3. 8. Um. no:
5 . an . an. 8. no. . «a. . 8. . 8. 3t 3. 3.. E. mo:
2.. a... 8. 8. «A. 9.. 3. an. 3. an. a. E “:2
3r . a. . z. . me. . «A. 8. «4. 2.. 3. 8. 8. a. uu<
aw . 8. a _. . «a. . 9. 3.. an. 8. 8. R. R. 25
am . 8. a... . 8. . 3. we. 8. 8. t. 8. z. 52.9
an . 3. m... . 8. an me. 8. 8. an mu. 8. 52.8
an . 8. . 3. 3r 3. 8. aw. :. 8 3. 5. saw
3 . R. . NP. 3. an. 8. mm. 8. mm a. no. 52.9.
R . an. . 8. 3. 8. en. 8. :. on. 5. mm. 52.8
a. own a. awe a. me: as mo: 3. as? $9.2 saw 52.9 52.8 53. 52-9 52.8.
guin— qu wm<
3.... .u ..:eu...... "4.. 3.2.13.2. 2.7.3"... ...:nu .e.n_..

v m.._m<._.

25

Thus, more affiliative persons were seen as less self-abasing by both peers and
self.

Table 4's 36 remaining correlations concerned linkages between the two
types of intrapersonal discrepancies and the three IBI-4 factor scores. Two-
thirds (24 of 36) of these correlations attained statistical significance, and about
60% were positive. Accounting for half (12 of 24) of the statistically significant
relationships, all discrepancy measures linked positively and significantly to
both self- and peer-based Affiliation. These correlations suggest that members
who viewed themselves as more self- and other-accepting than their peers
tended to be generally perceived as relatively Affiliative.

Self-ratings on Dependency correlated negatively and significantly with all
ARS and ARO discrepancies. However, peer-rated Dependency linked
negatively and significantly only with ARS discrepancies. Thus, members who
displayed a self-effacing intrapersonal rating style were also viewed by their
peers and self as more Dependent. The remaining three significant correlations
were negative and linked Hostility with all ARS discrepancies. This suggests
that members who place themselves above others on self-acceptance also

generally regarded themselves as more Hostile than others.

DISCUSSION

The representativeness of the present sample was evaluated due to its
modest size and the unique small group context. Mean differences between the
present sample and Flores' (1987) sample of 100 men and 100 women
previously enrolled in the same psychology course never exceeded .5 (g > .10)
on scales ranging from 0 to 36 and t-tests revealed no significant intersample
differences on the measures of self- and other-acceptance. Hurley (1989)
noted that the IBI-4 norms were absent from the literature as they were omitted
from the original article revising the IBI (Lorr 8. Suziedelis, 1969), and also in a
later examination of the lBl's factor structure (Bochner & Kaminski, 1974). The
present IBI-4's Hostility and Affiliation factors showed no statistically significant
differences from the available normative data, although the Dependency factor
did differ significamly (mean difference = 13.7, p 5.05). The 20-year gap since
the establishment of these norms, and the methodological differences (one rater
vs. five raters per target) likely contributed to this finding. The present data
appears reasonably representative of IBI-4 factor scores, given its
methodological advantages, and that two of the three IBI-4 factors did not differ
significantly from earlier reports.

This study's central findings supported the assertion (hypothesis 1) that
individuals who rate themselves markedly above others will be viewed as
relatively hostile by those others. Group members who rated themselves
considerably more favorably than they had rated others were viewed by these
peers as more hostile. This was especially true for individuals who manifested
other-disparaging intrapersonal discrepancies on the self-acceptance scale.

They were viewed by their peers as significantly (p 5.05) more hostile than

26

27

others both midway through the group, and near its end. The other-acceptance
scale yielded similar differences although these were not always statistically
significant.

Results from the second hypothesis regarding self-effacing intrapersonal
discrepancies also followed the expected pattern, although fewer of these
relationships attained statistical significance. Members who exhibited self-
effacing ARS and ARO based discrepancies were rated by peers, and by self,
as highest on the IBI-4's Dependency factor. All ARS discrepancies were
significant for peer-based ratings, but for self-ratings, only the summary ARS
discrepancy attained significance.

It had not been anticipated that intrapersonal discrepancies along the ARS
and ARO dimensions would relate differentially with the IBI-4's personality
factors. However, no ANOVA pertaining to other-acceptance or ARO
discrepancies attained statistical significance, while over two-thirds (7 of 9) of
the ARS related—discrepancies did ((15.05). Intrapersonal discrepancies
apparently play a larger role when an individual sees him/herself as more
assertive (i.e., expressive, active) vs. submissive (i.e., passive, guarded), as
opposed to more or less "nice“ (i.e., warm, gentle) than his/her peers.
Alternatively, it may be that the ARO discrepancies link to variables not fully
represented by the IBI-4 major factors. For example, individuals who rate self
higher than others for warmth and acceptance may be seen by their peers as
more compliant or superficial.

One area of study that has relevance to the present work is Brown's (1986)
concept of the “self-other bias.” Brown had individuals rate themselves and
others on various trait adjectives and the findings suggested a “pervasive
tendency to cast the self in more positive and less negative terms" in relation to

others (p. 370). However, Brown also identified three variables which he

28

believed would limit the magnitude of this ”self-other bias.“ He suggested that
the bias decreases as you become better acquainted with the person to be
rated. Conversely, he asserted that this bias would likely increase with greater
ambiguity of the trait to be rated. Finally, Brown noted that ”it is likely that
requiring individuals to appraise themselves and others publicly would reduce
the magnitude of the self-other bias” (p. 373). All of these features seem
pertinent to the present study, as all individuals were well-acquainted, the
“traits“ were relatively unambiguous, and it was known in advance that all
ratings would be shared publicly within the group. These considerations
suggest that the ”self-other bias" would be diminished in the present
circumstances.

Nevertheless, the present findings appear to partially support the self-other
bias. Individuals in the present population did rate themselves higher than they
rated others on self-acceptance (ARS mean discrepancy . 3.3). However, a
considerably smaller difference (ARO mean discrepancy a -.13) for the other-
acceptance scale was found. This interscale difference was statistically
significant (t = 2.00, 95 .05 one-tailed test). Closer inspection revealed that the
self-other bias tended to decrease as group participants became better
acquainted (AR823 mu, mean discrepancy = 1.9; self vs. other 1:- 2.00, g 5 .05
one-tailed test: ARS“ he... mean discrepancy = 1.5; self vs. other 1: 1.64, p >
.10 one-tailed test). Thus, there was modest support for the self-other bias on
measures of self-acceptance (people may tend to disparage others), however,
for scales of other-acceptance the self-other bias faded and its direction
reversed in the present sample.

One possible explanation is that these college students were concerned
with appearing ”nice“ or “social” and were therefore less likely to rate others as

more toward cold, harsh, unhelpful, and rejecting. Additional insight into this

29

question could be gained from looking at intrapersonal discrepancies on the
major IBI-4 factors. It may be that a self-other bias exists for the prosocial scales
that comprise the Affiliation factor, but not for the antisocial scales in the Hostility
factor. Further research in this area is needed to more meaningfully understand
these differences.

The correlations between the IBI-4 factors and the ARS and ARO
measures yielded mixed findings. Supporting the notion of disparagement
expressed by Sullivan (1953; 1970), and others (Crocker & Schwartz, 1985;
Crocker et. al., 1987), individuals with greater intrapersonal discrepancies on
self-acceptance were viewed as relatively hostile by their peers. Significant
negative correlations between discrepancies on self-acceptance with peer-
based ratings on the Dependency factor (i.e., scales of dependence,
abasement) suggested that individuals with self-effacing intrapersonal rating
styles are viewed as more Dependent by their peers. In addition, peers also
viewed raters whose style was toward self-effacing (on the self-acceptance
measure) as more hostile than others. While it may be true that these self-
effacers are attempting to ”beat the other guy to the punch,” they may also be
throwing a few punches in this process. Evidence for this was reflected in the
significant negative correlations between the IBI-4's Dependency and Hostility
factors. Group members who rated themselves (or were rated by peers) as
more toward dependent and abasing tended to be viewed by their peers as
higher in hostility.

There was surprisingly little overlap between intrapersonal discrepancies
on self- and other-acceptance. Thus, discrepancies on self-acceptance or ARS
accounted for about 13% of the variance of those on other-acceptance or ARO
after 23-hours of group interaction and only about four percent after 45-hours.

The relative independence of these measures ought to be considered in the

 

30

design and methodology of future studies. The IBI-4 factors of Affiliation,
Hostility, and Dependency showed moderate agreement between ratings by
pooled peers and self. Each of these intersource correlations attained statistical
significance (Affiliation = .72, p 5 .001; Hostility a .29, p 5 .05; Dependency =
.48, p 5 .01). This agreement was significantly stronger for the prosocial
Affiliation factor than for the relatively neutral Dependency factor and the
antisocial Hostility factor.

Results from the present set of correlations also produced some puzzling
findings. Among these were the positive and significant links between IBI-4's
self- and peer-based Affiliation factor (i.e., scales of affiliation, nuturance, etc.)
with all intrapersonal discrepancy scores. Thus, individuals who rated
themselves above their peers on measures of self- and other-acceptance were
also viewed as more affiliative by both self and peers. One possible
explanation for this relationship is that these individuals were actually more
accepting of themselves and others, resulting in more affiliative group
behaviors. .

The present work has addressed the methodological flaws inherent when
examining “trait” ratings and discrepancy phenomenon (Wylie, 1974; Kenrick &
Funder, 1988). This study also utilized a natural environment from which the
participants had nearly 50 hours of group interaction to make well-inforrned
ratings about their self and peers. However, several limitations must be noted.
First, the sample size was relatively small. Each category of intrapersonal
discrepancies (Other-Disparaging, Least Discrepant, and Self-Effacing) was
obtained by dividing the distribution of the total sample into thirds (yielding only
18 or 19 persons per category). Although this enhances confidence in the
present finding's generalizability, this process may not give a very clear picture

of more extreme disparagers or effacers (0.9., the bottom 5% of the distribution).

31

Another area of concern is that the participants were students enrolled in
an upper level psychology course at a large midwestem university.
Generalizing these results to other populations appears premature. Most
importantly, there appears a general lack of knowledge specifically addressing
intrapersonal discrepancies. Further work in this area might profitably focus on
the role of these discrepancies in other areas of life (i.e., job settings,

intelligence, judging the quality of works of art, music etc).

LIST OF REFERENCES

REFERENCES

Adams, H. B. (1964) “Mental illness” or interpersonal behavior? Amefigan
Wig, 12, 191-197.

Alperson, B. (1975). In search of Buber's ghost: Acalculus for
interpersonal phenomenology. 820mm 29, 191-97.

Bern, D. J. (1972). Self-perception theory. In L. Berkowitz(Ed.),Adyance§
W. New York: Academic Press.

Bochner, A. & Kaminski, E. (1974). Modes of interpersonal behavior: A
replication. W, 35, 1079-1083.

Brown, J. (1986). Evaluations of self and others: Self-enhancement bias
in social judgements. W A, 353-376

Buber, M. (1957). The \Mlliam Alanson White memorial lectures. fourth
series. 23mm 20, 95-129.

Carson, R. (1969). W. Chicago: Aldine.
Crocker, J. & Schwartz, I. (1985). Prejudice and ingroup favoritism in a

minimal intergroup situation: Effects of self-esteem. W
Win. 1.1.. 379-86.
Crocker, J., Thompson, L, McGraw, K. 81 lngerman, C. (1987). Downward

comparison, prejudice, and evaluations of others: Effects of self-

esteem and threat. W 52.
907-916. ‘

Cline, V. B. (1970). Interpersonal perception. In B. A. Maher (ed)

WWW. Vol. 1. New York:
Academic Press, pp. 221-284.

32

33

Dillavou, D. (1978). WWW-QM
UNPUbliShOd master's thesis, Michigan State University, East Lansing.

Donovan, D. M. &O'Leary, M. R. (1976). Relationship between distortions
in self-perception of depression and psychopathology. JgumaLQL
QllnieaLEsxcholooL 32. 16-19.

Egan. G. (1976). W
W. Monterey: Brooks/Cole.

Festinger, L. (1957). W. Evanston, IL: Row,
Peterson.

Flores. N. (1987). QenomitieeendeixeroeneieszeeeLxuieersbased

raflnoeountemersenaLbenaxieLinemalLomues Unpublished master‘s
thesis, Michigan State University, East Lansing.

Foa, U. G. (1961). Convergences in the analysis of the structure of
interpersonal behavior. W. 58. 341-53. A

Freud, S. (1923). Instincts and their vicissitudes. In J. Strachey (Ed. and
Trans), WM
Sigma (Vol. 14, pp. 109-140). London: Hogarth Press, 1957.

Funder, D. (1980). On seeing ourselves as others see us: Self-other
agreement and discrepancy in personality ratings. Jgumamt
W, 18, 473-493.

Gerstenhaber, L. M. (1975). W
W. (Doctoral dissertation, Michigan State
University 1974). W 35, 473-492.

Higgins, E. T. (1987). Self-discrepancy: A theory relating self and affect.
W. SA. 319-40.

34

Hurley, J. (1976). Helpful behaviors in groups of mental health

professionals and undergraduates. WWW
Esxcnothetaott 25173-189.

Hurley, J. (1978). Toward seeing ourselves as others see us. In L.
Wolberg, M. Aronoson, & A. Wolberg (Eds.), W (pp.
65-78). New York: anner/Mazel.

Hurley, J. (1983). Unpublished raw data.

Hurley, J. (1986a). Leaders' behaviors and group members' interpersonal
gains. (3391.18.19, 162-176.

Hurley, J. (1986b). Interpersonal behavior, range of ratings, and personal
security. EsxcheleoieaLBenens. 59, 219-228.

Hurley, J. (1988). Interpersonal correlates of discrepant ratings by self and
peers. W- 128. 653-665-

Hurley, J. (1989). Self-acceptance and other-acceptance scales for small
QVOUPS- WWW 1.1.5.
485-505 .

Hurley, J. 8 Rosenthal, M. (1978). Interpersonal perceptions within
AGPA's institute groups. Limp, 2, 220-238.

Hurley, J. & Rosenberg, D. (1990). Leaders' behaviors associated with group
members gains on scales of self-acceptance and other-acceptance.
Unpublished manuscript. Michigan State University, East Lansing.

James, W. (1948). Esmhglggy. New York: World. (Original work published
in 1890).

Johnson, J. A. (1981 ). The “self-disclosure” and "self-presentation”

views of item response dynamics and personality scale validity.

JoumaLoLEereenalitLandSoeiaLEemheloox. 4.0. 09- 761 -69.

35

Kenrick, D. 8 Funder, D. (1988). Profiting from controversy: Lessons
from the person-situation debate. W151. 53, 23-
34.

LaForge, R. 8 Suczek, R. F. (1955). The interpersonal dimensions of

personality: Ill. An interpersonal checklist. W24.
94-112.

Leary, T. (1957). W- New York:
Ronald Press.

Lemon, N. 8 Warren, N. (1974). Salience, centrality and self-relevance of
traits in construing others. W
mm, 13, 119-24.

Lewicki, P. (1983). Self-image bias in person perception. .IQumamL
EateenalihtandfieeiaLBsxchelmv. 9.5. 384-93.

Lorr, M. 8 McNair, D. M. (1963). An interpersonal behavior

circle. JeumaLetAbnennaLendSoeialfismnelooL 2. 823-830.
Lorr, M. 8 McNair, D. M. (1965). Expansion of the interpersonal behavior

circle. JournaleLEeLsenalinrendSeelalfisxeheleox. 2. 823-830.
Lorr, M. Bishop, P., 8 McNair, D. M. (1965). Interpersonal types among

psychiatric patients behavior circle. WWW
1Q, 468-72.

Lorr, M. 8 McNair, D. M. (1966). Methods relating to evaluation of
therapeutic outcome. In L Gotthard 8 A. Auerbach (Eds.), mum

Way, New York: Appleton-Century-Crofts.
Lorr, M. 8 Suziedelis, A. (1969). Modes of interpersonal behavior.

WW! 8. 124-132.
Markus, H. (1980). The self in thought and memory. In D. Wagner 8 R.

Vallacher (Eds.), W1 (pp. 102-130). New York:

36

Oxford University. Press.

Markus, H. 8 Fong, G. (1982). Self-schemes and judgements about others.
Winn, 1, 191 -204.

Markus, H. 8 Wurf, E. (1987). The dynamic self-concept: A social
psychological perspective. w, 38. 299-337.

May, 8. A. (1988). Images of self, peers’ perceptions, and reflexive self-images:
New support for the symbolic interactionists (Master's thesis, Michigan
State University,1988). WWW , 21.

McGreevy, J. C. (1962) Interlevel disparity and predictive efficiency.
WM 25. 80-87.

Norman, W. T. (1969). 'To see ourselves as ithers see usl": Relations
among self-perceptions, peer-perceptions, and expected peer-
perceptions of personality attributes. Wm
3mm, 4, 417-43.

0' Leary, M. R. 8 Donovan, D. (1974). Perception of depression in self and
others among male alcoholics. MW, 32,
16-1 9.

Reyher, J. (1981). Unpublished manuscript.

Schaefer, E. S. (1959). A Circumplex model for maternal behavior. Journal
W 5.9.. 226-35.

Schutz, W. C. (1958). FIRO: W
mum. New York: Rinehart.

Shrauger, J. 8 Patterson, M. (1974). Self-evaluation and selection of'

dimensions for evaluating others. W11, 42, 569-85.
Shrauger, J. 8 Schoeneman, T. (1979). Symbolic interactionalist view of

self-concept: Through the looking glass darkly. EmhologicaLBulletln.
£5, 549-73.

37

Small, D. 8 Hurley, J. (1978). Work effectiveness and personal attributes
of mental health paraprofessionals. In K. Nash, N. Litton, 8 S. Smith
(Eds.), W. New Haven: Advocate
Press. pp. 286-301.

Smith, J- A- (1979)- WWW
ratings. Unpublished master's thesis. Michigan State University, East
Lansing.

Spiegel, D. (1970). Psychological health-sickness and the perception of
self and others. Wm, 25, 274-76.

Sullivan, H. S. (1953). W. New York:

Nonon.

Sullivan, H. S. (1970). W. New York: Norton.
Taylor, S. 8Brown, J. (1988). Illusion and well-being: Asocial

psychological perspective on mental health. W,
103, 193-210.

\Mggins, J. (1982). Circumplex models of interpersonal behavior in
clinical psychology. In P. C. Kendall 8 J. N. Butcher (Eds.), flangmgkgt,
Leseamnmethndeinellniealmheleox New York: VWeY- pp-
183-221.

Willis, T. (1981). Downward comparison principles in social psychology.
EmbelooleaLBulletin. en. 245271.

Wylie, R. (1974). MW). Lincoln: University
of Nebraska Press.

Zuckerman, M. (1979). Attribution of success and failure revisited, or: The

motivational basis is alive and well in attribution theory. Journal 9L

Bersenalitv. .41. 245-287.

LIST OF APPENDICES

APPENDIX A

RATINGS OF BEHAVIORS IN GROUPS

INSTRUCTIONS: On this minibooklet's last page note that all group members'
Rate all group members, including self and leader(s). These ratings will be fully

shared with all group members later. Complete all ratings on each page before
turning ahead to the next. Unlike other scales which address behavior, the

represents your personal impression of each members' actual behavior within
Liked versus Disliked scale solicits your personal responses.

all group sessions up to now. These ratings will be most useful if you use the

names have been listed. Encircle your own name. Starting with the following
full range of possible ratings for each scale.

page, encircle the letter between the extremes of each scale that best

SHOWS FEELINGS

def. ghl ilklm n
detghijklmn
dellgh.| jklmn
dear ghi .Iklm n
dellghtljk|mn
dellghijk'mn
dOTgthklmfl
.OOI Chi .Iklm n
dallgh.l jkllmn
dOTghI jklMfl

HID—Es FEELINGS

DIS-LIKED

defghtl.lkllmn
dOfghIIkImn
dellghtljk'mn
defghljklmn
defgh.ljkl.mn
dafgh.l.lakimn
defghijk'mn
defghljklmn
defghljklmn
defghijklmn

LIKED

EXPRESSIVE

adeOfgthkimn
abc defgh.' clk|mn
abchfghIIklmn
adeellgh:|jkllmn
abodefghijkllmn
abode-elghijk'mn
abodallghijklmn
abchIQhI.Iklmn
adeefgthklmn
adeOTghijkIMH

GUARD—ED

COLD
adeOTgthkImn
abnvdetlghxljkllmn
adeefgh'jkI-mn
abodaTghIIRImn
abc dOTgh.I.I.kl.mn
adeellghijkI-mn
adeefghIlklmn
adeellghijk'mn
adeOTghIiJk'mn
abodellghxljkllmn

WARM

38

39

 

PASSIVE

b0
b0
b0
b0
b0
b0
b0
b0
b0
b0

dddddddddd

Of ghlirklmn
Of 9h.l.I.KImn
.I.I..Kl.mn
Bf 9h.l.IkImn
.Ijklrmn
afghiirklmn
afghlirklmn
9619.".I .Jklrmn
.Ijkimn
Sightiitknlmn

ACTIVE

HARMS OTHERS

bc
b0
b0
b0
b0
b0
bc
b0
b0
b3

dddddddddd

afgholl
6....th
afghtl
afghcl
OTghI
afghi
afgh.l.
afghi
Bighi
Ofghi

Iklm
.13-Kim
Iklm
irklm
.IkIm
Iklm
Ilk'm
Iklm
Iklm
Irh

nnnnnnnnnn

HELPS OTHERS

SUBMISSIVE

b0
b0
b0
b0
b0
b0
bC
bC
b0
b0

dddddddddd

9819.".Iiklrmn
afghijnK'mn
O ghijkImn
afghiirklmn

DOMINANT

GENTLE

deOTgh.I.lak|mn
bedallghtlilklrmn
deefgthk|mn
bedefghi.lakllmn
deafgthklmn
bodellghijkllmn
bedefghijkimn
bcdofghijklmn
bchfghIIkImn
deOllg-ntljkimn
MARSH

encircle your name\‘

8888888888

$8
to)
(d)

 

A

 

bbbbbbbbbbc

 

 

t?)
lei
(h)
(I)
(I)
(k)
(I)
(ml
(n)

 

 

 

 

 

 

 

 

 

 

CEPTS OTHERS

0.0

c
d

0d

see
ddd

0.0
0.0
cd
0.0

afghijklmn
afghijkllmn
OTgth-KImn
k

f

O ghljklmn
afghijklmn

afgho'j

REJECTS OTHERS

APPENDIX B

Interpersonal Behavior Inventory-4 (Instructions)

Consider what interpersonal behaviors you have observed and what the person
says. Discount anecdotes reported by others, or other second-hand
information.

Behavior manifested varies with the persons involved and with the individuals
role. Rate what is most typical of the person.

In arriving at a judgement consider the individuals attitude and interactions to
you along with other information.

As much as possible base your ratings on directly observable behavior.

Make no effort to present a consistent portrait. People may manifest, for good
reasons, seemingly contradictory behaviors.

If you cannot decide, go on to the next item and come back later to those items
you skipped.

If you feel uncertain about a judgement, record your best guess. Be sure to
judge every statement.

40

 

23-hour
43-hour

Summary

ABQ

23-hour
43-hour

Summary

APPENDIX C

41

1.85
1.45
3.30

.12
-.25
-.13

7.0
6.7
11.8

4.7
3.9
7.3

APPENDIX D

Full Descriptions of ANOVA's

H1: Other— disparaging raters will be viewed by their peers as highest on
the IBI-4's hostility factor, above both least discrepant based or self-

effaci ng raters.

Source
Between groups
Within groups
Total

Souroe
Between groups
Within groups
Total

Source
Between groups
Within groups
Total

Souroe
Between groups
Within groups
Total .

2
53
55

53
55

53
55

53
55

ARS'23 hour
of SumoLSouaree MeanSouare E 9

2876.6 1438.3 5.3 .008
14413.1 271 .9
17289.7

AF1343 hour

SumotSoueLesMeanSouete E0

3539.3 1769.7 6.8 .002
13750.4 259.4
17289.7

ARSSummary

SumotSoueLesMeanSouaLe En

3278.7 1639.4 6.2 .003
14010.9 264.4
17289.6

A8023 hour

Sum_ol_SoueLeeMeen_SoueLe E0

127.6
17162.0
17289.6

63.8 .20 .82
323.8

42

Source
Between groups
Within groups
Total

Source
Between groups
Within groups
Total

111
2

53
55

53
55

43

A6043 hour
SumotSouares W E 0
335.0 167.5 .52 .60
16744.3 322.0
17079.3
AROSummary
Sum.ot.$ouares W E a
140.3 70.1 .21 .81
17149.4 323.6
17289.7

H2: Self-effacers will be rated by peers as highest on the IBI-4's Dependency
factor, above other-disparagers and least discrepant based raters.

Source

Between groups
Within groups
Total

Source
Between groups

Within groups
Total

111
2

53
55

53
55

ARSZ3 hour
Sum.oi_Souares W E 9
1786.5 893.3 3.3 .05
14470.6 273.0
16257J
AF1543 hour
Wares MeenSouere E o
2473.0 1236.5 4.8 .01
13784.2 260.1

16257.2

 

44

ARSsUmr-y
Source or SumoLSouares W E o
Between groups 2 3694.3 1847.1 7.8 .001
Within groups 53 12562.9 237.0
Total 55 1 6257.2

ARO23 hour
Source or W W E 2
Between groups 2 27.3 13.7 .05 .96
Within groups 53 16229.8 306.2
Total 55 1 6257.1

ARO43 hour
Source of Wm W E u
Between groups 2 156.9 78.4 .26 .78
Within groups 53 15994.7 307.6
Total 55 16151.6

AROSumnary
Source of SumoLSouares W E 0
Between groups 2 198.9 99.4 .33 .72
Within groups 53 16058.3 303.0
Total 55 1 6257.2

H3: Self-effacers will rate self highest on the IBI-4's Dependency factor, higher
than both other-disparagers and least discrepants rate self.

Source

Between groups
Within groups
Total

Source
Between groups

Within groups
Total

Source

Between groups
Within groups
Total

Source

Between groups
Within groups
Total

Source
Between groups

Within groups

otSurnoLSouereeMeanSouere

2
44
46

44
46

44
46

44
46

44

45

AF623 hour

2499.9
17024.0
19523.9

ARS43 hour

SumoLSouaresMeanSouare

1761.1
17762.9
19524.0

ARSSummary

SumoLSouereeMeenScuere

953.0
18571.0
19524.0

A8023 hour

SumcISouaresMeanScuere

1828.0
17695.9
19523.9

ARO43 hour

Sum.ot.SouereeMean.Souere

1574.8
17714.7

1250.0
386.9

880.5
403.7

476.5
422.1

914.0
402.2

787.4
412.0

E
3.2

2.2

2.3

D.
.05

.13

.33

.12

.16

Total

Source
Between groups
Within groups
Total

46

44
46

46

1 9289.5
AROSurnmary

Sum.01_SouoresMean_Souare

1656.7
17867.3
19524.0

828.3
406.1

E
2.0

D.
.14

 

"‘lllllllllllllIll“