The role of identity style in reconstructing the self following self-discrepant information

MISSING IMAGE

Material Information

Title:
The role of identity style in reconstructing the self following self-discrepant information
Physical Description:
viii, 135 leaves : ill. ; 29 cm.
Language:
English
Creator:
Moore, Margaret Allison
Publication Date:

Subjects

Subjects / Keywords:
Identity (Psychology)   ( lcsh )
Self-perception   ( lcsh )
Genre:
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )

Notes

Thesis:
Thesis (Ph. D.)--University of Florida, 1994.
Bibliography:
Includes bibliographical references (leaves 127-134).
Statement of Responsibility:
by Margaret Allison Moore.
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 - 002027462
notis - AKL5061
oclc - 33017553
System ID:
AA00003633:00001

Full Text









THE ROLE OF IDENTITY STYLE IN RECONSTRUCTING THE SELF
FOLLOWING SELF-DISCREPANT INFORMATION








By

MARGARET ALLISON MOORE


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

UNIVERSITY OF FLORIDA


1994













This dissertation is dedicated to my mother, whose

strength and courage I have admired, and to Eva Parker,

whose altruism will forever be appreciated.












ACKNOWLEDGEMENTS

I would like to thank all who helped support me in my

development as a counseling psychologist. I am grateful to

my advisor and chair of my dissertation committee, Greg

Neimeyer, for introducing me to personal construct theory

and for providing wise words and timely turnarounds. Marty

Heesacker was a very valued member of my committee, and I

look forward to collaborating with him in the near future.

I am thankful to other members of my committee, Franz

Epting, Max Parker, Stephen Kraus, and Donna Webster, for

their invaluable comments and overall interest in my

professional development. I feel very fortunate to have

learned about psychotherapy from Harry Grater, Michael

Murphy, Paul Schauble, Jaquie Resnick, and Nancy Coleman;

they have contributed immensely to both my professional and

personal growth. I want to pay special acknowledgement to

my close friends, Tamara and Charles Martin, with whom I

shared a wealth of good times. I also want to express my

gratitude to Gray, my cat, who kept me company at the

computer and who waited until after my proposal was defended

to pass away. Finally, I wish to recognize my future

husband, Tom Britt, for his love, patience, and

understanding throughout all my years in graduate school.

iii













TABLE OF CONTENTS

ACKNOWLEDGEMENTS......................................... iii

ABSTRACT ............................................... vii

CHAPTERS

1 INTRODUCTION....................................... 1

2 REVIEW OF THE LITERATURE .......................... 7

Eriksonian Theory of Psychosocial Development.... 7
Marcia's Identity Status Paradigm................. 9
The Relationship Between Identity Statuses
and Identity Styles............................. 13
Social-Cognitive Correlates of Identity Style.... 16
Berzonsky's Process View of Identity Formation... 18
Individual Differences in the Processing of
Self-Discrepant Information.................... 20
Identity Status and Integrative Complexity... 20
Identity Styles and Coping with Stress....... 22
Identity Styles and Autobiographical
Memories................................... .. 24
The Present Research.............................. 26

3 METHODS .......................................... 31

Design ........................................... 31
Participants..................................... .. 31
Procedure......................................... .. 33
Measures ......................................... 38
Identity Style................................. .. 38
Value Dimensions................................ 40
Egalitarianism.................................. 41
Dependent Measures.............................. 41

4 RESULTS .......................................... 44

Manipulation Check................................ 44
Processing of Self-Discrepant Information........ 45
Interjudge Agreement............................ 46
Recall of Evaluative Information................ 46
Score Recall................................... .. 47
Recall of Feedback-Consistent Personal
Experiences.................................. .. 48







Recall of Positively Egalitarian Personal
Experiences.................................. .. 51
Evaluation of Self-Discrepant Information.......... 52
Internal Attributions for Test Performance..... 53
External Attributions for Test Performance ..... 54
Perceptions of the Test............ ............. 55
Utilization of Self-Discrepant Information....... 56
Intrinsic Interest in Taking Another
Personality Test.............................. 56
Request for Additional Evaluative Information
on Egalitarianism............................. 57
Post-Feedback Ratings of Egalitarianism.......... 59

5 DISCUSSION....................................... .. 62

Processing, Evaluation, and Utilization of
Self-Discrepant Information .................... 62
Processing of Self-Discrepant Information.... 63
Evaluation of Self-Discrepant Information.... 66
Utilization of Self-Discrepant Information... 68
Implications of the Results for Berzonsky's
Process Model of Identity Development............ 73
Information-Oriented Style ................... 74
Diffuse/Avoidant Style ......................... 76
Normative Style............................... 78
Potential Applications of the Present Results.... 80
Stress Management............................... 81
Discrepancy and Therapeutic Change............... 82
Prejudice Reduction............................. 84
Limitations of the Present Research............... 87
Conclusions...................................... .. 90

APPENDICES

A IDENTITY STYLE INVENTORY .......................... 92

B VALUE DIMENSIONS.................................. .. 96

C HUMANITARIANISM-EGALITARIANISM SCALE................ 98

D COVER STORY........................................ 100

E INFORMED CONSENT FORM ...... ...................... 101

F DRAW-A-PERSON TEST ADMINISTRATION INSTRUCTIONS.... 103

G PREPARATION FOR INTERPRETATION OF TEST SCORES..... 104

H INTERPRETIVE FEEDBACK............................... 106

I INFORMATION RECALL................................. 113

J SCORE RECALL....................................... 114
v







K EXPERIENCE RECALL ................................. 115

L INTEREST FORM ..................................... 117

M PERCEPTIONS QUESTIONNAIRE ......................... 119

N DEBRIEFING ........................................ 125

REFERENCE LIST ......................................... 127

BIOGRAPHICAL SKETCH .................................... 135













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


THE ROLE OF IDENTITY STYLE IN RECONSTRUCTING THE SELF
FOLLOWING SELF-DISCREPANT INFORMATION

By

Margaret Allison Moore

August, 1994

Chairman: Greg J. Neimeyer
Major Department: Psychology

The present study examined whether individual

differences in identity style predispose individuals to

differentially process, evaluate, and utilize self-

discrepant information. Participants in this study were

identified as utilizing either an information-oriented, a

normative, or a diffuse/avoidant identity style and received

bogus personality feedback based on a projective test. One

half of the participants were given feedback that

disconfirmed an attribute that they had indicated was both

descriptive of them and important to their sense of self.

The other half of the participants were given confirming

feedback on this attribute. Participants then recalled the

confirming or disconfirming information as well as personal

experiences relevant to the information, rated their

perceptions of the test and their test performance,

vii







indicated their interest in obtaining additional evaluative

information, and provided post-feedback self-ratings on the

attribute. Results provided mixed support for an individual

differences model of identity development. On the one hand,

the findings contradicted the primary hypothesis that

individuals utilizing different identity styles would

respond differently to discrepant information about the

self. On the other hand, the results helped to elucidate

the differential means by which the informational and

diffuse/avoidant identity styles, in particular, direct the

interpretation and integration of confirming versus

disconfirming self-relevant information. The implications

of these results for a process conceptualization of identity

formation as well as the areas of stress management,

therapeutic change, and prejudice reduction are discussed.


viii













CHAPTER 1
INTRODUCTION

Erikson (1959, 1968) has written extensively about the

development of a coherent sense of self-identity as the

primary psychosocial task facing adolescents. According to

Erikson (1975), identity ushers in adulthood, as it bridges

the gap between the experiences of childhood and the

personal goals, values, and decisions that permit each young

person to take his or her place in society. This self-

structure is thought to evolve through the exploration of,

and subsequent commitment to, those sexual, moral,

political, religious, and vocational ideologies considered

to be personally effective for construing life events

(Erikson, 1959). Likewise, experts in the area of social

cognition have conceptualized identity as an evolving, self-

constructed theory about oneself, or self-relevant

configuration, that serves as a framework for answering

questions about the purpose and meaning of life (see

Berzonsky, 1989a, 1990; Epstein, 1973; Marcia, 1980;

Schlenker, 1986; Swann, 1987).

Most researchers have employed James Marcia's (1966)

identity status paradigm in their attempts to provide

empirical validation for Erikson's (1959) theory of

identity development. Acknowledging the premium that

1









Erikson (1959) placed on one's reported experience with

self-exploration and commitment, Marcia (1966) identified

four identity classifications marked by high or low levels

of exploration and commitment: Foreclosure (commitment

without current or past self-exploration); Moratorium

(ongoing self-exploration in the absence of commitment);

Diffusion (absence of both self-exploration and commitment);

and Achievement (past self-exploration indicated and

commitment present).

Marcia's (1966) operationalization of identity

formation along differential levels or statuses emphasized

the relatively stable and distinct developmental positions

attained by individuals in their search for identity. In a

different vein, Berzonsky (1990) recently elaborated a model

of identity development that highlights the process by which

identity-relevant information and experiences are evaluated

and utilized in the development and maintenance of a

coherent self-structure. He predicted that individuals

grouped according to Marcia's (1966) taxonomy would differ

in the social-cognitive processes they use to make decisions

and negotiate identity issues.

According to Berzonsky's (1988, 1989a) model,

individuals differ in the orientation they employ to

construct, maintain, and revise their self-identities. He

postulated three different identity processing styles: 1)

Information-oriented or "scientific," 2) Normative-oriented








or "dogmatic", and 3) Diffuse/avoidant-oriented or "ad hoc"

(Berzonsky, 1989a).

Individuals utilizing an information-oriented identity

style tend to play a more active role in forging their

identities (Berzonsky, 1990, 1991b). Their orientation

disposes them to seek out self-diagnostic information, test

self-constructs, and make necessary revisions in their self-

structure (Sorrentino, Raynor, Zubek, & Short, 1990). This

orientation is hypothesized to be the primary approach

utilized by self-examining individuals who would be

categorized as identity achieved or in a state of moratorium

according to Marcia's (1980) identity status criteria.

Individuals employing a normative style in defining

themselves tend to devote less deliberate effort to the

elaboration and testing of self-relevant information.

Rather, they internalize the normative expectations and

dictates of significant others (e.g., parents) and reference

groups (e.g., peers; Berzonsky, 1992b). Moreover, their

primary concern with minimizing potential threats to this

co-opted self-structure motivates them to conserve existing

self-views. This style is characteristic of Marcia's (1980)

foreclosed identity.

Individuals utilizing the diffuse/avoidant processing

orientation are typically more passive in the construction

of their identities, reporting less control over and

personal responsibility for the identity they possess







4

(Berzonsky, 1991a). They tend to function as "ad hoc" self-

theorists who characteristically avoid dealing with identity

questions by procrastinating as long as possible until they

must rely on contextual demands and situational consequences

to generate a response (Berzonsky, 1990, 1991a). This

externally oriented approach is hypothesized to be

associated with Marcia's (1980) diffuse identity, which

characterizes individuals who lack firm convictions and who

have not engaged in self-exploration.

In the tradition of constructivism (Kelly, 1955),

Berzonsky's (1990) process model of identity assumes that

individuals construct their own psychosocial development in

such a way that they become more effective at organizing

their experiences, predicting future events, and guiding

their behaviors. Moreover, the model assumes that

individuals will be involved in an ongoing renegotiation of

their identities across the life span as feedback from

experience is continuously evaluated, processed, and

utilized (Berzonsky, 1990).

Much of the information and experiences that

individuals encounter in the course of daily living will be

consistent with their self-views and therefore construed and

assimilated with minimal attentional effort (Langer, Blank,

& Chanowitz, 1978). However, careful monitoring and

evaluation may be necessary, as well as subsequent

accommodative efforts, when the information is incompatible









with the existing identity structure. Thus, information

that necessitates a revision of self-conceptions is likely

to be perceived and processed differently from information

that implicates conserving the integrity of the self.

Central to Berzonsky's (1990) individual differences

account is the notion that the differential processing and

evaluation of confirmatory and disconfirmatory information

should depend on stylistic differences in self-construction.

According to Berzonsky (1990), individual differences in

identity style should predict variability in the propensity

for engaging assimilative versus accommodative processes in

response to self-relevant information and experiences.

Swann (1987) has noted that people are more likely to

seek and rely on evidence that confirms rather than

disconfirms their self-beliefs, presumably because they

regard confirming information as more diagnostic, or at

least more stabilizing, than disconfirming information.

Swann and his colleagues (Swann, 1987; Swann & Read, 1981a,

1981b) have observed that this drive to confirm self-

conceptions is all the more apparent when individuals are

threatened by identity-discrepant information or feedback.

However, the work of Berzonsky (1990, 1992c) suggests that

this type of dogged adherence to prior self-beliefs may be

more prevalent among certain types of individuals, depending

on the ways in which they process identity-relevant

information.









Berzonsky (1990) has proposed that individuals should

differ, depending on their identity style, in the degree to

which they will conserve the self in the face of self-

discrepant feedback. For example, individuals who utilize

an information-oriented style may be more likely to consider

actively the implications of invalidating information and

therefore adjust their self-beliefs based on the dissonant

feedback. In contrast, normative-oriented individuals

should operate as dogmatic self-theorists, defending against

and distorting information and experiences that might

potentially invalidate existing self-views (Berzonsky, 1990,

1992c).

The present research provides a test of Berzonsky's

(1990) theory that variability in identity style should be

associated with differential responsiveness to information

that contradicts conceptions of the self. In addition to

validating Berzonsky's (1990) process model of identity

development, the present study has important implications

for identifying stylistic differences in the effectiveness

with which individuals negotiate issues relevant to the

purpose and meaning of life.













CHAPTER 2
REVIEW OF THE LITERATURE

This chapter will begin with a brief overview of

Erikson's (1959, 1968, 1975) psychosocial theory of identity

development, followed by two conceptualizations that

highlight individual differences in identity formation, with

a special focus on Berzonsky's (1990) process model of

identity. Evidence supporting the relationship between

structural aspects of identity and processing orientations

will be reviewed, and empirically validated personality and

social-cognitive correlates of identity status and identity

style will be considered. Next, literature regarding the

relationship between identity style and the differential

means by which self-relevant information is processed,

evaluated, and utilized will be reviewed. Finally, a method

will be proposed for examining the direct implications of

social-cognitive identity styles for the processing of

information that contradicts personal values and self-

beliefs. This method will be detailed in the subsequent

chapter where issues of design, instrumentation, and

procedures are covered.

Eriksonian Theory of Psychosocial Development

Erikson's (1959) theory of psychosocial development

describes the impact of social experiences on the evolution

7








of personality across the life span. For Erikson,

personality develops through the negotiation of various

psychosocial crises that are encountered across a sequence

of developmental stages. Each stage emphasises a particular

aspect of transaction with the social environment, and each

crisis is a struggle to attain the psychological quality

that is necessary in order to advance to the next stage.

The central focus of Erikson's (1959, 1968)

psychosocial theory of personality is the development of a

consciously experienced sense of self, or identity. During

the adolescent stage of development, individuals are faced

with the crisis between developing feelings of identity

versus role confusion. Identity derives from an integration

of private and social self-conceptions that result in a

sense of personal continuity or inner congruence that allows

individuals to truly appreciate what their lives as adults

are to be about (Carver & Scheier, 1992). If the individual

fails to form an integrated identity, the result is role-

confusion, a self made up of multiple and conflicting

attributes that fails to provide the individual with a sense

of direction. A vast amount of research has emerged to test

Erikson's (1968, 1975) conceptualization of identity

development, with the majority of this work following

Marcia's (1966, 1980) identity status paradigm.









Marcia's Identity Status Paradigm

Marcia (1966) employed Erikson's (1959) psychosocial

concepts of self-exploratory crisis and commitment to

formulate his identity status paradigm. From Erikson's

(1959) perspective the process of self-construction consists

of an exploration of occupational, ideological, and

interpersonal choices (i.e., crisis) and an eventual

commitment to a system of beliefs and values. Marcia (1966)

utilized a structured interview format to assess the

presence or absence of self-reported crisis and commitment

within occupational, political, and ideological domains. In

operational terms, crisis refers to a self-perceived active

period of self-reflection and evaluation of identity-

relevant information. Commitment indicates the extent to

which an individual makes a firm investment in what he or

she values and believes and pursues goals consistent with

this resolution. The assessment of crisis and commitment

served to designate four identity statuses: Achievement

(crisis experienced and commitment made), Moratorium (crisis

experienced and commitment not made), Foreclosure (crisis

not experienced and commitment made), and Diffusion (crisis

not experienced and commitment not made). The existence of

these four identity statuses has been supported empirically

(see Marcia, 1980 for a review), and recent research (e.g.,

Marcia, 1980; Waterman, 1982) has been directed at examining








the social-cognitive and personality correlates of these

identity configurations.

Identity achievers are described as having experienced

a period of self-exploration, or crisis, that has resulted

in commitments to particular goals, values, and beliefs.

The moratorium status applies to individuals who are

currently exploring identity issues but have not yet made

commitments. In contemporary research, achievers and

moratoriums are often considered to represent the more

advanced or sophisticated identity statuses because of their

self-reported experiences of self-exploratory crisis (see

Josselson, 1973; Marcia, 1980; Read, Adams, & Dobson, 1984).

Slugoski, Marcia, and Koopman (1984) found that both

moratorium individuals and identity achievers demonstrated

significantly greater integrative complexity when dealing

with interpersonal problems than did individuals who were

foreclosed and diffuse. Integratively complex individuals

use more complex social-cognitive reasoning and therefore

are able to consider multiple and conflicting elements of

interpersonal feedback while maintaining a self-determined

perspective. Integrative complexity was identified by the

authors as a primary cognitive factor in identity

development because it derives out of interaction and

feedback from the environment and is presumed to influence

the degree to which individuals actively consider

alternative ideas and integrate their experiences.









Achieved and moratorium individuals have also

demonstrated the ability to process more extensive amounts

of interpersonal information than have those in the

foreclosed and diffuse statuses, while making fewer mistakes

(Read et al., 1984). In addition, self-exploratory

individuals (i.e., achieved and moratorium) tend to report a

higher sense of self-worth, reject authoritarian views, have

a strong sense of ethnic identification, be low in

prejudice, and endorse cultural pluralism (i.e., recognize

the unique contributions of multiethnic groups to the larger

society; Marcia, 1967, 1980).

Foreclosure is the status of individuals who have

committed themselves to particular goals, beliefs, and

values without considering implications or exploring

alternatives. Correlational findings indicate that identity

foreclosure is positively associated with a socially defined

identity, public self-consciousness, and other-directed

self-monitoring (Berzonsky, Trudeau, & Brenna, 1988), as

well as other-directed approaches to problem solving

(Grotevant & Adams, 1984). Read et al. (1984) found that

foreclosed individuals were least able to evaluate and

integrate multiple and conflicting sources of interpersonal

information and tended to exclude relevant information from

examination. The authors attributed these findings to a

less adaptive orientation toward interpersonal problems and

a relatively restricted attentional focus. In other









studies, foreclosures have been found to endorse

authoritarian views (Marcia, 1980), express an intolerance

of ambiguity (Schenkel & Marcia, 1972), and possess rigid,

change-resistant self-systems (Berzonsky & Neimeyer, 1988).

Interestingly, individuals who have prematurely foreclosed

their search for self-definition share a strong sense of

ethnic identification with achieved and moratorium statuses;

however, unlike these self-exploratory individuals,

foreclosures tend to be relatively high in prejudice

(Marcia, 1980), suggesting that they may maintain their

collective self-esteem by creating ingroup versus outgroup

comparisons.

In the fourth status, identity diffusion, individuals

have yet to engage in exploration of possible identities and

have yet to make commitments. Findings consistent with the

view that diffuse individuals tend to operate in an avoidant

and situation-specific, self-presentational manner have been

observed. For example, diffusions have reported a tendency

to avoid confronting personal problems and, like

foreclosures, tend to rely on other-directed problem-solving

strategies (Grotevant & Adams, 1984). Further, Berzonsky et

al. (1988) observed that identity diffuseness was inversely

correlated with a personally defined identity and private

self-consciousness and positively correlated with other-

directed self-monitoring and a tendency to act in a manner

consistent with current situational demands. Moreover,









Berzonsky, Schlenker, & McKillop (1987) provided empirical

support for the notion that diffusions are guided by self-

presentational concerns; diffusions changed their views of

themselves when role-playing in an actual interaction with a

confederate, but not in conditions that involved a written-

interview or anonymous presentation.

The Relationship Between Identity Statuses
and Identity Styles

In broad terms, the distinct characteristics that

typify individuals who have experienced an identity crisis

(i.e., identity achieved and moratorium statuses) suggest

that these individuals are actively self-reflective and tend

to approach identity decisions by both seeking new

information and considering others' opinions (Berzonsky,

1989b; Marcia, 1980). Compared with these self-reflective

statuses, evidence suggests that foreclosures have

inflexible self-systems, are likely to simplify complex

issues, and tend to be overly reliant on the normative

expectations of others. Further, findings serve to confirm

assumptions that identity diffused individuals refer to

social expectations and environmental circumstances when

deciding what to believe and how to act in any given

situation.

After reviewing the social-cognitive aspects of

identity status, Berzonsky (1989b, 1990) proposed that the

four identity configurations classified by Marcia's (1966)

paradigm may indicate different styles of structuring self-







14

relevant information into a sense of identity. He reasoned

that individuals classified as moratoriums and achievers

characteristically take an information-oriented approach to

forging their identity; they actively seek out and evaluate

self-relevant information before making decisions relevant

to their self-identity. Foreclosed individuals, on the

other hand, are likely to exercise more normative-oriented

strategies in self-construction; they internalize the norms

set for them by authority figures and significant others and

close themselves off from self-relevant information that

might contradict these normative prescriptions. Diffuse

individuals tend to enact an avoidant orientation or self-

presentational stance when dealing with identity-relevant

issues and events; they avoid confronting problems and

issues until they must refer to others or expected social

consequences in making self-relevant decisions.

In order to examine the relationship between identity

statuses and identity styles, Berzonsky (1989b) developed a

self-report identity style measure by separating the

commitment and self-exploration components that are combined

in objective measures of ego identity status (e.g.,

Grotevant & Adams, 1984). Correlational research using the

Identity Style Inventory (Berzonsky, 1989b) has supported

the notion of stylistic differences in the manner in which

individuals classified in terms of Marcia's (1966) identity









statuses make decisions, solve personal problems, and

process self-relevant information.

In the first study examining the relationship between

identity statuses and identity styles, Berzonsky (1989b)

administered the identity style measure and the Adams-

Grotevant measure of identity status (Grotevant & Adams,

1984) to late adolescent subjects. Findings revealed that

status and style were correlated in a theoretically

meaningful manner, indicating that distinct social-cognitive

processing orientations corresponded with the identity

statuses. For example, foreclosures had the highest

normative style scores and diffusions were highest on the

diffuse/avoidant scale. The pattern of findings for the

moratorium and achieved statuses was more complex. The

relationship between these self-exploring statuses and an

information-orientation appeared to be moderated by identity

commitment such that making firm commitments suppressed the

use of an information-oriented identity style (Berzonsky,

1990, 1992b).

Overall, these findings suggested that identity style

could not be regarded as independent from identity structure

in attempts to elaborate a valid and comprehensive

understanding of identity formation (Berzonsky, 1992a).

Berzonsky (1992a) proposed a reciprocal relationship between

processing orientation and identity structure whereby the

style one uses influences the identity that one has, which









in turn, determines the process that is utilized. This

conceptualization highlights the dynamic and evolving nature

of the identity structure.

Social-Cognitive Correlates of Identity Style

Assuming that differences in identity processing style

represent different approaches to self-construction,

Berzonsky and Sullivan (1992; see also Berzonsky, 1993a)

patterned their predictions about the social-cognitive

aspects of identity style after the observed

interrelationships between social-cognitive dispositions and

identity status. As expected, findings revealed that an

information-oriented style correlated significantly with a

number of information-seeking and self-reflective variables.

Specifically, the authors observed a relationship between an

information-orientation and openness to experience (Costa &

McCrae, 1978), a construct which depicts receptivity to new

ideas, a willingness to consider alternate values, and an

awareness of personal attitudes and feelings. In addition,

they found a significant relationship between an

informational style and need for cognition (Cacioppo &

Petty, 1982), a concept referring to the extent to which

individuals appreciate and are motivated to engage in active

information processing as well as in cognitive activities

involving alternative ideas. Information-oriented

individuals were also characterized by introspective







17

tendencies (Hansell, Mechanic, & Brondolo, 1986) and a focus

on their private self (Cheek & Briggs, 1982).

Berzonsky and Sullivan (1992) also observed that

individuals employing a normative identity style were

relatively rigid and narrowly focused; the normative

orientation was negatively associated with experiential

openness variables deemed to represent "core" areas of the

self, including values, actions, and fantasies.

Subsequently, they reasoned that these individuals might

defensively close themselves off from experiences and

information that might threaten or invalidate important

self-views (Berzonsky & Sullivan, 1992). Elsewhere,

normative scores on identity style measures were found to be

positively associated with a rigidly organized self-

structure (Berzonsky & Neimeyer, 1988), a greater fear of

negative evaluation (see Berzonsky, 1990), and a

collectively defined identity (Berzonsky, in press).

Additionally, individuals who reported adhering to this

protectionistic and other-defined stance tended to endorse

authoritarian values (Adorno, Frenkel-Brunswik, Levinson, &

Sanford, 1950) and thus can be distinguished by their social

rigidity, intolerance of ambiguity, and lack of adaptability

(Berzonsky & Sullivan, 1992).

Use of a diffuse/avoidant style was positively

correlated with a self-rated social identity emphasis as

measured by the Cheek and Briggs' (1982) Aspects of Identity







18
Scale, lending support to the view that a diffuse processing

orientation leads to an externally controlled approach to

solving problems and negotiating identity issues

(Berzonsky, 1991b) as well as greater reactivity to social

evaluation (Berzonsky, in press). Other research found

significant relationships between a diffuse/avoidant style

and procrastination tendencies as well as other-directedness

(see Berzonsky, 1990). Moreover, individuals with a

diffuse/avoidant identity style scored lowest of all

individuals on such information-seeking variables as

introspectiveness, need for cognition, and experiential

openness (Berzonsky & Sullivan, 1992). These findings are

consistent with the view of diffuse/avoidant individuals as

both avoidant and motivated by self-presentational concerns

to create the appropriate impression for immediate social

approval, even if that means sacrificing long-term

adaptation.

Berzonsky's Process View of Identity Formation

Based on the contructivist notion that a self-theory's

effectiveness depends on its utility for anticipating and

controlling life events (Kelly, 1955), Berzonsky's (1990)

process conceptualization of identity suggests that optimal

psychosocial adaptation will occur when individuals are

effectively responsive to the self-relevant experiences they

encounter in the course of daily life. In particular,

maximally effective responses to identity-relevant






19

information should be associated with a balanced utilization

of assimilative and accommodative processes. Perhaps most

central to Berzonsky's (see Berzonsky, 1988; Berzonsky &

Neimeyer, 1988) model is the hypothesis that individual

differences in identity style should determine the extent to

which individuals exercise this "dialectical interchange"

when processing self-relevant information, especially

information that disconfirms self-conceptions.

According to Berzonsky (1990), individuals who use both

assimilative and accommodative processing should make

dissonance-induced efforts to revise relevant aspects of the

self when confronted with self-discrepant information. In

contrast, individuals who characteristically favor

assimilation over accommodation in their processing of self-

relevant experiences should conserve self-views despite

invalidation. This individual differences account of the

identity negotiation process suggests interesting

implications for Swann's (1987) self-verification

formulation. Specifically, Berzonsky (1990) proposed that

the tendency to engage in systematically biased processing

in order to confirm self-conceptions may be moderated by the

characteristic style that one utilizes to construct his or

her identity. That is, differential reliance on self-

verification strategies (Swann, 1987) and self-serving

attributions (Ross, Lepper & Hubbard, 1975) when confronted








with identity-discrepant feedback should be a function of

identity style.

Individual Differences in the Processing of
Self-Discrepant Information

Research identifying the social-cognitive aspects of

identity statuses and identity styles has enabled

researchers to speculate about individual variability in the

processing of self-discrepant information. However, no

research has directly tested the impact of identity status

or identity style on the evaluation and utilization of

dissonant self-relevant information. The most relevant

studies include those that have examined the relationship

between identity status and integrative complexity in social

interaction (Slugoski et al., 1984) and between identity

style and stress management (Berzonsky 1989a, 1992a, 1992b).

Perhaps the most applicable study is one that has examined

the impact of self-consistent and self-discrepant

autobiographical recollections on the self-conceptions of

individuals who utilize different identity styles.

Identity Status and Integrative Complexity

A study conducted a decade ago by Slugoski et al.

(1984) examined the cognitive and social characteristics of

identity status. The authors used Marcia's (1966) Identity

Status Interview to classify male college students according

to identity status and had them complete the Paragraph

Completion Test (PCT; Schroder, Driver, & Streufert, 1967),

an instrument measuring integrative complexity relevant to








interpersonal interactions. They also categorized their

interactive behavior using Bales (1951) Interaction Process

Analysis system in a group situation involving moral

problem-solving tasks.

Results revealed that differential personal and

cognitive strategies for dealing with interpersonal feedback

and its sources reflected differences in identity status.

Foreclosures and diffusions showed impulsive decision styles

and the need for immediate closure, indicating that these

individuals defended against potential dissonance by

maintaining constrictive cognitive systems. These low-

identity statuses appeared to disregard or outright reject

the feelings and opinions of others and exposed obvious

signs of tension in the interactions. Moreover, the

foreclosed structure, in particular, appeared to be

supported by fairly distinctive interactional patterns.

Foreclosed individuals tended to respond to potentially

diconfirming information with either aggressive tactics such

as interruptions, sarcasm, and condescension or more

acquiescent techniques such as discounting and superficial

agreement. Subjects even resorted to deflating the status

of the source of information. Both of these types of

responses, antagonistic and acquiescent, achieved the larger

goal of closing off dissonant information that might induce

the need for self-modification. The more advanced statuses,

achieved and moratorium, produced PCT scores that indicated








cognitive flexibility and integrative complexity.

Observations of their group behaviors revealed that they

freely probed the opinions of other group members,

demonstrated higher levels of empathy and support, and

appeared comfortable discussing controversial issues.

Identity Styles and Coping with Stress

Berzonsky (1989a, 1992a, 1992b) recently extended his

process model of identity development to explain

dispositional differences in how individuals evaluate and

attempt to cope with stressors. In his series of

correlational studies, he conceptualized stressors as crises

or anxiety-arousing situations that have the potential for

invalidating or forcing revisions in individuals' self-

views. He proposed that individual variability in identity

style should be associated with differential utilization of

problem-focused and emotion-focused coping strategies

(Folkman & Lazarus, 1980, 1985; Lazarus, 1966).

Specifically, he predicted that information-oriented

individuals would perceive academic stress as manageable and

engage in more problem-focused coping efforts such as

seeking pertinent information, generating alternative

solutions, and identifying potential stress-inducing factors

in the environment and making the appropriate changes. In

contrast, diffuse/avoidant and normative-oriented

individuals were expected to engage in more emotion-focused

coping, characterized by the utilization of defensive








tactics such as detachment, wishful thinking, and

procrastination, all of which are aimed at dismissing the

cause of the stress and restricting the experience of

psychological tension (Folkman & Lazarus, 1980, 1985).

In a recent test of the hypothesized relationship

between processing orientation and coping strategies

(Berzonsky, 1992b), college students were administered the

Identity Style Inventory (Berzonsky, 1989b), the Ways of

Coping checklist (Folkman & Lazarus, 1985), and the Alpert

and Haber (1960) measure of facilitative and debilitative

test anxiety. In support of previous studies (Berzonsky,

1989a, 1992a), reported use of an information-oriented style

was positively related to problem-focused and social-

support-seeking coping efforts, as well as facilitative

anxiety responses. Information-oriented individuals

appeared to perceive stressors as manageable and actively

considered information that might facilitate resolutions of

problems.

Both diffuse/avoidant and normative styles, in

contrast, were positively associated with emotion-focused

coping strategies. By denying the problem or considering

only fanciful solutions, these individuals dodged the

immediate tension, but prevented long-term resolution by not

confronting the problem directly. Additionally, normative-

oriented individuals reported significantly more

debilitative test anxiety than the other statuses. These








findings suggested that diffuse/avoidant and normative-

oriented individuals preferred to ward off problems that had

the potential to threaten their self-constructions.

Identity Styles and Autobiographical Memories

In a recent study, Neimeyer and Metzler (1992) examined

identity style variability in the recall of self-discrepant

autobiographical memories and the differential impact of

these recollections on subsequent self-perceptions. They

hypothesized that because identity styles presumably

represent different approaches to the processing of self-

relevant information these differences should be reflected

in the relative ability or inability to generate personal

memories that were discrepant from self-conceptions

(Neimeyer & Rareshide, 1991). They also proposed that the

impact of autobiographical recollections that confirm or

disconfirm positive or negative self-perceptions should vary

directly with identity style.

Specifically, Neimeyer and Metzler (1992) expected the

memory recall of information-oriented individuals to

represent a balanced array of experiences that were

consistent and inconsistent with their self-beliefs.

Furthermore, individuals with an information-oriented style

were expected to be relatively more likely to revise their

self-images to accommodate image-discrepant recollections.

In sharp contrast, it was predicted that normative styles

would recall mostly those events that were consonant with








their self-views and summon few, if any, discordant

memories. Moreover, it was anticipated that they would

respond very differently to identity-discrepant

recollections, dismissing their informational value in an

effort to protect existing self-constructions. The authors

recognized that diffuse-oriented individuals' lack of

identity exploration and absense of commitment made it more

difficult to predict the nature (i.e., confirmatory versus

disconfirmatory) of the memories that they would recall.

Presumably, they would generate both types of memories and

base the alteration of their self-constructions on the

immediate demands of the situation or the perceived social

rewards.

Findings revealed that the nature (i.e., self-

consistent versus inconsistent) of autobiographical memories

recalled and the impact of that recall on subsequent self-

perceptions varied with identity processing orientation as

expected. While information-oriented individuals generated

significantly more confirmatory than disconfirmatory

memories, they still managed to generate more than twice the

number of self-discrepant memories than did normative

oriented subjects. Normative-oriented individuals not only

generated the least amount of invalidating memories but they

also showed the least change, again emphasizing their

preference for assimilation over accommodation. Consistent

with their presumed lack of self-definition and tendency to









conform to social dictates, diffuse individuals generated

the greatest balance in the number of consistent and

inconsistent memories recalled and showed the most extreme

changes in self-perceptions following recall of image-

inconsistent memories.

The Present Research

The research reviewed in this paper thus far concerns

the relationships between identity style or identity

structure and the social-cognitive dimensions relative to

evaluating self-discrepant information in personal and

interpersonal domains, different modes of coping with

identity-threatening stressors, and the recollection and

impact of self-discrepant memories on self-evaluations.

Taken together, these studies provide partial support for

Berzonsky's (1990, 1992c) theory that variability in

identity style should reflect differences in the processing

and evaluation of self-discrepant feedback, as well as the

degree to which such information should prompt a revision of

the self-structure.

The present research was designed to extend these

recent efforts by using an experimental paradigm to test

directly whether individuals with different identity

processing orientations are differentially responsive to

self-discrepant information. The present study involved

selecting individuals representing each of the three

identity styles who perceived themselves as highly









egalitarian and then systematically disconfirming these

self-perceptions with ego-discrepant feedback and measuring

the ways in which they processed, evaluated, and utilized

that information (see Methods for details). These

considerations regarding the relationship between identity

orientations and responses to self-discrepant information

gave rise to the following hypotheses:

1. Individuals with an informational style should be

relatively more willing than individuals utilizing the

normative or diffuse identity styles to seek, evaluate, and

utilize information that disconfirms their egalitarian self-

views. They should demonstrate this by achieving a greater

balance, relative to normative and diffuse/avoidant types,

in their recall of self-consistent and self-inconsistent

interpretive feedback as well as confirmatory and

disconfirmatory personal experiences. In addition,

informational types should show less bias in their recall of

percentile scores provided to them in the disconfirming

feedback condition. They also should demonstrate a

relatively greater intrinsic interest in volunteering to

take an additional personality test, and they should show a

greater preference for receiving further feedback pertaining

to egalitarianism after receiving disconfirming feedback on

this attribute. Finally, these individuals should perceive

the test as more credible and make fewer external

attributions and more internal attributions for their test







28

performance than either the normative or diffuse individuals

when they receive disconfirming feedback.

2. Individuals with a normative-oriented style should

devote less deliberate effort to the elaboration and testing

of self-invalidating information. They are expected to

recall a greater proportion of identity-consistent (versus

inconsistent) evaluative information and personal

experiences, as well as make more self-serving errors in

their recall of percentile score ratings than informational

individuals. Further, they should express less intrinsic

interest in taking an additional personality test after

receiving disconfirming feedback. Normative individuals

should also be less likely than informational types to

request additional information on the disconfirmed attribute

after receiving feedback that invalidates their self-

beliefs. Finally, normative types should show a greater

tendency than informational individuals to discredit the

validity of the test and seek out self-confirmatory

attributions for their test performance after receiving

disconfirming feedback.

3. Individuals with a diffuse identity style are

expected to be especially vigilant to social expectations

and external contingencies. Given that the attribute on

which they receive negative feedback, egalitarianism, is a

strong social norm, these individuals should bias their

attention away from recalling image-discrepant (i.e., anti-









egalitarian) information. They should be more likely than

normative or informational types to recall interpretive

feedback and personal experiences that are consistent with

their egalitarian self-concepts. Moreover, diffuse

individuals should be less intrinsically motivated, relative

to the other individuals, to take an additional personality

test and least likely to request additional feedback on

egalitarianism because such information might further

jeopardize the good impression that they have been trying to

manage. In addition, these individuals should be just as

likely as the normative types to question the validity of

the test and make external attributions for their test

performance after receiving identity-discrepant feedback,

although the reasons for doing so might be different (i.e.,

they will be motivated to reinforce existing self-beliefs in

order to gain social approval, whereas normative types will

be motivated to avoid the internal dissonance associated

with negatively discrepant feedback).

While shedding light on our understanding of individual

variability in responses to self-relevant information,

previous evaluations of Berzonsky's (1990) process model

have been compromised by the exclusive use of correlational

methodologies. The experimental design of the present study

allows for a more thorough and controlled analysis of the

tenets upon which Berzonsky's (1990) model is based and

therefore affords more conclusive interpretations.









In addition, the present study enables some insight

into the potential applications of Berzonsky's (1990)

social-cognitive model of identity development. In

particular, applications of these results to the areas of

prejudice reduction, stress management, and therapeutic

change are highlighted. In all these cases, individuals are

in the position of encountering information or experiences

that are discrepant from their current constructions of

reality and that necessitate an alteration of self-beliefs

in order to achieve optimal adaptation.













CHAPTER 3
METHODS

Design

The design of this study was a 3 (identity style:

informational, normative, or diffuse/avoidant) x 2

(feedback: confirmatory or disconfirmatory feedback) x 2

(gender: male or female) between subjects factorial.'

Participants

Berzonsky's (1989b) revised version of the Identity

Style Inventory (ISI2) was administered as a questionnaire

on "personal similarities" to a large sample (N = 758) of

undergraduate students who had volunteered in exchange for

experimental credit in their introductory psychology

classes. This instrument yielded scores for each of the

three identity styles (informational, normative,





1 Berzonsky (1993a) has suggested that the oft-cited
differential distribution of men and women among the identity
style classifications may be due to different socialization
patterns that may affect how individuals process self-relevant
experiences. Although a recent correlational study has
indicated that men and women with the same identity style do
not differ along social-cognitive reasoning variables (e.g.,
need for cognition), gender was included in the present
experimental design in order to evaluate whether men and women
utilizing the same identity style differed in their responses
to self-relevant information. Results revealed that gender
did not interact with identity style on any of the dependent
measures. Therefore, it will not be discussed further.

31









diffuse/avoidant) and a separate index of identity

commitment (see Measures, below, and Appendix A).

Participants also completed a set of 5-point scales

(0-4) that allowed them to rate themselves on three personal

value dimensions: Honesty, Responsibility, and

Egalitarianism. In addition, participants used a 5-point

scale to rate the degree to which they considered each of

these dimensions as important to their sense of self (see

Measures, below, and Appendix B). A more comprehensive

measure of egalitarianism, the Humanitarianism-

Egalitarianism Scale (Katz & Hass, 1988), was also

administered during pre-testing (see Measures, below, and

Appendix C).

In order to be selected for participation in this

experiment, individuals had to meet the following criteria.

First, they had to score above the median on the

Humanitarianism-Egalitarianism Scale (median = 29.29).

Second, they had to score above the midpoint (i.e., ratings

of 3 or 4) on the one-item measures of self-perceived

egalitarianism, responsibility, and honesty. Third, they

had to rate egalitarianism, responsibility, and honesty as

highly important to their sense of self (i.e., ratings of

3 or 4). Those individuals who did not meet these

requirements were excluded from participation. A total of

409 individuals from the larger sample were eligible (i.e.,

met the above criteria) for participation in this study.









When contacted by phone, 125 individuals (53 men and 72

women) elected to participate. This final sample (mean age

= 19.5) consisted of 80% European Americans, 7% Asian

Americans, 6% African Americans, 5% Hispanic Americans, and

2% Other. Of these individuals, 42 (22 men and 20 women)

were identified by the ISI as informational, 41 (15 men and

26 women) as normative, and 42 (16 men and 26 women) as

diffuse/avoidant.

Procedure

Each of the 125 participants was scheduled for an

individual session. A cover story was used in order to

enhance the credibility of the procedures and the feedback

manipulation (see Appendix D). When the participants

arrived at the experimental room they were told by a

confederate that their session had been purposefully

scheduled at the same time that a three-hour graduate

seminar on psychological assessment for doctoral psychology

students was meeting in another room. They were told that

the class had spent the major part of the semester being

trained in the administration, scoring, and interpretation

of projective personality inventories, particularly the

Rorschach, TAT, and DAP tests, and that they were soliciting

the help of undergraduate students to practice their

assessment skills under the direct supervision of their

professor, a licensed clinical psychologist. The

confederate then self-identified as a graduate student from









the assessment class whose role was to guide participants

through the administration, scoring, and interpretation

phases of a projective personality test. After this

introduction, participants provided their informed consent

for participation in this project (see Appendix E). The

participants were assured on the consent form that their

responses would be kept confidential by having a number

substituted for their name.

Participants were then given verbal assurance of

confidentiality and told that they would be taking the Draw-

A-Person test. Individuals were then administered an

adapted version of the Draw-A-Person picture test

(Goodenough, 1926; see Appendix F for instructions).

Participants were given five minutes to complete the first

picture, and while they were completing a second picture,

the confederate went to another room, designated the

assessment classroom, and (unbeknowngst to the participant)

turned on a tape recording of several people discussing

scoring-relevant issues before returning to the test

administration room. When the participant finished the

test, the confederate stated that they would take the

completed test to the assessment classroom together. The

participant was seated in the chair outside the classroom

door while the confederate went into the classroom to

ostensibly conduct the scoring with the rest of the class

under their professor's supervision. The tape recording was









playing so that it could be overheard by the participant.

After five minutes of discussion, a second tape recording

played the sounds of a printer, ostensibly generating an

analysis of the participant's drawings, and the confederate

emerged from the room with printouts in hand.

The confederate and the participant then returned to

the testing room for the test interpretation. At this time

participants were told that an important part of any

assessment procedure entailed the interpretation of scores

to the participant. They were then prepared for the

interpretation by reading along as the confederate spoke

aloud (see Appendix G). As part of this preparation, the

Draw-A-Person test was described as highly valid in order to

enhance the believability of the feedback that they were

about to receive.

The computer-printed interpretive analysis consisted of

both a narrative description of the participant's

performance and a line graph depicting actual percentile

scores (see Appendix H). The analysis had been prepared in

advance to communicate feedback that disconfirmed one-half

of the participants' self-perceptions as egalitarian while

confirming their self-perceptions as honest and responsible.

Individuals in these conditions were told that they scored

at the 28th percentile on egalitarianism and at the 90th and

89th percentile for the attributes of responsibility and

honesty, respectively.









The other half of the participants served as the

control conditions. Instead of getting disconfirmatory

feedback on the egalitarian construct, they received

uniformly consonant (i.e., confirmatory) feedback on all

three value dimensions (i.e., honesty, responsibility,

egalitarianism) that had been assessed as highly important

and descriptive of them earlier in the semester.

Individuals in these conditions were told that they scored

at the 88th percentile on egalitarianism and at the 90th and

89th percentile for the attributes of responsibility and

honesty, respectively. In all conditions, the attributes on

which individuals received feedback were counterbalanced.

Following this manipulation of feedback

(disconfirmation or confirmation), participants were told

that it was always necessary for a psychologist working with

an actual client to check with the client to see how the

interpretation was understood. They were asked to recall as

much as they could about the information they received

during the interpretation and to record what they remembered

on a sheet similar to a standard thought listing form

(Cacioppo & Petty, 1982) (see Appendix I). In addition,

they were asked to recall the percentile scores

(i.e., 1-100%) that they had been given on each of the three

dimensions (see Appendix J).

Participants were then told that it is not only

important for a psychologist to get a feel for what a client









retains from the feedback interpretation, but that it is

important to give the client a chance to communicate his or

her perceptions of the accuracy or inaccuracy of the

assessment analysis. The confederate stated that clinicians

often ask their clients to think about the feedback and how

it fits with their life experiences. The confederate then

acknowledged out loud that everyone at some time or another,

because of human nature, acts out the more favorable and

less favorable aspects of their personality. Participants

were asked to think about the interpretive feedback they

received and whether or not it seemed to correspond with

their thoughts and behaviors in actual life events. They

were then instructed to recall experiences that either fit

or did not fit with the interpretive feedback and to record

these experiences using a sheet similar to a standard

thought-listing form (Cacioppo & Petty, 1982, see Appendix

K).

Participants were then asked to complete a short set of

questions designed to assess their intrinsic and extrinsic

interest in taking another assessment test. Participants

were also asked to indicate which aspects of themselves

might they be interested in receiving feedback about if they

were to participate in another testing session. The

egalitarian-nonegalitarian dimension was couched in a list

of six dimensions (see Appendix L).









Finally, participants were asked to complete a

questionnaire consisting of 17 questions pertaining to the

perceived validity and reliability of the personality test,

perceived expertise of the interpreter and accuracy of the

interpretation, and perceptions of their test performance.

Three additional items (items 9, 10, 11) assessed post-

feedback self-evaluations. Three other items (items 21, 22,

23) were included to test the effectiveness of the feedback

manipulation (see Appendix M).

Participants were then carefully and thoroughly

debriefed. In addition to explaining the nature and purpose

of the experiment, the experimenter emphasized that the

"graduate student in the assessment class" was actually not

in such a class and had not evaluated the personality

measures. They were told that the scores and interpretation

they received were prepared before they took the test and in

no way reflected their actual performance on the inventory

(see Appendix N). When it was clear that the participants

were in no way adversely affected by the manipulation, they

were thanked for their participation and dismissed.

Measures

Identity Style

The revised version of the Identity Style Inventory

(ISI2; Berzonsky, 1989b) is a self-report measure used to

assess identity style. The inventory was developed by

uncoupling commitment and self-exploration components









contained in statements about identity status (see Adams,

Bennion, & Huh, 1989). The ISI2 contains 39 statements

which the participants rated on a Likert-type scale of 1 =

not at all like me to 5 = very much like me and yields a

score for each of the following: information-oriented style

("I've spent a good deal of time reading and talking to

others about religious issues"), normative style ("I've

always had purpose in my life; I was brought up to know what

to strive for"), diffuse/avoidant style ("Many times by not

concerning myself with personal problems, they work

themselves out") and commitment ("I know what I want to do

with my future").

Reported test-retest reliabilities for the individual

scales of the ISI2, over a two-month interval, varied from

.71 (diffuse) to .77 (commitment; Berzonsky 1989b). Alpha

coefficients were .67 for information, .66 for normative,

.78 for diffuse/avoidant, and .78 for commitment. Studies

that have focused on the psychometric properties of the

scales (e.g., Berzonsky 1989b, 1990, 1993a; Berzonsky &

Sullivan, 1992) cite evidence for the construct validity of

the ISI2 subscales. For example, scores on the

informational scale are correlated positively with openness

to experience and need for cognition. Scores on the

normative scale are positively associated with a rigidly

organized self-structure and negatively associated with

openness to experience. Scores on the diffuse/avoidant









scale are positively correlated with a self-rated social

identity emphasis and negatively associated with

introspectiveness. Evidence for criterion validity is

provided by Neimeyer and Metzler (1992) who showed that

identity processing orientation can predict differences in

how individuals respond to identity-discrepant

autobiographical memories.

In an attempt to classify the present sample according

to identity processing orientation, the procedure

recommended by Berzonsky (1992a) was employed. First, raw

scores for the 409 individuals who completed the three

identity style scales (and met the criteria for inclusion in

the study) were transformed into standardized z scores, with

means of 0 and standard deviations of 1. Then, an

individual's highest score on the three style subscales was

used to designate his or her identity style. This procedure

made it possible to categorize all potential participants

into one of the three identity processing orientations.

Value Dimensions

Participants' self-perceived ratings on three value

dimensions, honesty, responsibility, and egalitarianism, as

well as perceived self-importance of these dimensions were

measured using a six item Likert-type 5-point scale. Brief

descriptions of each of these dimensions were provided, and

after each description, participants were asked to rate the

degree to which they perceived the description as







41

characteristic of themselves and to rate the degree to which

this characteristic was important to their sense of self.

Honesty was described as being sincere, truthful, and

trustworthy. Responsibility was described as being

dependable, reliable, and conscientious. Egalitarian was

described as adhering to democratic ideals of equality,

social justice, and concern for others' well-being. This

scale was constructed for the present study. Therefore,

there is no existing validity or reliability information.

Egalitarianism

The Humanitarianism-Egalitarianism Scale (HE; Katz &

Hass, 1988) is a 10-item scale designed to assess an

egalitarian value orientation. Participants rated on a

Likert-type scale of 0 = strongly disagree to 4 = strongly

agree on items such as "Everyone should have an equal chance

and an equal say in most things" and "In dealing with

criminals, the courts should recognize that many are victims

of circumstances." Coefficient alpha for the 10 items was

.84, indicating that the instrument has adequate internal

consistency. Test-retest reliability information as well as

evidence for convergent and discriminant validity was not

available.

Dependent Measures

Participants were asked to recall the evaluative

feedback that they had received as well as personal

experiences relevant to that feedback. Two independent








raters coded each piece of information and each experience

as being relevant to either the egalitarianism,

responsibility, or honesty dimensions. Raters also judged

whether the information or experience recalled was

consistent or inconsistent with the feedback. In cases

where a disagreement existed, a third independent rater was

used to evaluate the information and experiences under

question and a majority (i.e., 2/3) vote was used. An

inspection of the data revealed that individuals rarely

recalled evaluative information that was inconsistent with

the actual feedback (i.e., the recall of self-consistent

information in the disconfirming feedback conditions or

self-discrepant information in the confirming feedback

conditions). Thus, the quantity of feedback-consistent

information recalled relevant to the egalitarian attribute

served as a dependent measure.

Regarding the recall of personal experiences,

individuals recalled virtually no identity-discrepant

experiences in the confirming feedback conditions. However,

identity consistent experiences were recalled in the

disconfirming conditions (i.e., participants recalled

identity-consistent experiences when they received

disconfirming feedback). Thus, the amount of feedback-

consistent egalitarian experiences as well as the number of

positive egalitarian experiences served as dependent

measures.









Participants indicated the percentile score (1-100%)

that they received for each of the three dimensions (i.e.,

honesty, responsibility, egalitarianism) during the

interpretive feedback session. Participants also indicated

their interest in receiving additional test feedback by

responding to two questions; one question assessed their

interest without any mention of compensation and the other

assessed their interest after mentioning the contingency of

receiving extra credit in their psychology course. If they

stated an interest in either of these conditions, they were

given the opportunity to indicate the psychosocial

dimensions about which they would like to receive feedback.

In addition, participants responded to 17 questions

designed to assess attributions for their performance on the

personality test. For example, participants were asked how

much effort they put into taking the DAP test and how

responsible for their performance were factors outside of

their own personality and effort. In addition, three

questions asked how discrepant the feedback on each of the

three dimensions was from their self-perceptions. Three

extra questions required the participants to state their

beliefs about how egalitarian, honest, and responsible they

are, irrespective of their test scores on each of the

dimensions.













CHAPTER 4
RESULTS

The dependent measures for the present research

consisted of the manipulation check and those measures that

assessed the ways in which individuals process, evaluate,

and utilize self-discrepant information.

Manipulation Check

The results demonstrate that the feedback manipulation

was highly effective in producing the desired effect.

Responses to a question asking how discrepant the

egalitarian feedback was from participants' self-perceptions

were subjected to a 3 (identity style: informational,

normative, or diffuse) x 2 (feedback: confirming or

disconfirming) x 2 (gender: male or female) between subjects

analysis of variance. The analysis revealed only the

predicted main effect of feedback, F(1, 83) = 73.38, p <

.001. As expected, those participants who received

disconfirming (anti-egalitarian) feedback said the feedback

was significantly more discrepant from their self-

conceptions (M = 4.71) than did those who received

confirming (pro-egalitarian) feedback (M = 2.21).

Further evidence of the effectiveness of the

manipulation check was demonstrated by showing that self-

perceptions of egalitarianism changed after the feedback was

44









given. Results from a 3 (identity style: informational,

normative, or diffuse) x 2 (feedback: confirming or

disconfirming) x 2 (gender: male or female) between subjects

analysis of variance on post-feedback self-evaluations

revealed that individuals who received disconfirming

feedback evidenced self-perceptions that were less

egalitarian (M = 4.87) than individuals who received

confirming feedback (M = 6.02), F(1, 113) = 39.64, p <

.0001. No other main effects or interactions were found.

Processing of Self-Discrepant Information

Individual variability in the processing of self-

discrepant information was examined by assessing the ways in

which identity style influenced the recall of evaluative

information and relevant personal experiences following

feedback that disconfirmed or confirmed existing self-

perceptions. The specific dependent measures relevant to

the processing of self-relevant information included: a) the

amount of feedback-consistent egalitarian information

recalled (i.e., the recall of pro-egalitarian information in

the confirming feedback condition or anti-egalitarian

information in the disconfirming feedback condition); b)

the discrepancy between the actual percentile ratings on the

egalitarian dimension and the recall of those ratings; c)

the number of feedback-consistent egalitarian experiences

recalled (i.e., the recall of pro-egalitarian experiences in

the confirming feedback condition or anti-egalitarian









experiences in the disconfirming feedback condition); and

d) the total number of positive egalitarian experiences

recalled (i.e., the number of pro-egalitarian experiences

recalled by individuals in both the confirming and

disconfirming conditions).

Interiudge Agreement

On a randomly selected sample of 414 pieces of

evaluative information recalled, the percentage of

interrater agreement was 93% for the classification of the

information as belonging to a given dimension (i.e.,

egalitarianism, responsibility, or honesty), and 99% for

whether the information was consistent or inconsistent with

the feedback. On a sample of 361 personal experiences

recalled, the percentage of interrater agreement was 91% for

the classification of the experience as relevant to a given

dimension, and 98% for whether the experience fit or didn't

fit with the feedback.

Recall of Evaluative Information

It was hypothesized that individuals with an

informational style would recall more self-discrepant

information than normative or diffuse/avoidant individuals

when they received disconfirming feedback. Additionally, it

was predicted that information-oriented individuals would

demonstrate a more balanced recall of self-consistent and

self-inconsistent evaluative information, relative to

normative and diffuse/avoidant individuals. Both normative








and diffuse/avoidant individuals were expected to recall

more identity-consistent than identity-discrepant

information.

In order to test this hypothesis, the number of

feedback-consistent pieces of egalitarian information

recalled was submitted to a 3 (identity style:

informational, normative, or diffuse) x 2 (feedback:

confirming or disconfirming) x 2 (gender: male or female)

between subjects analysis of variance. This analysis

revealed no significant main effects or interactions

involving the independent variables (p's > .10), indicating

that the amount of feedback-consistent information recalled

did not vary as a function of identity style or feedback.

Contrary to expectations, individuals utilizing different

identity styles did not differ in the amount of self-

discrepant information they recalled following the receipt

of disconfirming feedback. Also surprising was the

observation that identity style did influence the

preferential recall of identity-consistent over identity-

inconsistent information.

Score Recall

Individuals with an informational orientation were

expected to recall identity-discrepant percentile scores

with more accuracy than individuals with either a normative

or diffuse/avoidant approach to identity development.

Normative and diffuse/avoidant types were expected to









demonstrate a more pronounced confirmatory bias in their

recall of negatively discrepant percentile scores. A 3

(identity style: informational, normative, or diffuse) x 2

(feedback: confirming or disconfirming) x 2 (gender: male or

female) between subjects analysis of variance conducted on

discrepancy scores revealed only a marginally significant

main effect of feedback, F(1, 113) = 3.79, p = .05. In

fitting with the literature on confirmatory bias (e.g.,

Swann, 1987), a trend of responses emerged whereby

participants tended to make relatively more positively-

biased errors in their recall of percentile scores that

contradicted their self-evaluations (M = -2.14) relative to

those that confirmed their self-perceptions (M = .11).

Unexpectedly, individual variability in identity style did

not moderate this effect of feedback on the accuracy with

which percentile scores were recalled. This finding

indicates that individuals utilizing all three processing

modes were equally likely to engage in self-verification

when recalling feedback scores that contradicted their self-

evaluations.

Recall of Feedback-Consistent Personal Experiences

Individuals with an informational orientation to the

identity negotiation process were expected to recall

significantly more personal experiences relevant to the

negatively discrepant feedback than individuals utilizing a

normative or diffuse/avoidant style. Moreover, the personal









recollections of information-oriented individuals were

expected to reflect a greater balance of identity-congruent

and identity-incongruent experiences relative to the

recollections of normative and diffuse/avoidant individuals.

It was predicted that individuals with normative and

diffuse/avoidant processing styles would bias their recall

of personal experiences such that they would recall more

experiences that would serve to validate their self-

perceptions than invalidate them.

The number of feedback-consistent egalitarian

experiences recalled was submitted to a 3 (identity style:

informational, normative, or diffuse) x 2 (feedback:

confirming or disconfirming) x 2 (gender: male or female)

between subjects analysis of variance. The analysis

yielded a main effect of feedback, F(1, 113) = 12.86, p <

.01, and a marginally significant feedback X identity style

interaction, F(2, 113) = 2.89, p < .06. Confirming feedback

induced individuals to recall disproportionately more pro-

egalitarian experiences than the number of anti-egalitarian

experiences recalled following disconfirming feedback.

Unexpectedly, the simple effect of identity style was

not significant under the disconfirming feedback conditions,

F(2, 60) = 1.57, p > .21, indicating that individuals did

not differentially recall image-discrepant personal

experiences as a function of their identity style. The

means for the marginally significant feedback X identity








style interaction on recall of personal experiences are

presented in Table 1.

Table 1
Number of feedback-consistent egalitarian experiences
recalled as a function of identity style and feedback

Identity Style

Feedback Information Normative Diffuse

Confirm 1.45 1.79 2.23a

Disconfirm 1.05 1.18 .60b


Note. Means with different subscripts differ significantly
at P < .05 following a significant simple effects test.

Viewing the marginally significant interaction

differently revealed a simple effect of feedback that was

significant only for those participants with a diffuse

identity style, F(1, 40) = 17.71, p < .01. Diffuse

individuals were more likely to recall pro-egalitarian

experiences following feedback that confirmed their

egalitarian self-image than they were to recall anti-

egalitarian experiences following feedback that disconfirmed

their egalitarian self-image. The simple effect of feedback

was not significant for those participants with a normative,

F(1, 39) = 1.83, p > .18, or informational identity style,

F(1, 40) = 1.37, p > .25, indicating that informational and

normative participants were just as likely to recall

experiences that disconfirmed their self-image as

experiences that confirmed their self-image.









As predicted, the greatest balance in the recall of

image-consistent and image-discrepant experiences occurred

among information-oriented individuals (they recalled only

.40 more image-consistent than image-discrepant experiences,

compared to normative types at .61). Thus, although there

were no apparent differences among them in the recall of

self-discrepant experiences, individuals using different

identity processing styles did show a pattern of responses

indicating that they differed somewhat in the extent to

which they attended to and sought out evidence that

confirmed their self-beliefs, relative to that which

disconfirmed their self-beliefs.

Recall of Positively Egalitarian Personal Experiences

It was hypothesized that individuals with normative and

diffuse/avoidant identity styles would recall significantly

more positively egalitarian experiences than informational

individuals in both the confirming and disconfirming

conditions. The overall number of positive experiences

recalled relevant to egalitarianism was analyzed using a 3

(identity style: informational, normative, or diffuse) x 2

(feedback: confirming or disconfirming) x 2 (gender: male or

female) between subjects analysis of variance. No main

effects or interactions of the independent variables were

obtained (p's > .20), thus refuting the hypothesis that

diffuse/avoidant and normative-oriented individuals would

recall significantly more positively egalitarian (i.e.,








identity-confirming) experiences than informational

individuals, regardless of the type of feedback they were

provided.

Evaluation of Self-Discrepant Information

Regarding the evaluation of evaluative information,

individuals rated the perceived credibility of the test and

responded to questions that assessed attributions for their

test performance. The scales that were used to assess these

perceptions and attributions were derived by subjecting the

17 questionnaire items to a principal components factor

analysis. On inspection of the scree plot as well as an

"eigenvalue > 1" criterion, three factors were retained and

submitted to a varimax rotation. The results of the factor

analysis are presented in Table 2.

Table 2
Factor loadings for the three factor solution of the
perception questionnaire items

Item Factor 1 Factor 2 Factor 3

Interest in test .66 -.03 .25
Effort required .56 .34 -.23
Person at best .38 -.16 .01
Personal responsibility .64 -.06 .10
Importance to self .53 -.07 .17
Validity of test -.68 .06 -.29
Influence of external factors .19 .62 -.24
Influence of artistic ability .08 .57 -.06
Clarity of scoring guidelines .22 -.48 .11
Comfort taking test .42 -.60 -.21
Difficulty of test -.12 .64 .07
Reliability of test .32 -.16 .55
Accuracy of interpreter .03 .01 -.52
Accuracy of interpretation .10 .02 .73
Sensitivity of interpreter .30 -.11 .69


Note. Factor loadings that are underlined are > .38.









Three items loaded greater than .38 on more than one

factor. However, the item assessing degree of comfort

taking the test loaded predominately on one factor.

Therefore, this item was retained. Two other items, those

assessing comfort receiving feedback and perceived expertise

of the interpreter loaded approximately the same on two

factors, and therefore, these two items were deleted.

Factor 1 was labelled Internal, and reflected

attributions residing within the individual for his or her

test performance (e.g., How personally responsible do you

feel for your performance on the DAP test?). Factor 2 was

labelled External, and reflected factors outside the person

that may have influenced test performance (e.g., How

responsible were factors outside of your own personality and

effort for your performance on the DAP test?). Factor 3 was

labelled Test and reflected perceptions of credibility

regarding the test and test interpreter (e.g., If you were

to take a different test measuring the same aspects of your

personality, do you think you would do about the same?).

Therefore, the dependent measures relevant to the evaluation

of self-relevant information included: a) the scores on the

internal dimension; b) the scores on the external dimension;

and c) the scores on the test dimension.

Internal Attributions for Test Performance

It was anticipated that the self-reflective individuals

utilizing an informational style would be more likely than









normative or diffuse/avoidant individuals to attribute

causes for unfavorable feedback to characteristics of

themselves. Scores on the internal factor were submitted to

a 3 (identity style: informational, normative, or diffuse) x

2 (feedback: confirming or disconfirming) x 2 (gender: male

or female) between subjects analysis of variance. No main

effects or interactions of the independent variables were

obtained (p's > .13). Contrary to predictions, individuals

who differed by identity style did not exhibit differential

tendencies toward making internal attributions for identity-

disconfirming feedback (P's > .30).

External Attributions for Test Performance

It was hypothesized that diffuse/avoidant and

normative-oriented individuals, presumably operating from

self-presentational and defensive postures, respectively,

would be more likely than informational individuals to

externalize the cause for negatively discrepant test

feedback. A 3 (identity style: informational, normative, or

diffuse) x 2 (feedback: confirming or disconfirming) x 2

(gender: male or female) between subjects analysis of

variance was used to analyze scores on the external factor.

As with the internal factor, no main effects or interactions

of the independent variables were obtained (P's > .16).

Thus, contrary to expectations, individuals with a

diffuse/avoidant or normative identity orientation were no

more likely than individuals using an informational approach







55

to ascribe causality for identity-threatening test feedback

to factors residing outside of themselves.

Perceptions of the Test

It was hypothesized that information-oriented

individuals would evaluate the test on which they received

feedback as well as the source of feedback as significantly

more credible relative to normative or diffuse/avoidant

individuals. The scores for the test factor were analyzed

using a 3 (identity style: informational, normative, or

diffuse) x 2 (feedback: confirming or disconfirming) x 2

(gender: male or female) between subjects analysis of

variance. The analysis revealed only a main effect of

feedback for the test factor, F(1, 113) = 12.21, p < .01.

In support of previously documented findings (e.g., Crary,

1966; Markus, 1977; Shrauger & Lund, 1975), participants who

received disconfirming feedback rated the test and test

interpreter as less credible (M = 4.58) then did

participants who received confirming feedback (M = 5.08).

No other main effects or interactions were significant on

the test factor scale (p's > .18) Surprisingly, the use of

a diffuse/avoidant or normative processing orientation did

not augment self-serving tendencies as expected. Instead,

individuals utilizing all three of the identity styles

appeared equally likely to discount the credibility of

identity-discrepant information.









Utilization of Self-Discrepant Information

Responding to discrepant feedback about the self

involves more than simply attending to and evaluating it.

Individuals must also decide how they will use the

information to motivate their subsequent actions and alter

(or confirm) their self-constructions (Shrauger, 1975). The

dependent measures relevant to the utilization of self-

discrepant information involved assessments of: a) the

extent to which these individuals were intrinsically

motivated by the information to take an additional

personality test; b) whether or not they requested

additional information specific to the egalitarianism

dimension; and c) their post-feedback self-evaluations of

how egalitarian they are.

Intrinsic Interest in Taking Another Personality Test

It was predicted that self-exploratory, information-

oriented individuals would be more intrinsically interested,

relative to normative and diffuse/avoidant individuals, in

taking another personality test. Moreover, it was

anticipated that levels of intrinsic interest in taking

another test would be highest among informational

individuals, relative to other individuals, when feedback

was contrary to self-perceptions.

A 3 (identity style: informational, normative, or

diffuse) x 2 (feedback: confirming or disconfirming) x 2

(gender: male or female) between subjects analysis of









variance was used to examine the scores on the intrinsic

interest scale. Results yielded a main effect of identity

style, F(2, 113) = 3.70, p < .05, and although a significant

feedback X gender interaction emerged, F(1, 113) = 6.06, p <

.05, identity style did not interact with these variables.

As expected, those participants with an informational

identity style expressed the greatest degree of intrinsic

interest (M = 5.43), followed by those with normative (M =

5.05) and diffuse (M = 4.59) identity styles. Information-

oriented individuals evidenced significantly greater

intrinsic interest than diffuse individuals, F(1, 122) =

5.17, p < .05, with the normative types not differing from

either the informational or diffuse participants. However,

when informational, normative, and diffuse individuals were

compared on the degree of intrinsic interest they

demonstrated following identity-discrepant feedback there

were no significant differences (p's > .15). This latter

finding is at odds with the hypothesis that information-

oriented individuals would be more intrinsically stimulated

by identity-discrepant feedback to take an additional

personality test.

Request for Additional Evaluative Information on
Egalitarianism

Information-oriented individuals were hypothesized to

be more likely than either normative or diffuse/avoidant

individuals to request additional information pertaining to

egalitarianism after feedback contradicted their egalitarian









self-perceptions. A chi-square analysis was conducted

comparing the percentage of participants who chose to

receive feedback on the egalitarian dimension following

confirming and disconfirming feedback. The chi-square

analysis was significant for informational types, chi-

squared (1) = 9.72, p < .01, but not for normative, chi-

squared (1) = 1.15, p > .28, or diffuse types, chi-squared

(1) = 2.35, p > .12. When information-oriented participants

received confirming feedback, only 22% selected to receive

additional information about the egalitarian dimension.

However, when information-oriented participants received

disconfirming feedback, 78% chose to receive additional

evaluative information on the egalitarian dimension.

As expected, information-oriented individuals were more

likely to seek out additional information about themselves

relevant to egalitarianism when feedback challenged their

self-conceptions than when it did not. The nature of the

feedback did not make a significant difference in the

selection of egalitarian feedback for normative and diffuse

individuals. However, an interesting, although

nonsignificant, pattern of responses suggested that

individuals utilizing normative and diffuse styles were

especially likely to choose egalitarian-relevant information

after their self-beliefs were confirmed but reject

additional egalitarian-relevant information after their

self-concepts were threatened.









Post-Feedback Ratings of Egalitarianism

One long-standing interest among researchers examining

the effects of self-relevant feedback is how individuals use

such feedback to modify their self-beliefs (see Swann,

1987). It was hypothesized that information-oriented

individuals would be more likely than diffuse/avoidant or

normative-oriented individuals to lower their self-ratings

of egalitarianism following negatively discrepant feedback.

A 3 (identity style: informational, normative, or diffuse) x

2 (feedback: confirming or disconfirming) x 2 (gender: male

or female) between subjects analysis of variance was

conducted on post-feedback self-evaluations on the

egalitarianism dimension. Contrary to expectations,

identity style did not moderate the influence of the

feedback effect already discussed (see the Manipulation

Check section). Thus, an informational style was no more

likely than the other two modes of processing to facilitate

accommodative shifts in self-views in response to identity-

discrepant feedback.

A post-hoc analysis was conducted in order to identify

any possible indirect influence that self-relevant feedback

could have exerted on self-evaluations. A simulateous

multiple regression was used to examine whether the

relationship between the number of identity-consistent or

identity-inconsistent personal recollections that

participants generated and their self-ratings of








egalitarianism was greater for individuals with particular

identity styles. This regression analysis was conducted

separately for individuals with informational, normative,

and diffuse/avoidant identity styles. For each regression

the egalitarian self-rating was the criterion measure, and

the predictors were type of feedback, the number of

feedback-consistent experiences recalled, and the

interaction between type of feedback and the number of

feedback-consistent experiences recalled.

The regression for the diffuse/avoidant participants

revealed no significant effects of the predictors on self-

ratings of egalitarianism. The regression for those

individuals with a normative orientation revealed only a

significant effect of feedback, F(1, 37) = 4.33, p < .05.

Normative participants rated themselves as more egalitarian

following confirming feedback (M = 6.00) than following

disconfirming feedback (M = 4.59).

The regression for those participants using an

informational identity style revealed a significant effect

of the interaction between type of feedback and the number

of feedback-consistent egalitarian experiences recalled on

self-ratings of egalitarianism, F(1, 38) = 6.34, p < .02.

Simple effect tests showed that when informational types

received confirming feedback, there was no relationship

between the number of positively-egalitarian experiences

recalled and self-ratings of egalitarianism, t(19) = .37,







61

p > .37. However, when informational types received

disconfirming feedback, the greater the number of negative

egalitarian experiences recalled, the lower their ratings of

egalitarianism, t(19) = -2.76, p < .02. These results

suggest that information-oriented individuals who reflect on

personal experiences that verify self-discrepant information

may be more likely to revise their egalitarian self-

conceptions to incorporate this information.













CHAPTER 5
DISCUSSION

The discussion of the present research will be divided

into the following sections: (a) a discussion of the

influence of identity style on the processing, evaluation,

and utilization of confirming and disconfirming self-

relevant information, (b) the implications of the results

for Berzonsky's (1990) process model of identity development

(c) potential applications of Berzonsky's (1990) model,

taking into account the present results, (d) limitations of

the present research, and (e) conclusions.

Processing, Evaluation, and Utilization
of Self-Discrepant Information

In his process model of identity development, Berzonsky

(1990) proposed that stylistic differences in the

construction and reconstruction of the self should determine

distinctly different responses to self-discrepant

information. The present study examined how individual

differences in identity style influenced the processing,

evaluation, and utilization of information that either

confirmed or disconfirmed an important belief about the

self.









Processing of Self-Discrepant Information

The present section discusses those results that are

relevant to the processing of dissonant evaluative

information about the self. The literature on self-

verification has provided strong evidence that people are

more attentive to and recall more self-confirmatory

information than self-discrepant information (Crary, 1966;

Swann, 1987; Swann & Read, 1981a, 1981b). Drawing from

Berzonsky's (1990) proposal that individual differences in

the use of self-serving strategies should be a function of

one's identity style, it was hypothesized that the

preferential recall of self-confirmatory over self-

discrepant evaluative information and personal experiences

would be strongest for diffuse/avoidant and normative

individuals and weakest for informational individuals.

Surprisingly, individuals using normative and

diffuse/avoidant identity processing orientations were not

more likely than information-oriented individuals to recall

information that validated their egalitarian self-concepts

nor were they more likely to bias their recall of negatively

discrepant evaluative feedback in order to make it

consistent with initial self-perceptions of egalitarianism

(Shrauger, 1975). Oddly enough, feedback itself did not

affect the amount of confirming versus disconfirming

information recalled, a finding that is at odds with most of

the literature (e.g., Crary, 1966; Swann & Read, 1981b).








These observations indicated the possibility that factors

inherent in the experimental paradigm itself may have

minimized the impact of both feedback and style differences.

It is worth noting that the procedure of this study left

little room for recall bias in that it called for the

presentation of highly credible feedback by a highly

credible source followed immediately by the recall of that

information. Although feedback did have a marginal effect

on the accuracy with which percentile scores were recalled,

this finding may very well be an artifact of the greater

likelihood of error with the recall of continuous numerical

ratings than discrete narrative elements.

Although the interaction between feedback and identity

style achieved marginal significance for the recall of

personal experiences, individuals utilizing different

identity styles did not differentially recall personal

experiences that disconfirmed their egalitarian self-images.

This finding contradicted Berzonsky's (1990) proposal that

differences among individuals utilizing the various identity

orientations should be reflected in the extent to which they

engage self-serving biases when processing information that

challenges existing self-constructions.

Nevertheless, this marginally significant interaction

between feedback and identity style on the recollection of

personal experiences did reveal an interesting trend of

results suggesting that identity style may play some role in









differentially guiding the elaboration and testing of

identity-consistent versus inconsistent information. In

particular, individuals utilizing a diffuse/avoidant

approach to identity formation recalled somewhat more

identity-consistent experiences following confirming

feedback than identity-discrepant experiences following

disconfirming feedback. That is, these individuals recalled

personal experiences that served to maintain a socially

desirable self-image and suppressed potentially

incriminating personal recollections. Given the recent

evidence that diffuse/avoidant individuals are especially

motivated to strive for social approval and validation

(Berzonsky, 1994) it is not surprising that they would

refrain from attending to or elaborating on information that

could potentially discredit a socially-reinforced

egalitarian image.

This pattern of findings also suggested that

information-oriented individuals, unlike diffuse/avoidant

types, did not bias their recollection of personal

experiences in order to maintain an egalitarian self-image.

Rather, informational individuals were just as likely to

recall identity-discrepant experiences following

disconfirming feedback as identity-congruent experiences

following confirming feedback. This unbiased examination of

both identity confirming and disconfirming personal

experiences provides some support for the depiction of







66

informational types as introspective and willing to explore

aspects of their self-identity in response to self-relevant

feedback, even when it calls into question a valued self-

image (Berzonsky, 1990; 1993a).

Contrary to expectation, normative-oriented individuals

also appeared to be indiscriminate in their recall of

identity-consistent and identity-discrepant personal

experiences. However, it is worth noting that the

discrepancy in recall of self-consistent and self-discrepant

experiences was slightly larger, although not significantly

so, for the normative types than for the information-

oriented individuals. Nevertheless, given the strong

evidence suggesting that normative individuals would

defensively reject potentially invalidating information,

this finding suggests that researchers need to more clearly

specify the conditions under which these individuals would

presumably prefer self-serving, defensive maneuvers over

unbiased self-exploration (Berzonsky, 1993a).

Evaluation of Self-Discrepant Information

The following section addresses those results relevant

to how individuals utilizing the three styles of self-

construction evaluated information that confirmed or

disconfirmed important aspects of their identities.

Diffuse/avoidant individuals were expected to be especially

motivated to avoid the negative social ramifications of

being identified as non-egalitarian by employing self-







67

handicapping attributions for unfavorable test feedback and

making self-serving assessments of the test's validity.

Individuals with a normative orientation were expected to

use similar strategies to rigorously defend their self-

views, but for different reasons. Presumably, normative

individuals would engage this self-serving bias in order to

ward off the internal dissonance that might arise if their

internalized prescriptions were invalidated (Berzonsky,

1990). Finally, it was predicted that information-oriented

individuals would appreciate the potential implications of

both validating and invalidating information for their

developing sense of self and therefore make fewer strategic

attempts to discredit the test or rationalize their

performance when feedback challenged their self-conceptions.

Quite unexpectedly, identity processing style did not

affect perceptions of test validity. Participants with all

three identity styles were equally likely to ascribe less

credibility to the assessment measure, as well as the source

of feedback, after receiving information that was discrepant

with their egalitarian self-image (Crary, 1966; Markus,

1977; Shrauger & Lund, 1975). This finding suggests that

information-oriented individuals are not immune to the

preservation of the self through self-serving assessments of

image-discrepant feedback. At least one other study

(Berzonsky & Kinney, 1994b) has demonstrated that

informational types may deny or distort reality when







68

threatened with potentially invalidating feedback about the

self.

Surprisingly, the tendency to make internal versus

external attributions for test performance was independent

of both information processing style and feedback. Contrary

to previous findings (see Swann, 1987), individuals

receiving negative feedback were not more likely than those

receiving positive feedback to assign causality for the

feedback to external factors. As discussed in the

limitations section, the tendency to make self-serving

attributions may have been muted in the disconfirming

conditions as a result of receiving identity-confirming

feedback on two out of three attributes and then assessing

causality for all of the evaluative feedback together,

rather than independently for each attribute.

Utilization of Self-Discrepant Information

Responding to contradictory feedback about the self

entails not only evaluating the information, but also

deciding how to use the feedback to revise (or maintain)

self-evaluations, as well as determine one's level of

motivation for subsequent action relevant to the feedback

(Shrauger, 1975). Presumably motivated to seek information

and experiences that will facilitate the development of a

more effective self-theory, individuals with an

informational orientation were expected to respond to

disconfirming feedback with requests for additional self-









relevant information, as well as an openness to self-

reconstruction. In contrast, it was predicted that

normative individuals would defensively dismiss

opportunities to elaborate on identity-threatening feedback

or consider revisions in their self-structure. Finally, it

was expected that diffuse/avoidant individuals, relying on

an external orientation and lacking firm internal

commitments, would be motivated by the threat of social

rejection to conserve their egalitarian self-constructions

and reject any potentially incriminating information.

Seeking additional information. Regardless of the

feedback they received, information-oriented individuals

expressed the greatest intrinsic motivation to seek out

additional self-relevant information, with diffuse/avoidant

individuals showing the least intrinsic interest and

normative types scoring in the middle, not differing

significantly from either of the other two. This finding

supports the results of previous studies demonstrating that

need for cognition and openness to experience are positively

associated with an informational style and negatively

related to a diffuse/avoidant orientation (Berzonsky, 1993a;

Berzonsky and Sullivan, 1992). However, the observation

that normative individuals showed as much intrinsic interest

in seeking out additional self-relevant feedback as

informational types refutes previous findings that depict

individuals with a normative orientation as closed to







70

information that may challenge important aspects of the self

(Berzonsky, 1990; Berzonsky, 1993a; Berzonsky & Sullivan,

1992).

As expected, individuals using an informational style

showed a greater tendency to seek additional information

specific to the egalitarianism dimension following feedback

that disconfirmed this important attribute than following

feedback that confirmed this attribute. These results

provide some validation for Berzonsky's (1990) depiction of

informational types as generally skeptical about their self-

constructions and open to self-relevant information. Self-

discrepant feedback, in particular, stimulated information-

oriented individuals to seek additional information,

presumably in order to clarify the implications of the

feedback for their self-development.

In contrast, the nature of the feedback did not

influence the extent to which normative and diffuse/avoidant

individuals chose to receive additional information on the

egalitarianism dimension. Interestingly, however,

individuals relying on these two approaches showed a pattern

of responses directly opposite to that of information-

oriented individuals; diffuse/avoidant and normative

individuals were more likely to request additional

egalitarian information after receiving confirming feedback

than after receiving disconfirming feedback. This pattern

of responses for normative and diffuse/avoidant individuals







71

is better understood in light of the results from a study by

Swann and Read (1981a) in which individuals reported that

they could learn more about themselves by examining self-

confirmatory information as compared with self-

disconfirmatory information. It is possible that normative

and diffuse/avoidant individuals may regard information that

confirms their self-conceptions as more diagnostic than

information that disconfirms their self-conceptions and

therefore seek out the former over the latter.

Revising the self. Because self-appraisal is important

in determining subsequent behavior (Swann, 1987), it was

necessary to assess individual differences in the

willingness to revise self-perceptions in response to self-

discrepant feedback. Informational types were expected to

be more open, relative to the other identity styles, to

accommodating their self-views in light of dissonant

feedback (Berzonsky, 1990). Contrary to expectations,

although informational individuals evidenced lower

egalitarian self-ratings following disconfirming feedback

than confirming feedback, individuals with normative and

diffuse/avoidant identity orientations were just as likely

to demonstrate this modification in self-perceptions.

Although differences in identity style did not directly

moderate the effect of feedback on self-ratings as expected,

it was considered important to identify any indirect ways in

which identity style influenced individuals' self-







72

evaluations following self-discrepant information. Drawing

from Neimeyer and Metzler's (1992) observation that

autobiographical recollections influenced self-evaluations,

post hoc analyses were used to determine if the number of

identity-discrepant personal experiences recalled following

disconfirming feedback was positively related to lower self-

ratings of egalitarianism. Results indicated that this

relationship existed, but only for information-oriented

individuals. Thus, individuals who took an informational

approach to identity development not only actively explored

personal experiences that threatened their valued self-

image, but they also used these personal recollections to

further substantiate the need to revise their self-

constructions.

In contrast, the post-feedback self-ratings of

normative and diffuse/avoidant individuals were independent

of their personal recollections, indicating that the changes

they evidenced in self-ratings were probably not a function

of personal examination. Rather, the changes observed among

these individuals may have been more a function of the

highly structured situation in which they were unable to

influence or resist the feedback that they received. This

reasoning is supported by studies showing that laboratory

investigations of self-change may not be generalizable to

naturalistic settings because they do not afford the

opportunities that individuals ordinarily enjoy to resist









self-discrepant feedback (see Swann, 1987 for a review).

Moreover, given that their self-changes were not preceded by

personal examination, it is likely that the changes made by

the diffuse/avoidant and normative individuals may not only

be situationally induced but short-lived, as well.

Implications of the Results for Berzonsky's
Process Model of Identity Development

The process of self-definition has been conceptualized

as a life-long process reflecting a continuous examination

and re-examination of one's values, beliefs, goals, and

other identity-relevant issues (Berzonsky, 1990; Erikson,

1975). Berzonsky (1989a; 1990) theorized that individuals

exhibit characteristic differences in the ways in which they

approach or manage to avoid this identity negotiation

process. Moreover, it was presumed that these stylistic

differences would be most apparent when individuals

encounter information or experiences that necessitate a

revision of the existing self-structure (Berzonsky, 1990).

The results of the present study reveal that

Berzonsky's (1990) primary assumption is not upheld under

experimental scrutiny; individuals utilizing different

identity styles were not differentially responsive to self-

discrepant information. Given that Berzonsky's (1990)

predictions in this regard were based solely on

correlational evidence, further experimentation is necessary

to clarify the role of identity style in negotiating









information that disconfirms significant aspects of the

self.

These results do provide limited evidence indicating

that individuals with different identity styles may be

differentially responsive to confirming versus disconfirming

self-relevant information. An examination of the findings

relevant to each identity style highlights the potential

contributions of this work to our understanding of the

different ways in which each style directs the negotiation

of identity-consistent versus identity-inconsistent

information and experiences.

Information-Oriented Style

Information-oriented individuals showed evidence of

processing identity-relevant information in such a way to

preserve aspects of the existing self-structure even while

they remained open to self-elaboration and reconstruction.

On the one hand, adding credence to observations that

virtually all individuals will employ self-serving biases

(Greenwald, 1980; Taylor & Brown, 1988), informational

individuals were not immune to making self-serving

attributions and evaluations to justify unfavorable feedback

and biasing their recall of identity-discrepant feedback to

make it more consistent with their self-conceptions.

On the other hand, these individuals were also the most

intrinsically motivated of all individuals to learn more

about themselves after receiving self-relevant feedback.









Identity-discrepant information, as compared to identity-

consistent information, seemed especially likely to stir

informational individuals' skepticism regarding their self-

constructions and stimulate them to engage in deliberate

self-exploration and actively seek out new information

(Berzonsky, 1990). For example, they responded to self-

discrepant feedback with requests for additional evaluative

information relevant to the threatened attribute, with the

presumed purpose of this information search being to help

them clarify the implications of the inconsistent feedback

for their developing sense of self. Importantly, although

all individuals responded to identity-discrepant information

with changes in their self-conceptions, informational

individuals were the only ones who appeared to reorganize

the way in which they viewed themselves based on an

examination of their own personal experiences relative to

their existing convictions.

These findings provide some indication that, when

confronted with identity-relevant information, information-

oriented individuals may remain tentative about their self-

constructions and willing to test and revise their self-

views as needed in order to maximize effective adaptation

(Berzonsky, 1990). Their willingness to seek further

information and deploy accommodative processing strategies

in the face of self-discrepant information is consistent

with correlational evidence showing that information-









oriented individuals are introspective, open to experience,

and have a high need for cognition (Berzonsky, 1993a;

Berzonsky & Sullivan, 1992).

Diffuse/Avoidant Style

Diffuse/avoidant individuals were initially

characterized in the literature (Berzonsky, 1989a, 1989b,

1990) as taking a passive approach to identity construction.

In keeping with this description, the diffuse/avoidant types

in the present study expressed the least instrinsic interest

in learning more about themselves. However, rather than a

style marked by passivity or confusion as originally

proposed, the present study helps to verify more recent

depictions of diffuse/avoidant types as motivated to operate

in a strategic, self-serving manner intended to maintain

impressions for social approval (Berzonsky, 1994).

After receiving identity-consistent feedback,

individuals with a diffuse/avoidant orientation called to

mind past actions and experiences that served to reinforce

socially-approved beliefs about the self, whereas they

dismissed personal recollections that might lend credence to

contradictory, and therefore potentially incriminating,

information. These individuals' steadfast self-views of

egalitarianism led them to recall personal experiences that

would encourage the maintenance of these socially desirable

self-beliefs, a finding that highlights one way in which

self-conceptions exert a powerful channelling influence on









information-processing (for a review, see Higgins & Bargh,

1987). Thus, the present study provides some corroborating

evidence for the assumption that, rather than being directed

by internal commitments and convictions, diffuse/avoidant

individuals seem to be motivated primarily by past or

anticipated social consequences to preserve their existing

self-constructions (Berzonsky, 1990).

In addition to theorizing that diffuse/avoidant

individuals would deal with self-discrepant information by

displaying defensive avoidance and strategic social

monitoring, it was also proposed that diffuse/avoidant types

would be especially likely to use self-serving attributions

to rationalize unfavorable personality feedback (Berzonsky &

Ferrari, 1994a). While the diffuse/avoidant individuals in

this study were more likely to denigrate the validity of the

test and make external attributions for their test

performance after received disconfirming feedback, as

compared to confirming feedback, they were no more likely

than individuals with a normative or informational

orientation to rely on these self-serving strategies.

These results provide little support for Berzonsky's

(1990) original depiction of diffuse/avoidant individuals as

"ad hoc" self-theorists whose approach to identity

construction is primarily passive and accommodative.

Rather, the present findings provide some evidence in

support of Berzonsky's (1994) most recent depictions of








diffuse/avoidant individuals as motivated to use self-

serving strategies that will help them create or maintain

the impression of possessing socially desirable attributes.

However, while most of their responses to self-relevant

information suggested a tendency to verify socially

appropriate self-conceptions (Swann, 1987), others did not,

indicating that diffuse/avoidant individuals may only be

"strategic" when they feel that they can get away with it.

Normative Style

Perhaps the most theoretically and empirically

inconsistent findings were observed among those individuals

who utilized a normative approach to identity negotiation.

According to Berzonsky (1990), normative types, presumably

protective of the constructions they have borrowed from

close others and significant referent groups, should have

been especially motivated to dismiss identity-discrepant

feedback in a defensive effort to preserve their extant

self-perceptions. However, the present study showed no

evidence that distinguished normative individuals as

employing uniformly defensive, self-protective strategies.

In fact, the only support for the prediction that normative

types would be closed to information that threatened hard-

core areas of the self, such as important values (Berzonsky,

1990), was a nonsignificant pattern showing that they tended

to reject additional information pertaining to

egalitarianism once this attribute was threatened.









Individuals using a normative processing orientation

were no more likely than individuals relying on the other

identity styles to bias their recall of negatively

discrepant feedback or defensively blame unflattering test

feedback on external factors. Furthermore, normative

individuals were just as likely to recall personal

experiences that contradicted their self-conceptions as

those which confirmed them. Just as perplexing was the

finding that normative individuals expressed a level of

intrinsic interest in receiving additional evaluative

feedback that was not significantly different from that of

informational individuals.

These results are interesting in light of previous

inconsistent findings involving normative types. For

example, several recent studies have demonstrated that

individuals with a normative style use both favorable and

unfavorable coping mechanisms (Berzonsky, 1992a; 1992b) and

both adaptive and maladaptive defense mechanisms (Berzonsky

& Kinney, 1994b) when responding to stressful events.

Moreover, at least one other study (Berzonsky, 1993a) has

proposed that the defensiveness of normative individuals may

be more situation-specific and not a generalized reaction

like that of diffuse/avoidant types.

These findings, together with the results from the

present study, clearly challenge Berzonsky's (1990)

depiction of normative-oriented individuals as especially








threatened by and closed off to information that has the

potential to invalidate important self-views. At the very

least, Berzonsky's (1990) model needs to more clearly

specify the conditions under which normative individuals

prefer self-serving, defensive maneuvers over unbiased self-

exploration and revision.

It is possible that the unexpected findings in the

present study could be a function of normative individuals

in the disconfirming conditions selectively attending to the

identity-consistent feedback on the responsibility and

honesty dimensions. However, this possibility seems

unlikely considering that normative individuals in the

disconfirming conditions recalled dissonant personal

experiences relevant to egalitarianism, indicating that they

were at least somewhat responsive to the discrepant feedback

on the egalitarian dimension.

Potential Applications of the Present Results

Individuals encounter self-relevant experiences

everyday. The manner in which individuals negotiate life's

experiences, especially those which challenge existing

constructions of reality, has an impact on their ability to

adapt optimally in the environments in which they live

(Berzonsky, 1990). The present section entertains the

potential applications of the present results for three

areas in which the successful negotiation of self-relevant

information may be crucial.









Stress Management

Referring to Berzonsky's (1992a) definition of stress

as encountering a pressure to change in an undesired way,

one can see how the present results are immediately

applicable to the area of stress management. Applying his

process interpretation of identity development to the area

of stress management, Berzonsky (1992a; 1992b) proposed that

dispositional differences in identity style would influence

the ways in which individuals interpret and attempt to cope

with stressors. Using correlational methodologies,

Berzonsky (1992b) found evidence to support his hypothesis

that information-oriented individuals would confront

stressors with problem-focused, adaptive coping strategies

whereas diffuse/avoidant individuals would deal with

stressors by avoiding problem-relevant information and

engaging more emotion-focused, less effective strategies.

Normative individuals relied primarily on maladative coping

mechanisms but showed some tendencies toward healthier

coping skills, such as social support seeking (Berzonsky,

1992b).

The results of the present experimental study provide

some additional evidence to support suppositions that the

informational style may be marked by more adaptive and

effective psychological functioning. According to the

findings of this study, individuals with an informational

orientation are likely to respond to stressful events, or







82

events which challenge existing constructions, by gathering

problem-relevant information, calling to mind relevant

experiences from the past, and accommodating to change as

needed without sacrificing a stable self-structure.

In an interesting extension of this research on the

influence of identity style on modes of coping, Dusek and

Berzonsky (1993) examined whether processing styles interact

with event characteristics to determine how individuals

manage stress. Interestingly, the results of that study

showed that identity style was related to different optimal

means of processing specific types of stressful events. For

example, an informational style was preferred when dealing

with undesirable events for which individuals perceived

themselves as responsible (e.g., academic problems), perhaps

because they could be most effective at directly altering

the relationship between themselves and the stressor. An

experimental investigation of the relationship between

identity style and type of stressor should be the next step

in this line of research.

Discrepancy and Therapeutic Change

There are also implications of these results for

identifying client factors that might facilitate the change

process in counseling. Following the lead of cognitive and

behavior change theories (e.g., cognitive dissonance theory,

attribution theory) which proposed that encounters with

information and experiences that do not fit current









constructions of reality are a necessary condition for

change, research has revealed that communicating discrepant

ideas to clients may be a necessary condition for successful

counseling outcomes (Claiborn, 1982; Strong, 1968; Strong &

Claiborn, 1982). Kelly (1955) foreshadowed this theme some

forty years ago with his advice that therapists could effect

desired change by bringing alternative frames of reference

to bear on clients' constructions of the meanings of

personal and interpersonal issues related to their

difficulties.

Discrepant ideas are introduced in the counseling

process in the form of questions, reflections,

interpretations, counselor self-disclosures, confrontations,

personal feedback, and test interpretations (see Strong,

Welsh, Corcoran, & Hoyt, 1992). However, these types of

interventions threaten clients' constructions of reality and

may be experienced negatively, stimulating clients to

terminate the therapy relationship instead of changing

(Strong, et al., 1992). Systematic investigations of

interpretive discrepancy have attempted to determine the

factors involved in successfully managing these

interventions. Thus far, research has focused on the degree

of discrepancy between the feedback and the client's

conceptions, the timing of introducing discrepancy within

the treatment process, and the content of the discrepancy

(see Strong et al., 1992). However, what clients contribute







84

may be at least as important as these factors in determining

whether they accept discrepant feedback and make desired

changes.

The results of the present study indicate that the

characteristic manner by which individuals deal with

identity questions and decisions may predict their

willingness to examine and incorporate discrepant

interpretations into their self-constructions. For example,

even though information that confirms their self-views may

be more immediately appealing, information-oriented

individuals may be likely to remain open to alternative ways

of perceiving personal and interpersonal problems,

conflicts, and decisions because of the greater adaptability

that this flexible approach affords them. In sharp

contrast, diffuse/avoidant individuals may avoid dealing

with the discrepant feedback and eventually terminate

therapy altogether or dismiss the potential long-term

implications of the discrepant feedback by making short-term

changes that will gain approval from the therapist or

others. Normative individuals may examine information and

experiences but fall back on internalized prescriptions when

dealing with self-relevant information.

Prejudice Reduction

The present study also suggests implications of these

results for identifying individual differences in the

willingness to evaluate information relevant to personal









belief systems that maintain prejudice. Essed (1991) has

recently argued that the importance of the egalitarian value

system to Anglo-Americans' self-concepts has perpetuated the

denial of racism because people have co-opted this value

structure without critically examining their stereotypic

attitudes and prejudiced beliefs. The fact that

participants in this research reported that they espoused

egalitarian values and that these values were very

significant to their self-identity makes the results of this

study immediately applicable to the literature on prejudice.

Results from the present study revealed some evidence

indicating that identity style may differentially enable and

disable the seeking out and evaluation of self-discrepant

information, the personal exploration of actions and

experiences relevant to one's egalitarian self-image, and

the acceptance of feedback that is negatively discrepant

from one's egalitarian self-views. In particular,

individuals with an informational orientation to identity

construction appeared to operate from a self-determined

perspective within which conflicting sources of information

could be evaluated and integrated. These individuals'

willingness to test and revise aspects of their self-

identity when confronted with dissonant feedback suggests

that they might be more inclined to examine their personal

convictions of egalitarianism and recognize discrepancies







86

between them and their prejudiced attitudes and actions, the

first step in confronting prejudice (Rokeach, 1973).

The diffuse/avoidant individuals in the present study,

presumably driven by social incentives to maintain

appropriate impressions, seemed to more rigidly adhere to

their egalitarian self-beliefs and to insulate themselves

from information that threatened to invalidate these

critical self-views. Such tendencies are especially likely

to inhibit the awareness and reduction of racial bias

(Essed, 1991; Frey & Gaertner, 1986; Katz & Hass, 1988;

McConahay, 1986). The normative processing style, while not

especially open to potentially invalidating information, did

not appear to be marked by overriding tendencies toward

self-preservation either. It is possible that these

individuals may carefully monitor threats to their

egalitarian identity and respond according to the extent of

the perceived threat.

Overall, these findings suggest that appealing to the

ways in which people think rather than concentrating on the

content of what they think may be a successful approach to

prejudice reduction. For example, rather than training

individuals in what may be perceived as "politically

correct" forms of language and communication, it may be more

effective to encourage divergent, open, exploratory, and

integrative ways of thinking about and processing

information and experiences relevant to oneself and others.









Rather than learning mere tolerance of others for the sake

of social approval, this practice may lead to an

internalized desire to accept and appreciate one's own

culture as well as those that are different. Of course, a

direct investigation of the impact of this type of training

on prejudiced attitudes is necessary in order to derive any

firm conclusions.

Limitations of the Present Research

Having discussed the results and the theoretical

implications and potential applications of the present

research, it is also necessary to address the limitations of

this work. The present study involves delivering feedback

that is either consistent or negatively discrepant from

self-conceptions. Given that Berzonsky (1990) did not

specify what type of disconfirming information would augment

stylistic differences, it is possible that positively

discrepant feedback (i.e., feedback that disconfirms

negative conceptions of the self) may produce results that

are more consistent with Berzonsky's (1990) proposal. Thus,

this study would have benefitted from a design that included

informational valence.

In the present study, participants in the confirming

feedback conditions received uniformly positive feedback on

all three dimensions, whereas participants in the

disconfirming conditions were provided with identity-

consistent feedback on the responsibility and honesty









dimensions and identity-discrepant feedback on the

egalitarian dimension. This manner of providing feedback

was chosen in order to enhance the credibility of the bogus

egalitarian information in the disconfirming conditions

(Snyder et al., 1977). However, receiving negatively

discrepant feedback on only one attribute may have muted the

effects of the disconfirming feedback because individuals in

those conditions may have attended to the consistent

feedback. A clearer picture of the results and perhaps

stronger support for Berzonsky's (1990) individual

differences model may have been achieved had a less

conservative identity-threatening manipulation, such as

uniformly disconfirming feedback, been provided. In another

vein, providing individuals with ambiguous feedback (i.e., a

balance of confirming and disconfirming feedback) may have

allowed for a more judicious assessment of individual

tendencies toward confirmatory bias.

Another potential problem with this study involves the

possibility that the findings may be attributable to social

desirability factors rather than individual differences in

identity style. The use of a social desirability measure

would have allowed the experimenter to examine the extent to

which individuals endorse and defend egalitarian values for

social desirability reasons.

Regarding problems relative to the dependent measures,

the test perception and attribution questionnaire items were








constructed such that they collapsed across all three

feedback dimensions (i.e., egalitarianism, honesty, and

responsibility), yielding a general evaluation of the

feedback and making it impossible to discern how feedback

relevant to each particular attribute was evaluated. Thus,

in the disconfirming condition, where mixed (i.e., positive

and negative) feedback was delivered, it was not possible to

know for sure whether the individuals were responding to the

self-consistent or self-discrepant information, or both,

when they recorded their evaluations.

This study could have benefitted from other additional

assessments, as well. It would have been helpful to have

had questions designed to assess individuals' motivations

for responding to self-relevant feedback the way that they

did. Although the evidence gathered thus far in this line

of research suggests that informational individuals are

motivated to seek information in an attempt to better

understand themselves, it is impossible to make this

conclusion given the findings that are available. It is

possible that informational individuals requested additional

evaluative feedback in the hopes that the new information

would validate their preexisting self-beliefs. In addition,

this study could have benefitted from questions designed to

assess the processes that mediated the influence of self-

discrepant feedback on self-ratings. These questions are

left for future research.







90

Procedural factors may also have contributed to some of

the equivocal findings in this study. Optimally,

assessments of individuals' perceptions and attributions

relevant to the evaluative feedback would have been made

immediately after the information was received. However,

in the interim between the time the feedback was received

and the evaluations were made, individuals were given the

task of recalling the evaluative information, making

recollections of personal experiences, and indicating their

interest in additional information. This time lag may have

served to dissipate individuals' defensiveness, and

therefore, decrease any tendencies to make self-serving

attributions and evaluations.

Conclusions

Clearly, the results of the present study compromise

Berzonsky's (1990) assumption that individual variability in

identity style should predict distinctly different responses

to identity-discrepant information. However, these results

do afford some greater understanding of the ways in which

each identity style differentially drives the interpretation

and integration of the self-relevant experiences that

individuals encounter, depending upon whether those

experiences have the potential to validate or invalidate the

self.

Additional experimental research is needed in order to

make definitive conclusions regarding the similarities and







91

differences in how individuals utilizing different identity

styles process, evaluate, and utilize self-relevant

information. Importantly, the present study provides an

experimental paradigm that will be useful in this regard.

It will also be important to test Berzonsky's (1990) model

outside the laboratory. This type of investigation would

clarify the role of identity style in enabling or disabling

an individual's ability to make sense of personal

experiences, make useful predictions, and adequately adapt

and cope with the environment while still retaining some

sense of self-consistency over time and situations

(Berzonsky, 1993b). Finally, a necessary direction for

future work in this area includes a longitudinal examination

of the stylistic patterns through which individuals

interpret self-relevant information and redefine the self,

beginning with the early years and extending through

adulthood.













APPENDIX A
IDENTITY STYLE INVENTORY

Read each of the following statements carefully, then

use it to describe yourself. On the answer sheet, bubble in

the number which indicates the extent to which you think the

statement represents you. There are no right or wrong

answers. For instance, if the statement is very much like

you, mark a 4; if it is not like you at all, mark a 0. Use

the 0 to 4 scale below to indicate the degree to which you

think each statement is uncharacteristic (0) or

characteristic (4) of yourself.

0----------1---------- 2 ----------3----------4

NOT AT ALL VERY MUCH

LIKE ME LIKE ME

1. Regarding religious beliefs, I know basically what I

believe and don't believe.

2. I've spent a great deal of time thinking seriously about

what I should do with my life.

3. I'm not really sure what I'm doing in school; I guess

things will work themselves out.

4. I've more-or-less always operated according to the

values with which I was brought up.

5. I've spent a good deal of time reading and talking to

others about religious ideas.

92