Title: Memory retrieval latency and quantity as a function of objective self-awareness and self-schemata
CITATION THUMBNAILS PAGE IMAGE ZOOMABLE
Full Citation
STANDARD VIEW MARC VIEW
Permanent Link: http://ufdc.ufl.edu/UF00102772/00001
 Material Information
Title: Memory retrieval latency and quantity as a function of objective self-awareness and self-schemata
Physical Description: Book
Language: English
Creator: Landy, Cheryl R., 1956-
Copyright Date: 1986
 Record Information
Bibliographic ID: UF00102772
Volume ID: VID00001
Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
Resource Identifier: ltuf - AEP2524
oclc - 16243525

Full Text

MEMORY RETRIEVAL LATENCY AND QUANTITY
AS A FUNCTION OF OBJECTIVE SELF-AWARENESS
AND SELF-SCHEMATA




By

CHERYL R. LANDY


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


1986


















































Copyright 1986

by

Cheryl R. Landy





ACKNOWLEDGMENTS


I wish to express my special appreciation to my doctoral

chairman, Dr. Greg Neimeyer. I am grateful for Dr. Neimeyer's valuable

insight and support throughout my doctoral pursuit and for facilitating

my progress at every stage.

My gratitude is extended to each of the members of my doctoral

committee: Dr. Robert C. Ziller, Dr. James Archer, Jr., Dr. Mark Alicke

and Dr. Jeffery Braden. I am appreciative of their knowledge and of

their concerted efforts.

I wish to thank Dr. Keith White for kindly sharing his compu-

ter expertise during my data collection. I also thank Louis Dreblow for

transferring my data across computer systems. A warm thank-you is

extended to Perin Patel at Florida International University's Academic

Computer Services. Perin has so generously assisted me and I am parti-

cularly grateful. Finally, a special thank-you is extended to Mark

Matlock for typing my dissertation manuscript and to Trenesia Green for

typing the final revisions.

My deepest gratitude is extended to my dear husband Steve.

Steve's unwaning love, respect, and affection for me have meant so much

to me throughout my doctoral pursuit and throughout the nearly nine

years that we have shared our lives. I thank Steve with all my love.





















TABLE OF CONTENTS


Page

ACKNOWLEDGMENTS....................................... iii


LIST OF TABLES............................................ vi


LIST OF FIGURES........................................... vii


ABSTR~ACT. ........ r.............................vi



CHAPTERS


I INJTRODUCTI ON....................................... 1


Introduction to the Investigation....................... 1
Purpose of the Investigation.................... 5
Need for the Investigation............................. 5
Significance of the Investigation....................... 8


II REVIEW OF THE LIERTRE...................... 10



Memory Theory and Research.............................. 10
Self-Schemata Theory and Research....................... 18
Self-Awareness Theory and Research...................... 25



III METHODOL OGY........................................ 33


Introduction.................................... 33
Research Design...................................... 33
Independent Variables................................. 34
Objective Self-Awareness......,.................... 34
Schemata Congruence............................... 35
Schemata Evaluation............................... 36
Determination of Schemata......................... 37
Dependent Variables.................................. 39
Retrieval Latency................................. 39
Retrieval Quantity................................ 40
Research Hypotheses.................................. 40
Data Collection.................................... 42
Subjects ........ ....... ........ ....... ....... 42
Experimental Procedures........................... 43





IV RESULTS............................................ 50


Preliminary Tests for Gender Effects....................,..,... 50
Effects on Retrieval Quantity...........................,..... 51
Main Effects....................................... 52
Two-way Interaction................................. 53
Three-way Interaction..........................,...... 55
Effects on Retrieval Latency.................................. 57
Main Effects....................................... 58
Two-way Interactions................................ 60
Three-way Interaction................................ 63
Summary of Main Results..................................... 65



V DISCUSSION AND CONCLUSIONS.................,............... 68


Memory Retrieval: The Role of Self-Schemata.................. 68
Memory Retrieval: The Role of Self-Awareness................. 73
Implications of the Results.................................. 80
Limitations of the Investigation..................... 82
Recommendations for Future Research........................... 83

APPENDICES

A EXPERIMENTAL STIMULI.............................,........ 86

8 INFORMED CONSENT........................................ 90




BIOGRAPHICAL SKETCH..,....................................... 104
























Table Pg

1 Gender X Schemata Evaluation Interaction on Mean Retrieval
Quantity ... .. .. ... .. .. ... .. .. ..................... 50

2 ANOVA: Effects of Self-Awareness, Schemata Congruence,
and Schemata Evaluation on Mean Retrieval Quantity.......... 51

3 Main Effects of Self-Awareness, Schemata Congruence, and
Schemata Evaluation on Mean Retrieval Quantity.............,. 52

4 Self-Awareness X Schemata Congruence Interaction on Mean
Retrieval Quantity..................................... 54

5 Self-Awareness X Schemata Congruence X Schemata Evaluation
Interaction on Mean Retrieval Quantity.................,..... 56

6 ANOVA: Effects of Self-Awareness, Schemata Congruence,
and Schemata Evaluation on Mean Retrieval Latency........... 58

7 Main Effects of Self-Awareness, Schemata Congruence, and
Schemata Evaluation on Mean Retrieval Latency....,........... 59

8 Self-Awareness X Schemata Congruence Interaction on Mean
Retrieval Latency...................................... 61

9 Self-Awareness X Schemata Evaluation Interaction on Mean
Retrieval Latency...................................... 62

10 Self-Awareness X Schemata Congruence X Schemata Evaluation
Interaction on Mean Retrieval Latency....................... 64

11 Experimental Stimuli...................................... 87


LIST OF TABLES


















LIST OF FIGURES


Figure Page


1 Main Effects of Self-Awareness, Schemata Congruence, and
Schemata Evaluation on Mean Retrieval Quantity............. 53

2 Self-Awareness X Schemata Congruence Interaction on Mean
Retrieval Quantity..................................... 55

3 Self-Awareness X Schemata Congruence X Schemata Evaluation
Interaction on Mean Retrieval Quantity.......,.............. 57

4 Main Effects of Self-Awareness, Schemata Congruence,
and Schemata Evaluation on Mean Retrieval Latency.......... 60

5 Self-Awareness X Schemata Congruence Interaction
on Main Retrieval Latency.....................,............ 62

6 Self-Awareness X Schemata Evaluation Interaction
on Main Retrieval Latency.......................,.......... 63

7 Self-Awareness X Schemata Congruence X Schemata Evaluation
on Main Retrieval Latency...........,...................... 65















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




MEMORY RETRIEVAL LATENCY AND QUANTITY
AS A FUNCTION OF OBJECTIVE SELF-AWARENESS
AND SELF-SCHEMATA


By


Cheryl R. Landy

December, 1986


Chairman: Greg Neimeyer, Ph.D.
Major Department: Psychology


The experimental investigation examines the interactive influence

of three control processes on the quantity of and latency to memory

retrieval. Control processes examined are (a) the individual's state of

consciousness (low or high level of objective self-awareness) at

retrieval time, (b) the self-validating or self-invalidating implica-

tions of the recollection, and (c) the good-self or bad-self implica-

tions of the recollection. Undergraduates (n = 113) participated in (a)

an assessment of their self-theories and (b) a retrieval task that con-

cerned life experiences cuedd by self-schemata, which represent self-

perceptions). A 2 x 2 x 2 between-subjects design was used. Each

factor represents a respective control process. Hypotheses concerned

main effects of all factors, two-way interactions involving the first

two factors, and three-way interactions on both retrieval quantity and


viii















retrieval latency. All hypotheses were generally confirmed. Gender was

counterbalanced and, excepting one marginally significant two-way inter-

action, was not a source of variation on dependent values. The results

support the feasibility and effectiveness of using self-schemata as

retrieval cues and implicitly supported their role at the encoding

stage. The preferential access of self-validating experiences occurs

only in the high self-aware state, and conclusions concerning the pre-

ferential access of bad-self information also depend on a consideration

of the level of self-awareness of the recollector. The results indicate

that objective self-awareness, referring to that state of consciousness

in which one is acutely aware of being an object in the world and the

object of one's own consciousness, functions as a context that signifi-

cantly influences persons' process of retrieving life experiences. The

specific manner of its influence was found to depend on the other con-

trol processes. The results indicate that self-awareness governs the

elevating and lowering of thresholds of memory retrieval, resulting in

inhibition and facilitation effects, respectively. Self-awareness was

concluded to inhibit self-deceptive tendencies and to facilitate veridi-

cal and extensive retrieval. One practical implication of that conclu-

sion concerns the induction of that state during personality assessment

in order to increase the predictive validity of the procedure.















CHAPTER I
INTRODUCTION

Introduction to the Investigation

Psychologists long have noted that human consciousness creates and

projects upon itself a sense of having a distinctiveness, unity, and

continuity in identity, a complex projection that is referred to as the

"self" (Jame s, 1890). The self tends to be identified by persons

through the discrimination of its various aspects or characteristics,

which are referred to as self-schemata by contemporary cognitive

theorists (Markus, 1977). Markus and Smith (1981) clarified that self-

schemata are "generalizations or theories about the self, developed from

the repeated similar categorization and evaluation of behavior by one-

self and others, that result in a clearly differentiated idea of the

kind of person one is with respect to a particular domain of behavior"

(p. 240). Persons are necessarily and continually selective in pro-

cessing information pertaining to the self, and self-schemata provide a

framework within which selectivity is systematic and meaningful rather

than random (c.f., Neisser, 1967, 1976).

Self-schemata provide a framework within which a wide range of

information is structured, transformed, and interpreted. Information

encoded according to particular schemata appears to be decoded or

retrieved most efficiently according to the same schemata. If the sche-

mata used at the encoding stage are unavailable or altered at the time

of decoding or retrieval, a loss or reconstruction of the encoded infor-

mation may occur (Tulving & Thomson, 1973).













Much of the information that is processed by persons to develop and

maintain their sense of self is information about past life experiences

and behaviors. The amount and range of information is vast and indefi-

nite given that persons recall life experiences dated to approximately

age four (Crovitz & Quina-Holland, 1976). The information as it is

organized, reconstructed, and recollected at any given moment in time is

an integral constituent of the self. The act of recollecting the infor-

mation of life experiences and behaviors may be considered to be a kind

of selective exposure to the self. Persons, therefore, play an inher-

ently active role in self-definition and self-maintenance.

Particular reconstructions and varying frequencies of recollection

may be a means of elaborating upon a particular aspect of self, repre-

sented by a pole of a bipolar self-schema (e.g., kind-unkind, honest-

dishonest). The organization of memories according to self-schemata is

particularly apparent in persons' tendencies to recall readily many

instances when they acted in a manner to support a particular self-

conception. Kihlstrom and Nasby (1981) noted that the ease or speed of

recollection is, in itself, of significance because it will directly

affect the probability of that information being used in subsequent pro-

cessing. This type of self-referent information appears to be used in

evaluating oneself, in evaluating others, and in determining courses of

action.

Persons do not recall past actions in an exhaustive manner.

Rather, they implicitly and explicitly choose to recall certain exper-

iences, and they implicitly and explicitly select solely one of many

potential frames of reference in which to reconstruct any particular













experience. Selection invariably implies rejection and results in

biased processing, a fact that most persons tend to deny (Erdelyi,

1974). Even the most introspective of persons who attempt to examine

the nature of the selectivity of their memory processing are inherently

limited in their observation of the path and process of retrieval.

They, rather, can be aware solely of the products of their search pro-

cess. The products appear to be largely provided through an unconscious

process, referring to a process of which persons are unaware (Underwood,

1979).

The process of recollecting past behaviors and experiences, if gov-

erned by more general principles of laboratory memory recollection,

should depend on the context in which the process occurs (Tulving,

1983). The context refers to all factors, both from within and without

the individual, present at the time of retrieval. An incredibly diverse

range of contexts has been investigated concerning laboratory memory.

These include cognitive, physiological, motivational, and environmental

(e.g., color of the Laboratory walls) contexts. Scant experimental

research has been directed to contexts that determine the recollection

of life experiences, however.

One context that appears to be centrally meaningful in determining

the retrieval of personal life experiences is the individual's awareness

of his or her identity. This awareness is alternatively referred to as

self-awareness, objective self-awareness, or private self-conscious-

ness. Such self-awareness is considered to be the uniquely human capa-

city by which persons become aware of being an object in the world and

an object of their own consciousness (Duval & Wicklund, 1972).













Differences between persons, and within persons over time, have been

documented in the tendency to spontaneously and volitionally fluctuate

in self-awareness, and these fluctuations have been found to have a pro-

nounced and widespread impact on information processing (c.f., Carver &

Scheier, 1981). The theoretical nature of the mechanisms accounting for

the manifestations is controversial, however, and the impact of self-

awareness as a context of real-life memory retrieval has yet to be

examined.

Self-awareness can be viewed as a retrieval context that signif-

icantly determines the purposes and consequences of memory processing,

noting that persons are assumed to be motivated by anticipated con-

sequences. Memory reconstruction and recollection and, hence, self-

elaboration, is conceptualized to be determined by selective, biased,

and motivated processing in the interest and context of the self as an

object of consciousness. The role of self-related motivational proces-

ses in memory theory is considered undeniable by a diverse range of

theorists, including Rappaport (1971), a prominent psychoanalytic

theorist, and Neisser (1967), a prominent cognitive theorist. In his

classic, influential exposition of the cognitive approach, Neisser

(1967) asserted that the role of motivation could not be even temporar-

ily suspended in the development of memory theory. The purposes and

consequences of particular recollections clearly appear to interrelate

to varying degrees of self-enhancement, self-derogation, self-confronta-

tion, self-elaboration, and self-continuity. The purposes and conse-

quences of recollections for a sense of self should, therefore, depend

on the specific nature and intensity of persons motivations with respect












to self-elaboration and self-maintenance at the immediate moment of

memory retrieval. Self-awareness is, therefore, conceptualized to be a

context that significantly influences the process and content of memory

recollection.



Purpose of the Investigation

The present investigation attempts to examine the interactive

influence of various control processes on persons' recalling of life

experiences. These control processes include (a) the state of con-

sciousness (i.e., specifically, level of self-awareness) at the time of

retrieval, (b) the good-self or bad-self implications of the recollected

experience, and (c) the self-validating or self-invalidating implica-

tions of the recollected experience. The structure and elaboration of

the aforementioned self-schemata interrelate to the latter two processes.



Need for the Investigation

A vast amount of research on memory functioning has been conducted

since Ebbinghaus's (1885/1913) classic work over 100 years ago. During

this time, sophisticated paradigms have been developed in which to study

memory functioning. The research has resulted in sophisticated formula-

tions for explaining the processes of encoding, retaining, and retriev-

ing memories. These theories of memory functioning have, however, been

based on research that has used other than memories of life experiences

as targets of retrieval. The targets of retrieval instead have been

laboratory stimuli including words, figures, pictures, prose, and,

rarely, a more complex experience. As both Underwood (1979) and Neisser













(1982, 1985) have vigorously asserted, the experimental investigation of

memories as continually accessed contents of consciousness or mental

Life has been conspicuously neglected at the expense of the investiga-

tion of memories of laboratory stimuli in the context of the rules of

abstracted verbal learning. This has been due both to the dismissal of

the self in experimental research and to the misperception that experi-

mental rigor and control were difficult to achieve in research where

memories of life experiences were the retrieval targets. Consequently,

very little is known about the course of everyday memory for past life

actions and experiences. Prominent Laboratory memory researchers

increasingly have noted this striking research gap and have asserted

that it need not, and cannot justifiably, continue (Gruneberg, Morris, &

Sykes, 1978; Loftus, 1982; Schacter & Tulving, 1982). The experimental

investigation of the content and process of the recollection of exper-

iences that have occurred outside of the experimental laboratory is

widely regarded as long overdue within mainstream experimental memory

research.

There is also an acknowledged need for research concerning memories

of Life experiences within realms of research outside mainstream

memory. Memories are increasingly being recognized as pervasive

constituents of the ongoing flow of mental life. In fact, researchers

investigating the constituents of mental life have found that persons

spend only one-third of the time focused on the present, and that much

of the remaining two-thirds is focused on memories of life experiences

(Pope, 1978). Memories of life experiences constitute a significant

realm of human functioning that has been ignored as such and undoubtedly













interrelates in yet to be understood ways with cognitive, motivational,

personality, and interpersonal functioning. Psychologists have begun to

acknowledge the need for real-life memory research and have identified

specific directions of research exploration (Bower, 1981; Erdelyi, 1974;

Erdelyi & Goldberg, 1979; Kihlstrom, 1981; Kihlstrom & Nasby, 1981;

Luborsky, Sackheim, & Christoph, 1979). Kihtstrom (1981), a prominent

cognitively oriented personality psychologist, stipulated that memory

needs to be conceptualized as a process rather than a static entity and

recommended that the mechanisms of the process be investigated. Neisser

(1976) clarified the need for a full-fledged theory on memory of life

experiences that accounts for the role of both schemata and motiva-

tion. Kihlstrom and Nasby (1981) have proposed that recollections be

assessed as part of routine clinical personality assessment and that a

research base to help establish a data-gathering and interpretative

method be developed. Erdelyi and Goldberg (1979) and Luborsky et al.

(1979) have clarified the need for understanding the intrapersonal

context of memory retrieval with particular attention to the motivated

decision processes occurring on all levels of awareness that regulate

the raising and lowering of thresholds of cognitions. Cognitive

behavioral psychologists also have delineated the need for the syste-

matic, controlled investigation of the many levels of awareness within

which memories and other cognitions may operate and of the processes

regulating their access and use (Bowers & Meichenbaum, 1984; Clark &

Teasdale, 1982; Erdelyi & Goldberg, 1979; Hersen & Bellack, 1976;

Kendall & Korgeski, 1979; Nisbett & Wilson, 1977; Teasdale & Taylor,

1981).













Thus, both within and without mainstream memory, there is a clear

recognition that memory retrieval is a complex process that is likely to

be affected by (a) the total context present at the time of retrieval

and (b) the nature and use of schemata in the organization and recon-

struction of information. Each of the above variables is already well-

established in laboratory memory research as significantly determining

the content, speed, and quantity of memory retrieval, though their sig-

nificance in the retrieval of memories of life experiences has yet to be

established. There is a need to examine the impact of these variables

in a manner that allows both for the generalization of laboratory memory

experimental phenomenon and for the validation of the highly private,

idiosyncratic, and inherently self-involved nature of memories of life

experiences. An examination of the impact of objective self-awareness

(i.e., as a retrieval context) in interaction with self-schemata on the

content and process of the recollection of life experiences, therefore,

may extend and refine the literature.



Significance of the Investigation

The present investigation contributes to the building of a multi-

level, multiprocess model of memory functioning that encompasses Long-

term memory functioning associated with life experiences. Variables to

be included in the model are identified, and an explanation of the man-

ner in which they interact to impact on the retrieval process is pro-

posed. The ecological relevance of experimental memory research is

extended through the examination of memories of life experiences rather

than of memories of laboratory stimuli, and through the elaboration of













variables and theory related both to the workings of the self and to

states of consciousness (c.f., Hunt, 1985). An additional contribution,

and one that is particularly important given the preliminary state of

memory (i.e., concerning life experiences) research, is the development

of a research paradigm.















CHAPTER II
REVIEW OF THE LITERATURE

Memory Theory and Research

A vast amount of memory research has been conducted over the past

100 years (Ebbi nghau s, 1885 ) in an attempt to build a comprehensive

model of memory functioning that explains and predicts all types of

memory processing, including longterm memory functioning associated with

personal life experiences. A minimum of three levels of storage are

generally accepted: (a) a sensory memory with brief trace duration

(c.f., Sperling, 1960), (b) a short-term memory with limited capacity

(c.f., Broadbent, 1958), and (c) a long-term memory of indefinite

capacity (c.f., Frijda, 1975). Research attention predominantly has

been directed to the sensory and short-term memory. Research that has

been directed to long-term memory has investigated memory of laboratory

stimuli rather than memory of life events since the latter was perceived

to be difficult to examine within the experimental laboratory. Two

major contributions of Laboratory memory research to an investigation of

memory of Life experiences are (a) the development of experimental para-

digms and (b) the identification of robust principles governing memory

functioning.

The language used in memory theory is limited in its usefulness for

memory theory in which the retrieval targets are life events. It is

insufficiently powerful and comprehensive to allow for a description of

a working synthesis of the multiple interactions between the reconstruc-

tion and recollection of life experiences and processes defined as iden-

tity, mo t iva tion attent ion cogni t ion percept ion affect, and













neurology. The incredibly complex and idiosyncratic nature of memory

for experiences that have transpired during an entire lifetime will

require the introduction of theory and language from the personality and

psychoanalytic realms, in particular.

The majority of long-term memory research (i.e., concerning life

experiences) has been conducted within the personality and personality

assessment realms. One dominant research issue has concerned whether

persons recall more effectively pleasant or unpleasant memories (c.f.,

Baddeley, 1983; Matlin & Stang, 1978). In one type of research paradigm

that has been used since the early 1900s, subjects are requested to

construct lists of events from either their recent past, within a range

from one day to several weeks (Thompson, 1982, 1985; Turner & Barlow,

1951), or from their entire lifespan (Henderson, 1911; Meltzer & Ludwig,

1970; Washburn, Giang, Ives, & Pollack, 1925; Washburn, Harding, Simons,

& Tomlinson, 1925). Strongman and Beatson (1981) conducted one of the

more recent studies of this kind. They directed 29 subjects, aged 14 to

68 years, to describe into a tape recorder any personal emotional exper-

iences from their lives.

The aforementioned research has contributed minimally to an under-

standing of memory. This is partially because the experimental task was

of a public self-disclosure nature (i.e., the subjects were required to

disclose their memories to the experimenter). Thus, although the

researchers purported to assess what the subjects were able to, or

tended to, recall of their life experiences, they in actuality assessed

what the subjects tended to disclose publicly. This experimental pro-

cedure has resulted in a lack of theoretical clarity in the memory





literature because the distinction between private self-disclosure and

public self-disclosure is theoretically important and empirically justi-

fied.

Luborsky et al. (1979) recognized the importance of the foremen-

tioned distinction as a consequence of conducting research on fluctua-

tions in memory functioning during psychotherapy. They attempted to

determine the intrapersonal and interpersonal factors that contribute to

patients' memory lapses that occur within psychotherapy sessions. Ver-

batim accounts of psychotherapy sessions were analyzed according to a

theoretically sophisticated system. The variables that correlated with

patients' verbatizations that a memory lapse was occurring (e.g., "I

forgot what I was going to say," "I now can't remember what happened")

were identified. The researchers initially assumed that a verbalization

indicating a memory lapse reflected a true memory lapse but they later

recognized that the former does not necessarily reflect the Latter. The

former may also, or instead, reflect a withholding of the thought or

memory from the therapist and hence reflect a lack of public self-dis-

closure. Their methodology did not allow, however, for an analysis of

the role of solely intrapersonal factors in memory dysfunction. The

empirical analysis of the intrapsychic context of cognition, including

memory, is a largely unexplored, though emergent, area of investigation

(c.f., Bowers &r Meichenbaum, 1984; Erdelyi & Goldberg, 1979; Kihlstrom,

1982; Kihtstrom & Nasby, 1981).

Memory retrieval traditionally has been examined as a rather static

entity within personality assessment lite rature. This research has

examined the relationship between characteristics of earliest













recollections and personality characteristics (Bruhn, 1981; Langs, 1965;

Levy & Grigg, 1962; Mayman, 1968; Purcell, 1952). The most elaborate

study, conducted by Langs (1965), involved the correlational analysis of

62 earliest recollection memory variables with 76 personality varia-

bles. This type of research has established that earliest recollections

are of some diagnostic significance. Further research investigation of

earliest recollections must expand considerably on both the research

hypotheses and the research paradigm, however, to be justified as a sig-

nificant contribution to the literature. Explicit attention must be

directed to an examination of the processes underlying retrieval and to

the dimensions along which memories are organized.

Traditionally, there has been a strong research interest in the

role of the temporal dimension in the encoding and organization of

memory. The notion may be intuitively appealing though research indi-

cates that memories are not intrinsically dated or temporally ordered

(c.f., Underwood, 1977). Research indicates that when persons attempt

to temporally order or date life events they necessarily must conscious-

ly infer that information from idiosyncratic contextual information

associated with the trace (Friedman & Wilkins, 1985; Loftus & Marburger,

1983; Tzeng & Cotton, 1980; Wagenaar, 1986). Thus, the role of schemata

and retrieval context is supported. These variables are discussed in

the next two sections of the literature review.

There is a rapidly developing realm of research concerning memories

of life experiences that attempts to examine the process of retrieval

and the systematic organization of memories and, thus, to validate con-

temporary cognitive principles. In one subset of this realm,





researchers have used a case-study methodology in which the researchers

were the subject (Galt on, 1879, 1911; Linton, 1975, 1978; Wagenaar,

1986; White, 1982). Wagenaar (1986) conducted the most extensive

investigation of this kind. He recorded 2402 events (i.e., 1 or 2

events daily) from his everyday life during a six-year period. All

events were recorded according to five aspects: (a) what the event was,

(b) who was involved, (c) where the event happened, (d) when the event

happened, and (e) an additional critical detail. Recall was cued by

different combinations of the first four aspects and any given event was

recalled only once during the study. Despite the drawbacks of the

experimental design, the results were surprisingly consistent with

existing research and demonstrate the potential utility of the case-

study design in investigating relatively unexplored research territory

such as autobiographical memory. For instance, Wagenaar found the

"what" retrieval cue to be the most powerful single cue and "when" to be

nearly useless. This substantiates the aforementioned conclusion

concerning the temporal dimension. Additionally, he found that errors

made in the dating of salient events were in the direction of recency.

This substantiates the well-known telescoping phenomenon, which refers

to the recency dating error (c.f., Loftus & Marburger, 1983), as well as

a phenomenon recently documented by Brown, Rips, and Shevell (1985).

The latter found that the more that is known about an event, the more

recent the event seems to have occurred.

Another subset of the contemporary cognitive realm of autobiograph-

ical memory research includes studies conducted by Chew and Kihlstrom

(1980), Clark and Teasdale (1982) and Robinson (1976, 1980). Chew and













Kihlstrom (1980) found that personality characteristics were reflected

in memories of life experiences. They, therefore, refined the methodol-

ogy, though not the theory, of the earliest recollection personality

assessment Literature. Chew and Kihlstrom used a procedure for sampling

autobiographical memory based on Galton's (1879, 1911) chronometric

studies of word association and memory. Galton (1911) noticed that a

steady stream of recollections of life experiences came to mind evoked,

he believed, by the succession of objects he observed along the course

of his walk. Galton (1911), using himself as a subject, selected 75

words as stimuli for evoking memory recollection.

Crovitz introduced a variant of Galton's (1911) method into real-

life memory research (Crovitz, 1973; Crovitz & Quina-Holland, 1976;

Crovitz & Shiffman, 1974). The stimulus words used as retrieval cues in

Crovitz' paradigm are selected randomly and have included "meeting,"

"wood," "water," "machine," "shoe," "'window," and "dream" (Teasdale,

Taylor, & Fogarty, 1980). The stimulus words are presented sequentially

to the subject. The subject is typically allotted between 30 to 120

seconds per stimulus word to attempt to recall a Life experience promp-

ted by the stimulus word. The retrieval task is, thus, a projective,

free-associative task. The process and content of memory retrieval are

the dependent variables.

Robinson (1976) refined the above described paradigm by systematic-

ally selecting retrieval cues from syntactic categories. He selected 16

words from each of an activity category (e.g., make, throw, practice),

an object category (e.g., river, window, milk), and an affect category

(e.g., surprised, Lonely, doubt). In a subsequent investigation,





Robinson (1980) extended his investigation of affect words in order that

he could examine the relationship among pleasantness, intensity, and

speed of recall of personal experiences. Robinson (1980) requested that

subjects report the first Life experience recalled in response to a

series of fifteen pleasant or unpleasant affect words.

Memories of life experiences also have been examined in a systema-

tic manner with attention to the process of retrieval in an attempt to

investigate the development and maintenance of depression (Clark &

Teasdale, 1982; Ellis, Thomas, McFarland, &r Lane, 1985; Lloyd & Lishman,

1975; Teasdale & Fogarty, 1979; Teasdale, Taylor, &t Fogarty, 1980). The

paradigm used was the above described projective, free-associative stim-

ulus word paradigm adapted by Crovitz (1973). The independent variable

typically was the mood state of the subject; the dependent variables

typically were the latency to memory retrieval and the hedonic tone

(e.g., happy versus unhappy) of the memories retrieved. Memories were,

thus, conceptualized to be self-referent cognitions that functioned in

the development and maintenance of depression.

The research studies that have been conducted using Crovitz's

(1973) method of providing stimulus words to cue memories of life exper-

iences constitute the most sophisticated longterm memory research to

date (i.e., of that concerning memories of other than laboratory stim-

uli). They provide a background to the present investigation in some of

the very general principles examined, rather than in the specific hypo-

theses. The research clearly demonstrates that the real-life memory

retrieval process can be examined in a systematic manner within the

experimental laboratory. Some sy s tema ti c interrelationships between





characteristics of the retrieval cues, retrieval latency, characteris-

tics of the subject (e.g., personality, gender, and affective state),

and characteristics of the recollections have been observed. Theory

development lags far behind the few systematic relationships that have

been observed, as noted by Neisser (1985). Clearly, an understanding of

the real-tife memory retrieval process is rudimentary.

An understanding of the process seems to be most significantly

advanced by an examination of control processes. This was Johnson's

(1974) conclusion, in his review of the literature on the relationship

between personality and memory. Johnson declared that control processes

should receive primary concern in a model of personality and memory.

Control processes refer to "subject determined processes in memory"

including "both the strategies and techniques used by the subject, and

his decision to use them" (Johnson, 1974, p. 4). Control refers to the

strategies (c.f., Adams, 1985) and decisions by which subjects use their

memories. Control processes include the decision whether or not to

initiate a search, the time taken to make and implement the decision,

the particular organizing and grouping schemata that are implicitly or

explicitly used, the specification of particular criteria as to what

constitutes an adequate retrieval, and the selection of one particular

type of integration and interpretation of the recollected information.

The role of implicit and explicit selection processes and a "top-down"

reconstructive type of processing is, thus, validated. All control

processes presumably occur at varying levels of awareness, notably

including levels outside of conscious awareness. Additionally, differ-

ences across contexts and within individuals are expected to occur in













the use of control processes. This theoretical and experimental

approach is taken in the present investigation. The roles of self-

schemata and objective self-awareness as control processes are described

in the following two sections of the literature review.



Self-Schemata Theory and Research

Memory theorists and researchers (Bartlett, 1932; Schachtel, 1947;

Norman & Bobrow, 1975; Rappaport, 1971), personality theorists and

researchers (KelLy, 1955; Mischel, 1981; Markus, 1977), and cognitive

theorists and researchers (Neisser, 1967, 1976; Piaget & Inhelder, 1952)

have long recognized the importance of understanding the process and

structure of individuals' organization of their psychological world and

have advanced the notion of knowledge structures, referred to generi-

cally as schemata, to describe the active organizational processing of

information at all stages of input and output. Schemata are nonspecific

representations of prior experiences that are elaborated on at every

moment of ongoing mental activity and used to guide the comprehenders

interpretations, inferences, expectations, and attention (Neisser,

1967). Schemata are both necessarily and continually used to impose

order and meaning upon the data of life experiences (Neisser, 1967,

1976). Schemata are highly structured and highly integrated in their

representation of the components, attributes, and relationships that

typically occur in specific exemplars. Neisser (1967, 1976) clarified

that schemata are used in an embedded rather than successive manner in

that various schemata are simultaneously active and hierarchically

embedded in more extensive schemata.













The use of schemata in memory processing and retrieval will be

examined in this section of the literature review. Additionally, the

use of self-schemata in the processing and retrieval of information

pertaining to the self will be reviewed. Finally, the preliminary

research that explicitly investigates the role of self-schemata in the

processing and retrieval of memories of life experiences will be

reviewed.

Bartlett (1932) initiated the experimental investigation of the

role of schemata in memory functioning. He examined subjects recall of

narrative prose passages and he observed lawful omissions, transforma-

tions, reorganizations, and other types of distortions. Bartlett con-

cluded, "Remembering appears far more decisively an affair of construc-

tion than one of mere reproduction" (p. 205) and initiated, what is

referred to as, the constructivist approach to memory functioning. The

constructivist approach validates the use of schemata as organizing

principles in the processing and retrieval of memories.

The use of schemata is particularly apparent in the category clus-

tering phenomenon, which was identified by Bousfield (1953) using a

free-recall paradigm. In any of the many variations of the paradigm,

the experimenter presents information to the subjects and subsequently

requests the subjects to reproduce the information from memory. No

retrieval cues are provided. Bousfield found that when subjects are

presented with a series of randomly ordered words that have been syste-

matically selected from certain commonly identified categories, the sub-

jects regroup the words at retrieval time by the conceptual categories

rather than by the presentation order (c.f., Halpern, 1986). Subsequent













researchers found that the category clustering effect persists even when

the stimulus information is not selected from commonly identified con-

ceptual categories; the subjects then cluster according to their per-

sonal, idiosyncratic categories or schemata. These findings corroborate

Schachtel's (1947) observation that schemata function as memory cate-

gories. The accessibility of schemata specifying conceptual relations

among input information has been found to be a critical determinant of

free-recall performance (Anderson & Bower, 1973; Tulving, 1962; Tulving

& Thomson, 1973).

In a variant of the free-recall paradigm referred to as the inci-

dental free-recall paradigm, a variant popularized by Craik (Craik &r

Lockhart, 1972; Craik &r Tulving, 1975), the subjects are not informed

that they are participating in a memory study until the retrieval

stage. The quantity of recall as well as the specific, biased nature of

the content of recall is assessed. Mattin and Stang (1978) conducted a

multiple regression analysis of 99 of these selective word recall stud-

ies. They found that selectivity is more likely to be demonstrated if

there is a delay, even if only of a few minutes, between word presenta-

tion and word retrieval. Subjects presumably consolidate the target

information with self-schemata to an increasingly greater degree the

more elapsed time between information presentation and information

retrieval.

The use of schemata in a manner to retroactively affect recall of

behavioral personality trait information was demonstrated by Higgins and

King (1981). Undergraduate students were requested to read a paragraph

that ostensibly was about another student. The paragraph contained













evaluatively positive, negative, and ambiguous descriptive informa-

tion. Recall of the information was requested either 20 minutes or 1

week later and was cued by scales with either evaluatively positive or

evaluatively negative labels. Reproduction distortions were signifi-

cantly more positive when subjects responded on scales with positive

labels and significantly more negative when subjects responded on scales

with negative labels. Moreover, there was twice the reproductive memory

distortion after a 1-week delay compared with a 20-minute delay; Mattin

and Stang's (1978) above cited conclusion is thus substantiated.

Undoubtedly, when persons recall self-descriptive incidents that have

occurred during their lifetime, the reproductive distortion is tremen-

dous.

The specific use of schemata is frequently investigated with the

"release from proactive inhibition" (RPI) effect observed in the Brown-

Peterson paradigm. In this paradigm, three words are presented for

encoding but a mental arithmetic task is immediately presented to pre-

vent rehearsal of the words. Brown (1958) and Peterson and Peterson

(1959) independently reported the now robust finding that recall of the

words is significantly impaired after approximately 18 seconds of mental

arithmetic. Subsequent research has indicated, however, that recall is

unimpaired on the first few trials and that impairment occurs only if

all words constitute a single category (Keppel & Underwood, 1962). Pro-

active inhibition builds in a single category so that even a very short

period of mental arithmetic produces significant memory impairment

though it is released with category shifts. In the latter case, memory

for information in the new category is comparable to memory for













information in the first trial of the initial category. If information

from the new category is presented on subsequent trials, proactive

inhibition again builds to be released only by another category shift.

Subjects have been found to categorize stimulus words according to

taxonomic categories (e.g., types of trees or food), sensory modality

categories (e.g., visual or auditory information input modality), and,

relevant to the present investigation, connotative categories including

positive versus negative (Wickens, 1972), masculine versus feminine

(Kail & Levine, 1976), and self-descriptive versus nonself-descriptive

(Nasby, 1980). The RPI research demonstrated that persons continually

and implicitly sort information into categories or schemata for memory

encoding and retrieval.

Considerable research has demonstrated that the schemata that most

efficiently facilitate the retrieval of stored information are the sche-

mata that were used at the time of encoding the target information.

This principle, identified by Tulving and Thomas (1973), is referred to

as the encoding specificity principle. Tulving and Thomas stated, "[the]

specific encoding operations performed on what is perceived determine

what is stored, and what is stored determines what retrieval cues are

effective in providing access to what is stored" (p. 369). Implicit in

this well-validated principle is the notion that the retrieval of stored

information never occurs spontaneously but is initiated always by some

kind of cue, notwithstanding that the instigating cue cannot always be

identified (Tulving, 1976). In the present investigation, schemata were

provided by the experimenter to function as retrieval cues. Specific-

ally, self-schemata, which represents self-perceptions, were provided













since they have been found to be centrally used in the encoding and

decoding of information pertaining to the self (Markus, 1977; Mills,

1983; Kuiper & Rogers, 1979; Rogers, Kuiper, & Kirker, 1977).

Markus (1977) has provided preliminary support for the use of self-

schemata in the encoding and decoding of self-information. Markus clas-

sified subjects as high independent, Low independent, or aschematic

(i.e., the dimension was not very important or self-relevant). In part

of her investigation, she presented to subjects a list of eight indepen-

dent (e.g., individualized, leader) or dependent (e.g., conformist, fol-

lower) adjectives. She requested subjects to identify self-descriptive

adjectives and list reasons to support their choice. She requested the

subjects to cite specific evidence from their past behaviors. Markus

found that high independent subjects supplied, on the average, .93 exam-

ples per independent word, and more examples than did aschematic sub-

jects. This quite preliminary research directly supports the use of

self-schemata in the organization and retrieval of past behaviors.

Another part of Markus's (1977) investigation is relevant to the

present investigation. Markus provided the three groups of subjects

with bogus feedback about their degree of "suggestibility" (p. 73)

ostensibly on the basis of psychological assessment data. Independent

subjects were given dependent feedback (i.e., they were told they were

"suggestible") and dependent subjects were given independent feedback

(i.e., they were told they were "not suggestible"). Self-perceptions

were reassessed following the feedback according to both the extremity

of self-ratings on an adjective checklist and rating latencies. The

rating latency refers to the time taken to rate a particular self-per-

ception. The bogus feedback had no impact on the checklist self-ratings,













though it did impact on the rating latencies. Rating latency increased

for dependent and independent, though not aschematic, subjects on self-

ratings of independence and dependence, respectively. Thus, reaction

time assessment data proved to be a sensitive, discriminating measure of

a change in self-perceptions undetected by traditional adjective check-

list ratings. In the present investigation, self-ratings were assessed

using both checklist and rating latency measures.

Markus's (1977) study, among other studies in which self-schemata

were examined, has been criticized for various reasons related to the

inadequate assessment of schemata and the resultant lack of predictive

validity. As examination of the criticisms clarified that an assessment

of self-schemata in which rating extremity is not confounded with sche-

mata importance, and in which unipolar self-ratings in addition to

bipolar self-ratings are assessed, furthers the definitional, opera-

tional, and predictive precision of the self-schemata concept.

An understanding of the role of schemata in memory processing addi-

tionally entails an examination of the motivational retrieval context,

since schemata are used in a manner that is motivated by the anticipated

consequences of their use (c.f., Bartlett, 1932; Bruner, 1957; Neisser,

1967; Schachtel, 1947). Bartlett (1932) used the term attitude to refer

to the totality of the motivational context in which schemata are

used. He noted: "Here is the significance of the fact .. that when

a subject is being asked to remember, very often the first thing that

emerges is something of the nature of his attitude. The recall is then

a construction, made largely on the basis of the attitude" (p. 207).

Bruner (1957) noted that instances of particular schemata are actively













sought when a person's needs or ongoing activities require the informa-

tion. In pursuit of specific goals (e.g., to form an impression of self

as being a helpful person), instances of the particular self-schema

(e.g., assisting another person with his or her work, giving another

person advice) are actively sought and "come to mind." As Bruner (1957)

pointed out, this type of categorizing is often an implicit or uncon-

scious process. These categorizing and subcategorizing processes are

seen to be an intrinsic part of the multidimensional representation

termed a memory. In the present investigation, objective self-awareness

was viewed as a motivational context that significantly determines the

nature and extent of the elaboration of self-schemata.



Self-Awareness Theory and Research

The capacity for objective self-awareness is considered to be a

uniquely human characteristic. Persons become aware of being an object

in the world and an object of their own consciousness, hence, "objec-

tively" self-aware. Self-awareness, as the term is typically used, and

as the term is used in the present discussion, refers to objective self-

awareness. The terms are, thus, interchangeable. Self-awareness refers

to not simply the tendency to focus on some aspect of the self, but to

the overall awareness of, and reaction to, the duality of consciousness.

The phenomenon of self-awareness has long been recognized within

philosophical and psychological realms, though only recently has it been

recognized within experimental psychology. It was introduced as an

experimental variable by Duval and Wicktund in 1972 and it since has

been justified extensively both as a theoretical construct and as an













empirical variable. The rapidly increasing interest in self-awareness

appears to be due primarily to its pronounced consequences for informa-

tion processing. Self-awareness has been found to impact on information

processing whether the information is externally or internally generated

and whether the information pertains to the environment or to the

self. The consequences of self-awareness are, thus, pervasive. Persons

are active processors rather than passive recipients of information

(Neisser, 1967) and self-awareness is a significant factor affecting the

specific manner of processing. Although a multitude of factors undoubt-

edly affect the manner of processing, self-awareness is of particularly

central significance given its ongoing, fluctuating impact. Persons

fluctuate over time in their level of self-awareness, and they also

differ from each other in their disposition to volitionally induce self-

awareness.

Interindividual or dispositional differences in the tendency to

volitionally induce self-awareness have been detected and assessed.

Fenigstein, Scheier, and Buss (1975) developed a scale of self-con-

sciousness that has one subscale to assess dispositional differences in

private self-consciousness. The effects of high private self-con-

sciousness have been found to parallel the effects of experimentally

heightened objective self-awareness, thereby lending support and cred-

ibility to the overall self-attention construct. Carver and Scheier, in

their summary and integration of the objective self-awareness literature

(1981), concluded "Very few, if any, constructs in social psychology

have been more thoroughly validated than the self-attention construct"

(p. 50).





Many events and variables influence persons' level of self-aware-

ness. Two common events in everyday life that lead to increased self-

awareness, as well as to increased public self-consciousness, are (a)

the experience of eye contact and (b) the presence of an audience or

observer. The presence of physiological arousal may also affect self-

awareness, though researchers disagree on the presence and direction of

impact. Wegner and Giuliano (1980) found that it increased self-aware-

ness; McDonald, Harris, and Maher (1983) found no such increase and

interpreted the former result as an experimental artifact; Diener (1980)

theorized that physiological arousal may reduce self-awareness and

increase deindividuation. Deindividuation presumably lessens private,

internally governed self-regulatory processes (Diener, 1979, 1980).

Deindividuation experimental manipulations include dimmed lights, the

absence of name tags, the absence of addressing a subject by his or her

name, and the wearing of coveralls by subjects to diminish the sense of

identity personal clothes provide. The deindividuation manipulations

were developed in order to diminish subjects' awareness of self as an

individual in the public context of other persons, though subsequent

research has indicated that private self-awareness as well as public

self-awareness is diminished. Alcohol intake also has been found to

diminish self-awareness (Hull, 1981; Hull, Levenson, Young, & Sher,

1983; Hull & Young, 1983); alcohol consumption was found to inhibit the

encoding of self-relevant information in dispositionally high private

self-conscious male social drinkers to the extent that their responses

were similar to the responses of dispositionally low private self-con-

scious drinkers. Finally, hypnotic states have been hypothesized to













affect self-awareness, though no global statement can be made concerning

the direction of impact since hypnotic states are not considered to be

unitary or consistent states (Sheehan, 1979; Tart, 1975).

Wicklund and Duval (1971; Duval & Wicklund, 1972) were the first

researchers to intentionally and systematically vary subjects' level of

self-awar eness. They chose several manipulations, including a mirror

and a camera, that intuitively served as reminders of the self. Many of

their manipulations were found to enhance states, notably public self-

consciousness, in addition to self-focus, however. Public self-con-

sciousness refers to the state in which persons become aware of them-

selves as an object to others, thus, of their publicly displayed self-

aspects. The task of researchers following the lead of Wicklund and

Duval, therefore, was to validate a self-awareness manipulation that

solely heightened subjects' level of private objective self-awareness.

A mirror has been identified as the optimal experimental manipulation.

A considerable body of research evidence, reviewed by Carver and Scheier

(1981), has supported its construct and discriminant validity. Differ-

ences between persons in modes of information processing become more

pronounced. Significant, often extreme, effects in cognitive, affec-

tive, and behavioral domains have been demonstrated.

Gibbons (1978), for example, found that objectively heightened

self-awareness during exposure to erotica appeared to cause persons to

report reactions to sexual material that were more congruent with per-

sonal standards. Female subjects were selected on the basis of high or

low scores on Mosher's (1968) sex guilt scale and the dependent measure

was subjects' ratings of how sexually arousing, enjoyable, and well-

written passages from a pornographic novel were. Correlations between





subjects' pretest sex guilt scores and their ratings of the passages

were strong (r=.-57) in the high self-awareness (i.e., mirror presence)

condition and much weaker (r.0)in the low self-awareness (~.

mirror absence) condition. High guilt women reported considerably less

enjoyment and arousal under conditions of high self-awareness. Objec-

tive self-awareness, thus, impacted upon self-ratings of even self-per-

ceptions, ratings that might seem beyond the boundary conditions of

heightened self-awareness. The self-awareness literature has buttressed

theory concerning the dominant, though fluctuating, impact of the self-

system on the processing of information pertaining even to the self.

Another study, which was conducted by Carver (1975), is exemplary

of the many studies that have demonstrated the impact of heightened

self-awareness on behavioral self-regulation (Beaman, Klentz, Diener, &

Svanum, 1979; Carver, Blaney, & Scheier, 1979). Subjects were selected

on the basis of their high punitive or low punitive attitudes toward the

use of physical punishment as a teaching technique and on the basis of

their prediction concerning whether they would use punishment in this

manner. Several weeks later, in an ostensibly unrelated experiment, the

two groups of subjects were recruited to participate in an investigation

of the effect of physical punishment on learning. The severity of shock

level chosen by subjects to punish the confederate subject when the lat-

ter emitted incorrect responses was the dependent measure. In the low

self-awareness (i.e., mirror absence) condition, no differences between

the two groups were manifested. In the high self-awareness (i.e., mir-

ror presence) condition, however, clear differences between the two

groups were apparent: High punitive subjects chose even more intense





shocks and Low punitive subjects chose even less intense shocks. Sub-

jects presumably used their attitudes and self-concepts in choosing

shock levels. This study is exemplary of research that has demonstrated

that the impact of self-awareness is not global, but differentiated as a

function of the self-standards of the subject.

Although the impact of objective self-awareness on information pro-

cessing has been demonstrated in a variety of realms, the boundary con-

ditions (iethe depth and breadth) of its impact require further

delineation. The generalization from one realm of self-processing to

another is not necessarily direct. Specifically, the impact of self-

awareness on the process of reconstructing and recollecting life exper-

iences has yet to be investigated. Carver and Scheler (1981) noted that

a study conducted by Pryor, Gibbons, Wicklund, Fazio, and Hood (1977) is

the most direct research support of the relationship between self-aware-

ness and memory recollection. In that investigation, the impact of

objective self-awareness on the truthfulness of undergraduate students'

report of two objectively verifiable public facts was examined. The

public facts were (a) their high school combined SAT scores and (b)

their percentile ranking on their first psychology exam. The truthful-

ness of self-report of the SAT score was higher under the high self-

awareness (i.e., mirror presence) condition, at least for low scoring

subjects since high scoring subjects were truthful in either the high

self-awareness condition or low self-awareness condition. Comparable

differences were not found concerning the exam percentile data, but

because exam scores were never given to subjects in percentile terms

they were unlikely to have been encoded in that manner.













A close examination of the theoretical and empirical self-awareness

literature revealed that self-awareness is a significant phenomenon; its

impact is pronounced in a diverse range of processing realms. Consider-

able theoretical controversy was apparent concerning the manifestations

of its impact and concerning the nature of the mechanisms that account

for its impact. The various theories of self-awareness tend to differ

from each other in the relative attention directed to motivational pro-

cesses and to the relative acknowledgment of the various levels of con-

scious awareness at which information processing occurs (c.f., Carver &

Scheier, 1981; Duval & Wicklund, 1972; Holzman & Rousey, 1966; Hull,

1981; Hull & Levy, 1979; Sackheim & Gur, 1978). There is some consen-

sus, however, that self-awareness results in a heightened sensitivity to

self-standards (Carver & Scheier, 1981; Lewicki, 1983), and in a concom-

itant striving to reduce the discrepancies through various strategies.

In the present investigation, awareness of self as an object in the

world and as an object of ones' own consciousness is considered to func-

tion in a screening, regulating, monitoring manner in the processing of

information, specifically, memories of life experiences. Persons are

presumably guided by implicitly private concerns in processing self-

information. This type of self-regulation is distinctly different from

a public, interpersonal type of self-regulation where self-presentation

to other rather than to self is the control process; this important dis-

tinction is supported by theory and research (Fenigstein et al., 1975).

In the present investigation, self-awareness is hypothesized to be

a retrieval context that selectively influences persons' process of

retrieving memories of life experiences. When persons are highly self-

aware, they are expected to be more motivated to orient toward, select,













and elaborate on particular types of memories and also more motivated to

avoid, ignore, and not elaborate on other types of memories. Self-

awareness is, thus, hypothesized to have both facilitation and inhibi-

tion effects presumably attributable to an intensification of self-

regulatory efforts. The specific influence of self-awareness is hypo-

thesized to depend on the self-validating or self-invalidating implica-

tions of the recollection. Certain memories reflect an aspect of self

that one identifies with while others reflect an aspect of self that one

disowns. In the experimental paradigm, the schemata congruence variable

represents the self-validating or self-invalidating implications,

respectively. The specific influence of self-awareness is hypothesized

to depend also on the evaluative (positive, negative) implications of

the recollection for the self-theory. Certain memories reflect a fav-

orable attribute while others reflect an unfavorable attribute. Thus,

self-awareness (low, high), schemata congruence (congruent, incongru-

ent), and schemata evaluation (positive, negative) are examined insofar

as they simultaneously interact to influence two aspects of the memory

retrieval process: (a) retrieval quantity and (b) retrieval latency.

Self-schemata, which reflect self-perceptions that are hierarchically

organized within an overall self-theory, function as the retrieval cues

in the directed memory retrieval task.
















CHAPTER III
METHODOLOGY

Introduction

The operational research question in the present experimental

investigation was as follows: How, if at all, does the level of objec-

tive self-awareness interact with the nature of the memory recollection

task to influence the memory recollection process? The independent

variables schemata congruence and schemata evaluation were used to

manipulate the nature of the memory recollection task. Schemata

congruence refers to the self-congruence or self-incongruence of the

schemata used as retrieval cues. Schemata evaluation refers to the

positive or negative evaluative nature of the schemata used as retrieval

cues. A memory, thus, could be either positive or negative and either

self-congruent or self-incongruent. The impact of the variables was

assessed on two dependent variables: (a) retrieval latency and (b)

retrieval quantity. The experimental paradigm, thus, entailed three

independent variables and two primary dependent variables. A

description of these variables follows.



Research Design

A 2 x 2 x 2 factorial design was employed. The three between-

subjects factors were self-awareness, schemata congruence, and schemata

evaluation. Self-awareness had two levels (low, high); schemata con-

gruence had two levels (congruent, incongruent); schemata evaluation had

two levels (positive, negative). Thus, the eight experimental condi-

tions are identified as (a) Low self-awareness, congruent schemata,













positive schemata; (b) low self-awareness, congruent schemata, negative

schemata; (c) low self-awareness, incongruent schemata, positive

schemata; (d) low self-awareness, incongruent schemata, negative

schemata; (e) high self-awareness, congruent schemata, positive

schemata; (f) high self-awareness, congruent schemata, negative

schemata; (g) high self-awareness, incongruent schemata, positive

schemata; and (h) high self-awareness, incongruent schemata, negative

schemata. Gender was counterbalanced in the experimental design. The

two dependent measures were retrieval latency and retrieval quantity. A

three-way analysis of variance (ANOVA) constituted the primary data

analyses.



Independent Variables

Objective Self-Awareness

Persons differ across time and situation, and from each other, in

their awareness of self as an object in the world as an object of their

own consciousness. These differences in objective self-awareness have

been found to have a pronounced impact on emotions, cognitions, and

behaviors. In the present investigation, differences in objective self-

awareness were effected using the traditional mirror paradigm. In this

paradigm, an ordinary mirror is placed directly in front of the subject

in the high self-awareness condition. No mirror is present in the low

self-awareness condition. The experimenter does not mention the pres-

ence of the mirror and subjects reportedly do not express suspicion

concerning its presence (Carver & Scheier, 1981). In the present inves-

tigation, objective self-awareness was a between-subjects factor.













The mirror is physically placed in a manner that it cannot be per-

ceived as a two-way mirror. In the present investigation, the mirror

was, therefore, obviously leaned against the wall. A two-way mirror

implies observation by others and heightened awareness of the public

aspect of the self and is, therefore, not solely a private self-aware-

ness research manipulation. Considerable research evidence (c~f.,

Carver & Scheier, 1981) supports the distinction between public self-

awareness and private self-awareness as both theoretically meaningful

and empirically valid. This research evidence strengthens both the

discriminant validity of the mirror as a private self-awareness manipu-

lation and the construct validity of the private self-attention con-

struct. This research evidence is, however, only part of the large body

of research evidence in support of the mirror manipulation as a private

self-awareness manipulation and of the overall private self-attention

construct. The section "Self-Awareness Theory and Research" provided a

review of objective self-awareness both as a theoretical construct and

as an empirical variable.



Schemata Congruence

Certain recollections pertaining to the self are congruent with

self-perceptions. Presumably, retrieving this information should vali-

date a self-theory. Certain recollections pertaining to the self are

incongruent with self-perceptions. Presumably, retrieving this informa-

tion should invalidate a self-theory. The congruent or incongruent

nature of the memory recollection task was manipulated by the indepen-

dent variable, schemata congruence. Schemata congruence was a between-

subjects factor with two levels: (a) congruent and (b) incongruent.





Each subject was presented with four schemata that were all either

congruent or incongruent with subjects' self-perceptions on particular

bipolar dimensions within their self-theory. For example, if a subject

rated "shy" as self-descriptive, then "shy" was categorized as schema-

congruent. If a subject rated "shy" as not self-descriptive, however,

it was categorized as schema-incongruent. The determination of stimuli

for the schemata congruence variable was individualized and derived from

subject's self-ratings. Appendix A contains the experimental pool of

schemata, and the section "Determination of Schemata" describes the

schemata derivation process.



Schemata Evaluation

Certain recollections pertaining to the self reflect an aspect of

self that is evaluatively positive. Certain recollections pertaining to

the self reflect an aspect of self that is evaluatively negative. Sche-

mata evaluation was a between-subjects factor with two levels: (a) posi-

tive and (b) negative.

Each subject was presented with four schemata that were all of

either a positive or a negative evaluative nature. For example, "truth-

ful" is of a positive aspect evaluative nature and "untruthful" is of a

negative evaluative nature. The determination of the evaluative nature

of the trait was taken from Anderson's (1968) normative data (see Appen-

dix A). Anderson's (1968) Likableness ratings were used. The higher

rated trait of the bipolar trait pair was selected to function as the

positive schema.





Determination of Schemata

The entire assessment of a subject's self-theory was conducted on a

Commodore 64 computer. The subject completed various rating tasks,

which are described Later in this chapter (i.e., schemata ordination

ratings, schemata self-description ratings). The computer was pro-

grammed to (a) hierarchically organize all traits according to their

importance in a subject's self-theory, (b) sort the subject's self-

ratings into either a congruent or an incongruent category, and (c) sort

the traits into either a positive or a negative category. The computer

was programmed to then select, from those traits ranked superordinate in

importance, four schemata that were (a) either congruent or incongruent

with a subject's self-perceptions and (b) of either a positive or a

negative evaluative nature. The assigned experimental condition determ-

ined which selections were made, noting that the computer was programmed

to randomly assign the experimental condition and proceed in the determ-

ination of schemata accordingly.

Random selection processes occurred concerning the selection of the

four most superordinate schemata in those instances in which a subject

rated more than four traits at a given ceiling rating (i.e., in the

schemata ordination ratings). Similarly, if only three traits were at a

given ceiling rating then the fourth trait was randomly selected from

among those at the next highest rating. Note that subject A's ceiling

rating may be a "6" and subject B's ceiling rating may be a "5." There

is, thus, not necessarily consistency across subjects concerning the

absolute ceiling rating though there is consistency across subjects con-

cerning the relative position of the ceiling rating. Within any





particular subject, the schemata presented as retrieval cues were always

those traits in the particular category that the subject perceived to be

most important relative to all other traits. Across-subject consistency

concerning the absolute ceiling rating was not built into the computer

program because a subject selection bias would have invariably occurred

(e.g., all subjects with less than 4 traits rated a "6," for instance,

would be barred from participation).

The determination of schemata congruence was made on the basis of

subjects' ratings on the self-description ratings task. The computer

was programmed to classify a trait as schema-congruent if a subject

rated the trait as self-descriptive (i.e., a rating of "3," "4," or "5")

and as schema-incongruent if a subject rated the trait as not self-

descriptive (i.e., a rating of "1," "2," or "3"). The computer was pro-

grammed such that the degree of schemata ordination prevailed over the

degree of schemata congruence or incongruence; the latter was not con-

trolled for because of the inevitable subject selection bias that would

have resulted. Consequently, a moderately congruent schema (e.g.,

rating of "5" on the trait trustworthy) on a superordinate dimension

(e.g., the trustworthy-untrustworthy dimension was the most important

dimension) was selected over a very congruent schema (e.g., rating of

"6" on the trait outgoing) on a less important or unimportant dimension

(e.g., the outgoing-shy dimension is among the Least important dimen-

sions). The determination of schemata evaluation was based on normative

data (see Appendix A) and, thus, was independent of a particular sub-

ject's ratings. The evaluative valence was built into the trait in the

programming and that information was used simultaneously in conjunction

with information pertaining to ordination and congruence in the deriva-

tion of schemata.













Dependent Variables

Retrieval Latency

Retrieval latency refers to the time elapsed between the presenta-

tion of the retrieval cue and the subject pressing a key on the computer

indicating that a memory had been retrieved. The latency associated

with each retrieval was recorded by the computer. The mean retrieval

latency constituted the dependent variable. Retrieval latency was

assessed in half-seconds since this level of discrimination has proven

sufficient in memory research. It has been demonstrated that subjects

take approximately 6 to 15 seconds to retrieve a memory of a life exper-

ience (Clark & Teasdale, 1982; Robinson, 1980). This is an extremely

large range in the context of laboratory memory and information proces-

sing literature where milliseconds is the normative Level of discrim-

mnation.

Retrieval latency has a well-established tradition and validity as

a measure of the accessibility of memories in experimental memory

research. Within nonmemory research, but self-schemata research more

generally, reaction-time measures are also quickly becoming widely

recognized as sensitive, discriminating dependent variables that detect

intrasubject and intersubject differences undetected by more traditional

measures (Markus, 1977; Turner, 1978b; Mills, 1983).





Retrieval Quantity

Retrieval quantity refers to the total number of memories

retrieved. The subject was directed to press a key on the computer each

time a memory within the specified category was retrieved; the quantity

retrieved was recorded by the computer. Retrieval quantity has a well-

established tradition and validity as a dependent variable in experi-

mental memory research.



Research Hypotheses

1. A main effect of level of schemata evaluation on retrieval

quantity. Subjects in a positive schemata condition demonstrate an

increased retrieval quantity compared with subjects in a negative sche-

mata condition.

2. A main effect of level of schemata evaluation on retrieval

latency. Subjects in a positive schemata condition demonstrate a

decreased retrieval Latency compared with subjects in a negative sche-

mata condition.

3. A main effect of level of schemata congruence on retrieval

quantity. Subjects in a congruent schemata condition demonstrate an

increased retrieval quantity compared with subjects in an incongruent

schemata condition.

4. A main effect of level of schemata congruence on retrieval

latency. Subjects in a congruent schemata condition demonstrate a

decreased retrieval Latency compared with subjects in an incongruent

schemata condition.





5. A main effect of level of self-awareness on retrieval quant-

ity. Subjects in a high self-awareness condition demonstrate an

increased retrieval quantity compared with subjects in a low self-

awareness condition.

6. A main effect of level of self-awareness on retrieval

latency. Subjects in a high self-awareness condition demonstrate a

decreased retrieval Latency compared with subjects in a low self-

awareness condition.

7. A two-way interaction between level of self-awareness and

Level of schemata congruence on retrieval quantity. Subjects in a con-

gruent schemata condition demonstrate an increased retrieval quantity if

they also are in a high self-awareness condition compared with a low

self-awareness condition. Subjects in an incongruent schemata condition

demonstrate a decreased retrieval quantity if they also are in a high

self-awareness condition compared with a low self-awareness condition.

8. A two-way interaction between Level of self-awareness and

level of schemata congruence on retrieval latency. Subjects in a con-

gruent schemata condition demonstrate a decreased retrieval latency if

they also are in a high self-awareness condition compared with a low

self-awareness condition. Subjects in an incongruent schemata condition

demonstrate an increased retrieval Latency if they also are in a high

self-awareness condition compared with a low self-awareness condition.

9. A three-way interaction between level of self-awareness,

Level of schemata congruence, and level of schemata evaluation on

retrieval quantity. Subjects in a schemata condition that is both

congruent and positive demonstrate a particularly increased retrieval













quantity if they also are in a high self-awareness condition compared

with a low self-awareness condition. Subjects in a schemata condition

that is both incongruent and negative demonstrate a particularly

decreased retrieval quantity if they also are in a high self-awareness

condition compared with a low self-awareness condition.

10. A three-way interaction between level of self-awareness,

level of schemata congruence, and level of schemata evaluation on

retrieval latency. Subjects in a schemata condition that is both con-

gruent and positive demonstrate a particularly decreased retrieval

latency if they also are in a high self-awareness condition compared

with a low self-awareness condition. Subjects in a schemata condition

that is both incongruent and negative demonstrate a particularly

increased retrieval latency if they also are in a high self-awareness

condition compared with a low self-awareness condition.



Data Collection

Subjects

Description

One hundred and thirteen (n = 56 male, n = 57 female) subjects

participated in the experiment. The modal age of the subjects was 18

years (n = 38); the mean age was 20.64 years. Eighty-two percent of the

subjects were white, 6% were black, and 5% were of Hispanic ethnic

background. Forty-eight percent of the subjects were first-year stu-

dents, 20% were second-year students, 18% were third-year students, 10%

were fourth-year students, and 2% were post baccalaureate students.

Twenty-one percent of the subjects were undecided concerning their













college major, 13% were business majors, 13% were journalism majors, 11%

were psychology majors, and 9% were health and physical education

majors. Groups of subjects each comprising less than 5% of the total

sample represented a wide range of other college majors.



Sampling procedures

The subjects were recruited from introductory psychology classes.

Credit in partial fulfillment of the psychology course requirement was

provided to participants; monetary compensation was not provided.

Potential subjects were informed that the investigation concerned the

organization of memories of life experiences. They were also informed

that they would not be required to disclose the content of any recollec-

tion.



Experimental Procedures

Overview

The subjects were randomly assigned to the experimental condi-

tions. Separate randomization schedules were used for male and female

subjects because gender was counterbalanced in the design. All subjects

were assured of anonymity by the use of a numerical identification sys-

tem. All subjects were run individually. The subjects took approxi-

mately 40 minutes to complete the experiment. Participation involved

completing three primary tasks: (a) schemata ordination ratings, (b)

schemata self-description ratings, and (c) memory retrieval. The tasks

are described below. All tasks were completed on a Commodore 64 compu-

ter; an 11 inch screen was used. The experimenter was not present in





the experimental room (and the door to the room remained closed) during

task completion to assure the subjects of privacy. At the completion of

all tasks, the experimenter addressed inquires concerning the investiga-

tion and requested that the subject not discuss the experiment with

potential subjects.



Schemata ordination ratings

After reading and signing the informed consent form (see Appendix

B), the subjects were seated in front of the computer and given practice

to an equal response latency (i.e., within 500 milliseconds) in pressing

one of six keys numbered from 1-6. The stimulus was the 6-point rating

scale used in all subsequent rating tasks. The rating scale was in the

same screen position for all rating tasks. As a first primary task, the

subjects rated the importance of 28 bipolar trait pairs. The directions

to this schemata ordination rating task were presented on the screen as

follows:

Persons form impressions of others according to various
dimensions such as ambitious-lazy or polite-impolite.
Persons differ concerning how important they believe
any particular dimension is in forming an impression of
another. One dimension at a time will be presented on
the screen. Please rate each dimension according to
how important you believe it is in forming an impres-
sion of a person. The rating scale is:

not at all .. .. very
important 1 2 3 4 5 6 important.

Rate according to your own private opinion.
You will press a key from 1-6.

Twenty-eight pairs of traits were presented, one pair at a time.

The presentation order was randomized across subjects. Unbeknownst to

the subjects, the initial three pairs functioned solely to orient them





to the task. The three practice pairs (i.e., ambitious-lazy, forgiving-

unforgiving, and conformist-nonconformist) were taken from the same

stimulus pool as the subsequent twenty-five pairs (see Appendix A).

Each pair remained on the screen until the subject pressed a rating key;

the next pair then was presented on the screen. The computer recorded

the pair of traits, the associated rating, and the associated rating

latency (in milliseconds). The task functioned to assess subjects'

ordination of schemata on an individualized basis in order that traits

ranked as most important could be selected to function as schemata in

the subsequent memory retrieval task. At the completion of the ratings,

a screen message said, "Please get the experimenter now," so that the

experimenter could verify that the directions to the self-description

ratings task were understood.



Schemata self-description ratings

As a second task, the subjects rated themselves along a set of 53

traits. The directions to this schemata self-description rating task

were presented on the screen as follows:

Now, please rate each of the following traits according to
the degree that you privately believe that the trait
describes you. The rating scale is:

not me .. .. me
---1 2 3 45 6--

You will press a key from 1-6.

The fifty-three traits were presented, one trait at a time. The

presentation order was randomized across subjects. Unbeknownst to the

subjects, the initial three traits functioned solely to orient them to

the task. The three practice traits (i.e., logical, unforgiving, and





impolite) were taken from the same stimulus pool as the subsequent 50

traits (see Appendix A). Each trait remained on the screen until the

subject pressed a rating key, which advanced the next trait onto the

screen. The computer recorded the trait, the evaluative valence, the

associated rating, and the associated rating latency (in milliseconds).

Thus, an individualized data base was created for each subject from

which the schemata congruence could be determined. At the completion of

the ratings, a screen message said, "Do not press any key at thi s

time. Please get the experimenter. Thank-you," so that the experimen-

ter could verify that the directions to the memory retrieval task were

understood.



Memory retrieval

As a third task, the subjects recalled memories of life experi-

ences. The directions to the memory retrieval task were next presented

as follows:

A trait will be presented on the screen. Your task is
to attempt to recall specific incidents in your life
when you exemplified or demonstrated the particular
trait. Press the red key each time you recall an inci-
dent. Please note the following guidelines:

(a) You yourself must have exemplified the trait.

(b) The incident may have occurred quite recently or
many years ago.

(c) You must be able to recall something that makes
the incident a distinct memory. If the same type
of incident happened more than once, press the
key for each incident only if you can recall
something that makes you certain that the other
incident happened on a different occasion.





(d) It doesn't matter if another would agree with you
as to whether the incident "counts" -- your opin-
ion is all that matters.

(e) Press the red key as soon as you recall an inci-
dent, but not before.

(f) It is important that you keep trying to recall
incidents during the allotted time. A total of 4
traits will be presented. The computer is pro-
grammed to give you 2 1/2 minutes per trait and a
30-second rest period between traits. You will
know when time is up for a particular trait
because "STOP" messages will appear all over the
screen.

The written copy of the directions remained with the subjects so

that they might reread them before they pushed a key to initate the

retrieval task. The subjects were requested to wait until the experi-

menter left the room and closed the door before they initiated the

retrieval task. The experimenter did not cite any examples of task com-

pletion in order to prevent priming effects. Additionally, the experi-

menter did not establish a standard for the subjects to recall "as many

incidents as possible."

The subjects were also given a secondary task. They were requested

to jot down no more than a word or so about each incident (subsequent to

pressing the red key). They were provided with 4 blank sheets of

untined paper (one for each trait) and several pencils. The subjects

were informed: "(a) the task is secondary in importance to the memory

retrieval task, (b) don't write down the trait that was presented -- the

computer will select the trait and the experimenter need not know which

trait was selected, (c) jot down only enough so that you yourself can

identify the incident for the purpose of answering an objective, stan-

dardized question about the incident after the retrieval task is over,





and (d) no one will ever see your notes -- you'll discard them." The

question that the subjects were to answer concerned the date of the

recollected incident. The subjects were not informed of this, however,

to prevent any retrieval bias.

The computer was programmed to administer the entire memory

retrieval task. One schema was presented every 3 minutes. A total of

four schemata were presented. Each schema remained on the screen for

2.5 minutes at which time "STOP" messages appeared on the screen and the

screen changed background colors. Any key presses that may have been

made during the .5 minute intertrial interval were not counted. The

computer was programmed to record the retrieval quantity and the

retrieval latency.

The retrieval task is a variant of the traditional category gen-

eration task. In that kind of task, the subject is provided with a

category such as "animals," and is requested to produce as many items

from that category as possible within a standard period of time

(Bousfield & Sedgewick, 1944). The task is most frequently used to

examine the organization of semantic memory. Semantic memory, as

distinguished from episodic memory by Tulving (1972), refers to memory

for stored information of an impersonal, general nature such as the

capital of France, the dictionary definition of a word, or the colors of

the rainbow. Episodic memory receives and stores information about

temporally dated episodes or events, and temporal-spatial relations

among these events. It is always stored in terms of its autobiograph-

ical reference to the already existing contents of the episodic memory

store (p. 385). The emphasis of episodic memory, the memory of interest





in the present investigation, is on the subjective flavor of the exper-

ience and the meaning and personal significance of the information for

the target individual.

Subsequent to the completion of the entire memory retrieval task,

the subjects were requested to date each recollected event as to month

and year of occurrence. This is normative data in real-tife memory

research. The event-age then may be calculated by the experimenter.

Event-age refers to the number of months from the present time of memory

retrieval to the time at which the recalled event occurred (Robinson,

1976). Robinson (1976) analyzed the test-retest reliability of event

age dating. He based his analysis solely on information pertaining to

month and year of occurrence. He assessed the consistency of event-age

dating on an individual basis by correlating the event-age reported in

session one with the event-age reported in session two, which occurred

one week later. The product-moment correlations obtained were uniformly

high, averaging +.94. The modal error in redating was 12 months, indi-

cating that subjects were more readily confused about specific year of

occurrence than they were regarding relative time within a calendar

year. Additionally, discrepancies were as likely to antedate as to

postdate the event-age reported in session one. This principle was val-

idated for events reported as occurring within a few weeks of the exper-

imental sessions as well as for remote incidents (Robinson, 1976).













CHAPTER IV
RESULTS

Preliminary Tests for Gender Effects


Four-way ANOVAs were conducted (iegender as the fourth

independent variable in 2 X 2 X 2 X 2 designs) to discern any effects of

gender on variation in mean retrieval quantity or latency values. The

results indicated no significant main effects of gender and no signifi-

cant two-, three-, or four-way interactions involving gender, excepting

one marginally significant (F = 3.58; p < .062) Gender X Schemata

Evaluation interaction on retrieval quantity values. Table 1 provides

the concomitant mean retrieval quantity and standard deviation values.

Note that all reported statistical analyses were conducted using the

second edition of the Statistical Package for the Social Sciences (SPSS)

(Nie, Hull, Jenkins, Steinbrenner, & Bent, 1975).


Table 1

Gender X Schemata Evaluation Interaction on Mean Retrieval Quantity



Experimental Mean retrieval Standard
condition quantitya deviation


Positive
Femaleb 30.10 16.45
Malec 24.79 13.52
Negative
Femaled 16.81 8.88
Malec 18.93 10.43


a n. of memories.
bn = 29.
cn = 28
d -)
n = 27











































Source of Sum of Df Mean F g
variation squares square


Main effects 6467.08 3 2155.69 19.89 .000
Self-awareness (SA) 1381.50 1 1381.50 12.75 .001
Schemata congruence (SC) 2476.15 1 2476.15 22.85 .000
Schemata evaluation (SE) 2629.57 1 2629.57 24.27 .000
Two-way interactions 1642.57 3 547.52 5.05 .003
SA X SC 1532.16 1 1532.16 14.14 .000
SA X SE 24.82 1 24.82 .23 .633
SC X SE 89.14 1 89.14 .82 .367
Three-way interaction 1125.32 1 1125.32 10.38 .002
SA X SC X SE 1125.32 1 1125.32 10.38 .002
Explained 9234.97 7 1319.28 12.17 .000
Residual 11378.80 105 108.37
Total 20613.76 112 184.05


Effects on Retrieval Quantity


Table 2 summarizes the results of the 2 X 2 X 2 between-subjects

ANOVA. The first factor reflected the levels of self-awareness (low,

high), the second factor reflected the Levels of schemata congruence

(congruent, incongruent), and the third factor reflected the levels of

schemata evaluation (positive, negative). Mean retrieval quantity was

the dependent variable.




Table 2

ANOVA: Effects of Self-Awareness, Schemata Congruence, and Schemata
Evaluation on Mean Retrieval Quantity













Main Effects

Main effects on retrieval quantity occurred for all three fac-

tors. The effects indicated a greater retrieval quantity (a) in the

high compared with the low level of self-awareness, (b) in the congruent

compared with the incongruent level of schemata congruence, and (c) in

the positive compared with the negative level of schemata evaluation.

Table 3 presents the mean retrieval quantity values and concomitant

standard deviation values for each of the two levels of each of the

three factors. Figure 1 depicts the main effects.



Table 3

Main Effects of Self-Awareness, Schemata Congruence, and Schemata
Evaluation on Mean Retrieval Quantity



Experimental Mean retrieval Standard
condition quantitya deviation


Self-awareness
Lowb 19.32 10.25
Highc 26.14 15.62
Schemata congruence
Congruentb 27.35 15.70
Incongruentc 17.96 8.86
Schemata evaluation
Positiveb 27.49 15.18
Negativec 17.82 9.60


a no. of memories.
bn = 57.
c n = 56.










































Self- Schemata Schemata
Awareness Congruence Evaluation





Figure 1

Main Effects of Self-Awareness, Schemata Congruence, and Schemata
Evaluation on Mean Retrieval Quantity


Congruent


High


Mean
Retrieval
Quantity
(no. of
memories)


Two-way Interaction

Self-awareness interacted with schemata congruence to influence

retrieval quantity. The interaction indicated a greater retrieval

quantity in the congruent compared with the incongruent level of

schemata congruence only in the high compared with the low level of


Positive





Self-Awareness X Schemata Congruence Interaction on Mean Retrieval
Quantity


self-awareness. Table 4 presents the mean retrieval quantity values and

concomitant standard deviation values for each of the four Self-

Awareness X Schemata Congruence experimental cells. Figure 2 depicts

the interaction.


X Congruent Schemata
O Incongruent Schemata


38-


34-


Mean Retrieval
Quantity (no.
of memories)


O O


Low


High


Level of Self-Awareness


Figure 2














Table 4

Self-Awareness X Schemata Congruence Interaction on Mean Retrieval
Quantity



ExperientalMean retrieval Db Standard
condition quantitya deviation


Low self-awareness
Congruent schematac 20.41 2.23 10.41
Incongruent schematad 18.18 10.13
High self-awareness
Congruent schematae 34.54 16.79 17.14
Incongruent schemated 17.75 7.56


a no. of memories.
b refers to the difference between the congruent and incongruent means
in the particular level of self-awareness.
C n = 29.
d ,,
n = 28


Three-way Interaction

Self-awareness interacted simultaneously with both schemata

congruence and schemata evaluation to influence retrieval quantity. A

particularly increased retrieval quantity occurred in the condition that

represented both congruent and positive Levels (i.e., of the schemata

congruence and schemata evaluation factors, respectively) if also the

level of self-awareness was high compared with low. A particularly

decreased retrieval quantity tended to occur in the condition that

represented both incongruent and positive levels if also the level of

self-awareness was high compared with low. Additionally, a particularly

increased retrieval quantity tended to occur in the condition that































on Mean Retrieval Quantity


I


Experimental Mean retrieval Standard
condition quantitya Db deviation


Congruent schemata
Positive schemata
Low self-awarenessc 22.53 21.40* 12.23
High self-awarenessd 43.93 17.50
Negative schemata
Low self-awarenessd 18.14 7.00 7.86
High self-awarenessd 25.14 10.66
Incongruent schemata
Positive schemata
Low self-awarenessd 24.86 5.86 9.40
High self-awarenessd 19.00 5.72
Negative schemata
Low self-awarenessd 11.50 5.00 5.37
High self-awarenessd 16.50 9.08



a no. of memories.
refers to the difference between the means in the Low compared with
the high level of self-awareness.
Cn =15.
n =14.
* refers to a significant difference as tested by Newman-Keuls pre-
planned comparisons (i.e., alpha = .05; critical value = 7.76; mean
square error term used = 108.37).


represented either (a) both incongruent and negative levels or (b) both

congruent and negative levels if also the level of self-awareness was

high compared with Low. Table 5 presents the mean retrieval quantity

values and concomitant standard deviation values for each of the eight

Self-Awareness X Schemata Congruence X Schemata Evaluation experimental

cells. Figure 3 depicts the three-way interaction.

Table 5

Self-Awareness X Schemata Conaruence X Schemata Evaluation Interaction




















































Figure 3

Self-Awareness X Schemata Congruence X Schemata Evaluation Interaction
on Mean Retrieval Quantity



Effects on Retrieval Latency

Table 6 summarizes the results of the 2 X 2 X 2 between-subjects

ANOVA. The first factor reflected the levels of self-awareness, (low,

high), the second factor reflected the levels of schemata congruence


X Congruent, Positive Schemata
x Congruent, Negative Schemata
0 Incongruent, Positive Schemata
o Incongruent, Negative Schemata


46-

42-

38-

34-

30-

26-

22-

18-

14


Mean Retrieval
Quantity (no.
of memories)


Low High

Level of Self-Awareness
























Evaluation on Mean Retrieval Latency



Source of Sum of Df Mean F E
variation squares square


Main effects 13839.13 3 4613.04 20.50 .000
Self-awareness (SA) 868.64 1 868.64 3.86 .052
Schemata congruence (SC) 7588.05 1 7588.05 33.72 .000
Schemata evaluation (SE) 5351.40 1 5351.40 23.78 .000
Two-way interactions 2077.93 3 692.64 3.08 .031
SA X SC 1150.17 1 1150.17 5.11 .026
SA X SE 764.26 1 764.26 3.40 .068
SC X SE 177.80 1 177.80 .79 .376
Three-way interaction 700.14 1 700.14 3.11 .081
SA X SC X SE 700.14 1 700.14 3.11 .081
Explained 16617.20 7 2373.87 10.55 .000
Residual 23630.06 105 225.05
Total 40247.27 112 359.35


(congruent, incongruent), and the third factor reflected the levels of

schemata evaluation (positive, negative). Mean retrieval latency was

the dependent variable.


Table 6

ANOVA :


Effects of Self-Awareness, Schemata CongrecadShmt


Main Effects

Main effects on retrieval Latency occurred for all three factors.

The effects indicated a lesser memory retrieval latency (a) in the high

compared with the low level of self-awareness, (b) in the congruent

compared with the incongruent level of schemata congruence, and (c) in






























































Figure 4

Main Effects of Self-Awareness, Schemata Congruence, and Schemata
Evaluation on Mean Retrieval Latency


the positive compared with the negative level of schemata evaluation.

Table 7 presents the mean retrieval latency values and concomitant

standard deviation values for each of the two levels of each of the

three independent variables. Figure 4 depicts the main effects.


Incongruent


Negative


54-


51-


Mean
Retrieval
Latency
(half-
seconds)


Self-
Awareness


Schemata
Congruence


Schemata
Evaluation













Table 7

Main Effects of Self-Awareness, Schemata Congruence, and Schemata
Evaluation on Mean Retrieval Latency


Experimental Mean retrieval Standard
condition latencya deviation


Self-awareness
Low 48.62 18.30
High 43.34 19.40
Schemata congruence
Congruent 37.85 13.99
Incongruent 54.31 19.83
Schemata evaluation
Positive 39.14 15.76
Negative 53.00 19.50


Shalf-seconds.
bn = 57.
c n = 56.



Two-way Interactions

Self-awareness interacted with schemata congruence to influence

retrieval Latency. The interaction indicated a Lesser memory retrieval

latency in the congruent compared with the incongruent Level of schemata

congruence only in the high compared with the Low level of self-

awareness. Self-awareness additionally tended to interact with schemata

evaluation to influence retrieval latency. The interaction indicated a

greater retrieval latency in the negative compared with the positive





level of schemata evaluation only in the low compared with the high

level of self-awareness. Table 8 presents the mean retrieval latency

values and concomitant standard deviation values for each of the four

Self-Awareness X Schemata Congruence experimental cells. Figure 5

depicts the respective interaction. Table 9 presents the mean retrieval

latency values and concomitant standard deviation values for each of the

four Self-Awareness X Schemata Evaluation experimental cells. Figure 6

depicts the respective interaction.




Table 8

Self-Awareness X Schemata Congruence Interaction on Mean Retrieval
Latency


Experimental Mean retrieval Db Sadr

condition latencya deviation


Low self-awareness
Congruent schematac 43.55 10.32 14.10
Incongruent schematad 53.87 20.79
High self-awareness
Congruent schematad 31.94 22.80 1.3
Incongruent schematad 54.74192



a half-seconds.
befsto difference between the congruent and incongruent
schemata means in the particular level of self-awareness.
c n = 29.
d ,
n = 28












































Self-Awareness X Schemata Congruence Interaction on Mean Retrieval
Latency



Table 9

Self-Awareness X Schemata Evaluation :Interaction on Mean Retrieval
Latency


Experimental Mean retrieval Db Standard
condition latencya deviation


Low self-awareness
Positive schematac 39.22 19.13 12.58
Negative schematad 58.35 18.37
High self-awareness
Positive schematad 39.04 8.6187
Negative schematad 47.64 19.43


a half-seconds.
b refers to the difference between the positive and negative schemata
means in the particular level of self-awareness.
c n = 29.
d ,
n = 28


X Congruent Schemata
0 Incongruent Schemata


45-


40-


35-


30-


Mean Retrieval
Latency (half-
seconds)


High


Low


Level of Self-Awareness


Figure 5





X Positive Schemata
0 Negative Schemata
59-


55-


51-
Mean Retrieval
Latency (half-
seconds) 47-


43-


39- X X


Low High

Level of Self-Awareness

Figure 6

Self Awareness X Schemata Evaluation Interaction on Mean Retrieval
Latency


Three-way Interaction

Self-awareness tended to interact simultaneously with both schemata

congruence and schemata evaluation to influence retrieval latency. A

particularly decreased retrieval latency occurred in the condition that

represented both congruent and positive levels (i.e., of the schemata

congruence and schemata evaluation factors, respectively) if also the

level of self-awareness was high compared with low. A particularly

increased retrieval latency occurred in the condition that represented

both incongruent and positive levels if also the level of self-awareness

was high compared with low. Additionally, a particularly decreased





retrieval latency (a) occurred in the condition that represented both

congruent and negative levels and (b) tended to occur in the condition

that represented both incongruent and negative levels if also the level

of self-awareness was high compared with Low. Table 10 presents the

mean retrieval latency values and concomitant standard deviation values

for each of the eight Self-Awareness X Schemata Congruence X Schemata

Evaluation experimental cells. Figure 7 depicts the interaction.

Table 10

Self-Awareness X Schemata Congruence X Schemata Evaluation Interaction
on Mean Retrieval Latency


Experimental Mean retrieval Db S dr

condition latencya deviation


Congruent schemata
Positive schemata
Low self-awarenessc 37.93 11.55* 12.25
High self-awarenessd 26.38 8.23
Negative schemata
Low self-awarenessd 49.57 1.8 38
High self-awarenessd 37.49 11.53
Incongruent schemata
Positive schemata
Low self-awarenessd 40.61 1.9 32
High self-awarenessd 51.70 17.79
Negative schemata
Low self-awarenessd 67.14 93 85
High self-awarenessd 57.78 20.72



a o.. of memories.
b refers to the difference between the means in the low compared with
the high level of self-awareness.
cn =15.
n =14.
* refers to a significant difference as tested by Newman-Keuls pre-
planned comparisons (i.e., alpha = .05; critical value = 11.17; mean
square error term used = 225.05).





Level of Self-Awareness


Figure 7

Self-Awareness X Schemata Congruence X Schemata Evaluation Interaction
on Mean Retrieval Latency


X Congruent, Positive Schemata
x Congruent, Negative Schemata
0 Incongruent, Positive Schemata
o Incongruent, Negative Schemata


Mean Retrieval
Latency (half-
seconds)


High


Low


Summary of Main Results


Predictions (i.e., hypotheses 1, 2, 3, 4, 5, and 6) concerning the

presence of main effects of each of the three independent variables on

both retrieval quantity and retrieval Latency were confirmed. As





predicted, subjects in a high self-awareness condition demonstrated an

increased retrieval quantity and a decreased retrieval latency compared

with subjects in a low self-awareness condition. As predicted, subjects

in a congruent schemata condition demonstrated an increased retrieval

quantity and a decreased retrieval latency compared with subjects in an

incongruent schemata condition. As predicted, subjects in a positive

schemata condition demonstrated an increased retrieval quantity and a

decreased retrieval latency compared with subjects in a negative

schemata condition.

Predictions (i.e., hypotheses 7 and 8) concerning the presence of

an interaction between self-awareness and schemata congruence on both

retrieval quantity and retrieval latency were generally supported. As

predicted, subjects in a congruent schemata condition demonstrated an

increased retrieval quantity and a decreased retrieval latency if they

also were in a high self-awareness condition compared with a low self-

awareness condition. Subjects in an incongruent schemata condition did

not, however, demonstrate a decreased retrieval quantity or an increased

retrieval Latency if they also were in a high self-awareness compared

with a low self-awareness condition. Within the incongruent schemata

condition, differences were not demonstrated solely as a function of

level of self-awareness.

Predictions (i.e., hypotheses 9 and 10) concerning the presence of

a three-way interaction between self-awareness, schemata congruence, and

schemata evaluation on both retrieval quantity and retrieval Latency

were generally supported. As predicted, subjects in an schemata





condition that was both congruent and positive demonstrated a

particularly increased retrieval quantity and a particularly decreased

retrieval Latency if they also were in a high self-awareness condition

compared with a low self-awareness condition. Subjects in a schemata

condition that was both incongruent and negative did not, however,

demonstrate a particularly decreased retrieval quantity or a

particularly increased retrieval Latency if they also were in a high

self-awarenenss condition compared with a low self-awareness

condition. In fact, subjects in a schemata condition that was both

incongruent and negative tended to demonstrate an increased retrieval

quantity and a decreased retrieval latency if they also were in a high

self-awareness condition compared with a low self-awareness condition.

Two additional results are noteworthy concerning the three-way

interactions: (a) Subjects in a schemata condition that was both

congruent and negative demonstrated a decreased retrieval Latency and

tended to demonstrate an increased retrieval quantity if they also were

in a high self-awareness condition compared with a low self-awareness

condition, and (b) subjects in a schemata condition that was both

incongruent and positive demonstrated an increased retrieval latency and

tended to demonstrate a decreased retrieval quantity if they also were

in a high self-awareness condition compared with a low self-awareness

condition.





CHAPTER V
DISCUSSION AND CONCLUSIONS

Memory Retrieval: The Role of Self-Schemata

It has been established in research concerning memories of

laboratory stimuli that (a) the retrieval of a memory is always

triggered by a retrieval cue even if the cue (referred to as a schema

because of its categorical function) cannot be identified (Tulving &

Thomson, 1973); (b) the schemata used at the retrieval stage are

effective in triggering recall only if they were used at the encoding

stage (this principle is referred to as the "encoding specificity"

principle, Tulving and Thomson, 1973, p.359); and (c) the detecting of

recurrences of events according to schemata is a fundamental cognitive

process developed early in life (c.f., Hasher & Zacks, 1984). The

results of the present investigation support the generalization of these

contextual principles to the realm of memories of Life experiences. The

feasibility and effectiveness of using self-schemata as retrieval cues

in a category generation task that involved the retrieval of a stream of

recurrences of life events was demonstrated. Consequently, the role of

self-schemata as a organized system of internal cues that is used in the

encoding of memories can reasonably be inferred. These general

principles have been supported in only scant, quite preliminary research

to date (Bellezza, 1984; Markus, 1977; Warren, Hughes, & Tobias, 1985).

The present results extend beyond the validation of general

principles concerning self-schemata; factors that interact to influence

the specific elaboration of self-schemata in the memory retrieval

process are identified. One factor is the congruence or incongruence of





the memory to the self-theory. The global, superordinate congruent-

incongruent standard of categorization has been found to be an important

variable in laboratory memory research (Underwood, 1979), in self-

schemata research (Markus, 1977; Mills, 1983), and in research

concerning the relationship between implicit personality theory and

memory for information about persons (i.e., hypothetical persons defined

by the experimenter) (c.f., Srul1, Lichtenstein, & Rothbart, 1985; Wyer,

Bodenhausen, & Srull, 1984). The present results support its

contribution additionally to an understanding of the process of

recollecting life experiences that represent aspects of self.

The finding that persons overall access self-congruent experiences

more quickly and in greater quantity than they access self-incongruent

experiences is consistent with some findings in related realms. Num-

erous researchers have found that persons are more likely to notice and

to recall confirming, as opposed to disconfirming, instances (Einhorn &

Hogarth, 1978; Snyder & Cantor, 1979; Snyder & Swann, 1978; Taylor,

Crocker, & D'Agostino, 1978; Tvers ky, 1977). Other researchers have

found, however, that incongruent or disconfirming instances are more

likely to be recalled (Hastie, 1980; Wyer &r Gordon, 1982; Wyer,

Bodenhausen, & Srull, 1984). Existing research is, thus, conflictual

and indicates that the type of preferential access that occurs depends

on numerous variables. The present investigation introduces one intra-

individual variable, that of the individual's state of consciousness at

the time of retrieval, that appears to influence pronouncedly the pref-

erential access process. In the state of consciousness in which persons

are not acutely aware of being both the subject and the object of the













retrieval search (iethe low self-aware state), no preferential

access of self-congruent experiences occurs. In the highly self-aware

state of consciousness, however, considerably more self-congruent access

is facilitated. The present results indicate that the discrepancy

between the self-congruent and self-incongruent retrieval quantities in

the Low self-aware state was approximately 2 memories, while the

respective discrepancy in the high self-aware state was approximately 17

memories. An analogous process occurs concerning the time taken to

recall a self-congruent as opposed to a self-incongruent memory. The

discrepancy between the self-congruent and self-incongruent retrieval

latencies in the Low self-aware state was approximately 10 half-seconds

while the respective discrepancy in the high self-aware state was

approximately 23 half-seconds.

These results indicate that a superordinate categorizing dimension

such as "like self-unlike self" that might appear to be elaborated on in

a rather constant manner, rather seems to be elaborated on in a

fluctuating manner depending on the background context. In this

instance the background context is that of self-awareness. The

perceived relationship between what is similar to self and what is

dissimilar to self appears to become more salient in the high self-aware

state. The identification of process variables, such as Level of

objective self-awareness, that influence the manner of elaboration of

self-schemata is a research priority, according to Schmittdiel (1983).

Schmittdie1 observed that process variables have been conspiciously

neglected in the interest of structural dimensions.





Another, perhaps structural, dimension whose selective elaboration

depends on the process variable self-awareness is the evaluative

(positive-negative) dimension. It is centrally used by persons in

information processing generally (Osgood, Suci, & Tannenbaum, 1957) and

in the processing of information pertaining to the self more specifi-

cally (Mischel, Ebbesen, & Zeiss, 1973). Kim and Rosenberg (1980) found

in their analysis of personal construct systems, that the good-bad

evaluative dimension was the sole dimension that emerged as a superor-

dinate standard of categorization in all persons' systems. The present

results support its importance in accounting for variability in the

retrieval of personally relevant information. The specific result that

persons overall access experiences that reflect unfavorably on the self

in smaller quantity and more slowly than they access experiences that

reflect favorably on the self is generally consistent with related

research. Mattin and Stang (1978) concluded on the basis of summarizing

research on selectivity in recall that persons tend to both rehearse and

recall the pleasant more than the unpleasant or neutral. Markus and

Smith (1981) concluded that research indicates that the self works to

maintain high self-esteem and attends selectively to the good aspects of

behaviors. They stated that self-structures influence memory of events

about the self so that good, correct, responsible, consistent, and

successful activities are much more likely to be recalled than the bad,

incorrect, irresponsible, or unsuccessful ones.

A generalization concerning the preferential retrieval of good-self

experiences is a misleading simplification, however, according to inter-

action effects of the present investigation. As discussed in the next





section, the three-way interaction indicates that when both the indivi-

dual's state of consciousness and the self-congruence or self-incongru-

ence of the memory are also considered, certain bad-self experiences are

actually accessed faster than are certain good-self experiences. Addi-

tionally, a two-way interaction indicated that an individual's state of

consciousness at the time of retrieval tended to influence his or her

ease of accessing bad-self experiences. Experiences that reflect the

bad-self tend to surface to a conscious level of awareness more quickly

in the high self-aware state than they do in the low self-aware state.

Persons took approximately 19 half-seconds longer to recall a bad-self

versus a good-self experience when not highly self-aware though they

took only approximately 9 half-seconds longer to recall the respective

bad-self experience when highly self-aware at the time of memory

retrieval. This marginally significant (E<.068) effect is discussed

because it is, in fact, consistent with preliminary findings in the

self-awareness literature concerning information processing (~.

concerning the processing of other than memories of life experiences).

The result substantiates the research conclusion that the high self-

aware state of consciousness facilitates the more efficient processing

of information that reflects unfavorably on the self. For instance,

Turner (1978b) found that high scores on the Private Self-Consciousness

subscale of the Self-Consciousness Scale (Fenigstein et al., 1975) were

correlated with quick judgements (i.e., on a 2-point "me-not me" rating

scale) concerning the self-descriptiveness of socially undesirable trait

terms.





One additional finding concerning an interaction effect that

involves the evaluative dimension may be noteworthy. There is a

tendency (p<.062) for females to access experiences according to a good-

self/bad-self ratio of 1.76/1 specificallyl, 30 memories/17

memories) while males access according to a respective ratio of 1.32/1

specificallyl, 25 memories/19 memories). Females, thus, may

access information reflecting on the self in a tess balanced manner than

do males. This tendency of females may have its adaptive aspects in

maintaining high esteem though it may also have its maladaptive aspects

in shielding self from reality-based information pertaining to the

self. Existing research does not specifically clarify the present

gender finding. Some researchers that have investigated autobiographi-

cal retrieval have found gender differences (Meltzer, 1931; Robinson,

1976). Other researchers conclude that there are no substantial gender

differences in memory retrieval (Maccoby & Jacklin, 1974; Mattin &

Stang, 1978). Note that gender was not significantly involved in any

other interaction (i.e., two-way, three-way, or four-way); nor was there

a main effect of gender. The result that the impact of self-awareness

does not depend on gender is consistent with the self-awareness

literature (Carver & Scheier, 1981).



Memory Retrieval: The Role of Self-Awareness

It has been established in research concerning memories of Labora-

tory stimuli that the retrieval context, referring to all aspects from





both within and without the individual present at the time of retrieval,

significantly affects the nature of the retrieval process (Tulving,

1983). The present results indicate that objective self-awareness

functions as a retrieval context that significantly affects persons'

process of retrieving life experiences. Objective self-awareness,

referring to that state of consciousness in which persons are acutely

aware of being an object in the world and the object of their own

consciousness, influences both the amount of memories retrieved and the

amount of time taken to retrieve any particular memory.

According to the present results, persons process of accessing

various types of information reflecting the self becomes more variable

when persons are made to become self-aware; intraindividual modes of

responding become more pronounced. The results, therefore, indicate

that a consideration of self-awareness as a control process may be

useful. Control processes refer to processes that account for persons'

invariably selective, biased information processing (Underwood, 1979).

Control processes govern the elevating and lowering of thresholds of

cogni t ion, re sul ting in inhibition and facilitation effects, respec-

tively (Erdelyi & Goldberg, 1979). The present results indicate that

self-awareness has both inhibition and facilitation effects. Specific-

ally, inhibition occurs in the high self-aware state when persons recall

experiences that relate to a centrally important positive though incon-

gruent self-schemata. Persons take longer to recall, and tend to recall

fewer, of these memories if they are highly self-aware. For example,

consider a prototypical person who privately believes she is not a

generous person though believes that the generous-selfish self-schema is





an important, central dimension in judging persons. According to the

present results, she should not exhibit difficulty in recalling

incidents that reflect her generosity if she is not highly self-aware,

but would be relatively inhibited if made to be highly self-aware at the

time of retrieval. Thus, persons may be relatively inhibited in

invalidating their perceived liabilities when highly self-aware. One

explanation for this effect is that persons more acutely experience the

invalidation as self-deception while in that state due to their

increased and simultaneous awareness of their opposing bad-self (e.g.,

selfish) tendencies.

One assumption on which the aforementioned explanation is predicated

is that self-schemata have an inherently bipolar nature. Persons are

assumed to elaborate on one pole of a bipolar self-schema with an

awareness of the opposing pole. The simultaneous awareness is

considered to be inevitable though it may be present on various levels

of awareness. The present results appear to necessitate a consideration

of the assumption of bipolarity, an assumption that is solidly grounded

in self-concept, personal construct, and some of the self-schemata

theory.

Self-awareness additionally has, in contrast to its inhibition

effects, facilitation effects. One particularly significant and

pronounced facilitation effect concerns the recollection of experiences

that are congruent with persons' central and positive self-schemata.

Persons are apparently able to be considerably more self-enhancing when

highly self-aware if the enhancement is of a self-perception that is

congruent with their self-theory. Persons are more able to seek out





information that validates their most valued attributes if they are

highly self-aware. For example, consider a prototypical person who

believes, for instance, that he is trustworthy and that trustworthy-

untrustworthy is an important dimension in judging people. The present

results suggest that he would take considerably less time to recall, and

recall many more, experiences that validate his perceived trust-

worthiness if he is made to be highly self-aware. The increased

awareness of the bad-self (i.e., untrustworthy) tendencies is presumed

to facilitate the desire and tendency to validate good-self (i.e.,

trustworthy) tendencies, noting that because the latter is a strong,

integrated personal asset self-deception is not involved.

A simultaneous consideration of the inhibition effect and the

facilitation effect (and given consideration to all other effects)

clarifies that (a) self-awareness does not globally facilitate the

retrieval of any and all information solely because the information

pertains to the self, (b) self-awareness does not globally facilitate

the retrieval of any and all kinds of information solely because the

information is self-enhancing, and (c) self-awareness does not solely

result in an increased manifestation of persons' self-diminishing or

self-critical tendencies. Duval and Wicklund's (1972) theory relates to

the latter aspect. They theorized that self-awareness results in self-

critical tendencies and in lowered self-esteem attributable to the

increased perception of discrepancies between the present state and

standards of correctness.

A second condition in which facilitation occurs in high self-

awareness is when persons recall experiences that relate to a centrally













important congruent through negative self-schema. For example, consider

a prototypical person who privately believes that she is selfish and is

requested to recall experiences in which she reflected that trait. The

present results suggest that she would recall more of these self-

congruent experiences that reflect unfavorably on herself significantly

faster, and tend to recall more of those experiences, if she is made to

be highly self-aware. It can be presumed (i.e., due to experimental

assessment procedures) that the person has a strong, integrated sense of

self as selfish since it is more difficult due to social desirability

and centrality-positivity biases (Lewicki, 1983) to rate self unfav-

orably on a superordinate self-schema (c.f., Adams-Webber, 1982). The

simultaneous awareness of the good-self tendencies of the person would

not be sufficient in this case to inhibit the awareness and retrieval of

experiences relating to the bad-self. This result is consistent with

the explanation that self-awareness inhibits self-deceptive tendencies

and facilitates veridical (i.e., honest and truthful) tendencies; given

the apparent strength of the bad-self tendencies, self-deception would

be involved should a facilitation effect not occur. It may be noted

that this explanation is consistent with preliminary evidence in the

self-awareness literature indicating that self-awareness increases the

veridicality of self-report (Scheier, Buss, & Buss, 1978; Turner,

1978a).

The notion of increased veridical access during high self awareness

warrants a consideration of one additional facilitation tendency, noting

that the tendency did not meet a .05 level of statistical significance.

Specifically, there was a facilitation tendency in the high self-aware





state for the retrieval of experiences that reflect unfavorably on the

self even if the experiences relate to a disavowed self-perception. For

example, consider a prototypical person who believes that the

trustworthy-untrustworthy dimension is central to his self-theory but

who does not consider himself to be untrustworthy. When he attempts to

retrieve bad-self untrustworthyt) experiences, he tends to

retrieve the experiences more quickly and in greater quantity if he is

made to be highly self-aware at the time of retrieval. Presumably,

persons have the bad-self memories available on some level and self-

awareness tends to facilitate their accessibility. Persons are less

able to disavow the negative-self while highly self-aware or, conver-

sely, are more able to disavow the latter while not highly self-aware.

This facilitation tendency is consistent with the explanation that self-

awareness inhibits self-deceptive biases and facilitates veridical and

extensive retrieval.

The present results question theory concerning a dichotomous state

of consciousness; theory that is supported by Duval and Wicklund

(1972). Duval and Wicklund assert that persons' consciousness is

directed either toward external objects in the subjectively self-aware

state or toward one's conscious state, personal history, physical being

or other personal aspects of self in the objectively self-aware state.

They state that "increased objective self-awareness" simply denotes a

greater proportion of time spent in the objective state and "should not

be taken to mean that the objective-subjective self-awareness dimension

is continuous" (p.3). This theory poses important theoretical problems

because it does not allow for the greater or lesser functioning of the













self as a control process that regulates the accessing of information,

even that pertaining to the self. The present results support the

notion that there is a continuum of involvement of the self even on an

inherently self-involving and self-objectifying information search. The

retrieval task used in the present investigation necessarily required

the subject to access information about self as an object in the world

and was, to this extent, self-objectifying. A ceiling effect did not

occur in the low self-aware condition, however, as may have occurred if

the notion of "increased" self-awareness was not viable.

The present results buttress the notion of the self as a fluctua-

ting control process regulating the accessibility of information. This

view of the self has been relatively neglected in experimental litera-

ture, which has tended to study the self as solely a concept or a sche-

mata rather than also as a level of control. At any given moment num-

erous control processes interact to regulate persons' cognitions and

behaviors; self-awareness as one salient control process is of ecolo-

gical importance since fluctuations in self-awareness recurrently occur

both spontaneously and volitionally during the ongoing flow of con-

sciousness. The present results indicate that fluctuations in self-

awareness influence the ease of accessibility to memories of life

experiences. The specific nature of the impact of the fluctuations in

self-awareness depend on both the self-evaluative and self-validating or

self-invalidating implications of the recollected experience. Overall,

the high self-aware state can be concluded to facilitate veridical

retrieval efforts and inhibit self-deceptive retrieval efforts.













Implications of the Results

One practical implication of: the present results may be in the

ream of personality assessment. During the completion of personality

assessment instruments and interviews, persons invariably review past

tendencies and aspects of self. The amount and type of information

recalled and the ease with which it is recalled has implications for the

validity of the assessment procedure, noting that the predictive vali-

dity is of central concern. If persons are more Likely to engage in a

veridical, extensive review of past behaviors and tendencies and are

less likely to engage in self-deceptive portrayals of themselves while

highly self-aware then the induction of that state during personality

assessment may result in an increase in the predictive validity of the

particular procedure. Preliminary research supports the conceptual

relationship between the review of past behaviors, the expressing of

attitudes, the stating of future intentions, and behavioral actions

(Fazio & Zanna, 1976; OLson & Zanna, 1978), as well as the conceptual

relationship between self-awareness, self-reports, and behavioral ten-

dencies (Scheier et al., 1978; Turner, 1978a). Scheier et al. (1978)

and Turner (1978a) found that self-reports of dispositional tendencies

(toward hostility and dominance, respectively) were better predictors of

actual behavior among subjects high in the disposition toward private

self-consciousness than among subjects lower in the disposition toward

private self-consciousness. The present results considered in conjunc-

tion with preliminary related research, thus, support future research

investigation into the role of self-awareness in increasing the

predictive validity of personality assessment procedures, perhaps

including interview procedures.













The role of self-awareness in influencing persons' accessibility to

memories reflecting the self additionally may have implications in the

realm of psychotherapy. The present results suggest that psychotherapy

patients who are made to be highly self-aware will access certain memo-

ries with greater ease and other memories with greater difficulty than

those patients who are not made to be highly self-aware. No primary

focus on remote or childhood memories is implied because it is intui-

tively clear that much of a patient's stream of thought during a psycho-

therapy session invariably relates to memories of quite recent, daily

life experiences. The present results suggest that the aspects of life

experiences that the patient will allow himself or herself to recall

depends at least partially on his or her immediate Level of self-

awareness (i.e., the patient's self may be more or less an object of the

patient's own disclosure). The present results, thus, highlight the

importance of the intraindividual context of accessibility to cognition,

a context that has been neglected in psychotherapy research relative to

the interpersonal context of disclosing cognition to the psychotherapist

(c.f., Luborsky et al., 1979). Self-awareness is undoubtedly one of

many complex, interacting control processes regulating thresholds of

cognition within the psychotherapy session, though the present results

suggest that it is one control process worthy of future investigation.













Limitations of the Investigation

One limitation of the present investigation concerns the lack of a

more extensive investigation of cognitive processes that occurred during

the retrieval period. Targeted processes may have included (a) the

setting of response criteria, which is related to the category bandwidth

and the defining features of the prototypical instance (c.f., Barsalou,

1985), (b) categorical search processes that occur subordinate to the

self-schemata categorical search process (c.f., Whitten, W.B., II, &r

Leonard, 1981), (c) conscious and unconscious decision processes whether

to initiate or continue the search process, and (d) interfering

processes including distractibility and task-irrelevant cognitions.

Many of these processes are intrinsic to the multidimensional phenomenon

termed "memory" and, thus, are part of what was examined. A more

extensive investigation, however, would clarify the nature of the impact

of the experimentally manipulated control processes.

The limitation relates to another limitation of the present

investigation: The experiment used a research paradigm that is not yet

well-established in memory research concerning the retrieval of life

experiences. This was unavoidable because of the preliminary state of

that research. Fortunately, and to the benefit of future researchers in

the area, the numerous exploratory aspects of the paradigm did prove to

be workable. The degree of generalization from memory research

involving contrived laboratory stimuli, therefore, appears promising.

An additional limitation of the present investigation concerns the

lack of a more extensive investigation of characteristics of the

recalled experiences that are of interest to the study of personality













and undoubtedly inextricably interrelate to the above mentioned

processes, including (a) attributional processes, (b) other persons

involved, (c) specific content and themes, and (d) emotional valence.

This was avoided to both protect subjects' privacy and to limit the

scope of the investigation.

Finally, a quite specific limitation of the investigation concerns

the assessment of subjects' self-theories. The present investigation

clearly extended the level of the assessment of persons' self-theories

compared to that level used in existing experimental memory research.

The process of eliminating plausible rival theoretical explanations,

however, would have been aided by an even more thorough and specific

assessment. Characteristics that may have been examined include (a) the

overall positive or negative evaluative nature of the self-system (i.e.,

high or low self-esteem), (b) the overall like self-unlike self nature

of perceptions, referred to as the complexity of the self-concept (c.f.,

Ziller, 1973), (c) interrelationships among the self-schemata and the

overall tightness or looseness of the self-theory (c.f., Mancuso &

Adams -We bbe r, 1982), and (d) interrelationships among perceptions of

self and perceptions of others concerning the descriptiveness of a

particular attribute.



Recommendations for Future Research

Researchers may continue to contribute to the building of a

multiprocess, multilevel model of memory functioning that encompasses

memories of life experiences by conducting experimental research where

control is exerted over the retrieval context. Retrieval contexts of













high ecological relevance may include those impacted by the following

interrelated overlapping factors: (a) physiopharmacological states

(c.f., Nelson, McSpadden, Fromme, & Marlatt, 1986), (b) affective

states, (c) interpersonal states (c.f., MacWhinney, Keenan, & Reinke,

1982) including the immediate therapist-patient interaction (c.f.,

Luborsky et al., 1979), (d) antecedent activation of memories, (e)

antecedent activation of cognition at varying levels of conscious

awareness (i.e., using priming manipulations; c.f., Seifert, Abelson,

McKoon, & Ratcliff, 1986); and (f) states of consciousness (including

hypnotic states; c.f., Kihtstrom & Evans, 1979; Kihlstrom, Brenneman,

Pistole, & Shor, 1985; Stager & Lundy, 1985), deindividuation or lowered

self-awareness states, and public self-consciousness states. Regarding

the Latter, the present results justify the continued exploration of the

impact of the objective high self-aware state of consciousness speci-

fically as induced by the mirror manipulation. The manipulation was

found to be sufficiently powerful to influence the memory retrieval

process. It is important that researchers avoid attempting to identify

the most important retrieval context and instead attempt to specify the

impact (and the interactive impact) of particular contexts on particular

types of cognitive processes and on particular types of information

searches. Although the resultant theory is extremely complex, existing

research (particularly that concerning the physiology of memory; c.f.,

Mayes, 1981) indicates that parsimonious theory may be misleading.

The specific results of the present investigation additionally

justify the continued exploration of the role of self-schemata in the













organization and retrieval of life experiences. Computerized assessment

can be used to a much greater advantage by conducting even more

differentiated assessments of persons' self-theories, thereby enabling

more comprehensive theoretical formulations to be developed. Theory and

research in personal construct theory appears to offer a most sophis-

ticated contribution to an assessment of persons' construal of them-

selves and others (c.f., Neimeyer & Neimeyer, 1981). Dimensions in

addition to schemata congruence and evaluation that interrelate to the

elaboration of self-schemata may be identified. Researchers may seek to

identify effective retrieval cues in addition to self-schemata, inclu-

ding visual representations (c.f., Tversky & Baratz, 1985) and concep-

tual categories that represent interpersonal experiences.

Finally, researchers additionally may explore the current unex-

plored but fertile domain of how persons use their memories (at varying

levels of conscious awareness) to modify self-perceptions, perceptions

of others, and courses of action. Memories of life experiences consti-

tute a vast internal source of self-information that significantly

functions in the stream of consciousness but until recently largely has

been ignored in research as such. Overall, researchers are encouraged

to take a comprehensive view of the retrieval process because the

concomitant research is at such an exploratory stage.















APPENDIX A
EXPERIMENTAL STIMULI



Fifty trait adjectives, which are organized and conceptualized as

25 bipolar trait adjectives, comprised the pool of trait adjectives from

which schemata were derived. The adjectives were selected from

Anderson's (1968) list of 555 personality trait words. Each was rated

according to its likeability and meaningfulness as a personality

characteristic. Anderson carefully constructed this list through

extensive screening and normative work. An initial pool of 18,000

traits complied by Allport and Odbert (1936) was eventually narrowed to

555 trait adjectives. Anderson (1968) specifically constructed the List

for use in research related to information integration in personality

impression tasks. Anderson (1968) provided normative data on the like-

ableness and meaningfulness of the words on the basis of ratings by 100

male and female college students. Anderson's (1968) analyses indicated

that the main effect of gender and the Gender by Word interaction were

nonsignificant. Subsequent researchers have extended Anderson's (1968)

normative data. The use of the list is widespread, particularly in

social and personality psychology.

For the purposes of the present investigation, the meaningfulness

normative ratings were particularly useful. Bipolar trait adjectives

provided the basis for the experimental paradigm and it was important

that both poles were high in meaningfulness. Anderson's (1968) meaning-

fulness ratings were based on 100 subjects ratings of 555 words on a

five point scale, which varied from (0) "I have almost no idea of the













meaning of this word" to (4) "I have a very clear and definite

understanding of the meaning of this word" (p. 277). The intermediate

steps also were verbally categorized. The subjects were instructed to

rate each of the words according to how well they knew its meaning as a

description of persons. The subjects were encouraged to spread their

ratings over the entire 0-4 range as much as possible. Various cutoff

points for high meaningfulness have been recommended for research pur-

poses (e.g., Anderson, 1968, proposed a 3.45 cutoff point). A perusal

of the relevant research indicated, however, that no precise point has

widely agreed upon significance. Rather, the research utility of the

meaningfulness ratings lies in the ability to exclude low meaningful

traits.

Only high meaningful traits (i.e., meaningfulness ratings within a

range from 3.58 to 3.86), which were paired into bipolar dimensions,

were used in the present investigation. Those traits whose bipolar

opposite was low in meaningfulness, indistinct, or difficult to define

in a word were excluded as candidates for the experimental stimuli

list. The 48 traits that were selected are presented in Table 11. Note

that the traits "attractive" and "unattractive" were not on Anderson's

(1968) List; Anderson screened all traits pertaining to physical char-

acteristics. The aforementioned traits were included in the present

investigation because it is now recognized that physical characteristics

are important aspects of the development of the self-concept.





r


88







Table 11
Experimental Stimuli


Mb

384

368

370

366

374

376

370

364

382

374

372

380

374

364

360

368

382

380

380

362

370


La

545

537

459

471

455

529

510

412

466

428

461

373

447

503

427

503

427

476

519

433

539


Trait

truthful

intelligent

generous

courageous

independent

thoughtful

unselfish

outgoing

punctual

modest

tolerant

obedient

competent

broad-minded

decisive

gentle

tidy

cooperative

friendly

skilled

trustworthy


Trait

untruthful

unintelligent

stingy

cowardly

dependent

thoughtless

selfish

shy

unpunctual

boastful

intolerant

disobedient

incompetent

narrow-minded

indecisive

forceful

untidy

stubborn

unfriendly

unskilled

untrustworthy


L

43

168

143

110

254

77

82

291

192

122

98

128

110

80

219

263

175

196

92

224

65


M

380

364

368

374

360

366

384

376

366

380

362

378

364

374

376

358

386

380

386

360

376






89











TABLE 11 continued


Trait


Trait


reliable

popular


374

368

368


unreliable

unpopular

hostile


386

362


kind


91 372


Source: Selected from Anderson's (1968) list of 555 trait adjectives.


Note: The traits "attractive" and "unattractive," which were not on
Anderson's (1968) list, were two additional traits in the pool of
experimental stimuli.


Refers to the likeability ratings of the trait adjectives. The
higher the rating, the more favorable or desirable the trait as a
description of a person (Anderson, 1968).
b eest h ennflesrating of the trait adjective. The
higher the rating, the more meaningful the trait in the description
of a person (Anderson, 1968).





APPENDIX 8
INFORMED CONSENT


The experimental research project concerns the identification of
factors which affect and regulate the process of recollecting life
experiences. The organization of memories, the cues which trigger the
recollection, and the relationship of memories to self-perceptions are
the central research concerns. You will be requested to rate various
personality traits according to designated criteria. You will also be
requested to recall memories of experiences that you have had during
your life in response to a series of words. You will be requested to
press a designated key on the computer each time a memory is recalled.
You will not, at any time, be requested to disclose the content of your
memories. Your privacy will, thus, be maintained.

Benefits expected from participation in the experiment may include
an increased sense of self-exploration and self-knowledge. Sometimes,
however, persons prefer to avoid self-exploration since they find it to
be uncomfortable. Should you find the experimental procedures to be
uncomfortable, you are free to withdraw your consent and to discontinue
participation in the research project at any time without prejudice. If
necessary, follow-up counseling can be made available.

Monetary compensation will not be awarded for participation in the
experiment.

I am willing to answer any inquiries you may have concerning the
experimental procedures. You may contact me at my home residence:
11393 S.W. 112th Street, Miami, Florida 33176, (305) 595-7874. You may
alternatively arrange to meet with me while I am in Gainesville for the
running of the experiment. You may leave a message with Dr. Greg
Neimeyer at 392-6768 for me to contact you to arrange an appointment.




I have read and I understand the procedure described above. I agree to
participate in the procedure and I have received a copy of this descrip-
tion.



Subject Date Witness Date



Relatioship i other Date Principal Investigator's Dt
than subject name














REFERENCES


Adams, L.T. (1985). Improving memory: Can retrieval strategies help?
Human Learnin 4, 281-297.

Adams-Webber, J.R. (1982). Assimilation and contrast in personal
judgement: The dichotomy corollary. In J.C. Mancuso & J.R. Adams-
Webber (Eds.), The construing person (pp. 96-112). New York:
Praeger.

A11port, G.W., & Odbert, H.S. (1936). Trait-names: A psycho-lexical
study. Psychological Monographs, 47 (1, Whole No. 211).

Anderson, N.H. (1968). Likeableness ratings of 555 personality trait
words. Journal of Personality and Social Psychology, 9, 272-279.

Anderson, J.R., & Bower, G.H. (1972). Recognition and retrieval
processes in free recall. Ps chological Review, 79, 97-123.

Anderson, J.R., & Bower, G.H. (1973). Human associative memory.
Washington, D.C.: V.H. Winston.

Baddeley, A.D. (1983). Your memory. Middlesex, England: Penguin.

Barsalou, L.W. (1985). Ideas, central tendency, and frequency of
instantiation as determinants of graded structures in categories.
Journal of Experimental Psychology: Learning, Memory, and Cognition,
11, 629-654.

Bartlett, F.C. (1932). Remembering: A study in experimental and social
psychology. Cambridge, England: Cambridge University Press.

Baxter, M.F., Yamada, K., & Washburn, M.F. (1917). Directed recall of
pleasant and unpleasant experiences. American Journal of Psychology,
28, 155-157.

Beaman, A.L., Klentz, B., Diener, E., & Svanum, S. (1979). Self-
awareness and transgression in children: Two field studies. Journal
of Personality and Social Psychology, 37, 1835-1846.

Bellezza, F.S. (1984). The self as a mnemonic device: The role of
internal cues. Journal of Personality and Social Psychology, 47,
506-516.

Bousfield, W.A. (1953). The occurrence of clustering in the recall of
randomly arranged associates. Journal of General Psychology, 49,
229-240.




University of Florida Home Page
© 2004 - 2010 University of Florida George A. Smathers Libraries.
All rights reserved.

Acceptable Use, Copyright, and Disclaimer Statement
Last updated October 10, 2010 - Version 2.9.9 - mvs