A DEVELOPMENTAL STUDY OF ‘THE
ORGANIZATION OF IMPRESSIONS WRITTEN:
FROM INCONSISTENT INFORMATION

Thesis for the Degree of Ph. D.
MICHIGAN STATE UNIVERSITY
DONALD STEPHEN BISKIN
197 1

I»)

 
 

 

 

 

This is to certify that the

thesis entitled
A DEVELOPMENTAL STUDY OF THE

ORGANIZATION OF IMPRESSIONS WRITTEN
FROM INCONSISTENT INFORMATION

presented by

Donald Stephen Biskin

has been accepted towards fulfillment
of the requirements for

Ph. D. degreein Educational Psychology

fig Lute/IQ? 2°:ng

M» W /

DateleA/j/
/ /

0-7839

LIBRARY

Michigan State
University

 

- éinomc BY

    

IIIIAG 8: MRS

BOOK BINDERY INC. .

LIBR

me! .
L A

an N DERS
, mum!

 

ABSTRACT

A DEVELOPMENTAL STUDY OF THE
ORGANIZATION OF IMPRESSIONS WRITTEN
FROM INCONSISTENT INFORMATION

BY

Donald Stephen Biskin

Purpos e

This study was conducted in order to investigate dif-
ferences between ages hithe organizationcmfimpressions formed
from inconsistent information. The research questions and

strategies were based on cognitive developmental theory.

Literature Review

Three areas of investigation were reviewed in
Chapter II: studies of impression formation from incon-
sistent information, developmental studies of impression
formation, and studies of the ontogenetic changes in moral
judgment. The results of these investigations clearly
support the research questions.

In summarizing the studies of the organization of
impressions from inconsistent information about another it
should be noted that different organizational patterns are

related to the cognitive complexity of the subject. What

Donald Stephen Biskin

is of paramount importance is the fact that the definition
of complexity utilized in the reviewed studies is similar
to the definition of the developmental process as employed
by Piaget, Werner, and the present investigator. In addi-
tion evidence was presented to support the existence of
sex differences in organizational patterns.

The results of the developmental studies of im-
pression formation indicate that differences in the or-
ganization of impressions vary as a function of age; and
with increasing age an individual's written impressions of
another become more differentiated and integrated. There
is also evidence to support an age related shift from con-
crete to abstract differentiation. In addition evidence
exists to support the hypothesis that females and subjects
with higher I.Q.'s employ more advanced organizational
patterns.

Additional support for the present investigation
came from studies of the development of moral judgment.

A similar theoretical orientation and two other factors
made these investigations relevant to the present study.
The experimental paradigms are comparable and the results
bear a great similarity to the results of the studies of

the organization of impressions.

Donald Stephen Biskin

Des—ign of the Study

One hundred and forty-eight subjects, categorized
on the basis of grade, 1.0. level, and sex were presented
two video-taped vignettes. Each vignette was composed of
five scenes. The first scene of both vignettes was
neutral. Two of the remaining scenes showed an actor
portraying socially acceptable behaviors, while the other
two showed him portraying socially unacceptable behaviors.
Upon the completion of each movie the subject was asked to
write his spontaneous impression of the stimulus person.

The impressions were content analyzed according to
the scoring procedure outlined by Crockett (1970). Using
the derived dependent variables a multivariate analysis of
variance was employed to test the research hypotheses.

Examination of the effects of grade level on the
organization of the written impressions provided evidence
relating to the main purpose of the study. Also examined
were the effects of 1.0. level and sex on the organiza-

tional patterns.

Results

The major findings can be summarized as follows:
(1) There was-a statistically significant main effect
for grade for each dependent variable.
(2) There was a statistically significant main effect

for 1.0. level for the degree of abstract

Donald Stephen Biskin

differentiation, level of organization, and the
degree of total differentiation.

(3) There was a statistically significant main effect
for sex of subject for the degree of connectedness,
and the degree of total differentiation.

Post hoc analyses were conducted in order to

determine the cause of a significant hypothesis test.

A DEVELOPMENTAL STUDY OF THE
ORGANIZATION OF IMPRESSIONS WRITTEN

FROM INCONSISTENT INFORMATION

BY

Donald Stephen Biskin

A THESIS

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

DOCTOR OF PHILOSOPHY

Department of Counseling, Personnel Services and
Educational Psychology

1971

To Suzanne

LIST OF

LIST OF

Chapter

I.

II.

III.

TABLE OF CONTENTS

TABLES . . . . . . .

FIGURES O O O I O 0

STATEMENT OF THE PROBLEM

Introduction . .

Theoretical Rationale

Research Questions

Overview . . . .

REVIEW OF THE LITERATURE

Introduction . .

Impression Formation from
Contradictory Information . . . . .

Developmental Studies

Formation . . . .

Developmental Changes

Judgment . . . .
Summary . . . . .

DESIGN OF THE STUDY .

Sample . . . . .
Design . . . . .
Procedure . . . .

Stimulus Materials
Dependent Measures

Reliability of the Scoring Procedure

Research Hypotheses

Analysis of the Data

Summary . . . . .

iii

of Impression

in Moral

Page

vi

10
10

12

12
12
19

28
29

32

32
35
37
38
41
48
49
51
52

Chapter

IV. ANALYSES AND RESULTS . . . . .

Introduction . . . . . . .
The Effects of Grade on th
Organization of Impressions

The Effects of I.Q. Level on the

Organization of Impressions

The Effects of Sex of Subject on

Organization of Impressions
Summary of the Results . .

V. SUMMARY AND CONCLUSIONS . . .

Purpose . . . . . . . . .
Literature Review . . .
Design of the Study . .
Results . . . . . . . . .
Discussion . . . . . . . .

The Effects of Grade on the Organization

of Impressions . . . . . .

The Effects of 1.0. on the Organization

of Impressions . . . . . .

The Effects of Sex on the Organization

of Impressions . . . . . .
Implications . . . . . . .
Limitations . . . . . . .
REFERENCES 0 O I O O O O O O O O O 0

Appendix A Rules for Scoring . . . .

Appendix B Means of Each Dependent Variable

iv

Page

54
54
55
62

65
70

73
73
73
74
75
76
76
81
83
85
88
90
93

101

LIST OF TABLES

Table Page
301 sample Size 0 O O O O O O O O O O O O O O O O O 33
3.2 Average sample age . . . . . . . . . . . . . . 34

3.3 Average sample IQ level . . . . . . . . . . . . 36

3.4 Interrater reliabilities of the two inde-
pendent judges for each dependent
variable 0 O O I O O O O O O O O O O O O O O 48

3.5 Inter-impression reliability for each
dependent variable . . . . . . . . . . . . . 52

4.1 Multivariate analysis of variance for the
independent variables . . . . . . . . . . . . 55

4.2 Univariate analyses of variance for the
effects of grade for each dependent
variable 0 O O O O O O O O O O O O O O I O I 56

4.3 Univariate analyses of variance for the
- effects of 1.0. level for each dependent
variable 0 o o o o o o o o o o o o o ' o o o o 62

4.4 Univariate analyses of variance for the
effects of sex for each dependent
variable 0 O O O C O O O O O O O I I O O O O 65

5.1 Frequency of the levels of organization in
eaCh grade 0 O O I O O O O O I O O I O O O 0 79

Figure

4.1

4.9

4.10

LIST OF FIGURES

Interaction of grade and sex on the degree
of concrete differentiation . . . . . .

Interaction of grade and sex on the degree
of abstract differentiation . . . . . .

Interaction of grade and sex on the degree
of connectedness . . . . . . . . . . . .

Interaction of grade and sex on the level
Of organization 0 O O I O O I I O O O 0

Interaction of grade and sex on the degree
of total differentiation . . . . . . . .

Interaction of sex and grade on the degree
of concrete differentiation . . . . . .

Interaction of sex and grade on the degree
of abstract differentiation . . . . . .

Interaction of sex and grade on the degree
of connectedness . . . . . . . . . . . .

Interaction of sex and grade on the level
of organization . . . . . . . . . . . .

Interaction of sex and grade on the degree
of total differentiation . . . . . . . .

vi‘

Page

57

58

59

60

61

66

67

68

69

70

CHAPTER I

STATEMENT OF THE PROBLEM

Introduction

 

Recently there has been an increasing consideration
of cognitive factors in the study of social development.

An area of investigation where this orientation has gained
some popularity is the study of the organization of impres-
sions that one person forms from observing the appearance
and behaviors of another.

Most investigations of the organization of impres—
sions have been oriented around the construct of cognitive
complexity.

Crockett defines cognitive complexity as a cogni-
tive system that contains a relatively large number of
elements which are integrated hierarchically by relatively
intensive bonds of relationship. This definition is de-
rived from Warner's (1957) orthogenetic principle: "When-
ever deve10pment occurs it proceeds from a state of rela-
tive globality and lack of differentiation to a state of
increasing differentiation, articulation, and hierarchic

integration." (Werner 1957, p. 126)

The construct of.cognitive complexity has been
generally impotent in generating hypotheses that would
result in increased understanding of theEnature of social
cognitions, and specifically the organization of impressions
of others. Kaplan & Crockett (1968) support this criticism
in stating that investigations of the organization of
impressions formed from inconsistent information have been
more concerned with superficial products rather than the

actual organizing process. He states:

It will be seen that the situation here (investigation
of impressions formed from inconsistent information)
is a special case of one in which the end is that of
forming a judgment or impression about a social object
and in which there are various means or routes to the
end without anyone necessarily ensuring a greater cor-
rectness or veridicality than the other. An analysis
and genetic ordering of these means must, therefore,
ignore the secondary and here non-adjudicable issue of
veridicality or correctness and concern itself with
the ways in which they differentiate, weigh and inte-
grate the diverse factors pertinent to the formation

of such an impression. (Kaplan and Crockett 1968, p. 665).

Kaplan and Crockett's concern for placing cognitive com-

plexity within atheoretical model has not been extremely pOpular.

It is generally assumed that cognitive complexity is a
very Specific construct. Bieri (1968) a prolific investi—

gator of cognitive complexity states:

In contrast with some approaches to cognitive structure,
cognitive complexity is fairly specific in its sc0pe,
being concerned only with social stimuli and with the
dimensional versatility of the person in his social
judgments. Such specific focus should render the cog-
nitive structure of complexity - simplicity more
directly relevant to differences in processing infor-
mation. (p. 635)

It seems naive to assume that our system for organizing
stimuli is different for social than physical stimuli, let
alone different for inconsistent information about another
than other social stimuli. Crockett (1970) assumes a more
logical position concerning the nature of the organization
of stimuli when stating that it does not matter what the
social stimuli are, for there is some underlying process
that organizes them in a similar fashion.

The narrow theoretical perspective assumed in
studies of impression formation has resulted in virtually
no attention being paid to the development of the struc-
tures or organizing prOperties underlying one's impressions
of another. -Studies examining cognitive complexity have
almost exclusively employed adult subjects, thus not at
all considering the possibility that the nature of what
they call cognitive complexity varies as a function of
development.

It is the purpose of this investigation to explore
how the organization of impressions formed from inconsis-
tent information about others change as a function of age.
The results will allow us to make more valid inferences
about the processes that determine an impressions

organization.

Theoretical Rationale

The hypotheses to be investigated are generated
from a cognitive developmental theoretical orientation.

This theory makes the following assumptions:

(1) Basic deve10pment involves transformations of
cognitive structures which cannot be defined by the
parameters of associationistic learning (contiguity,
repetition, reinforcement, etc.) and which must be
explained by parameters of organizational wholes or
systems of internal relations. These transformations
are more pOpularly labeled stages and have the follow-
ing general characteristics:

(a) Stages imply distinct or qualitative differ-
ences in children's modes of thinking or of
solving the same problem at different ages.

(b) These different modes of thought form an in-
variant sequence, order, or succession in
individual development. While environmental
factors may speed up, slow down, or stop
develOpment, they do not change its sequence.

(c) Each of these different and sequential modes
of thought forms a structural whole. A given
stage response on a task does not just repre-
sent a specific response determined by knowl-
edge and familiarity with that task or tasks
similar to it. Rather, it represents an-un-
derlying thought-organization.

(d) Cognitive stages are hierarchical integrations.
Stages form an order of increasingly differen—
tiated and integrated structures to fulfill a
common function, which is the maintenance of
an equilibrium between the organism and the
environment.

(2) Development of cognitive structure is the result
of processes of interaction between the structure of
the organism and the structure of the environment,
rather than being the direct result of maturation or
the direct result of learning.

(3) Cognitive structures are always structures of
action, while cognitive activities move from the

sensori-motor to the symbolic, to the verbal proposi-
tional modes, the organization of these modes is always
an organization of actions upon objects. (Kohlberg
1968b,p. 348)

The above assumptions hold for cognitive develop-
ment in general; the development of ways of thinking about
both physical and social stimuli. Their application to
social deve10pment is made more concrete by the following

additional assumptions:

(4) Social development and functioning, and cognitive
development and functioning are not distinct realms.
Social and cognitive development are parallel; they
represent perspectives and contexts in defining
structural change.

(5) There is a fundamental unity of personality or-
ganization and development termed the ego of the self.
While there are various strands of social develOpment
these strands are united by their common reference to

a single self-concept in a single social world. Social
develOpment is in essence the restructioning of the
self-concept in its relationship to concepts of other
people, conceived as being in a common social world
with social standards.

(6) All the basic processes involved in physical cog-
nitions, and in stimulating developmental changes in

these cognitions are also basic to social development.
(Kohlberg 1968b,p. 349)

A first step in attempts to elucidate the nature
of the organizational structure of impressions is to ex-
amine the direction or sequence of development. Cognitive
develoPmental theory prOposes that change is characterized
by two indices. One is increased independence of psycho-
logical processes from direct physical characteristics.
The other is increased subordination of concrete experience

to abstract integrative principles.

More specifically a lack of differentiation between
self and the world characterizes the early phases of
sensori-motor behavior. During this period cognitive
operations depend upon concrete manipulations of objects.
The next stage, concrete Operations of the perceptual
level, is marked by increased subject-object differentia-
tion. At this stage the individual is less dominated by
immediate concrete situations, but is still bound to that
which is directly given by the senses. At the conceptual
level or formal Operations, the individual symbolically
manipulates abstract representation between self and the
world; the concrete experience of objects during this
stage is subordinate to more abstract cognitions. This
develoPment is characterized by increased freedom from the
domination of the immediate situation and decreased de-
pendence on the immediate stimulus. This "permits a more
accurate assessment of others. The adult is more able
than the child to distinguish between the motivational
dynamics and the overt behavior of personalities."
(Werner 1957, p. 127)

This increased freedom from the environment is a
manifestation of the increased ability to differentiate
and integrate one's physical and social world. Langer

(1969) states that

Progressive differentiation and specification of
action systems, their functions and structures is one

side of develOpment. Functions are fused with each
other or they have no operative relationship to each
other and operate in rigidly fixed and unstably fluct-
uating fashion at the onset of development. They
progressively differentiate into increasing discrete
yet internally integrated means of action and intended
ends that operate in a flexible and stable fashion.
Structures are indefinite in boundaries and diffuse in
character and relationship to each other early in
develOpment. In the course of development they become
differentiated into increasing definite and well
articulated parts that are internally integrated.
Progressive hierarchic integration and centralization
of differentiated parts into the total mental organi-
zation is the other side of development. The evolving
relationship between action systems is directed toward
a structural hierarchization and functional integration
of parts to each other and to the whole mental organi-
zation. The result is the progressive formation of a
holistic mental organization such that most central
and develOped action systems increasingly subordinate
and regulate the less developed systems. (Lange 1969,
p. 92)

Harvey, Hunt and Schroder (1961) support these
prOpositions in their discussion of conceptual development.
For them concepts, in their matrix of interrelatedness,
serve the cognitive function of providing a system of or—
dering by means of which the environment is broken down
and organized, is differentiated and integrated into its
many psychological facets. Like Werner (1957) and Langer
(1969) they view differentiation and integration as the
psychological process or activity which is integrally
related to greater freedom from perceptual constraints.
This freedom is defined in terms of the level of concrete-

ness or abstractness of the individual conceptual system.

Variations in the level of concreteness-abstractness
results in differences in stimulus boundness, the

extent to which the receiving and responding individual
is restricted or can go beyond the physical character-
istics of the immediately impinging stimulus in or-
ganizing his evaluation of a situation. The greater
one's abstractness, the greater is his ability to
transcend immediacy and to move more into the tempor-
arily and spatially remote and the more capable he is
of abstracting relationship from objects of his ex-
perience and organizing them in terms of their inter-
relatedness. The greater one's concreteness on the
other hand, the greater the degree of stimulus "ought-
ness" in dictating his response, with concomitants
almost opposite to greater abstractness. Furthermore,
whereas the more abstractly functioning individual
tends toward differentiating his world into many
facets and integrating them holistically, but inter-
dependently, the more concretistically functioning
person is more likely to make only few differentia-
tions of his environment and to leave these cognitive
elements in a greater state of isolation. Thus, al-
though both the more concretely and more abstractly
functioning persons might display holistic cognitive
structures and subsequent responses, the underlying
basis of the two would be vastly different, the former
resting on an undifferentiated totality whereas the
latter underlaid by a differentiated and integrated
holism. (Harvey, Hunt, and Schroder 1965, p. 25)

Investigations of the organization of impressions
formed from inconsistent information could greatly benefit
from adapting a model similar to those previously dis-
cussed. Crockett (1970) has made this attempt and his
efforts are manifest in a scoring procedure for analyzing
the organization of impressions formed from inconsistent
information. He states his "system was designed to analyze
the degree of differentiation and the level of hierarchic
integration of impressions that have been formed from

contradictory information about another person." (Crockett

1970, p. 2.)

There are two aspects of organization in Crockett's
(1970) system, the degree of differentiation, and the
degree of relationship or integration. The scoring sys-
tem, based on these two aspects and other specified cri-
teria, identifies five general levels of organization,
each with three levels within, resulting in a fifteen
point ordinal scale which varies from unintegrated to

fully integrated. The five general levels are:

Aggregation.-—The representation of both desirable
and undesirable qualities without recognition of the in-
consistancy and without reconciling that inconsistancy.

In such impressions inconsistent information is differen-
tiated but not integrated.

UnivaZenee.--The inclusion of only undesirable or
only-desirable qualities in the impression. It is placed
higher than aggregation because such impressions usually
achieve a considerable degree of unity and integration
even though the inconsistency is not faced and the incon-
sistent elements are not reconciled.

Recognition of inconsistencH,--The inclusion of
both desirable and undesirable qualities and the recogni-
tion of their inconsistency without that inconsistency
being resolved.

Partial resolution.—-The inclusion and integration

of some but not all of the inconsistent qualities.

lO

Resolutions.--The inclusion of inconsistent quali-

ties and their integration within an organized whole.

Research Questions

The primary concerns of the present investigation
are:

(a) Does the organization of impressions of others
change as a function of age?

(b) Are there differences in the frequency and/or
nature of the degree of differentiation and inte—
gration as a function of age?

(c) What are the other variables that cause differences
in the organization of impressions within age

groups?

Overview

In Chapter II most of the previous developmental
studies of impression formation are reviewed and critiqued.
In addition related research in the area of impression
formation and ontogenetic changes in moral judgment will
be discussed.

The sample description, experimental design, de-
pendent variables, research hypotheses and statistical
analyses are presented in Chapter III. In addition Chapter
III contains descriptions of the stimulus material used in

the investigation.

11

The results of the statistical tests are presented
in Chapter IV and in Chapter V the results are discussed,
interpreted, implications are considered, and conclusions

and summaries are presented.

CHAPTER II

REVIEW OF THE LITERATURE

Introduction

 

This study, while not an exact replication of any
previous, is nevertheless based on the other research
findings. The literature is far from saturated with stu-
dies investigating the relationship of age to impression
formation or person perception.

Two areas of investigation are directly related to
the present study, while a third provides validity for the
.assumed theoretical orientation. Studies of impressions
formed from contradictory information and developmental
studies of impression formation and person perceptions
provide direct support for the present study; while stu-
dies of the ontogenetic changes in moral judgment support

the general theoretical rationale.

Impression Formation from
Contradictory-Information

 

 

The general experimental paradigm utilized in
developmental studies of impression formation presents

subjects with contradictory information regarding another

12

13

person and asks them to describe that stimulus person.
This model is based on the assumption that when given
information about another which is both positive and neg-
ative some amount of conflict or disequilibrium occurs.
When asked for their impression of the stimulus person,
the subject may employ different means to resolve the
conflict. It is the different organizational patterns of
these modes of resolution that can be ordered develop-
mentally. There are studies which utilize this paradigm,
but where age is not varied. They serve the heuristic
function of generating hypotheses and explanations that
might be employed in developmental investigations.

One experiment of this type was conducted by Gollin
(1954) who showed male undergraduates a silent motion
picture portraying a young woman in five different scenes.
In two scenes she was shown behaving in a manner which
suggested that she was promiscuous and immoral, while in
the other two scenes she was shown behaving in a manner
which suggested that she was kind and considerate. The
fifth scene which was neutral portrayed the women walking
along talking to a female companion. Subjects descriptions
of the young woman were analyzed into three predetermined
categories. In the least mature category subjects included
only one of the major character qualities. In this study
it was either promiscuity or kindness. Intermediately

complex impressions were aggregated and contained both

14

major character qualities, but did not in any way attempt
to relate them. The most mature impressions were inte-
grated, and characterized by the presence of both major
character qualities and explicit attempts to account for
their presence in the stimulus person.

Similar results were obtained by Haine and Grunes
(1950) and Hayden (1955). In all three investigations
(Gollin 1954, Haine and Grunes 1950, and Hayden 1955)
subjects most frequently formed simplified and aggregated
impressions, with the former occurring more frequently
than the latter. Besides providing clues to help define
different organizational patterns of impressions, these
results indicate modal responses at a specified age, which
would support a developmental hypothesis. Those subjects
who employed more advanced organizational patterns could
possibly have been more highly developed, in an ontogenetic
sense, than those who employed simplified impressions.

Additional clarification of the processes and fac-
tors involved in forming impressions from contradictory
information comes from studies concerned with the effects
of cognitive complexity. In these studies a cognitive
system is considered complex in structure when it (a)
contains a relatively large number of elements and (b)
the elements are integrated hierarchically by relatively
extensive bonds of relationship (Crockett 1965). The

possible relationship between cognitive complexity and

15

cognitive development adds further importance to these
investigations.

The procedure generally employed in these investi-
gations is to initially pretest subjects to determine
their level of cognitive complexity; this is then followed
by the presentation of some form of contradictory infor-
mation about another person. The subjects are requested
to describe the stimulus person in either a spontaneous
impression or a more structured response. The descriptions
are analyzed and categorized according to specific cri-
teria, and then related to the subjects level of cognitive
complexity.

Mayo (1959) and Rosenkrantz and Crockett (1965)
used similar designs. Their subjects heard information
about another in two univalent blocks. After each block
they were requested to write their impressions of the
stimulus person. Mayo (1959) found differences between
high and low complex subjects in regard to changes from
initial to later impressions. High cognitively complex
subjects employed less univalence in the final impression
and changed fewer of their initial impressions when the
conflict producing information was added. Low cognitively
complex subjects generally solved the conflict by either
ignoring part of the information given to them or by dis-
crediting the source of it (cited in Rosenbach 1968).

Thus high complex subjects seemed to modify their original

l6

impressions by assimilating the new information to their
earlier impressions while low complex subjects were bound
by the original block of stimulus. Rosenkrantz and Croc-
kett (1965) received similar results for males but not
females.

After pretesting for complexity Nidorf (1961)

presented subjects with the following story.

A young man was described by six people who knew him
well. The following are the salient characteristics
that have been abstracted from those descriptions;
pessimistic, competitive, intelligent, kind, self
centered and sensitive.

After hearing the story each subject was requested to
write an impression of the young man. The impressions
were then classified into integrated and non-integrated
categories. A point biserial correlation between com-
plexity and category of impression (unintegrated being
low and integrated high) indicated a significant positive
relationship.

Inga more recent study showing differences in the
organization of impressions, Meltzer, Crockett, and Rosen-
krantz (1966) varied cognitive complexity, level of relig-
ious values, and congruity of values with the stimulus
person. For aIl subjects a young man was described as
having both positive and negative characteristics; but for
one group the young man's religious values were congruent

with those of the subject while in a second group these

17

values were incongruent. When subjects received informa-
tion about the stimulus person who supposedly had similar
religious values complex subjects provided more highly
differentiated impressions, however when the information
was incongruent with the subjects values complexity was
unrelated to the number of constructs used to describe the
stimulus person.

In a study-by Crockett, Gonyea, and Delia (1970)
sex, type of stimulus, and complexity of subject were
varied. Each subject received a set of eight statements
characterizing a young man, and attributed to eight indi-
viduals who knew him well. In one condition each statement
consisted of an abstract personality trait with a brief
dictionary-type definition of its meaning. In the other
condition each statement described behaviors, which ac-
cording to the independent judges represented a concrete
realization of one of the abstract traits. After reading
their appropriate stimulus material subjects were instructed
to write an impression of the young man.

The descriptions were analyzed according to two
criteria. The degree of differentiation which is the
total number of traits a subject used to describe the
young man, and the level of organization which are an
ordered sequence of fifteen organizational patterns, which
a subject might employ in impression formation. Variance

in differentiation scores produced main effects for all

18

three independent variables. Descriptions from abstract
traits were more differentiated than those from concrete
traits, women produced more differentiated impressions
than men, and complex subjects had more highly differen-
tiated impressions than less complex subjects. In addition
there was an interaction between sex and complexity re-
flecting the fact that the difference between complex and
noncomplex subjects was greater for females than males.
The level of organization scores also showed a significant
main effect for the type of stimulus and complexity; with
the directions being the same as they were for the degree
of differentiation.

In summarizing these studies of the organization
of impressions from inconsistent information about another
it should be noted that different organizational patterns
exist and that these patterns are related to cognitive
complexity. What is of paramount importance is the fact
that the definition of complexity utilized in the reviewed
studies is similar to the definition of the developmental
process as employed by Piaget, Werner and the present
investigation. In addition evidence was presented to
support the existence of sex differences in organizational

patterns.

l9

Developmental Studies of
Impression Formation

After reviewing this section it will become appa-
rent how little attention has been paid to the investiga-
tion of the development of impressions formation with age.
This state of affairs is surprising since the theoretical
rationale for such studies is very rich. As was discussed
previously both Piaget and Werner have provided the neces-
sary theoretical base from which to launch such investiga-
tions.

The present section will be primarily concerned
with reviewing and critiquing those studies that have
investigated the changes in impression formation with age.

Gollin (1958) asked three groups of subjects with
mean chronological ages of 11.7, 13.6, and 16.2 to write
a description of the personality of a boy whose behavior
they had observed in a five scene movie. The five scenes
included a neutral scene showing the boy's full face and
profile, two scenes in which the boy is exhibiting socially
approved behavior and two scenes where he is exhibiting
socially disapproved behavior. The spontaneous descrip-
tions were analyzed according to two criteria; inference
or the attempt to go beyond just the action shown in the
movie and concept or the attempt to account for the diver-

sity of the behavior.

20

The results indicated age and sex differences.
The oldest subjects exceeded both groups of younger sub-
jects, and the thirteen year olds exceeded the ten year
olds in both inference and concept. Girls exceeded boys
in both the use of inference and concept at all ages, with
the exception of the youngest age group.

There were no main effects due to I.Q. level,
however certain interactions became manifest. In the
oldest group differences in I.Q. produced little vari-
ability in the use of inference. While in the thirteen
year old group females in both intelligence categories
exceed males in the same categories in the use of infer—
ence. In the youngest group subjects in the upper I.Q.
category exceeded subjects in the lower category in the
use of inference.

There was little use of concept in all groups of
the youngest subjects, subsequently producing little var-
iance as a function of I.Q. Of the thirteen year old
subjects the only group that frequently used concept was
the high intelligence females, while the other three
groups differed little from the younger subjects. In the
oldest age group subjects in the high intelligence group
exceeded like sexed subjects in low intelligence categor-
ies, with the low intelligence females exceeding high

intelligence males in the use of concept.

21

Gollin's (1958) results complemented by a fairly
strong methodology provide support for further investiga-
tions into the development of organization of impressions.
There are, however, improvements that should be made in
subsequent investigations. The range of ages must be ex-
tended so as to help provide a full description and under-
standing of the developmental trends in impression forma-
tion. This would also help to elucidate any develOpmental
sequence that may exist.

Gollin (1958) only employed two levels of I.Q.,
thus seriously restricting the possibility of finding I.Q.
differences. The importance of this point lies in the
fact that, if as assumed, social and cognitive develOpment
are not distinct realms, and there is a .70 correlation
between summed scores on Piaget tests and Binet scores
(Kohlberg l968a);there should then be a difference in the
organization of an impression as a function of I.Q.

Kohn and Fiedler (1961) investigated how differ-
ences in the age and sex of the perceiver affect their
judgment about themselves and significant others. Three
groups of subjects; high school freshmen, college freshmen
and college seniors were asked to describe themselves and
six significant others on a twenty six-point bipolar scale.
Two measures were obtained from the descriptions; the simi-
larity of ratings of each pair of persons and the favor-

ableness with which the subject described the person.

22

The results indicated significant main effects for
age and sex. Post-hoc analyses showed that older subjects
perceived more differences in personality traits among
their significant others than the younger subjects.
Females on the other hand perceived fewer differences in
personality traits than males.

From the vieWpoint of the present research this
study suffers from two serious limitations. First it
relied exclusively upon the subjects judgments on isolated
dimensions which were provided by the experimenter in the
checklist. This procedure made it impossible to determine
the natural differentiation a subject employs in his per-
ceptions of another. It also did not allow an investiga-
tion of how the subject would relate these traits to one
another. This latter point prevented a full understanding
of the organizational patterns of the different groups of
subjects' descriptions. Second, the age span of the sub-
jects did not include age groups in which the develop-
mentally early types of impressions were likely to occur.

In a position paper Kaplan and Crockett (1968)
provided the necessary theoretical organization for future
investigations of changes in organizational patterns of
impressions with age. They postulated three basic modes
of resolving inconsistent information about another person.
The three modes bare more than superficial resemblance to
a possible developmental sequence that could be supported

by cognitive developmental theory.

23

The three modes are; the primitive mode character-
ized by an exclusive concern with a single factor, namely
the various attributes or behaviors asserted by the stimulus
person with no utilization of nonmanifest factors that
might play a role. The intermediate mode which is charac-
terized by going beyond the information given, is an at-
tempt to resolve the manifest incompatibility of the
traits. The advanced mode explicitly considers a multi-
plicity of factors other than those that are directly
given. And in an attempt to resolve the inconsistency,
the subject may bring into play factors such as; the con-
textual variability of behavior, the perspectival select-
ivity of sources, deeper lying personality factors, etc.
Moreover these factors are not simply brought in side by
side, but are considered with respect to each other, so
that the sources and contexts likely to evoke the behavior
eliciting a particular characterization are more likely
articulated than in less advanced modes (Kaplan and
Crockett 1968, pp. 666-667).

A direct result of the Kaplan and Crockett (1968)
article is the scoring manual (Crockett 1970) employed in
the present study. The scoring procedure described in the
manual (Crockett 1970) provides criteria by which to con-
tent analyze the organizational structure of an impression
formed from inconsistent information. The quantified

variables derived from the content analysis in conjunction

24

with other criteria specified in the manual provide the
necessary scale by which to judge to which of fifteen
levels the organization of an impression belongs. The
organizational patterns represented by the levels range
from developmentally primitive to developmentally advanced.

Two previous studies of the ontogenetic development
of person perceptions have employed some edition of
Crockett's (1970) scoring procedure. Rosenbach's (1968)
investigation comes closest in methodology and hypotheses
to the present investigation. In individual interviews
three groups of subjects aged 6-7, 12-13, and 18-19 were
asked to give their impression of the personality of a
stimulus person. In addition they replied to probing
questions designed to clarify their impression of the
other person. The stimulus material was a silent movie
similar to that used by Gollin (1958). In two of the
scenes the person to be described portrayed two socially
approved behaviors, while in the remaining two scenes he
displayed socially disapproved behaviors. After the movie
each subject was requested to give a spontaneous descrip-
tion of the stimulus person, after which the series of
probing questions were administered.

The impressions and responses torthe questions
were then content analyzed according to three criteria;
inference or the number of constructs that go beyond the

information actually presented in the movie, the number

25

of links that connected different elements of the impres—
sion (as defined in Crockett's scoring manual). There was
a significant difference between ages in the frequency of
inference. The number of links and the level of organi-
zation both correlated positively with age.

There was no difference between the subject's
level of performance on the spontaneous description and
the probing questions. This result substantiates the
assumption that a group administered procedure is just as
effective in collecting spontaneous impression from speci-
fied stimulus materials as one that is individually
administered.

Unfortunately one severe methodological error
invalidates Rosenbach's (1968) results. Instead of em-
ploying the same stimulus material for each age, a differ-
ent film was presented to each group. The result was a
complete confounding of age and stimulus material. Another
less severe criticism is that I.Q. was not controlled.

The two younger groups were enrolled at a local Hebrew
School, while the older group was college freshmen.

Scarlett, Press, and Crockett (1971) conducted the
most recent investigation of developmental changes in
person perception. They asked three groups of boys from
grades 1, 3, and 5 to describe four persons; a liked boy,
a disliked boy, a liked girl, and a disliked girl. The

subjects were also requested to listen to taped recorded

26

stories about two boys and then relate the story back to
the interviewer, making sure to include descriptions of
the boys.

In support of their hypotheses there was a highly
significant increase with age in the average number of
constructs used to describe their peers. This was ac-
companied by a shift from concrete to abstract constructs
from the third to the fifth grade. There was also a sig-
nificant increase with age in the total number of con-
structs subjects used in their descriptions of the fic-
tional boys. Even though the interaction of age with the
type of construct used (concrete vs abstract) was not
significant; a supplementary analysis of the proportion of
abstract to total number of constructs showed a highly
significant effect for age.

What is of further significance is the fact that
at every age high differentiating subjects employed more
constructs in their impressions of the fictional characters
than did low. But in the youngest and oldest group the
difference between high and lows was accounted for entirely
by the constructs most typical of each group. First grade
highs used more concrete constructs than did lows, but
first grade highs and lows did not differ in the number of
concrete constructs. In contrast, fifth graders, highs
used more abstract constructs than did lows but highs and

lows did not differ in the number of concrete constructs.

27

For the third graders the differences between highs and
lows resulted predominately from differences in the number
of concrete constructs, the most typical type for that
group, but there was also a small difference in the number
of abstract concepts with highs giving more than lows as
expected (Scarlett, gt_§l., 1971, p. 449). These results
support a developmental stage hypothesis, by showing common
modes of response within specific age groups.

Like Kohn and Fielder (1961), Scarlett, Press, and
Crockett (1971) employed an extremely restricted sample of
ages, and do not deal with the integration of the descrip-
tion and organizational levels. However Scarlett, Press,
and Crockett's (1971) results did provide further support
for developmental changes in the organization of
impressions.

The results of the studies reviewed in this section
indicate that differences in the organization of impres-
sions are a function of age. And that with increasing age
an individual's written impression of another becomes more
differentiated and integrated. There is also evidence to
support age differences in levels or categories of impres-
sion formation. In addition, evidence exists to support
the possibility of sex and I.Q. differences within age

groups.

28

Developmental Changes in Moral

Judgment

Additional support for the present investigation
comes from Piaget's (1948) and Kohlberg's (1958, l963a)
studies of the development of moral judgment. Both these
investigations base their research on the assumption that
universal and regular age trends of development are found
in moral judgment and that they have a formal cognitive
base.

Besides the similar theoretical orientation, two
other factors make the above mentioned research relevant
to the present investigation. The experimental paradigm
is comparable. In moral judgment research subjects are
presented moral conflicts and asked how they would resolve
them. The results also bear great similarity to develop-
mental studies of the organization of impressions formed
from inconsistent information.

While Piaget (1948) attempted to define two stages
of moral development; heteronomous and autonomous, Kohlberg
(1958) indicated that these formulations were not com-
plete. Kohlberg (1958, l963a)studied the development of
moral judgment in the years ten to sixteen by utilizing an
interview containing ten hypothetical stories that posed
moral conflicts. The purpose of this procedure was to
obtain responses that would reveal the cognitive organiza-

tion behind a moral judgment. The children's responses,

29

in terms of the reasoning of their choices and their ways
of defining the conflict stories, shed much light on modes
of moral thinking. Evidence indicated that there were age
differences in modes of moral thinking and led to a devel-
0pmenta1 typology in which six sequential stages were
identified.

The earlier stages of moral judgment are based on
external happenings or physical needs expressed by charac-
ters in the stories, while in the intermediate stages
moral value resides in performing good or right roles in
maintaining conventional order and the expectancies of
others. Moral develOpment, as defined by the most highly
develOped stages, is a function of conformity by the self
to shared or sharable standards, rights or duties.

The stimulus boundaries, as exhibited during the
earlier stages of moral development, is similar to that
exhibited by the younger subjects in the develOpmental
studies of impression formation. The older subjects in
both cases are able to go beyond the information presented
and employ abstract concepts to resolve the inconsistency

or the dilemma.

Summary
Three areas of investigation were reviewed in this
chapter; studies of impression formation from inconsistent

information, developmental studies of impression formation,

30

and studies of the ontogenetic changes in moral judgment.
The results of these investigations clearly support the
rationale of research questions posed in Chapter I.

In summarizing the sgudies of the organization of
impressions from inconsistent information about another it
should be noted that different organizational patterns
exist and that these patterns are related to cognitive
complexity. What is of paramount importance is the fact
that the definition of complexity utilized in the reviewed
studies is similar to the definition of the developmental
process as employed by Piaget, Werner, and the present
investigation. In addition evidence was presented to sup-
port the existence of sex differences in organizational
patterns.

The results of the develOpmental studies of im-
pression formation indicate that differences in the organi—
zation of impressions vary as a function of age, and with
increasing age individual's written impressions of others
becomes more differentiated and integrated. There is also
evidence to support an age related shift from concrete to
abstract differentiation. In addition evidence exists to
support the hypothesis that females and subjects with
higher I.Q.‘s employ more advanced organizational patterns
than do males and subjects with lower I.Q.‘s.

Additional support for the present investigation

comes from studies of the development of moral judgment.

31

Besides the similar theoretical orientation two other
factors make these investigations relevant to the present
study. The experimental paradigm is comparable and the
results bear a great similarity to the results of devel-

opmental studies of the organization of impressions.

CHAPTER III

DESIGN OF THE STUDY

Sample

The sample for this investigation was drawn from a
central Michigan school district. Two fourth, eighth, and
twelfth grade classes were randomly selected from all
available classes of those respective grades. Even though
the unit of selection was class, the student still quali-
fied as the unit of analysis. This is supported by three
facts; (1) all classes within the sampling frame were
heterogeneously grouped on the basis of ability and sex,
(2) the stimulus material was recorded on video tape and
therefore constant for all six groups, and, (3) there were
no systematic differences as a function of class that
theoretically could have differentially affected the
results.

A total of 148 subjects participated in the present
study. They included 57 fourth, 48 eighth, and 43 twelfth
grade students. The average ages were 10 years, 13 years
7 months, and 18 years for the fourth, eighth, and twelfth
graders respectively. Tables 3.1 and 3.2 summarize the

information regarding sample size and age.

32

33

Table 3.1. Sample size.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

GRADE
4 8 12
E
3 s E x E
o
O B
H
M F M F M F
H 11 8 8 8 6 9 50
M 11 11 8 7 7 5 49
L 9 7 7 10 11 5 49
TOTAL 31 26 23 25 24 19 148
i

 

 

Table 3.2.

Average sample age.

 

GRADE

 

 

 

 

 

 

 

 

 

 

 

GRADE

 

 

 

 

 

 

12
E
.‘3 x E
O
O! E-1
H
M F M F M F
H 10-0 10-2 13-7 l3-9 17-11 17-10 13-10
M 10-3 10-1 13-6 13-6 18-0 17-11 13-11
L 10-6 10-4 13-7 13-8 18-5 18-0 14-1
AVERAGE 10-3 10-3 13-7 13-7 18-2 17-11
AVERAGE
FOR 10-3 13-7 18-0

 

 

35

Subjects were also classified on the basis of I.Q.
and sex. These variables were included because of their
relationship to cognitive development (Kohlberg 1968a) and
general development (Garai and Schenfeld 1968) and their
hypothesized relationship to the organization of impres-
sions formed from inconsistent information (Gollin 1958).

I.Q. scores were available for eighth and twelfth
graders from school records. The Otis Quick Scoring
Mental Abilities Test: Alpha Short Form was administered
to the fourth graders in order to obtain similar informa-
tion. The subjects were then classified into three levels
of intelligence, high = 110 or above, middle = 109 to 101
and low = 100 and below. The average I.Q. scores were
117.1, 105.4 and 90.9 for the high, middle and low sub-
jects reSpectively. The arbitrary selection of the out
off points was based on the total distribution of I.Q.
scores rather than on the distribution of the age and sex
subsamples. This provided identical I.Q. score cutting
points for all the groups. Table 3.3 summarizes the in-

formation regarding intelligence test scores.

Design
The design of the study was a 3 x 3 x 2 factorial

design; with subject being classified on the basis of

grade, I.Q., and sex as shown in Table 3.1.

36

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Table 3.3. Average sample IQ level.
GRADE
‘ 8 12
I—‘l
E . .4
{3 s E x g
0! E!
H 2
M F M F M F
H 113.2 114.2 121.8 115.4 115.8 122.3 117.1
M 104.7 105.5 106.9 105.7 104.0 105.6 105.4
L 91.2 90.6 88.9 89.0 91.0 94.8 90.9
rVERAGE1103.4 103.0 105.8 103.4 103.6 107.6 104.5
VERAGE
FOR 103.2 104.6 105.6 104.5

GRADE

 

 

 

 

 

 

37

Procedure

The data collecting procedure was conducted ac-
cording to the following schedule.

The subjects came to their regularly scheduled
classroom, where they were met by the experimenter and an
associate. The experimenter introduced himself as a
teacher and researcher from Michigan State University,
interested in how children form impressions of other
peoPle's personalities. The subjects were then instructed

by the experimenter in the following manner.

I am going to show you two short movies about two real
people. I want you to watch both movies closely. At
the end of each movie you will be asked to describe
the person, telling me what kind of person he or she
1s.

During these initial instructions the experimenter's

associate was placing answer sheets on the students' desks.

Please put your name, age and sex on the front page of
the paper that has been handed out. Are there any
questions? The first person you will see is Joel; he
will be the one I want you to tell me about. He is
wearing the striped shirt, and will be the only person
in every scene. So watch the movie closely. Please
don't say anything out loud; whatever you are thinking,
keep it to yourself, and tell it later when you will
be asked to describe the person. Are there any ques-
tions? O.K. here's the movie about Joel.

The movie is then shown and the experimenter states:

That's the end of the movie about Joel. Now I want
you to think about Joel and on the paper that was
handed out, write your impression of him. Tell me all
about him, what kind of child he is, so that I will
get to know him from your description. Take as much
time and write as much as you want.

38

This schedule was repeated twice, once for each
movie. Upon completing the description of the second
stimulus child, each subject was thanked for his or her

cooperation.

Stimulus Materials

The stimulus material consisted of two video-taped
vignettes. Each vignette is composed of five scenes. The
first scene of both vignettes was neutral. It showed the
object of the description in a portrait pose. This scene
was included so as to familiarize the subjects with the
object of their descriptions. Two of the remaining scenes
showed the object portraying socially acceptable behaviors,
while the other two showed him portraying socially unac-
ceptable behaviors. The order of the presentation of the
scenes was neutral, positive, negative, positive and
negative. This order was maintained for both vignettes.
The order of the presentation of scenes is not considered
relevant. Gollin (1959, p. 150) found that "neither
structure of the impression nor its affective quality is
altered by the order of presentation."

The first vignette observed by all the subjects
was about a boy named Joel. The following were the scenes

that comprised this movie:

Scene 1: This was the neutral scene. Joel enters the

field and looks straight at the camera. He is

Scene 2:

Scene 3:

Scene 4:

39

wearing a distinctively striped shirt, so that
he will be easily recognizable. During the
duration of Joel's presence on the screen the
narrator points out the fact that this is the

boy who they are to describe.

This scene was intended to depict unselfishness.
Joel and another boy are standing in front of a
Coke machine. Joel buys a coke, after which he
notices that the other boy does not purchase one.
He asks why and is told that the other boy does
not have any money. Joel then unselfishly offers
to give the other boy money to purchase a coke.

He does so, and the two boys walk off.

This scene was intended to depict dishonesty.

Two boys are in front of their lockers apparently
getting ready to go to gym. They finish changing
their clothes, close their lockers and walk away.
Joel comes in, looks around to see if anyone is
watching. He sees no one looking. He opens up
one of the boy's lockers and takes out his pants
removes the wallet from the rear pocket pulls

out the bill and then runs away.

This scene was intended to depict kindness.

Joel is seen sitting on a couch petting and

comforting a hurt dog.

Scene 5:

40

This scene was intended to depict lying. Joel
and another boy are sitting and talking when the
phone rings. The phone call is from another
friend of Joel's, who wants to know if he can
come out to play catch. Joel says he cannot
because he is sick and has to stay home. Joel
hangs up and says to his friend, "Let's go out
and play ball." Both boys pick up their base-

ball mits and caps and exit laughing.

The second vignette observed by all the subjects

was about a girl named Laura. The following are the

scenes that comprise the movie.

Scene 1:

Scene 2:

This was the neutral scene. It was exactly the
same as the first scene in the first movie, ex—
cept Laura was the actress. She was also wearing
a striped shirt, so that she would be easily

recognizable.

This scene was intended to depict helpfulness.
Laura and a boy are sitting in a waiting room.
The boy gets up to pick up a box of checkers.
The bottom part Of the box slips from his hands
and the checkers go all over the floor. He
begins to pick them up, but while doing so, he
hits his elbow on the chair standing nearby.
Laura comes over and offers to pick up the re-

maining checkers, which she does.

41

Scene 3: This scene was intended to depict rudeness. A
number of children are waiting at a water foun-
tain for a drink. Laura is at the end of the
line looking somewhat impatient. After a few
seconds she goes up and breaks into the front of
the line--much to the dismay of the other

children.

Scene 4: This scene was intended to depict consideration.
Laura is seen bringing books and flowers to her
sick friend. She stays for awhile and talks,

wishing her friend good health as she exits.

Scene 5: This scene was intended to depict greediness.
Laura and a little boy are seated on a couch
talking about school when a woman comes in and
offers them some candy. Laura is the first of-
fered. She grabs the bowl and refuses to let

the little boy have any.

Where it was apprOpriate a dialog was included in
the scene.
The stimulus material was presented via an audio-

video recorder and monitor.

Dependent Measures

The subjects' impression of the stimulus persons
were transcribed and coded by individuals other than those

who did the content analysis.

42

The impressions were content analyzed according to
the scoring procedure outlined by Crockett (1970). The
procedure was developed for the expressed purpose of
analyzing the organization of impressions which had been
formed from contradictory information about another person.
Additional scoring procedure is explained more fully in
Appendix A. The impressions were analyzed in regard to
five dependent variables:

(1) Degree of Concrete Differentiation. Any verbal
unit in the subjects descriptions that referred to charac-
teristics of the stimulus person which was either given
visually in the movie or directly seen in the situation.
These included activities performed, clothes and physical
traits.

(2) Degree of Abstract Differentiation. Constructs
which referred to abstract qualities of the stimulus per-
son, that is, to qualities that were not limited to a
specific context presented in the stimulus material.

(3) Degree of Connectedness is the proportion of pairs

 

of constructs and relevant conditions that are connected
by a linking pathway relative to the total numbers of
pairs of constructs and relevant conditions. A linking
pathway is said to exist between a pair of constructs, if
beginning with one of them it is possible to find a set of

links and constructs by which the second can be reached.

43

(4) Levels of Organization. Crockett (1970) defines
fifteen levels of organization. Content analyses of the
impressions in relation to descriptions of the levels were
conducted. The fifteen levels are:

Level 1: Minimal Aggregation. An aggregated im-
pression is one in which the stimulus information
is included without any recognition of the incom-
patibility, with no central theme for the impres-
sion, and without conveying the picture of a
unified person. An impression at Level 1, minimal
aggregation, is one which includes only a part of
the stimulus information and no inferences to

other domains.

Level 2: Typical Aggregation. Impressions scored
at Level 2 differ from the preceding only in that

they include all of the stimulus material but add

no additional inferences beyond the original

information.

Level 3: Extended Aggregation. Impressions scored
at this level go beyond the original information

by inferring other qualities in the stimulus per—
son, but they remain aggregated in that the incon-

sistency is neither recognized nor reconciled.

Level 4: Minimal Univalence. Univalent impres-

sions manage to resolve the inconsistency by

44

emphasizing only one side or the other of the
stimulus information. However, they do not rep-
resent the inconsistent qualities in any way in

the impression. From reading such impressions one
would think that only desirable or only undesirable
information had been available in the first place.
Univalent impressions at Level 4 are those which
include only a part of one side of the original

information.

Level 5: Simple Univalence. Impressions scored
at Level 5 differ from the preceding only in that
they include all of the desirable or the undesir-
able qualities, yet do not infer qualities beyond

those that were given in the stimulus information.

Level 6: Extended Univalence. Impressions at
this level present only desirable or only undesir-
able qualities, but they contain inferences to
attributes beyond those that were available in the

original information.

Level 7: Minimal Recognition of Inconsistency.

 

Impressions are scored at Level 7 when the recog-
nition of inconsistency is implicit, as indicated
by the use of "but," "nevertheless," and similar
expressions to connect one part of the information

to the other.

45

Level 8: Explicit Recognition of Inconsistency.
Impressions at this level explicitly recognize and
label the incompatibility, by such expressions as
"inconsistent," "contradictory," and the like—-but
do not go beyond labeling the inconsistency to
reconcile it. A Special case of impressions at
this level is rejection of the task, in which the
subject says no other person could actually display

such marked inconsistency in behavior.

Level 9: Recognition and Inference of Additional

Qualities. Impressions scored at Level 9 differ

 

from the preceding in the extent to which they
dwell upon the inconsistency and go beyond the
information to infer additional qualities. Fre-
quently, such impressions employ pseudo-personality
traits such as "split personality" or "schizo—

phrenic" to describe the inconsistency.

Level 10: Minimal Resolution of Inconsistency
Without Overall Organization. An impression at
this level provides a general theme which accounts
for the inconsistency; they may even include im-
plicit links among constructs, but they do not
have explicit links which provide a pathway be-

tween constructs of opposite valence.

46

Level 11: Some Resolution of Inconsistency Without
Overall Organization. An impression is scored at
Level 11 if there is an explicit linking pathway
from a positive construct domain to a negative
construct domain, but if (a) a substantial part of
the inconsistency is left unresolved and (b) there
is no central theme around which the inconsistency

is organized.

Level 12: Extensive Resolution of Inconsistengy_
Without Overall Organization. Impressions at this
level differ from those at Level 11 in having at
least two desirable and undesirable construct
domains of Opposite valence that are joined by

linking pathways.

Level 13: Some Resolution of Inconsistency With
Overall Organization. For an impression to be
scored at Level 13 or higher, there must be some
organizing theme or some integration of the various
links in the impression by means of some higher
order linking principle. To say that there must
be an organizing theme does not mean that the im-
pression must be accounted for by a cluster of
related constructs. Such constructs are often
motivational, referring to processes in the other

person that account for his varied behavior; they

47

may, however, be sociological, calling upon complex
role structures as the source of the apparent-in—
consistency in behavior. Such themes differ from
the use of vague personality dispositions and
delimited role behaviors--as in the "schiZOphrenic"
of subject L or the "in public . . . with compan-
ions" of subject 0--in being both more specific
about the processes that mediate the inconsistency
and more general in the range of constructs that
those processes account for. Impressions are
scored at Levels 13, 14, or 15 depending upon (a)
the number of stimulus domains that are reconciled
by the theme and (b) the extent to which the theme
is developed. Impressions which reproduce only a
small part of the original information and unify

it with a central scheme are scored at Level 13.

Level 14: Much Resolution of Inconsistency With
Overall Organization. Impressions scored at Level
14 differ from the preceding principally in the
prOportion of the inconsistency that is reconciled:
a good many, but not all, of the original qualities

are tied together by the organizing theme.

Level 15: Complete Resolution of Inconsistency
Without Overall Organization. The highest level

of organization within this system is found in

48

impressions which include all of the original in-
formation and which organize that information

through some central theme.

(5) Total Differentiation. This variable is the sum

of the concrete and abstract differentiations.

Reliability of the Scoring Procedure

Because of the nature of the variables, an inter-
rater reliability check was conducted. After all impres-
sions were scored by judge l, the impressions of 20% of
the total sample were randomly selected and subsequently
scored by judge 2. The relationship of the two judges,
on each dependent variable was obtained by using a Pearson
Product Moment Correlation. The interrater reliabilities
are; .78 for concrete differentiation, .94 for abstract
differentiation, .73 for degree of connectedness, .74 for
level of organization and .88 for total differentiation.

The reliability are summarized in Table 3.4.

Table 3.4. Interrater reliabilities of the two independent
judges for each dependent variable.

 

Concrete Differentiation .78
Abstract Differentiation .94
Degree of Connectedness .73
Level of Organization .74

Total Differentiation .88

 

49

Research Hypotheses

The following hypotheses were examined.

Hypothesis

Hypothesis

Hypothesis

Hypothesis

Hypothesis

Hypothesis

Hypothesis

1.

la.

1b.

1c.

1d.

le.

There will be a difference in the dependent

variables as a function of grade.

Subjects in lower grades will employ a
higher degree of concrete differentiation
in their impressions than subjects in

higher grades.

Subjects in higher grades will employ a
higher degree of abstract differentiation
in their impressions than subjects in lower

grades.

Subjects in higher grades will employ a
greater degree of connectedness in their

impressions than subjects in lower grades.

Impressions of subjects in higher grades
will represent higher levels of organiza-

tion than subjects in lower grades.

Impressions of subjects in higher grades
will be more highly differentiated than

subjects in lower grades.

There will be a difference in the dependent

variables as a function of I.Q. level.

Hypothesis

Hypothesis

Hypothesis

Hypothesis

Hypothesis

Hypothesis

Hypothesis

Hypothesis

2a.

2b.

2c.

2d.

2e.

3a.

3b.

50

Subjects with lower I.Q.‘s will employ a
greater degree of concrete differentiation
in their impressions than subjects with

higher I.Q.‘s.

Subjects with higher I.Q.‘s will employ a
greater degree of abstract differentiation
in their impressions than subjects with

lower I.Q.‘s.

Subjects with higher I.Q.‘s will employ a
greater degree of connectedness in their

impressions than subjects with lower I.Q.‘s.

Impressions of subjects with higher I.Q.‘s
will represent higher levels of organization

than subjects with lower I.Q.‘s.

Impressions of subjects with higher I.Q.‘s
will be more highly differentiated than

subjects with lower I.Q.‘s.

There will be a difference in dependent

variables as a function of sex.

Females will employ a greater degree of
concrete differentiation in their impres-

sions than males.

Females will employ a greater degree of
abstract differentiation in their impres-

sions than males.

51

Hypothesis 3c. Females will employ a greater degree of
connectedness in their impressions than

males.

Hypothesis 3d. Impressions written by females will repre-
sent higher levels of organization than

those written by males.

Hypothesis 3e. Impressions written by females will be more
highly differentiated than those written by

males.

Analysis of the Data

 

The relationship between the two impressions was
computed for each dependent variable. Pearson Product
Moment correlations for concrete differentiation, abstract
differentiation, and level of organization were .68, .72,
and .83 respectively. Because of the highly skewed dis-
tribution of the degree of connectedness scores, a phi
coefficient was computed for that variable. The coeffi-
cient was .54. A summary of the relationship between the
two impressions for each dependent variable is presented
in Table 3.5.

Because of the relatively high correlations between
the two impressions for each dependent variable, scores
were averaged across impressions. The averaged scores

were then used in the analysis of the data.

52

Table 3.5. Inter-impression reliability for each dependent

 

variable.
Concrete Differentiation .68
Abstract Differentiation .72
Degree of Connectedness .54
Level of Organization .83

 

A multivariate analysis of variance was employed
to test the main research hypothesis. The subhypotheses
were tested using a univariate analysis of variance; where
significant differences were observed, Scheffe post hoc

analyses were used to determine the locus of significance.

Summary

One hundred and forty-eight subjects categorized
on the basis of grade, I.Q. level, and sex were presented
two video taped vignettes. Each vignette was composed of
five scenes. The first scene of both vignettes was neu-
tral. Two of the remaining scenes showed an actor por-
traying socially acceptable behaviors, while the other two
showed him portraying socially unacceptable behaviors.
Upon the completion of each movie the subject was asked to
write his spontaneous impression of the stimulus person.

The impressions were content analyzed according to

the scoring procedure outlined by Crockett (1970). Using

53

the derived dependent variables a multivariate analysis

of variance was employed to test the research hypotheses.
Examination of the effects of grade level, on the

organization of the written impressions was the main pur-

pose of the study. Also examined were the effects of I.Q.

level and sex on the organizational patterns.

CHAPTER IV
ANALYSES AND RESULTS

Introduction

The results of the study will be presented in this
chapter. The three main research hypotheses were tested
using a multivariate analysis of variance, while a uni-
variate analysis of variance was used to test the sub-
hypotheses. For all the multivariate tests of the main
hypotheses .05 alpha level with the apprOpriate degrees of
freedom was used. Because the five univariate tests were
conducted simultaneously the .01 alpha level with the ap-
prOpriate degrees of freedom was used in order to compen-
sate for the inflation in the significance level. The .01
alpha level was also employed in the Scheffe post hoc
analysis of the significant univariate tests.

The results of the multivariate analysis of var-
iance is presented in Table 4.1. As can be observed a
significant interaction exists between grade and sex. The
presence of this interaction prevents a direct interpreta-
tion of any of the hypothesized multivariate main effects.
Therefore the univariate tests of the subhypotheses must

be considered. This will allow an examination of the main

54

55

effects of each independent variable. The remainder of
this chapter will present the results of the univariate
tests and any justified post hoc analyses that will assist

in elucidating the cause of a significant test.

Table 4.1. Multivariate analysis of variance for the
independent variables.

 

 

Significance

Effect Multivariate F df Level
Grade 23.8 10,252 P>.0000
IQ 3.1 10,252 P>.001
Sex 4.5 5,126 P>.0009
Grade x Sex 2.8 10,252 P>.0028
Grade x IQ 1.1 20,418 P>.3930
IQ x Sex 1.7 10,252 P>.0789
Grade x Sex x IQ .49 20,418 P>.9689

 

The individual cell means will be frequently cited
throughout the chapter. For convenience, all cell means,
for each dependent variable, will be presented in Appendix

B.

The Effects of Grade on the
Organization of Impressions

Hypothesis 1. There will be a difference in the dependent

variables as a function of grade.

56

As was previously stated the significant inter-
action between grade and sex prevents any-direct interpre-
tation of the multivariate tests. Consequently the
univariate tests will be considered. The results of these

tests are presented in Table 4.2.

Table 4.2. Univariate analyses of variance for the ef-
fects of grade for each dependent variable.

 

Between
Dependent Mean Significance
Variables Squares F df Level

 

Degree of concrete
differentiation 188.92 63.39 2,130 P>.0000

Degree of abstract
differentiation 373.44 69.96 2,130 P>.0000

Degree of

connectedness 0.14 14.24 2,130 P>.0001
Level of
organization 443.62 144.90 2,130 P>.0000

Degree of total
differentiation 175.34 8.57 2,130 P>.0004

 

Because of the significant multivariate interaction
between grade and sex, graphic presentations and descrip-
tions of grade cell means will be made relative to sex of
subject. The presentation of these ordinal interactions

will help elucidate the existence of the main effects.

57

Hypothesis la. Subjects in lower grades will employ a
higher degree of concrete differentiation

than subjects in lower grades.

The F ratio of 63.39 (df 2,130) was statistically
significant. The data supported the hypothesis of differ—
ences in the degree of concrete differentiation. As can
be observed in Figure 4.1 the differences are in the pre-
dicted direction. Irregardless of sex fourth graders
(X=4.84) exceed both eighth (X=2.07) and twelfth (X=l.11)
graders in the utilization of concrete differentiation.
However the results of the Scheffe post hoc analysis indi-
cated that only the differences between twelfth and fourth
graders (-3.73 i 2.13) and eighth and fourth graders

(-2.77 i 2.07) were statistically significant.

 

 

6+
8 Fourth
4J9:
010 5*
H~4
8‘5
O«4 4»
UJJ
.48
OH 3;-
0
332: Eighth
a3 21- //
g Twelfth
lb /
0 l l
Males Females

Figure 4.1. Interaction of grade and sex on the degree of
concrete differentiation.

58

Hypothesis 1b. Subjects in higher grades will employ a
higher degree of abstract differentiation
in their impressions than subjects in lower

grades.

The difference between grades in the use of ab—
stract differentiation was statistically significant
(F=69.96). The relationship between the three grades
relative to sex of the subject is presented in Figure 4.2.
The apparent differences between the grades, irrespective
of sex, are obvious. However only the differences between
the twelfth (X=7.17) and fourth (X=l.57) graders (5.60 i
2.86) and eighth (X=4.44) and fourth graders (2.87 i 2.77)

reach statistical significance.

 

 

9.
" 8
88 " Twelfth
Hw47
$1"; '
fi~45.
‘é
44:115- Eighth
OH
0’4- /
831
H-H3-
U‘QZ
g ' ggg 1 Fourth
1.
0 .L I

 

Males Females

Figure 4.2. Interaction of grade and sex on the degree of
abstract differentiation.

59

Hypothesis 1c. Subjects in higher grades will employ a
higher degree of connectedness in their

impressions than subjects in lower grades.

The difference in grade means for the degree of
connectedness was found to be statistically significant
(F=14.24 df 2,130). Therefore this hypothesis was ac-
cepted. As-can be observed in Figure 4.3, twelfth graders
(X=.101) exceeded both eighth (X=.077) and fourth (X=.007)
graders. Even though the univariate test was significant
none of the post hoc comparisons between means of students

at different grades were significant.

0 0
rd H
o +4
I

O O

O 0
co \0
If

Twelfth

. C
O O
U1 m
r I

Eighth

Degree of Connectedness
OO O

wc- \l

I I

—.01_ __4 Fourth

7——

I l

Males Females

 

Figure 4.3. Interaction of grade and sex on the degree of
connectedness.

60

Hypothesis 1d. Impressions of subjects in higher grades
will represent higher levels of organiza-

tion than subjects in lower grades.

This hypothesis was supported by the data (F=l44.90
df 2,130). The apparent differences between grade relative
to sex are presented in Figure 4.4. These differences are
supported by post hoc analyses between each possible com-
bination of grades. Twelfth graders (X=9.00) utilized a
statistically significant higher level of organization
than eighth (X=6.53) graders (2.47 i 2.25) and fourth
(X=3.08) graders (5.92 i 2.16). While the impressions of
eighth graders represented a statistically significant
higher level of organization than those of fourth graders

(3.45 i 2.10).

 

 

€310 _
"-19.
.|.J
g 8.. Twelfth
'3 7.
(U '\ o
3' 6.. Eighth
o 5_
“a

4 ’ Fourth
H
a) 3)- /
5
1-1 2'

1.

0 i 44

Males Females

Figure 4.4. Interaction of grade and sex on the level of
organization.

61

Hypothesis 1e. Impressions of subjects in higher grades
will be more highly differentiated than

subjects in lower grades.

The data also supported this hypothesis (F=8.57 df
2,130). Figure 4.5 depicts the differences between grades
relative to sex. There is little difference between fourth
(X=12.78) and eighth (X=12.95) graders, while twelfth
graders (X=16.50) exceed both younger groups. A Scheffe
post hoc analysis indicated that there was no difference

between grades in any of the possible grade comparisons.

21.

Twelfth
18.
15.

Eighth

/ Fourth

0‘ \O
W T

Degree of Total Differentiation
w '5
1 T

 

O 1 l
Males Females

 

Figure 4.5. Interaction of grade and sex on the degree of
total differentiation.

62

The Effects of I.Q. Level on the
Organization of Impressions

Hypothesis 2. There will be a difference in the dependent

variables as a function of I.Q. level.

As was previously stated the interaction between
grade and sex prevents any direct interpretation of the
multivariate tests. Instead the subhypotheses will be
considered. Where justified, Scheffe post hoc analyses
were conducted in order to determine the cause of each
significant test.

The univariate tests for the effects of I.Q. level

for each dependent variable are presented in Table 4.3.

Table 4.3. Univariate analyses of variance for the effects
of I.Q. level for each dependent variable.

 

 

Between
Dependent Mean Significance
Variables Squares F df Level
Degree of concrete ,-
differentiation .03 .11 2,130 P>.8933
Degree of abstract
differentiation 25.98 4.87 2,130 P>.0092
Degree of
connectedness .02 2.41 2,130 P>.0935
Level of
organization 25.74 8.41 2,130 P>.0004

Degree of total
differentiation 125.05 6.11 2,130 P>.0030

 

63

Hypothesis 2a. Subjects with lower I.Q.‘s will employ a
greater degree of concrete differentiation
in their impressions than subjects with

higher I.Q.‘s.

The differences in means between I.Q. levels were
not found to be statistically significant (F=.ll df 2,130).
The actual means were virtually identical; high = 2.76,
middle = 2.63 and low = 2.63, therefore the hypothesis was

not supported.

Hypothesis 2b. Subjects with higher I.Q.‘s will employ a
greater degree of abstract differentiation
in their impressions than subjects with

lower I.Q.‘s.

The obtained F ratio of 4.87 (df 2,130) was sta-
tistically significant. Subjects in the high I.Q. group
(X=5.12) exceeded subjects in the middle (X=4.10) and low
(X=3.96) I.Q. categories in the use of abstract differen-
tiation. However the post hoc analyses indicated that
none of the possible differences between I.Q. groups,
independent of other factors, were statistically

significant.

Hypothesis 2c. Subjects with higher I.Q.‘s will employ a
greater degree of connectedness in their

impressions than subjects with lower I.Q.‘s.

64

For the high I.Q. group, the mean degree of con—
nectedness was .071, while the mean for the middle I.Q.
group was .075 and .038 for the low I.Q.“group. The dif-
ference between these means was not statistically signifi~

cant (F=2.4l df 2,130).

Hypothesis 2d. Impressions of subjects with higher I.Q.‘s
will represent higher levels of organization

than subjects with lower I.Q.‘s.

The data showed that there were statistically sig-
nificant differences (F=8.41 df 2,130) in the levels of
organization employed by the different I.Q. groups. The
mean level of organization for high I.Q. subjects (X=6.81)
was significantly different (1.49 i .99) than the mean of
low I.Q. subjects (X=5.32). In addition the difference
between the means of middle I.Q. subjects (X=6.73) and low
I.Q. subjects was also statistically significant (1.05 i

.99).

Hypothesis 2e. Impressions of subjects with higher I.Q.‘s
will be more highly differentiated than

subjects with lower I.Q.‘s.

The differences between I.Q. groups, for the degree
of differentiation was found to be statistically signifi-
cant (F=6.ll df 2,130). However the post hoc comparisons
between high (X=15.67) and middle (X=13.44), high and low
(X=13.1l) and middle and low I.Q. groups were not

significant.

65

The Effects of Sex of Subject on
the Organization of Impressions

Hypothesis 3. There will be a difference in the dependent

variables as a function of sex.

The significant interaction between grade and sex
prevents any direct interpretation of the multivariate
test. Instead the univariate tests of the subhypotheses
will again be considered. The results of the univariate
tests for the effect of sex on the dependent variables are

presented in Table 4.4.

Table 4.4. Univariate analyses of variance for the effects
of sex for each dependent variable.

 

 

Between
Dependent Mean Significance
Variables Squares F df Level
Degree of concrete
differentiation 14.93 5.01 1,130 P>.0270
Degree of abstract
differentiation 26.33 4.93 1,130 P>.0281
Degree of
connectedness .07 6.89 1,130 P>.0098
Level of
organization 9.58 3.13 1,130 P>.0793

Degree of total
differentiation 325.94 15.93 1,130 P>.0002

 

66

Because of the grade x sex interaction graphic
presentations and subsequent descriptions of sex group

mean will be presented, relative to grade.

Hypothesis 3a. Females will employ a higher degree of con-
crete differentiation in their impressions

than males.

This hypothesis was not supported by the data
(F=5.01 df 1,130). However the means of females at each,
grade exceeded the means of males of the same grade. This
is graphically presented in Figure 4.6. Fourth, eighth,
and twelfth grade females had means of 5.37, 2.18 and 1.38
respectively, while the means for the fourth, eighth, and

twelfth grade males were 4.30, 1.95 and .84 respectively.

DJ :5 U1
l 1 1

Differentiation

N
I

Females

Degree of Concrete

1 ~ Males

 

0 . 1 I . IJ
4th 8th 12th

Figure 4.6. Interaction of sex and grade on the degree of
concrete differentiation.

67

Hypothesis 3b. Females will employ a higher degree of ab-
stract differentiation in their impressions

than males.

The obtained F=4.93 (df 1,130) was not statistically
significant. As can be observed in Figure 4.7 fourth grade
males (X=l.57) and females (X=l.56) had almost identical
means. But the higher grade females (8th X=4.80, 12th
X=8.24) manifested a greater degree of abstract differen-

tiation than males (8th X=4.07, 12th X=6.10).

 

 

# 9 .
338- Females
333'
m to 7F
£136. Males
%;§ESL
$444.
9311
3‘33 .
Q 2 .

1L

4 l 9
0 4th 8th 12th

Figure 4.7. Interaction of sex and grade on the degree of
abstract differentiation.

Hypothesis 3c. Females will employ a greater degree of
connectedness in their impressions than

males.

68

There was a statistically significant difference
(F=6.89 df 1,130) between males and females in the degree
of connectedness employed. However the difference was not
in the predicted direction as shown in Figure 4.8. Eighth
(X=.ll3) and twelfth (X=.l37) grade males exceeded females
(X=.041, X=.064) in those respective grades; while there
was very little variation between fourth grade males

(X=-.005) and females (X=.12).

.14 . Males
.13.
.12.
.11.
.10.
.09.
.08.
.07.
.06. Females
.05.
.04.
.03.
..02.
.01.
.00.
-'01F

Degree of Connectedness

 

l 1 .1
V 7

4th 8th 12th

Figure 4.8. Interaction of sex and grade on the degree of
connectedness.

69

Hypothesis 3d. Impressions written by females will repre-
sent higher levels of organization than

those written by males.

The F ratio of 3.13 (df 1,130) was not significant,
therefore this hypothesis was not accepted. There were
however apparent sex differences, which are represented in
Figure 4.9. At grade four females (X=3.34) employed a
higher level of organization than males (X=2.82). While
eighth (X=6.89) and twelfth (X=9.84) grade males exceeded

eighth (X=6.l6) and twelfth (X=8.16) grade females.

 

 

10. Males
g 9-
38 Fl
3 . ema es
N 7.
'36
mi
3‘5.
0

4r
‘H
O3_
82
>‘b
311.

0

1 1 !_1
4th 8th 12th-

Figure 4.9. Interaction of sex and grade on the level of
organization.

Hypothesis 3e. Impressions written by females will be more
highly differentiated than those written by

males.

70

The test of this hypothesis showed that there was
a statistically significant difference between sexes on
the degree of total differentiation (F=15.93 df 1,130).
As can be observed in Figure 4.10 the hypothesized direc-
tion was supported. Females at all grades (4th X=13.84,
8th X=13.89, 12th X=l9.l9) exceeded males (4th X=11.71,

8th X=12.00, 12th X=13.81) in the same grade.

 

 

 

2l_
Females

was 13.
3.9.
o+JlS_ -
Pug ::;———fl—fl—*/”””,,; Males
733 12_
m3}
010 9.
833
0w: 6
on: '

3.

0 | l ——I

4th 8th 12th

Figure 4.10. Interaction of sex and grade on the degree
of total differentiation.

Summary of the Results

 

The presence of a significant multivariate inter-
action between grade and sex prevented a direct interpre-
tation of any of the hypothesized multivariate main ef-
fects. Instead the univariate tests of the subhypotheses

were considered.

71

There was a statistically significant main effect
for grade for each dependent variable; and all differences
between grade levels were in the hypothesized direction.
Post hoc analyses indicated significant grade differences
for the following comparisons.

(a) Degree of Concrete Differentiation.

(l) twelfth graders vs fourth graders

(2) eighth graders vs fourth graders

(b) Degree of Abstract Differentiation
(l) twelfth graders vs fourth graders
(2) eighth graders vs fourth graders

(c) Degree of Connectedness--none

(d) Level of Organization ‘

(l) twelfth graders vs eighth graders

(2) twelfth graders vs fourth graders

(3) eighth graders vs fourth graders

(e) Degree of Total Differentiation--none

There was a statistically significant main effect
for I.Q. level for the degree of abstract differentiation,
the level of organization, and the degree of total differ-
entiation; all significant hypotheses tests were in the
predicted direction. Post hoc analyses indicated signifi-
cant I.Q. level difference for the following comparisons.

(a) Degree of Abstract Differentiation—-none

(b) Level of Organization

72

(1) high I.Q. vs middle I.Q.
(2) high I.Q. vs low I.Q.
(3) middle I.Q. vs low I.Q.

(c) Degree of Total Differentiation--none

There was a statistically significant main effect
for sex of subject for the degree of connectedness and the
degree of total differentiation. However, of the signifi-
cant hypotheses tests, only the one concerned with the
degree of total differentiation was in the predicted

direction.

CHAPTER V

SUMMARY AND CONCLUSIONS

P11122086

This study was conducted in order to investigate dif-
ferences between ages in the organizationcmfimpressions formed
from inconsistent information. The research questions and

strategies were based on cognitive developmental theory.

Literature Review

 

Three areas of investigation were reviewed in~
Chapter II: studies of impression formation from incon—
sistent information, develOpmental studies of impression
formation, and studies of the ontogenetic changes in moral
judgment. The results of these investigations clearly
support the research questions.

In summarizing the studies of the organization of
impressions from inconsistent information about another it
should be noted that different organizational patterns are
related to the cognitive complexity of the subject. What-
is of paramount importance is the fact that the definition
of complexity utilized in the reviewed studies is similar

to the definition of the developmental process as employed

73

74

by Piaget, Werner, and the present investigator. In addi-
tion evidence was presented to support the existence of
sex differences in organizational patterns.

The results of the developmental studies of im-
pression formation indicate that differences in the or-
ganization of impressions vary as a function of age; and
with increasing age an individual's written impressions of
another become more differentiated and integrated. There
is also evidence to support-a; age related shift from con-
crete to abstract differentiation. In addition evidence
exists to support the hypothesis that females and subjects
with higher I.Q.‘s employ more advanced organizational
patterns.

Additional support for the present investigation
came from studies of the development of moral judgment.

A similar theoretical orientation and two other factors
made these investigations relevant to the present study.
The experimental paradigms are comparable and the results
bear a great similarity to the results of the studies of

the organization of impressions.

Design of the Study

 

One hundred and forty-eight subjects, categorized
on the basis of grade, I.Q. level, and sex were presented
two video-taped vignettes. Each vignette was composed of

five scenes. The first scene of both vignettes was

75

neutral. Two of the remaining scenes showed an actor por-
traying socially acceptable behaviors, while the other two
showed him portraying socially unacceptable behaviors.
Upon the completion of each movie the subject was asked to
write his spontaneous impression of the stimulus person.
The impressions were content analyzed according to
the scoring procedure outlined by Crockett.(l970). Using
the derived dependent variables a multivariate analysis of
variance was employed to test the research hypotheses.
Examination of the effects of grade level on the
organization of the written impressions provided evidence
relating to the main purpose of the study. Also examined
were the effects of I.Q. level and sex on the organiza-

tional patterns.

Results

The major findings can be summarized as follows:

(1) There was a statistically significant main effect
for grade for each dependent variable.

(2) There was a statistically significant main effect
for I.Q. level for the degree of abstract differ-
entiation, level of organization, and the degree
of total differentiation.

(3) There was a statistically significant main effect
for sex of subject for the degree of connectedness,

and the degree of total differentiation.

76

Post hoc analyses were conducted in order to

determine the cause of a significant hypothesis test.

Discussion

The Effects of Grade on the
Organization of Impressions

The grade-related hypotheses of the present in-
vestigation appear substantially confirmed. Analyses of
the subject's descriptions showed that with increasing
grade or age there was a decrease in the degree of con-
crete differentiation and an increase in abstract differ-
entiation. In addition impressions written by the older
subjects were more integrated and represented higher
levels of organization. These results are completely
congruent with the cognitive developmental theories of
Werner (1957) and Piaget (1960).

According to cognitive developmental theory two
processes define development; the degree of differentiation
and integration. Development, as a function of these
processes is most succinctly defined by Werner's onto-
genetic principle: "Whenever development occurs it pro-
ceeds from a state of relative globabity and lack of
differentiation to a state of increasing differentiation
articulation and hierarchic integration." (Werner 1957,

p. 126)

77

If total differentiation is considered the major
dependent measure of differentiation, the prediction of an
increase in differentiation with age is only partially
substantiated. There was virtually no difference between
the fourth and eighth graders in the degree of total dif-
ferentiation, while twelfth graders differed substantially
from the two younger groups.

There was however a qualitative difference between
the fourth and eighth graders. The modal response was
concrete differentiation for the former and abstract dif-
ferentiation for the latter. The quantitative difference
between the twelfth and eighth graders was predominately a
function of increased abstract differentiation by the
older group.

Cognitive develOpmental theory assumes that children
in the early stages of development do not differentiate
between themselves and the external world. Consequently
there is no representation of objects or people, forcing
the child to understand the world by direct manipulation.
As he grows older the child develOps cognitive Operations
that are characterized by increasing subject-Object dif-
ferentiation. Initially this differentiation is based on
information directly given by the senses. At some later
stage this heavy reliance on sensory acquired information
disappears and the individual can transcend the immediate

bounds of the existing stimuli and begins to employ

78

abstract phenomena in his attempt to differentiate and
subsequently understand his world.

The develOpmental process outlined in the previous
paragraph describes exactly what occurred in the present
investigation. The younger subjects exceeded the older
subjects in the degree of concrete differentiation. They
represented the stimulus person on the basis of their ac-
tivities, clothing and physical appearance to a greater
degree than the older subjects. While the older subjects
employed a greater degree of abstract differentiation.

In their impressions the concrete characteristics of the
stimulus person were subordinate to motivational, dispo-
sitional and other abstract constructs.

Cognitive developmental theory also predicts a
progressive hierarchic integration or centralization of
differentiated parts into a total mental organization.
This postulate was also substantiated by the data. However
none of the groups utilized a substantial amount of ex-
plicit links. It might possibly be concluded that the
explicit integration of diverse social constructs might
not become manifest in overt behavior until very late in
develOpment.

The degree of differentiation and the amount of
integration along with other specified criteria define
Crockett's (1970) different levels of organization. Be—

sides providing an opportunity to examine a subject's

79

organizational pattern as a whole, the level of organiza-
tion also Operationalizes the cognitive mode utilized in
processing the inconsistent information.

There were statistically significant differences
between grade pairs in all possible pairwise grade com-
parisons for the variable of level of organization. How-
ever the frequency of responses for each grade level as
presented in Table 5.1 might provide a clearer indication

of normative response patterns for particular ages.

Table 5.1. Frequency of the levels of organization in
each grade.

 

Levels of Organization

 

Grade 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

 

4 18 41 32 - - 3 20 - — - _ _ - _ _
8 - 2 18 - — 12 42 1 4 16 - 1 - - -
12 - - 2 - - 5 8 1 10 54 6 - - - -

 

As can be observed in Table 5.1 the frequencies of
the levels of organization for the fourth grade are highly
skewed. Seventy-nine percent of the impressions written
by fourth grade subjects represented either levels one,
two or three. These three levels represent aggregated
organizations; the chief characteristic of which is a lack

of recognition of the inconsistency. The subjects repeat

80

what they saw in the stimulus material without any attempt
at trying to synthesize the information. Their inability
to integrate prevents them from realizing that the infor-
mation is incongruent. They are bound by the immediate
and forced to function in an associative fashion, respond-
ing only to what they had concretely experienced.

Forty-nine percent of the impressions written by
the eighth graders utilized levels seven, eight, and nine.
The salient characteristic of these levels is the recogni-
tion of the inconsistency, but with no attempt to resolve
it.

Of the eighty-six impressions written by the sen-
iors, seventy percent were either levels ten or eleven.
Both of these levels represent attempts at resolution,
but with no overall organization.

The effects of abstract differentiation and in-
creased integration probably account for the advanced
levels of organization manifested by the older subjects.
With increased freedom from the physical constraints of
the stimuli, more accurate impressions of the individual
will be made. Instead of just reacting in an associative
manner the subject is able to symbolically represent the
information, and subsequently compare the abstract quali-
ties characterized by the concrete behaviors. The com-
parison leads to the conclusion that an inconsistency

exists.

81

The eighth graders functioned at this level, while
the oldest group went one step beyond and attempted to
resolve the contradiction. An obvious explanation for the
advanced mode employed by the twelfth graders is that they
possessed a greater ability to integrate the contradictory
information.

Based on these data it may be concluded that
children of different ages treat inconsistent information
differently. And that the modes employed are a direct
function of the organization of the information, which in
turn is, to a great extent, determined by the development
of the child's cognitive operations.

The question as to whether the differences in.
organization are part of an invariant sequence of stages
cannot be answered by this study. .However the conditional
sequence outlined by the different age groups might provide
support for a stage hypothesis. There cannot be resolution
without recognition, while recognition must obviously fol-

low nonrecognition.

The Effects of I.Q. on the
Organization of ImpressiOns

 

I.Q. was included in the present investigation as
an external criterion of cognitive develOpment. If social
and cognitive development are not distinct realms, and

there is a .70 correlation between summed scores on Piaget

82

tests and Binet scores (Kohlberg 1968a) there should then
be a difference in the organization of an impression as a
function of I.Q. Three of the dependent variables differed
significantly between I.Q. groups; the degree of abstract
differentiation, the level of organization, and the degree
of total differentiation. All differences were in the
predicted direction, of high I.Q. subjects exceeding mid-
dle and low subjects and middle subjects exceeding low
subjects. There were virtually no differences between
I.Q. groups on the degree of concrete differentiation,
while high and middle I.Q. subjects differed from low I.Q.
subjects on the degree of connectedness. These results
can possibly be attributed to the fact that concrete dif-
ferentiation represents a relatively low level ability
which subjects at all I.Q. levels are able to manifest.
While the degree of connectedness may require more advanced
abilities which are not characteristic of low I.Q. sub-
jects. If I.Q. is assumed to be a criterion of cognitive
develOpment, and if it is further assumed that there is a
relationship between cognitive and social development,
then these results are consistent with cognitive develop-
mental theory.

The variation within grades as a function of dif-
ferences in I.Q. should not be interpreted to mean that
individual differences play a large role in determining

organizational patterns of impressions. Rather it should

83

be construed that the variation in cognitive develOpment
within grades, as measured by I.Q. scores results from the
same differences in develOpment that produces variation in

the components-and levels of organization.

The Effects of Sex on the
Organization of Impres§iOns

The interaction between grade and sex became
clearly manifest when the sex main effects were considered.
It was predicted that females would out perform males on
all indices of organization. However the data indicated
that females exceeded males on only the degree of concrete
differentiation, the degree of abstract differentiation,
and the degree of total differentiation with the latter
being the only statistically significant difference.

The only other significant main effect was on the
degree of connectedness. The interpretation of this test
is complicated by the existence of a disordinal interac-
tion. Eighth and twelfth grade males exceeded the re-
sPective females, while fourth grade females exceeded
fourth grade males. The same pattern repeated itself on
the level of organization, but without significance.

Female superiority on the measures of differentia-
tion substantiates similar findings (Crockett, Gonyea, and
Delia 1970, Gollin 1958). Theoretical support for these

results can be drawn from three sources, the advanced

 

84

maturation of females, the social orientation of females,
or the greater verbal fluency of females (Garai and
Scheinfeld 1968).

Even though females mature at a faster rate than
males, it is questionable what effect this has on the
development of cognitive competencies. In fact females
tend to be superior on quick perception of detail, verbal
fluency and rote memory, while males tend to excel on the
higher order abilities of reasoning and problem solving
(Garai and Scheinfeld 1968, p. 251). Consequently ad-
vanced maturation does not lead to more highly developed
Operations that might be instrumental in causing females
to be superior on differentiation measures.

There is a considerable amount of documentation
supporting the greater social orientation and verbal
fluency of females. In a review of sex differences Garai

and Scheinfeld (1968, p. 252) state:

The earlier speech develOpment and greater verbal
fluency of girls appear to be related to the earlier
maturation of their speech organs, their closer con-
tact with their mothers, and their greater interest
in people. . . . Furthermore the social environment
tends to encourage responsiveness to parents and com-
pliance with their demands at an earlier age and more-
insistently with regard to girls than with regard to
boys.

The difference in differentiation scores between
males and females might not be a function of advanced
development but rather the ability of females to articulate

more of their differentiated world.

85

In the interactions on the degree of connectedness
and level of organization fourth grade females exceed
fourth grade males, while an increasing superiority of
males begin sometime between the fourth and eighth grades.
The initial superiority of females is perplexing and defies
any rational explanation. However the same abilities that
allow males to excel in problem solving and abstract
reasoning most likely account for their later superiority
on these two measures. Integration is Operationalized by
the degree of connectedness and essential for the utiliza-
tion of higher levels of organization, just as it is nec-
essary for bringing together the diverse elements in

correct problem solving.

Implications

 

The most obvious implication derived from the
present investigation is that it provides further support
for the fact that individuals at different times in their
lives process information differently. And that the pos-
sible number of abilities and behaviors that these differ-
entials effect might be extremely pervasive.

The investigation and validation of these differ-
ences promote a fuller understanding of children. The
results should assist anyone who is charged with the

responsibility of facilitating socialization.

86

There are specific recommendations that can be
derived from this study that could assist educators in
their attempts to promote socialization. Traditionally,
the schools, and especially the elementary grades, have
placed a great emphasis on the recognition and recall of
information, with less attention paid to higher level
abilities. This situation could be the result of a lack
in the availability of assessment instruments. The vig-
nettes and procedure employed in this study might provide
a method for assessing higher level abilities, Specifically
the ability to synthesize diverse information.

A logical result of such an assessment should be
attempts to promote skills or abilities that are necessary
to process information at higher levels. Teacher training
programs might include units on strategies that would
assist students in increasing their ability to differen-
tiate and integrate.

Assuming that the student is developmentally ready
these strategies could be modeled after those utilized in
efforts to accelerate cognitive and moral development. In
such efforts a state of dis-equilibrium is generated in
the subject by bringing to his attention the possibility
that there are ways, other than his, to look at a partic-
ular piece of information. In relation to the task per-
formed in this investigation, subjects could be asked

questions that would require more abstract differentiation.

87

"Why did Joel steal the money from the locker?" Or, "Why
do you think Laura was so nice to her girl friend after
being so rude to others?" Increased integration could be
promoted by asking the student to explain how the same
child could exhibit two diverse types of behavior.

The task performed by the subjects in this study
approximated an essay type test. The data indicated that
because of their ability to respond to information in dif-
ferent fashions some children might perform more favorably
on this form of examination. Consequently, teachers
should be aware of their student's modal response levels
so as to insure that what he is attempting to measure is
not confounded by a student's inability to respond in the
required way.

A less direct implication has to do with the
school's responsibility to assist in areas of socializa-
tion, other than those directly concerned with formal
education. Kohlberg (1968b)states that social develOpment
tends to lag behind cognitive development. If this is
true, and it is assumed that the School's responsibility
lies in areas other than those traditionally defined as
the charges of the educational establishment, then teachers
should be responsible for closing the gap between social
and cognitive development. This could possibly be accomp-
lished through the use of the strategies previously

discussed.

88

A further implication might be derived from the
relationship between an individual's level of organization
and stereotypic perceptions of others. If younger children
utilize concrete characteristics to form impressions of
others, and if these characteristics carry Specific conno-
tations, then a child's feeling for another might be com-
pletely determined by an individual's observable features.

Certain physical qualities might carry such strong
connotations that they are still processed at less advanced
levels. Concrete characteristics such as skin color, sex
or age may remain paramount in impressions of older per-
sons, even though their modal organizational pattern is
much higher.

If such a situation exists, it could possibly be
eliminated by training at the specific times when an indi-
vidual is moving from a concrete dominated level to a more
abstract pattern. The training procedures could follow
the paradigm utilized by efforts to accelerate cognitive'

and moral develOpment.

Limitations

 

The major limitation of this study is that it does
not provide data that would support a cognitive develop-
mental theory as compared to a learning theory interpreta-
tion. The ideal method for testing the difference between

the two theories would be a longitudinal investigation.

89

But even then there are no methodological procedures that
would provide unequivocal support of a cognitive develOp-
mental theory.

However, a longitudinal investigation would provide
the opportunity to observe whether differences in the
organization of impressions are gradually acquired and can
be explained by previous environmental contingencies. Or
whether the differences occur suddenly and without any
previous justification. Results could be further validated
by attempting to accelerate subjects from one level to the
next or by attempting to extinguish naturally acquired
responses. Additional support can be gleaned by adminis-
tering conceptually Similar tasks such as those employed
by Piaget and Kohlberg and correlating the subject's score
on these tasks with his level of organization.

Within the confines of this study, the inclusion
of a wider age range would have provided a better indica-
tion of the complete Span of organizational patterns.
However, if younger subjects were tested the procedure
would have to be modified so as to take into consideration
differences in the ability to write.

There are also specific limitations in the general-
ization of the results. Interpretations from the results
must be qualified according to the age and other demo-
graphic characteristics of the subjects, the type of

stimulus material employed, and the scoring procedure.

REFERENCES

REFERENCES

Bieri, J. Cognitive complexity and judgment of incon-
sistent information. In R. P. Abelson, E. Aronson,
W. McGuire, T. Newcomb, M. Rosenbery, and T.
Tannenbaum (Eds.), Theories of Cognitive Complex-
ipy: A Sourcebook. Chicago: Rand McNally and
Co., 19681

 

Crockett, W. Cognitive complexity and impression forma-
tion. In B. R. Maher (Ed.), Pro ress in Experi-
mental Personality Research. VOi. 3. New York:
Academic Press, 1965.

 

Crockett, W. Structural Analysis of the Organization of
Written Impressions. Mimeographed,il970.

 

Crockett, W., Gonyea, A. and Delea, A. Cognitive complex-
ity and the formation of impressions from abstract
qualities or from concrete behaviors. Presented
at the American Psychological Association 1970.

Flavell, J. H. The Developmental Psychology of Jean
Piaget. Princeton: Van Nostrand, 1963.

 

Garai, J. and Schenfeld, A. Sex differences in mental and
behavioral traits. Genetic Psychology Monographs,
77, May 1968, 169-299.

Gollin, E. S. Forming impressions of personality.‘ Journal
of Personality, 1954, 23, 65-76.

4 Gollin, E. S. Organizational characteristics of social
judgment: A developmental investigation. Journal
of Personality, 1958, 26, 139-154.

Haire, M. and Grunes, W. F. Perceptual defenses: Processes
protecting an original perception of another per-
sonality. Human Relations, 1950, 14, 649-659.

 

Harvey, J. J., Hunt, D. E. and Schroder, H. M. Conceptual
Systems and Personality Organization. New York:
John Wiley and Sons, Inc., 1961.

 

90

91

Hayden, R. Some evidence for conflict resolution in im-
pression formation. Journal of Abnormal and
Social Psychology, 1955, 5177302-307.

 

Hays, W. L. Statistics for Psychologists. New York:
Holt, Rinehart and Winston, 1963.

 

Kaplan, B. and Crockett, W. DevelOpmental analysis of
modes of resolution. In R. P. Abelson, E. Aronson,
W. McGuire, T. Newcomb, M. Rosenberg and P.
Tannenbaum (Eds.), Theories of Cognipive Con-
sistency: A Sourcebookl Chicago: Rand McNally
and Co., 1968.

 

Kohlberg, L. The development of modes of moral thinking
in the years ten to sixteen. Unpublished doctoral
dissertation, University of Chicago, 1958.

Kohlberg, L. Moral develOpment and identification. In
H. W. Stevenson (Ed.), Child Psychology. 62nd
Yearbook of the National Society for the Stpéy of
Education. Chicago: University of Chicago Press,
l963a.

 

Kohlberg, L. The development of children's orientation
toward a moral order: Sequence in the development
of moral thought. Vita Humana, l963b, 6, ll-33.

 

Kohlberg, L. Preschool education: A cognitive develop-
mental View. Child Development, 1968a, 39, 1013-
1062.

 

inpKohlberg, L. The cognitive develOpmental approach to

3 socialization. In D. A. Goslin (Ed.), Handbook of
Socialization Theory and Research. Chicago: Rand
McNaIly and Co., 1968b.

Kohn, A. and Fielder, F. Age and sex differences in the
perception of persons. Sociometry, 1961, 24,
157-164.

 

Langer, J. Theories of Development. New York: Holt,
Rinehart and Winston, 1969.

Mayo, C. Cognitive complexity and conflict resolution in
impression formation. Unpublished doctoral dis—
sertation. Clark University, 1959.

. 4'
a
r
n
o
'5
3

92

Meltzer, B., Crockett, W., and Rosenkrantz, P. Cognitive
complexity, value incongruity, and the integration
of potentially incompatible information in impres-
sions of others. Journal of Personality and Social
Psychology, 1966, 4, 338-343.

 

Nidorf, L. J. Individual differences in impression forma-
tion. Unpublished doctoral dissertation, Clark
University, 1961.

Piaget, J. The Moral Judgment of the Child. Glencoe:
Free Press, 1948.

Piaget, J. The Psychology of Intelli ence. Patterson:
Littlefield, Adams and Co., 1 0.

Rosenbach, D. Some factors affecting reconciliation of
contradictory information in impression formation:
an ontogenetic study in person perception. Unpub-
lished doctoral dissertation, Clark University,
1968.

Rosenkrantz, P. and Crockett, W. Some factors influencing
the assemelation of disparate information in im-
pression formation. Journal of Personality and
Social Psychology, 1965, 2, 397-402.

 

.Scarlett, H., Press, A., and Crockett, W. Children's

descriptions of peers: A Wernerian developmental
analysis. Child Development, 1971, 42, 439-453.

 

Werner, H. Comparative Psychology of Mental Dcvelopment.
New York: International Universities Press, 1948.

 

Werner, H. The concept of development from a comparative
and organismic point of view. In D. B. Harris
(Ed.), The Concept of Development. Minneapolis:
University of Minnesota.Press, I957.

APPENDICES

APPENDIX A

RULES FOR SCORING

Rules for Identifying Concrete Characteristics

Any verbal unit in the subject's impression that
referred to characteristics of the stimulus person which
was either given visually in the movie or directly seen in
the Situation. These included activities performed,

physical traits and clothing.

Rules for Identifying_Constructs

Rule 1: Whenever it is debatable whether a phrase
Should be scored as one construct or several constructs,
or when two nearly synonomous, but not identical constructs
are mentioned in an impression, the subject should be
given the benefit of the doubt and multiple constructs
should be scored instead of just one.

Rule 2: When the subject uses an adverbial or
adjectival qualifier which as used appears to be an in-
trinsic part of the noun it modifies, contrary to Rule 1,
the two words are to be scored as one construct.

Rule 3: Identical, repeated words or phrases are
scored only once. Words, which are very similar in mean-
ing, but which are not identical are scored twice.

Rule 4: Idiomatic constructions which run to

several words are usually scored as one construct.

93

94

Rule 5: Only qualities which are relevant to the
task the subject has been set should be scored as
constructs.

Rule 6: General statements about what people
Should do or about the nature of mankind are not scored
as constructs unless they are specifically tied to char-

acteristics of the person who is being described.

Relevant Conditions

 

Relevant conditions Specify-the conditions under
which one or another aspect of the object person's behavior
will be displayed. Verbiage is scored as a relevant con-
dition under two circumstances:

(1) When it refers to some observation, condition, or
circumstance that is not an inference about the
other person is like nor about the reaction he
elicits from others, but which answers the question
"Under what circumstances will he act," thus, or
"When will he Show this quality?"

(2) When it refers not to the other person's behavior
as such, but to aspects in the reporters which

affect their interpretations of the behavior.

These expressions often play an important role in
relating constructs to one another in an impression. They
are not counted when the differentiation of the impression

is scored. Their omission from the score for degree of

95

differentiation is based on the fact that the constructs
they tie together are each recorded independently while
the relevant conditions, themselves, refer not to the
qualities of the other person but to the conditions within
which these qualities are displayed or the circumstances

that lead to their inference by some reporters.

Integration of Constructs

The relationship among constructs define the
manner in which the differentiated parts are organized.
Links, as they are labeled by Crockett (1970) can either
connect constructs of the same or Opposite valences.
Links do not manifest themselves as explicitly as con-
structs, consequently, two rules must be kept in mind when
scoring relationship.

Rule 1: A link Should not be scored between con-

structs unless the connection is explicitly made

in the subject's impression. Links are never
scored when the connection is implicit.

Rule 2: A link is scored only if it connects two

constructs which the subject used to describe the

other person.

Types of Links

Crockett (1970) recognizes five types of links.

They are:

96

Reinterpretation, or the subsumption of one construct to

 

another. This type of link usually accomplishes much the
same purpose as would an outright denial of one of the
qualities. However, in reinterpretation the quality is
not denied so much as its apparent manifestation is ex-
plained in terms of some other construct, usually of the

opposite valence.

Differences amopg Observers. Subjects frequently ascribe
the various reports of the other person's behavior to dif-
ferences in the perspectives, personalities, or role posi-
tions of those who made the reports. Links of this type
use expressions such as "Friends . . . enemies . . .", or
"those who know him . . . casual acquaintances. . . ." For
a link of this type to be scored, it is essential that the
subject be specific as to the differences that have pro-

duced the differential reports.

Variation in Contexts. Under this heading fall links
which prOpose that the other person's behavior varies with
the situation or the context in which he finds himself.
For such a link to be scored, it is necessary that those
aspects of the context be identified which elicit the

different behavior patterns.

Role Relationships. Links in this category differ from

 

those differences among observers and variation in context

by specifying the social role relationships which set the

97

contexts for the other person's differential behavior.
Again, a link is scored only when the impression unambig-

uously identifies the role relationships that are involved.

Dispositionalgpalities. The final type of link involves
the explicit accounting for one or more other constructs
by some underlying motivational or dispositional quality.
Once more, the connection between one construct and
another must be explicit in the impression. Isolated
motivational or dispositional statements such as "he wants
to succeed" or "he has an inferiority complex" do not
qualify as links because they are not connected explicitly

to other constructs.

 

9 8
EXAMPLES

The score sheets of the following two impressions
are presented as examples of the previously described
rules.

Impression 1

He lend his friend some money to buy some pop. Then he
went into the room with his friend when they left he ran
into the room again he went in the locker and got some
money then he went home and play with his puppies then the
phone rang Joel answer it it was his dad Joel said he
wasn't felling Good then he left.

Impression 2

He's thoughtful and kind toward his friends and his dog,
yet he is a thief. At times he's very sure of himself;
however other times he seemed insecure. He wants approval
from his peers. I didn't dislike him, but I didn't like
him either--he's "just another" kid.

Impression one was written by a fourth grader and
represents level 2 or a typically aggregated impression.
Other scores are 4 for degree of concrete differentiation,
0 for the degree of abstract differentiation, 0 for degree
of connectedness, and 4 for the degree of total
differentiation. .

Impression two was written by a senior and exemp-
lifies level 10. It also received a 0 on the degree of
concrete differentiation, 7 on the degree of abstract
differentiation, .14 on the degree of connectedness and
7 on the degree of total differentiation.

Share

    
   

lend his friend
some money
to buy pOp

 

 

went into room
with friend

 

99

Dishonest

 
  

  
   
   

nent in the
locker and got
some money

 

Kind

went home and

played with
his puppies

 

Liar

     

said he wasn't
feeling good

 

 

 

 

 

 

 

100
+ 0 -
Considerate Greedy
very sure
of himself
(r1123 approval
.from his peers
{jggtianother
kid
Helpful Rude

thoughtful 43——

o/

; [toward friends]

 

 

 

I insecure)

APPENDIX B

MEANS OF EACH DEPENDENT VARIABLE

101

 

 

 

 

 

 

 

 

 

 

 

 

Table B-1. Means for the degree of concrete differentiation.
GRADE
4 8 12

A

1 m
E o
'4 SEX g
<5l <2
H

M F M F M F

H 4.39 5.45 2.04 2.26 .93 1.46 2.76
M 4.26 5.33 1.91 2.14 .80 1.33 2.63
L 4.26 5.33 1.91 2.14 .80 1.34 2.63
m
(D
g 4.30 5.37 1.95 2.18 .84 1.38 2.67
d
m
0
g 4.84 2.07 1.11 2.67
d

 

 

 

 

 

 

 

Table B—2.

102

Means for the degree of abstract differentiation.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

GRADE
8 12

04

g m

m 3
*4 SEX E
o': a:
H

M F M F M F

H 2.30 2.29 4.80 5.53 6.83 8.97 5.12
M 1.27 1.26 3.77 4.51 5.81 7.95 4.10
L 1.13 1.12 3.63 4.37 5.67 7.81 3.96
til

0

E 1.57 1.56 4.07 4.80 6.10 8.24 4.39
E

m

U

E 4.44 7.17 4.39
d

 

 

103

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Table B—3. Means for the degree of connectedness.
GRADE
12
A
E I'd
w is”
"' SEX m
a E
H
M F M F M F

H .010 .015 .124 .053 .149 .075 .071
M .013 .019 .128 .056 .152 .079 .075
L -.028 —.022 .087 .015 .111 .038 .038
El
(9
g -.005 .012 .113 .041 .137 .064 .061
a
[:1
0
g .007 .077 .101 .061

_;§

 

 

 

104

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Table B-4. Means for the level of organization.
GRADE
4
A
9 8 '
F11
'4 SEX g
c? m
H .
M F M F M F
H 3.43 3.95 7.49 ”6.77 10.45 8.77 6.81
M 3.08 3.61 7.14 6.43 10.10 8.43 6.47
L 1.94 2.46 6.00 5.28 8.96 7.28 5.32
El
(9
g 2.82 3.34 6.89 6.16 9.84 8.16 6.20
a
El
(9
g 3.08 6.53 9.00 6.20
.33

 

 

 

105

 

 

 

 

 

 

 

 

 

 

 

 

Table B-5. Means for the degree of total differentiation.
f
GRADE
12
A
g M
m E
'4 SEX E
O: .4:
H
M F M F M F
H 13.31 15.44 13.60 15.49 15.40 20.78 15.67
M 11.08 13.21 11.37 13.26 13.18 18.56 13.44
L 10.75 12.88 11.04 12.93 12.85 18.23 13.11
m
o
g 11.71 13.84 12.00 13.89 13.81 19.19 14.07
a
m
(D
g 12.78 12.95 16.50 14.07
é

 

 

 

 

 

 

 

 

”71111111111311!flififlifliflliflfllfliflllflmfi’