Citation
Temperamental and developmental dimensions of self-control in middle childhood

Material Information

Title:
Temperamental and developmental dimensions of self-control in middle childhood
Creator:
Waugh, Mark Howard
Copyright Date:
1982
Language:
English

Subjects

Subjects / Keywords:
Chess ( jstor )
Child psychology ( jstor )
Clinical psychology ( jstor )
Cognitive psychology ( jstor )
Ego ( jstor )
Factor analysis ( jstor )
Intelligence quotient ( jstor )
Self control ( jstor )
Temperament ( jstor )
Temptation ( jstor )

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
09320736 ( oclc )
ABU6015 ( ltuf )
0028558400 ( ALEPH )

Downloads

This item has the following downloads:


Full Text












TEMPERAMENTAL AND DEVELOPMENTAL DIMENSIONS OF
SELF-CONTROL IN MIDDLE CHILDHOOD








BY



MARK HOWARD WAUGH


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



UNIVERSITY OF FLORIDA


1982





























COPYRIGHT 1982

BY

MARK HOWARD WAUGH
















ACKNOWLEDGMENTS
Acknowledgments I wish to express must be located within my educa-

tional experiences in clinical psychology. First of all, for first

indicating a sense of the excitement of psychology, I thank George

Welsh, Ph.D., who taught and inspired me as an undergraduate at the

University of North Carolina. For providing lasting influence, I

express deep gratitude to the committee members, teachers, supervi-

sors, role models, colleagues, and good people who are Drs. Hugh

Davis, Jacquelin Goldman, Roger Blashfield, and Mary McCaulley. Dr.

Saundra Damico has assisted well in my labors in the dissertation

process. I also must thank my friends and colleagues R. Peter Fortney

and Leslie Morey, Ph.D., for their friendship and intellectual stimu-

lation. And most importantly, to my family, especially my wife and

son, Marcia S. Waugh and Robert M. Waugh, my heartfelt thanks for

years of encouragement and tolerance.


iii


I












TABLE OF CONTENTS


ACKNOWLEDGMENTS . . .

LIST OF TABLES . . . .

ABSTRACT . . . . .


CHAPTER

I INTRODUCTION AND FRAMEWORK .. ..


. . . iii

. . . vi

. . . vii


. . . 1


A Proposed Model of Personality
Self-Control . .
Temperament . . .
Ego Development . .


II METHODOLOGY . . .


. . . 17


Subjects . .
Measures . .
Activity Level .
Self-Control .
Independent Measures
Procedure . .
Statistical Analyses. .
Factor Analysis of De


. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
Dendent Measures .


Multiple Regression of Independent and
Dependent Variables . .


. . 25


III RESULTS. . . . . ... . .... 26

Preliminary Analyses. . . . . ... 26
Interrater Reliability Assessment of the SCT
Ratings . . . .... .26
Psychometric Assessment of the Activity Schedule 27
Dimensionality of the Set of Dependent Variables. . 27
Relations of Age, IQ, and Ego Level to Factor
Dimensions . . . . ... ... .34
Post Host Factor Analyses . . . ... .34


i j i i















ly DISCUSSION . . .


Principal Findings . .
Evaluation of Methodology .
Implications of the Study .


REFERENCE NOTES . . . .


REFERENCES . . . . .

APPENDICES . . . . .


DRAW-A-LINE . . .

SENTENCE COMPLETION TEST FOR BOYS.

ACTIVITY SCHEDULE . .

PARENT RATING FORM . .

RESISTANCE TO TEMPTATION OBJECTS .

COMPLETE CORRELATION MATRIX. .

DESCRIPTIVE STATISTICS . .


. . . 47


. . 48

. . 55


. . . . 55

. . . . 56

. . . . 57

. . . . 59

S . . . 63

S . . . 64

S . . . 65


BIOGRAPHICAL SKETCH . . . . .... ....... b


. . . .













LIST OF TABLES



Table Page


1. Major Systems of Temperament Dimensions. . . 8

2. Some Milestones of Ego Development . . .. 11

3. Principal Axes Factor Analysis of Dependent Variables with
Varimax Rotation. . . . .... .29

4. Principal Axes Factor Analysis of EASI-III Scales with
Varimax Rotation. . . . .... .32

5. Summary Statistics of Regressions of Age, IQ, and ego
level on Factor Scores. . . . ... 35

6. Principal Axes Factor Analysis of EASI-III Scales and
Dependent Variables with Varimax Rotation. . .. .37

7. Principal Axes Factor Analysis of EASI-III Scales,
Dependent, and Independent Variables with Varimax
Rotation. . . . . ... .... 39

















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


TEMPERAMENTAL AND DEVELOPMENTAL DIMENSIONS OF
SELF-CONTROL IN MIDDLE CHILDHOOD

by

MARK HOWARD WAUGH

August 1982

Chairman: Hugh C. Davis, Jr., Ph.D.
Major Department: Clinical Psychology

Psychological study of self-control was reviewed in terms of a

proposed model of personality. This model of personality considers

developmental and dispositional constructs to be orthogonal and super-

ordinate sectors of personality. Precedents to this model in the

literature include similar models and the distinction between develop-

mental and trait variance. When applied to the phenomena of self-

control, this model suggested that the temperamental dimension of

activity level would be dispositional and that other measures of

impulsivity would be developmental. The major developmental dimension

of self-control was hypothesized to be ego development. Previous

research on self-control has indicated a multidimensional character,

but the earlier research lacks a comprehensive conceptual base and

suffers methodological shortcomings. The present study measured seven

variables thought to reflect either activity level or developmental


vii













self-control on seventy-eight normal male school children of middle

childhood (ages six to twelve). A principal axes factor analysis

resulted in three dimensions. The first factor was interpreted as

either a "difficulty" of parentally perceived temperament or a method-

specific dimension. The second factor was seen as an activity level

dimension. The third factor was interpreted as a self-control dimen-

sion. Independent variables of age, intellectual ability, and level

of ego development were used to predict factor scores on the factor

dimensions to test for their developmental status. Factor one was

independent of age, IQ, and ego level. Factor two was associated with

IQ, but post hoc analysis suggested IQ was a correlate as opposed to

an intrinsic component of the activity level dimension. Factor three

was related to both age and level of ego development, but ego develop-

ment contributed minimal independent variance. Factor three was seen

to be a developmental component of self-control. Results of the study

were discussed in terms of the importance of the developmental and

dispositional distinction, the importance of the proposed model of

personality sectors, and implications to personality factors in child

psychopathology.


viii

















CHAPTER I

INTRODUCTION AND FRAMEWORK

The aim of this investigation is to explore the construct of

self-control within a proposed model of personality. Self-control is

taken to be a reflexive property of the self (Mead, 1964), which

involves volitional regulation of impulse or wish, usually in the

service of goal attainment. In recent years, self-control and im-

pulsivity have become rapidly expanding topics of research (e.g.,

Goldfried and Merbaum, 1973; Meichenbaum, 1979; Bandura, 1977; Karoly,

1977; Kendall and Finch, 1978; 1979). This explosion of research is

characterized primarily by a cognitive-behavioral orientation. Langer

(1969) has grouped theories of behavioral development in terms of

their allegiance to three general models: the psychoanalytic, the

mechanical mirror (behavioral), and the organic lamp (cognitive-

developmental). Although isolated research efforts have been made

into self-control from the psychoanalytic perspective (e.g., Davids,

1969) and the organic lamp model (e.g., Rozek, Wessman, and Gorman,

1977), it is clear that the area of self-control or impulsivity re-

search is dominated by the mechanical mirror image of man. One pur-

pose of the present study is to further the marriage of diverse litera-

tures and orientations within the study of self-control.













Loevinger's (1966a; 1966b; 1976) theory of ego development is a

theory of self and self-development which is explicitly psychoanalytic

and cognitive-developmental in character. Loevinger (1976) has articu-

lated and synthesized much of the common ground explored by Freud

(1953), Sullivan (1953), Piaget (1932), and Kohlberg (1964). Loevinger

(1976) incorporates self-control as one of a number of central dimen-

sions of ego development. Therefore, in considering that ego develop-

ment includes self-control, that ego development preserves the most

central principles of psychoanalysis (Loevinger, 1966b), and that ego

development possesses the breadth of scope intrinsic to cognitive-

social developmental models (Davison, King, Kitchner, and Parker,

1980), it is reasonable that the theory of ego development holds high

promise of providing an understanding of self-control as an aspect of

an integrated theory of personality development.



A Proposed Model of Personality

The conceptual framework for this study of self-control consists

of a proposed model of personality. This model posits two major and

orthogonal dimensions of personality. The horizontal dimension or

axis includes such constructs as temperament, interpersonal style, and

cognitive style. The vertical dimensional set consists of primary de-

velopmental constructs such as ego and cognitive development. This

model is not entirely original; there are precedents. For example,

Swensen (1980, Note 1) invokes a similar notion in his personality

scheme consisting of ego development and Leary's (1957) interpersonal












style system. Additionally, Royce and colleagues seem to be utilizing

a similar model in their conceptual and empirical work on individuality

theory, which involves integration of self and cognitive style theory

(see Wardell and Royce, 1975; 1978). Another precedent to this pro-

posed model is implicit in Weisz, O'Neill and O'Neill (1975) and

Achenbach and Weisz (1975). These investigators have drawn the dis-

tinction between cognitive ability and cognitive style. The former is

considered a developmental variable and the latter a trait variable.

They have shown independent developmental and trait components of

variance within the domains of reflectivity-impulsivity and field de-

pendence-independence. This distinction between development and trait

is identical to that made between horizontal and vertical sectors of

personality.

This proposed model of personality is not finely articulated nor

has the model been subject to programmatic research. In fact, given

the relative absence of scientific confirmation, the model's value is

best seen as heuristic. Nevertheless, some preliminary support for

the model exists in other empirical studies. For example, Lorr and

Manning (1978) found very few significant correlations between a large

set of interpersonal style variables and the sentence completion test

of ego development. Furthermore, MaCrae and Costa (1980) found ego

development unrelated to neuroticism and extraversion (measured by the

Eysenck Personality Inventory), and also unrelated to the temperament

dimensions of emotionality, sociability, and activity level (using

Buss and Plomin's (1975) EASI-III temperament survey). MaCrae and










Costa (1980) also found ego development and the inhibition of impulse

subscale of the EASI-III impulsivity scale to be independent, although

they do not provide the results for the full impulsivity scale.

Accordingly, insofar as interpersonal styles and temperament dimen-

sions appear orthogonal to ego development, the proposed model gathers

credibility. Fleshing out the full structure and ramifications of

this proposed model will require programmatic research beyond the

scope of the research project herein. A portion of this model, how-

ever, is used as the conceptual framework to guide this study. That

is, the independence of ego development, as a vertical dimension of

personality, and temperament dimensions, as a horizontal axis, is

accepted.



Self-Control


The terms self-control and impulsivity are taken to be inter-

changeable. The psychological literature on self-control is quite

diverse (Karoly, 1977), and often this literature is segregated con-

ceptually and empirically into foci such as child versus adult or

treatment versus assessment. The present study seeks to establish

self-control as an aspect of ego development and thereby making arti-

ficial such distinctions as child versus adult approaches to the study

of self-control. Self-control is considered an aspect of the theory

of ego development and is thus. hypothesized to be orthogonal to tempera-

ment dimensions. Self-control, then, is viewed as qualitatively dif-

ferent from the temperamental variable of activity level.














Self-control or impulsivity has often been regarded as multidi-

mensional and implicated in accounts of hyperactivity, conduct dis-

order, and learning disability (Block, Block, and Harrington, 1974;

Homatidis and Konstantareas, 1981; Paulsen and Johnson, 1980). The

proposed model of personality described previously would suggest that

previous investigators have confounded measures of temperament (activity

level) with ego development (self-control). Confounding activity

level with self-control assures a spurious multidimensional character.

Paulsen and Johnson (1980) studied the multidimensionality of a set of

impulsivity measures using a sample of 55 preschool children. Scales

such as delay of gratification choices, the Matching Familiar Figures

Test (MFFT), teacher ratings, and the Porteus Maze test were given and

factor analyzed by a principal components procedure. The authors

found evidence for factors of sex (defined by activity level), age

(delay of gratification and MFFT errors), and IQ (Porteus Maze).

Notice that activity level and age-related measures (probable ego

developmental aspects) emerged as separate factors, a finding con-

gruent with the reasoning that self-control and activity level are

artificially confounded. However, Paulsen and Johnson's (1980) study

is weakened in import because their small sample size seriously limits

the reliability of their factor analysis (i.e., 10 variables were

measured on 55 subjects).











Temperament


References have been made to notions of temperament. Before pro-

ceeding, some of the ideas about and research on temperament are pre-

sented.

In psychology, there is a long history of interest on this topic,

and interest has intensified recently. Historical approaches include

theories of bodily humors, Pavlov's (1952) studies of nervous system

types, and Sheldon's (1942) theory of correspondences between somato-

type and personality. The single greatest impetus in the recent ex-

pansion of interest in temperament is the work of Thomas, Chess, and

Birch (Thomas, Chess, and Birch, 1968; Thomas, Chess, Birch, Hertzig,

and Korn, 1963; Thomas and Chess, 1977; Thomas and Chess, 1980). How-

ever, before presenting the Thomas and Chess approach, it is helpful

to consider definitions of temperament. Allport (1937) makes the dis-

tinction between temperament and other aspects of personality as fol-

lows:

Temperament refers to the characteristic phenomena
of an individual's emotional nature, including his
susceptibility to emotional stimulation, his cus-
tomary strength and speed of response, the quality
of his prevailing mood, and all peculiarities of
fluctuation and intensity in mood, these phenomena
being regarded as dependent on constitutional make
up, and therefore largely hereditary in origin.
(Allport, 1937, p. 54)

Allport's (1937) remains the core from which contemporary defini-

tions are derived. For example, at the 1980 Temperament Research Sym-

posium in New Haven, the following definition was proffered:












Temperament involves those dimensions of personality
that are largely genetic or constitutional in origin,
exist in most ages and in most societies, show some
consistency across situations, and are relatively
stable, at least within major developmental areas.
(Plomin, 1982, p. 269)

The work of Thomas and Chess represents the major force in con-

temporary temperament research. Using their clinical experience and

parental interviews of a subsample of 22 infants from their sample of

133 infants in the New York Longitudinal Study, which now has been

followed into young adulthood, Thomas and Chess induced 9 basic cate-

gories of temperament. They are presented in Table 1. Buss and

Plomin (1975) represent another significant current in temperament

study. They utilized criteria of heritability and stability to opera-

tionally define personality dimensions as candidates for the status of

temperament dimensions. Through factor analytic studies, Buss and

Plomin (1975) ascertained 4 core dimensions of temperament, which are

listed in Table 1 also. Rowe and Plomin (1977), in a factor analytic

study of items written from both the Thomas and Chess (1977) and Buss

and Plomin (1975) approaches and administered to parents of 91 twins

of ages 5 months to 9 years (total of 182 children), found that 4

factors accounted for 93% of the common variance and 53% of the total

variance in the item ratings. Rowe and Plomin (1977) called these 4

factors emotionality, activity level, sociability, and impulsivity.

These factors involved Thomas and Chess (1977) and Buss and Plomin

(1975) temperament categories in the expected pattern. Therefore,

there is considerable overlap between these two temperament systems as












































,C
0



0


.)o .

U 0

O -4
0 u


0
u
z, 0
O U








OO
1. 4-
0
02




4-1 .q
0 0





02 44

H H


0
O

0


O

4-4


,4





o

0


.C
0


02



co







ol
co
0- *IS
302
02
*02 4
042


1-1


c l

H3 *-



4l H
*., 4 ..0 *-4

> 0 41

4J 4J O. 0C

u 11:1


00

0


C




*J
0



--4
02








.i



o





-,4
0
44


.-


4

CO

0

CO

4J



U(
*H-




0





0)


-j >-
o 4-
0 -4
*'-
-4
,4-1 .X
0 *--

4 02
4J (4
4H 4
cd a
3 *-I
0- -
0r C


0
(U

0
U
4-1






"n


.4
U













0
'4-I




0c
-lI




CO
-4

0
a4








O
0
1a




















r0
0
4-4
3

c)






0
(U

0










0
4-4









-4
t0



















Sa
fi i-
o) e
4J il
4-1 c
*< FL


r-
02
4 02 > >,

-4 *4 -4

0 U) 0 02
- 4 > H-
4J *r 4
0 41 U 0.
1 u 0 S
r X 4 cn H


0l


a



or

0

E-4










shown in empirical results, the major difference being that Thomas and

Chess (1977) temperament categories are more specific and numerous.

One major problem in temperament research has been a lack of

sophisticated measuring instruments (Plomin, 1982). Until very re-

cently, upon the publication of Lerner, Palermo, Spiro, and Nesselroade's

(1982) psychometrically excellent Dimensions of Temperament Survey,

only Buss and Plomin (1975) have published a systematically constructed

and validated survey measure of temperament. As the Buss and Plomin

C1975) EASI-III Temperament Survey has been used in the present study,

their instruments are reviewed.

Buss and Plomin (1975) have developed 3 instruments which can be

modified slightly, by minor word substitutions, for child or adult

ratings by either self report or observer methods of rating. These

scales are termed EASI-I, -II, and -III (emotionality, activity level,

sociability, impulsivity). The scales have been shown to satisfy

standard psychometric criteria of internal consistency, factorial

homogeneity, and test re-test reliability (Buss and Plomin, 1975);

across a number of studies using children and adults, reliability

coefficients always exceed .7, usually substantially higher. Their

major sample of subjects involved a group of 137 families with twins

in the 2 to 6 year age range (in addition to college student samples).

They have demonstrated a moderately sized heritability component to

the temperament dimensions of emotionality, activity level, and soci-

ability in their twin study (Buss and Plomin, 1975). However, the

impulsivity dimension did not show heritability, but the authors

consider impulsivity a temperament dimension on theoretical grounds.










As noted in the above section on self-control, this view is challenged;

that impulsivity is properly considered a developmental construct is

consistent with Kopp's (1982) theoretical review of the development of

self-regulation in children. More specifically, impulsivity is hypothe-

sized to be a developmental construct within the purview of the major

developmental dimension of ego development or self development (Loevinger,

1976), and therefore fails the temperament criteria of stability--and

heritability (Buss and Plomin, 1975).


Ego Development

Loevinger's (1966a; 1976) theory of ego development is a high

level abstraction along which developmental characterologies are

ordered. Ego development is separated conceptually from other de-

velopmental traces such as physical, psychosexual, and intellectual

growth. A particular level of ego development, according to Loevinger

(1976), determines, or more precisely, is the characteristic manner of

making sense of and imparting order to reality. The flavor and sub-

stance of Loevinger's (1976) conception is seen in the descriptions of

the various levels of ego development.

The first period of growth includes the presocial and symbiotic

stages (see Table 2). Ego is considered absent at birth, the infant's

task being to differentiate self from the non-self of the physical

surround. Following elementary differentiation of self, the mothering

figure becomes the center of the infant's relationship to the external

world in the symbiotic phase. Language development is thought to con-

solidate emergence from symbiosis. Hence, these early stages of ego











TABLE 2, Some Milestones of Ego Development


Impulse Control Interpersonal Style
Stage Code Character Development


Presocial
Symbiotic


Autistic
I-i Symbiotic


Impulsive


Self-Protective



Conformist



Conscientious-
Conformist


Conscientious





Individualistic


Autonomous


Integrated


Impulsive, fearo.f
Retaliation


A Fear of being caught,
externalizing blame,
opportunistic


1-3




I-3/4


1-4


Conformity to external
rules, shame, guilt
for breaking rules

Differentiation of
norms, goals


Self-evaluated stan-
dards, self-criticism,
guilt for consequences
long-term goals and
ideals


I-4/5 Add: Respect for indi-
yiduals


Add; Coping with con-
flicting inner needs
toleration


Add: Reconciling inner
conflicts, renuncia-
tion of unattainable


Receiving, dependent,
exploitative

Wary, manipulative,
exploitative


Belonging, superfi-
cial niceness


Aware of self in re-
lation to group,
helping

Intensive, responsi-
sible, mutual, con-
,cern for communica-
tion


Add: Dependence as an
emotional problem

Add: Respect for au-
tonomy, interdepen-
dence



Add: Cherishing of
individuality


applying to the


Source; Loevinger (1976., pp, 24-25).


Note: "Add" means in addition to the description
previous level,
















Conscious Preoccupations


Self vs. non-self


Bodily feelings, especially sexual
and aggressive

Self-protection, trouble, wishes, things
advantage, control


Appearance, social acceptability, banal
feelings, behavior


Adjustment, problems, reasons, opportu-
nities (yague)


Differentiated feelings, motives for
behavior, self-respect, achievements
traits, expression



Add: Development, social problems, differ-
entiation of inner life from outer

Vividly conveyed feelings, integration of
physiological and psychological, psycho-
logical causation of behavior, role con-
ception, self-fulfillment, self in so-
cial context


Stereotyping, conceptual
confusion


Conceptual simplicity,
stereotypes, cliches


Multiplicity



Conceptual complexity,
idea of patterning




Add: Distinction of
process and outcome

Increased conceptual
complex patterns,
toleration for ambigu-
ity, broad scope,
objective


Add: Identity


Cognitive Style










development are inaccessible to study by verbal methods such as Loevinger's

(,1916 sentence completion test of ego development.

In the impulsive stage the world is viewed with full egocentricity.

Others are seen in terms of good or bad, often construed as "good to

me" and "bad to me." Conscious preoccupations are physical and, when

age appropriate, sexual or aggressive. Action is valued solely in

terms of reward and punishment. Incommoded states are attributed to

places or situations rather than within oneself.

The self-protective stage is marked by opportunistic hedonism.

First steps toward impulse control occur, but delay of gratification

is minimal. The concept of rules is recognized but self-interest and

expediency override. When responsibility for wrong is acknowledged,

it tends to be disclaimed through blame on body or other minimizing

aspect.

In the conformist stage rules are accepted and obeyed precisely

because they are rules. Compliance with group consensus rather than

the consequences of action characterizes the morality of the stage.

Social approval is a powerful motivator.

In the transition to the conscientious from the conformist stage,

a dawning capacity for introspection and relativism marks the self-

aware level. This appears to be the modal level for adults in Ameri-

can society (Loevinger, 1976); Holt (1980) has confirmed that the

self-aware level is the modal ego level in our society in his study of

3,522 subjects in a national probability sample. Rudimentary self-

criticism, understanding of psychological causation, and appreciation

of multiplicity are experienced at this ego level.













In the conscientious stage long-term goals and achievement for

its own sake become salient. Choice, as opposed to fate, is a pre-

occupation. Inner life is differentiated; others are seen in terms of

traits, motives, and patterns of behavior. Time perspective increases

and often events are perceived in a broader social context.

In the individualistic level there occurs a concern for depen-

dence as an emotional necessity despite the hallmark theme of indi-

viduality. Awareness of inner conflict grows, but here the conflict

is seen as only partially internal as opposed to the strong sense of

personal determination at the conscientious stage. Instead of re-

ducing contradictions to polarities, toleration of paradox begins to

emerge. Interpersonal relations are often seen as in conflict with

drives for achievement and responsibility.

The autonomous stage of ego development is conceptually complex.

Inner conflict is readily acknowledged and coped with. Poignancy,

existential humor, and paradox are often found on test protocols of

subjects at this stage. The limitations of autonomy are recognized

and this ushers in a sense of emotional interdependence. Self-ful-

fillment is salient, taking priority over achievement.

The integrated stage appears only rarely. Maslow's (1970) de-

scription of the self-actualizing individual is isomorphic with

Loevinger's (1976) integrated level of ego development.

The standard method for assessing level of ego development is a

sentence completion test (SCT) of 36 items. Loevinger and Wessler

(1970), using several hundred SCT protocols, have worked out scoring

procedures and manuals for the instrument. There have been several












reviews of the SCT of ego development (Holt, 1974; Hauser, 1976;

Waugh, 1979; Loevinger, 1979); therefore, only portions of research

bearing on the construct validity of the SCT are presented.

Reliability study of the SCT includes interrater and internal

consistency analyses. Across diverse samples and raters, Loevinger

and Wessler (1970) report high indices of reliability. Typical fig-

ures are 60% or greater for total protocol rating agreements and 95%

for agreement within a half step. Median interrater correlations are

.85. These levels of reliability have been matched by independent re-

searchers (e.g., Frank and Quinlan, 1976; Sullivan et al., 1970;

Redmore and Waldman, 1975; Waugh, 1981). Internal consistency study

repeatedly has shown coefficient alphas in the .85 to .90 range, and

principal component analysis has demonstrated a single dimension

(Loevinger and Wessler, 1970).

Convergent validity of the SCT with similar measures of constructs

is also favorable. For example, Frank and Quinlan (1976) found that

lower levels of ego development were highly associated with delin-

quency in female teenagers. Sullivan et al. (1970) found significant

relationships with moral and conceptual level development. Groesch

(1978) showed convergence with cognitive complexity, as did Lamb

(1971) with role taking perspective. Hoppe and Loevinger (1977)

demonstrated the theoretically predicted curvilinear relation of

conformity with ego development in a sophisticated study. Waugh and

McCaulley (1981) demonstrated discriminant validity with broad types

of psychopathology. These studies are representative of the many

reviewed in the major review papers referenced previously.












Loevinger's (1976) assumption of sequential, hierarchical develop-

ment of stages has seen only minimal empirical study, the segment of

contract validation most lacking regarding the SCT. There is, however,

one important longitudinal study in the literature. Redmore and

Loevinger (1979) report on several longitudinal samples. Briefly,

for their 8 subsamples, there was a mean rise in ego level and each

test retest correlation was positive (10 of 14 significant). Loevinger

(1979) reasons that these results approach the maximum strength of as-

sociation possible given the reliability of the SCT and problems in re-

peated measurement. However, more molecular longitudinal study of ego

development within homogenous groups and comparisons with cognitive de-

velopment are possible fruitful ways to studying the sequentiality

assumption further.

In summary, the construct validity of the SCT of ego development is

favorable to the conception and methodology. The solid theoretical and

empirical bases of Loevinger's (1976) measure provide a powerful frame-

work within which to study the interface of self-control, temperamental

activity level, and psychological development.
















CHAPTER II

METHODOLOGY

It is hypothesized that self-control is a function of ego de-

velopment independent of the temperament dimension of activity level.

This set of relations is believed to hold across all age levels. How-

ever, in this investigation, the age range is restricted to make the

study practical. Accordingly, the age range chosen is that of middle

childhood (Lidz, 1976); that is, the years within which children begin

to broaden their social world beyond the family is called middle

childhood. This socialization process largely takes place in elemen-

tary schools and the peer group becomes highly significant to the

child (Parsons, 1964; Sullivan, 1953).

This study follows a two-pronged procedure. First, a set of im-

pulsivity measures is examined for evidence of multidimensionality, in

the expectation that a self-control and activity level dimension will

emerge. Second, ego development is compared to each dimension. It is

hypothesized that ego level will be related to self-control and not

related to activity level. These relations are expected to hold in-

dependent of the chronological age and intellectual level of the in-

diyiduals.

Subjects

Male school children are chosen for study in the interest of

controlling for the variable of sex because problems in self-control

17












are more common in boys than girls. Subjects were recruited from the

P.K. Yonge Laboratory school of the University of Florida in Gaines-

ville, Florida. The P.K. Yonge Laboratory school is especially ap-

propriate for this study because the composition of the student body is

balanced with respect to race and socio-economic status, making the

students roughly representative of children in the population of the

Gainesville, Florida, community in terms of demographic variables.

The subjects were recruited by sending a packet of material to

the parents) of all the male students in grades 1 through 5. The

packet included an informed consent letter and a parent rating form of

the EASI-III. This material (see appendix) was sent to the parents)

of 141 children. Those potential subjects whose parents did not

return the material were recruited two additional times over a period

of weeks. The final sample contained 78 subjects, representing a 55%

recruitment rate. The final sample for study includes 20 (25%) first

graders, 17 (22%) second graders, 21 (27%) third graders, 12 (15%)

fourth graders, and 8 (10%) fifth graders.

Measures

Seven measures of dependent variables and 3 measures of inde-

pendent variables are used. There are 3 dependent variables of activity

level and 4 dependent variables of self-control. Each is described

below.

Activity Level

Sharron and Weller's (1971) Draw-A-Line test provides a measure

of motoric activity level. The time taken for a subject to draw a












line between two horizontal lines (a distance of 6 inches or 15.5 cen-

timeters) is considered an index of fine motor activity level. This

measure is the first measure of activity level. The second measure is

derived from a structured interview used by Scarr (1966). Scarr

(1966) used a structured interview covering the kinds of activities

children engage in to demonstrate that activity level possesses a ge-

netic component and thereby qualifies as a temperament dimension.

Scarr (1966) found that the total number of activities endorsed as en-

gaged in was a function of heritability. In this research project an

Activity Schedule is tailored from Scarr's (1966) structured inter-

view. The Activity Schedule (see appendix) consists of a list of 33

activities (e.g., jump rope, football, pinball, reading, bicycling)

which is read to the subject. The subject is to state for each activity

whether he engages in it. Each activity is considered in three situa-

tions: at home after school, at school, and on the weekends. The

score on this measure is the sum of endorsements of all activities for

all situations. The third measure of activity level is the activity

level scale of Buss and Plomin's (1975) EASI-III temperament survey.

Self-Control


Sharron and Weller's (1971) Draw-A-Line test directly yields a

measure of motoric self-control. After the subject draws a line be-

tween two horizontal marks, the time taken to draw another line as

slowly as possible is recorded. The difference between the two times

is an index of fine motor inhibition. This procedure has also been












used by Maccoby, Dowley, Hagen, and Degerman (1965) and Homatidis and

Konstantareas (1981). The second measure of self-control is the im-

pulsivity scale of the EASI-III (Buss and Plomin, 1975). The third

measure is a resistance to temptation experimental procedure. Resis-

tance to temptation situations has a long history of use in studies of

self-control (e.g., Aronfreed and Reber, 1965; Grinder, 1962), but no

standardized measure is available; investigators typically construct

one appropriate to their study. In this study, the following items

are used as stimulus objects: a high quality metallic, moving-part toy

steam roller truck; two attractive animal-shaped pencil sharpeners;

three working miniature tools (wrench, hammer, pliers); and a furry,

touchable toy animal. See appendix E for illustration of the items.

These objects are placed in front of the subject on a desk and he is

asked to remain seated and not to touch the toys as the examiner

leaves the room for a few minutes. The examiner observes through a

one-way window; if the subject does not touch the objects within five

minutes, he has passed the procedure. The fourth measure of self-

control is a delay of gratification procedure. Delay of gratification

experimental methods have been the topic of scores of studies in

psychology (see Mischel, 1974). The most typical delay of gratifica-

tion situation has the subject choose between receiving a smaller

reward immediately or a larger reward in the future. Selection of the

larger reward is considered an index of self-control. In this study,

subjects are presented the choice of either one balloon or colored

pencil at the conclusion of the testing session, or if they wait for

one week, a choice between either three colored pencils or three











larger balloons. The delayed reward choices are obviously more at-

tractive and valuable.

Independent Measures


Three independent variables are measured in this study. They are

chronological age, recorded to the nearest month, a short-form intel-

lectual ability measure, and a short-form test of level of ego de-

velopment. Intellectual ability is estimated from the sum of the

scale scores of the Vocabulary and Block Design subtests of the Wechsler

Intelligence Scale for Children-Revised. Kaufman (1979) recommends

this sort of short-form from the WISC-R.

Level of ego development is typically assessed by a well-studied

sentence completion test (Loevinger and Wessler, 1970). In this

study, however, subjects are younger than subjects with whom the

instrument has heretofore been used. Therefore, the SCT requires

extension downward in age. As the theory of ego development covers

life-span development and evaluations of the SCT are favorable, there

is reason to believe such an extension is feasible. Moreover, the

investigations of Lucas (1971), Farrell (1975), and Brinkerhoff (1971)

have shown that ego development can be judged from interview material

similar to that from the SCT. Thus, the construct appears general-

izable across methods of measurement. In addition, McCammon (1979)

found that for sixth grade boy subjects, oral and written forms of the

SCT were equivalent in ego level scores and correlation with IQ.

Given these precedents, it is reasonable to adopt an oral interview

form of the SCT in this research study.











Each subject is giyen a structured interview in which 23 items

from the 10-68 form of the SCT are stimulus items. Because of time

constraints and difficulty level of some of the items in the full 36

item SCT, a short form has been used. Holt (1980) has developed a 12

item SCT which retains excellent psychometric properties. Therefore,

the 23 item short form should not compromise the psychometric quality

of the SCT. Applying Cronbach's (1970) procedure for estimating ef-

fects of changes in test length on coefficient alpha to the SCT shows

that a 23 item SCT should reduce the alpha value from .91 to .87.

This sacrifice of internal consistency reliability is considered mini-

mal relative to the gains in decreasing difficulty level of the items

and ease of administration.

Procedure

Pilot study with the above described measures has shown that

administration time involves about 40 minutes per subject. Therefore,

administration has been conducted in one assessment session. The

following order of administration has been used: (1) Draw-A-Line

(normal and slow as possible conditions); (2) oral 23 item SCT inter-

view; (3) activity schedule; (4) Vocabulary and Block Design subtests

of the WISC-R; (5) resistance to temptation experimental procedure;

(6) delay of gratification choice. Subjects' parents completed the

parent rating form of the EASI-III. Subjects were tested individually

in a private room near the library of P,K. Yonge Laboratory School

during the school day during Spring 1981. The large window in the

door to the room was equipped with one-way viewing plastic to permit

observation of the subject during the resistance to temptation











procedure. Ten subjects (5 second and 5 fourth graders) were re-tested

with the activity schedule 7 to 10 days subsequent to testing to

provide data for test re-test reliability assessment of the measure.


Statistical Analyses

Preliminary statistical analyses involve test re-test reliability

assessment of the activity schedule and interrater reliability assess-

ment of the SCT. Total scores of the activity schedule for 10 sub-

jects re-tested are compared with a product moment correlation. The

five dimensions of the activity schedule (like, afternoon, school,

weekend, and total) are subjected to a principal components factor

analysis to ascertain the dimensionality of the modified instrument (n

= 78). The SCT protocols were typed onto separate lists for each stem

of the SCT. This permitted raters to evaluate the responses blind

from other responses of each subject. Two raters were used, myself

and another graduate student at the University of Florida. The first

rater (myself) is self-trained following Loevinger and Wessler's

(1970) procedures, trained at the 1978 ego development scoring work-

shop at Washington University in St. Louis, and has demonstrated

acceptable levels of interrater reliability in other samples (Waugh,

1981). The second rater is self-trained and trained with the first

rater on other SCT data two years ago. The first rater scored the

entire SCT data set; the second rater scored all of the responses for

one half of the subjects. A product moment correlation was computed

between the raters' total scores of 40 subjects to permit evaluation

of interrater reliability. Total scores were formed by assigning the

following numerical values to each item rating and summing the values











of the 23 items, of each protocol: 1-2 = 1, I-2/A = 2; I-A = 3, I- A/3

- 4; 1-3 = 5, 1-3/4 = 6; 1-4= 7; 1-4/5 = 8, 1-5 = 9, 1-6 = 10. SCT

rating was accomplished using the manuals of Loevinger, Wessler, and

Redmore (1970) and Redmore, Loevinger, and Tamashiro (1978, note 2).

Factor Analysis of Dependent Measures

The dimensionality of the set of 3 putative activity level and 4

self-control measures is evaluated factor analytically. The sample

size of 78 subjects meets the minimum requirement of 10 subjects per

variable (Skinner, 1981). It is hypothesized that 2 factor dimensions

exist in the data; thus the factor analysis is a hypothesis testing

procedure. As the number of subjects is relatively small, the number

of variables is relatively small, and the purpose is hypothesis testing,

a principal factor method is considered the method of choice (Gorsuch,

1974). The iterated principal factor method with rotation to varimax

solution of the Statistical Package for the Social Sciences was used

(Nie, Hull, Jenkins, Steinbrenner, and Bent, 1975). Two criteria were

used to evaluate the number of factors to retain. First, the typical

criterion of all eigenvalues greater than 1 was used to establish the

number of factors. This number was confirmed by Cattell's (1978)

scree test. The scree test involves plotting the eigenvalues of all

variables from a principal component analysis against the number of

variables and drawing a line between the points. The point at which

the curve veers sharply upward is associated with the number of factors

to retain. Following Gorsuch (1974), because the factor study is

relatively mitigated by a minimum number of subjects, variables











loading greater than or equal to the absolute value of .35 on retained

factors are interpreted. This rather stringent interpretational

criterion increases the likelihood that results are reliable. Follow-

ing the factor analysis of the 7 variable set of dependent measures,

factor scores of each subject on each retained factor dimension were

computed. These factor score computations permit evaluation of rela-

tions with the set of independent variables.

Multiple Regression of Independent and Dependent Variables


It is predicted that ego level is associated with self-control

and not activity level, independent of the effects of age and intel-

lectual ability level. These relations are tested by multiple regres-

sions of age, IQ, and item sum SCT scores on each of the sets of fac-

tor scores for the retained factor dimensions. The SPSS multiple re-

gression statistical procedure was used.















CHAPTER III

RESULTS

Preliminary Analyses


Preliminary analyses include interrater reliability assessment of

SCT ratings, test-retest reliability assessment of the activity sched-

ule, and factor anlaytic assessment of the activity schedule.



Interrater Reliability Assessment of the SCT Ratings


A Pearson correlation coefficient was computed between the two

raters' sets of SCT item sum scores for 40 subjects. The resulting

correlation coefficient of .73 is considered of sufficient magnitude

for research purposes but lower than values typically obtained with

the SCT (Loevinger and Wessler, 1970). The most obvious explanation

for this relatively lower level of agreement is that the rating manuals

were developed from samples that did not include subjects of the age

range in this study. Therefore, specific verbal content character-

istic of children in middle childhood is not represented in the manuals

as aids to rating, and raters must make discrimination more on the

basis of theory and extrapolation from the language style of older but

low ego level subjects. Thus, use of the SCT with subjects in middle

childhood seems to entail a sacrifice, although not severe, in degree

of rater agreement.











Psychometric Assessment of the Activity Schedule


Test-retest reliability and factor analytic study of the activity

schedule were conducted to ascertain whether the minor modifications of

Scarr's (1966) procedure were deleterious to the psychometric char-

acteristics of the structured interview. The test-retest Pearson cor-

relation coefficient between the initial and 7 to 10 day retest activity

schedule total scores for 10 subjects gave a value of .73. This is

within acceptable psychometric limits. The principal components

factor analysis of the 5 variables in the activity schedule (like,

afternoon, school, weekend, and total) yielded 1 factor with an eigen-

value of 3.5 accounting for 70% of the total variance. The following

factor loadings were obtained: like = .56; afternoon = .92; school =

.72; weekend = .92; total = .99. Since the total score is a composite

of the afternoon, school, and weekend scores, it is not independent

and tends to inflate the degree of intercorrelation. However, it is

obvious that a single source of variance underlies the activity schedule

and the best expression of this variance is the total score variable.

These analyses confirm that the minor modifications of Scarr's (1966)

procedure in developing the activity schedule have not compromised its

psychometric credibility.

Dimensionality of the Set of Dependent Variables

An iterated principal factor method of factor analysis was ap-

plied to the set of 7 dependent variables measured on 78 subjects.

This analysis required 18 iterations and yielded a 3 factor solution











accounting for 59.3% of the total variance. Both the eigenvalue

greater than or equal to 1 and Cattell's (1978) scree test criteria

indicated that a 3 factor solution, as opposed to the hypothesized 2

factor solution, was appropriate (see Table 3 and Figure 1). Table 3

shows the loadings of the variables on the 3 factors subsequent to

Varimax rotation. Using the criterion of interpreting only those

loadings greater than .35, it is seen that the first factor (Fl) is

comprised of high and positive loadings for the EASI-III impulsivity

and activity level scales. The second factor dimension (F2) has a

high and positive loading for the activity level total score and a

moderate and negative loading for the resistance to temptation measure.

The third factor (F3) is defined only by a high and positive loading

for the delay of gratification measure.

These results are contrary to the predicted number of factors and

pattern of loadings. Although exact predictions were not supported,

the meanings of the results are consistent with the hypotheses, as

discussed below.

Interpretation of the Fl dimension is problematic. There are two

ways in which to understand this dimension. First of all, because

only EASI-III variables define this dimension, and both variables

share a common method of measurement (parental ratings), it is reason-

able to assume that this dimension simply is a method-specific factor

for parental rating. However, a second possibility is that Fl repre-

sents a "difficult temperament" dimension. The literature on tempera-

ment supports the idea that a small set of second order factors subsume













TABLE 3. Principal Axes Factor Analysis of Dependent Variables with
Varimax Rotation.



Variables Communality Eigenvalue Factor 1 Factor 2 Factor 3


resistance to
temptation

delay of grati-
fication

normal draw rate

draw slow differ-
erence score

activity schedule

EASI-III impul-
sivity

EASI-III activity
level


.196


.275


.094

.140


.655

.991


.219


1.628


1.464


1.057

.948


.853

.624


.426


-.054


-.060


-.092

.265


.124

.995


.452


-.363


.015


.177

-.238


.797

.000


.115


.248


.521


-.233

.116


.071

-.012


-.043






30


.65
1.60

.55
.50
.45
.40

.35
.30
.25
.20

.15 Scree at
.10
.05 %Factors

1.00
.95
.90
.85
.80
.75
.70
.65
.60
.55
.50
.45
.40



1 2 3 4 5 6 7

Number of Factors



FIGURE 1


Scree Test Indicating 3 Factor Solution












the Thomas and Chess (1937) constellation of temperament dimensions

(Ploipn, 1982). Thomas and Chess (1968) have delineated three groups

of children defined by different patterns of temperament. They are

termed the "difficult," "easy," and "slow to warm up." The "diffi-

cult" pattern is characterized by low rhymicity, negative mood, low

approach, low adaptibility, and high intensity (Thomas and Chess,

1977). In order to investigate the possibility that the Fl dimension

reflects the EASI-III version of a "difficulty" second order factor,

the full EASI-III was subjected to a principal factor analysis on a

post hoc basis. After 15 iterations, 2 factors with eigenvalues

greater than 1 emerged and accounted for 76.1% of the total variance.

The first EASI-III factor is defined by emotionality, impulsivity, and

activity level (see Table 4); the second factor is composed of socia-

bility and, to a lesser extent, activity level. This first factor

appropriately can be called "difficulty" and the second factor appears

to be an "easy" dimension. Therefore, it would be reasonable to re-

gard the Fl dimension, with its impulsivity and activity level scale

composition, as an EASI-III "difficulty" dimension. Unfortunately, in

the present study, there is no independent means of ascertaining

which--if either--interpreation of Fl is justified. The most cautious

interpretation is to consider Fl to be a method specific rating factor.

The second factor (Fl) is defined by the activity schedule total

score and the resistance to temptation score (high scores on this di-

mension are related to touching the toys when alone after being asked

not to). The activity schedule is thought to assess a heritible


1













TABLE 4. Principal Axes Factor Analysis of EASI-III Scales with
Varimax Rotation.



Variables Eigenvalue Factor 2 Factor 2


Emotionality 1.831 .729 -.243

Activity Level 1.212 .5.14 .469

Impulsivity .519 .728 .191

Sociability .438 -.036 .647












temperament of activity level (Scarr, 1966), and this variable most

saliently defines F2. Although the resistance to temptation measure

was hypothesized to reflect self-control, in view of its pairing with

activity level, it is more properly seen as an index of amount of

motoric activity rather than ability to inhibit motoric response.

Therefore, the F2 dimension appears to be an activity level factor.

The third factor (F3) is defined by the delay of gratification

experimental procedure. Persons scoring high on F3 demonstrate the

ability to wait for a preferred reward as opposed to accepting a less-

preferred one immediately. This dimension most closely reflects the

hypothesized self-control factor.

In summary, the factor analytic results show that a three factor

solution is appropriate to explain the interrelations in the set of

dependent variables. The first factor is ambiguous and open to in-

terpretation as either a method-specific or "difficulty" dimension; in

the interest of caution, the method-specific interpretation is chosen.

The second factor is considered a temperamental dimension of activity

level. The third factor, although defined only by delay of gratifica-

tion, appears to be a self-control dimension. In the next section,

these dimensions are related to age, intellectual level, and level of

ego development to determine whether Fl, F2, and F3 are developmental

dimensions, specifically if they are ego developmental dimensions

independent of age or level of intellectual ability.











Relations of Age, IQ, and Ego Level to Factor Dimensions


Relationships of age, IQ, and level of ego development to the

three factor dimensions were examined by computing multiple regressions

of the independent variables on the factor scores of Fl, F2, F3.

Table 5 shows summary statistics. The Fl dimension is not related to

age, IQ, or ego level singly or in combination. Therefore, Fl can

not qualify as a developmental factor. Dimension F2, called activity

level, is predicted mostly by IQ; the contributions of age and ego

level are minimal. F2 seems to be associated with lower intellectual

performance. In view of the fact tht IQ tests, such as the WISC-R,

assess scholastic ability, it is not surprising that individuals less

scholastically able, and by inference less scholastically-minded and

reflective, tend to be more action-oriented. The F3 dimension is sig-

nificantly predicted by both age and ego level, but the independent

contribution of ego level is negligible. Therefore, this delay of

gratification/self-control dimension qualifies as a developmental

dimension. Although level of ego development is associated with this

developmental dimension of self-control, ego development is less

important than simple chronological age.


Post Hoc Factor Analyses

Previous hypothesis testing methods have yielded results which

support the hypotheses, in general; however, ambiguity was found in

the interpretation of the first factor in the set of dependent vari-

ables. In addition, the second factor, called activity level, was

found to be related to intellectual ability. The problem with this
















TABLE 5.


Summary Statistics of Regression of Age, IQ, and Ego
Level on Factor Scores.


Dependent Independent Multiple Simple F test of
Variable Variable(s) R R Model


Factor 1 Age .023 -.023 .038

Factor 1 Age+IQ .096 .096 .351

Factor 1 Age+IQ+Ego Level .193 -.122 .957



Factor 2 Age .035 -.035 .093

Factor 2 Age+IQ .330 -.310 4.567 **

Factor 2 Age+IQ+Ego Level .342 -.045 3.263 *



Factor 3 Age .408 .408 15.130 ***

Factor 3 Age+IQ .420 .001 8.013 ***

Factor 3 Age+IQ+Ego Level .425 .223 5.429 ***


Note. = p < .025; ** = p < .01; *** = p < .001










interpretation is, that other research (Thomas and Chess, 1977; Plomin,

1982) finds no association between IQ and temperament, and the present

putative activity level dimension thus becomes open to question. In

order to provide further means of responding to these ambiguities, two

factor analyses are presented. Both analyses are tentative in impli-

cation because they are post hoc and severely strain their reliability

because the ratio of measures to subjects is so large.

The first method uses a principal axes factor analysis of the set

of 7 dependent variables combined with the two temperament dimensions

of the EASI-III not included in the hypothesis-testing portion of this

study. If the first factor obtained in the original factor analysis

indeed is a difficulty factor, then this factor should include the

variable of emotionality (intensity of response). If this result

fails to occur, or all of the EASI-III variables themselves define the

factor, then the interpretation of it as a method-specific dimension

gains credibility. Table 6 presents the results from this analysis,

which required 25 iterations and accounted for 66.1% of the total

variance among the 9 variables. The number of factors to retain was

determined by the eigenvalue greater than or equal to 1 criterion.

Four factors were retained. As can be seen, the first factor in this

analysis appears to be a "difficulty" dimension defined by impulsivity,

activity, and emotionality. Therefore, although this result is tenta-

tive, more weight is added to the argument that this dimension is not

simply a method-specific rating factor. The second factor is nearly

identical to the previous dimension called activity level. The third

factor is defined only by sociability and probably can be called an











TABLE 6. Principal Axes Factor Analysis of EASI-III Scales and
Dependent Variables with Varimax Rotation.



Variable Factor 1 Factor 2 Factor 3 Factor 4


Emotionality .494 .117 -.287 .467

Activity Level .590 .108 .115 -.093

Sociability .259 .017 .758 -.196

Impulsivity .821 .046 -.090 .158

Normal Draw Rate -.042 .121 -.103 -.609

Draw Slow Dif-
ference Score .278 -.236 .045 .056

Delay of Grati-
fication -.075 -.011 .314 .121

Resistance to
Temptation -.063 -.355 .276 .149

Activity
Schedule -.098 .842 .109 -.005











"easy" temperament dimension, The fourth factor is comprised of the

Draw-A-Line activity level measure and emotionality. A tentative

interpretation is that this represents an intensity of response, as

the Draw-A-Line task requires subjects to draw without instructions,

and those who are temperamentally intense responders would be expected

to execute the task quickly; this is in keeping with the strong loading

of emotionality. It is interesting that delay of gratification failed

to emerge in the definition of a factor. Most likely, the develop-

mental variance in this set of variables is minimal and thereby swamped,

preventing it from emerging as a dimension. This reinforces the

notion that developmental variance is not strong in temperament,

activity level, and putative self-control measures.

The second post hoc method of analysis is even weaker statistically

than the first because 12 variables measured on 78 subjects are used.

Nevertheless, tentative influences of the independent variables (age,

IQ, ego level) are observable directly in their contributions to the

factor structure, and can help in evaluating the importance of de-

velopmental variance and intellectual ability variance, especially re-

garding the putative activity level dimension. The principal axes

factor analysis required 25 iterations and accounted for 67% of the

total variance. Table 7 shows the results. Factor one again is the

"difficulty" of temperament dimension. Factor two is a solid develop-

mental dimension made up of age, delay of gratification, and ego

development, respectively. Thus, insofar as self-control represents a

developmental phenomenon, only delay of gratification appears to

qualify as a developmental/self-control dimension. Factor three is












TABLE 7. Principal Axes Factor Analysis
and Independent Variables with


of EASI-III Scales, Dependent,
Varimax Rotation. *


Variable Factor 1 Factor 2 Factor 3 Factor 4 Factor 5


Emotionality .709 -.032 -.038 -.209 -.003

Activity Level .444 -.096 .035 .379 .151

Sociability -.014 .107 .022 .799 -.145

Impulsivity .786 -.086 .083 .180 .051

Normal Draw Rate -.240 -.049 .076 .081 .320

Draw Slow Dif-
ference Score .261 .140 .288 .063 -.164

Delay of Grati-
fication -.045 .528 .029 .140 -.068

Resistance to
Temptation -.077 .126 .026 .150 -.513

Activity Schedule .133 .063 -.305 .153 .554

Age .081 .814 -.193 -.100 -.117

IQ .034 -.043 .912 .025 -.065

Ego Level -.146 .468 .299 .004 .037











defined only by IQ. Factor four is the "easy" temperament dimension.

Factor five is the putative activity level dimension, consisting of

resistance to temptation and activity schedule variables. It is

important to note that, in this analysis when IQ was allowed to covary

among the complete set of variables, IQ did not contribute to the

activity level factor. Thus, although IQ predicts the factor dimen-

sion of activity level in multiple regression analysis, IQ cannot be

considered an intrinsic component of the dimension. This finding,

weak and unreliable as it is, serves to dampen the criticism of the

interpretation of the dimension as activity level because of its

correlation with IQ, which after all, amounts to only about 9% of the

variance in activity level.















CHAPTER IV

DISCUSSION

Principal Findings


This study began from the perspective that two major classes of

personality constructs, called horizontal and vertical dimensions, can

offer conceptual power in understanding personality. Specific use of

this framework was made in studying the hypothesized independence of

self-control and activity level temperament dimensions through factor

analytic procedures. It was found that common measures of activity

level and self-control could not be separated into two major dimensions

analogous to a developmental and trait distinction; rather, three

dimensions were necessary in this particular set of measures. However,

two of these dimensions appeared to conform to the developmental and

trait distinction. The other factor was viewed as either a method-

specific or "difficulty" of temperament factor. Relations of these

dimensions to predictors of age, IQ, and level of ego development

revealed both expected and unanticipated results. The first factor,

probably a "difficulty" of temperament factor, but cautiously labelled

a method-specific factor, was not predicted by age, IQ, or ego development;

this is consonant with theory. The second factor, considered an

activity level dimension, was unexpectedly predicted by intellectual

ability. This is discrepant with previous temperament research. A










post hoc factor analysis suggested that IQ is not an intrinsic compo-

nent of this activity level factor but a correlate. This correlation

is understood as a function of the temperamentally active child's

action orientation as opposed to a more reflective, scholastic style.

However, this interpretation is not established in the present study

because of the lack of external confirmation. The third factor was

related to both age and level of ego development, but the contribu-

tion of ego level was insignificant compared to simple chronological

age. This third factor is considered a self-control dimension which

is solidly developmental in nature, an interpretation partially but-

tressed by the tentative post hoc analysis.



Evaluation of the Methodology

Methodologically, this study possesses relative strengths and

weaknesses. Factor analytic procedures played an important role in

the study. The major factor analysis used 7 variables measured on 78

subjects. This ratio approaches the limits of providing results that

can be considered statistically reliable (Skinner, 1981). On the

other hand, care was exercised in extracting an appropriate number of

factors through the use of two reputable techniques and factor loadings

were interpreted by a stringent criterion of loadings greater than

.35. These safeguards enhance the reliability of the factor results

even though no cross-validation was possible. Moreover, an approp-

riate factor model was employed, the principal axes model. This is a

poor comment on the state of multivariate study of self-control,











impulsivity, and activity level constructs to claim as an advantage the

appropriate use of factor models, but most previous research (e.g.,

Paulsen and Johnson, 1980; Homatidis and Konstantareas, 1981) has used

an inappropriate model (principal components), which serves to inflate

spuriously the magnitude of the relations. In considering the methodo-

logical strengths and shortcomings of the present study, the balance

indicates that this study represents a significant advance in methodo-

logical sophistication.

Another possible criticism of this research concerns the specific

measures. The Buss and Plomin (1975) EASI-III temperament survey is

apt to be replaced by the Lerner, Palermo, Spiro, and Nesselroade

(1982) Dimensions of Temperament Survey in terms of providing a stan-

dardized temperament measure applicable to a broad age range. Thus,

the present findings may not be directly comparable to future research

employing the more sophisticated Dimensions of Temperament Survey;

however, this instrument was not available when the present study was

conducted. More damaging, however, is the use of the Draw-A-Line

procedure. This measure was not found to relate to any other; thus it

is likely that the Draw-A-Line technique is a poor choice in studies

of impulsivity and activity level (the Draw-A-Line activity level

score did relate to an intensity temperament dimension in a very

tentative post hoc analysis). It is possible that this absence of

effect results from the fact that the Draw-A-Line test consisted of 1

"item" and therefore is an underestimate of a possible "true score"

which might manifest upon multiple item assessment. At any rate, the











delay of gratification and resistance to temptation variables showed

solid inter-variable relations even though they too are 1 "item"

measurements (and scored on a dichotomous basis). Thus, these experi-

mental procedures should be considered superior construct measurements

for use in further research, which could be enhanced by multiple

"item" measurement.


Implications of the Study

This factor analytic study of temperamental and developmental di-

mensions of self-control in middle childhood has many implications for

study of self-control, temperament, ego development, and by inference,

behavior disorders of middle childhood. These implications are indi-

cated below.

Prominent constructs in personality theory, particularly within

the realm of developmental personality theory, include self-control,

temperament, and ego development. In this study, it has been seen

that the proposed horizontal and vertical dimensions distinction is

teneable. This idea strengthens the trait and developmental distinc-

tion of Achenbach and Weisz (1975) and Swensen (1980, Note 1). An

important implication is that a host of personality variables hereto-

fore considered traits may be due evaluation in terms of horizontal

and vertical dimensional status. In addition, the importance of

temperament theory is evident both in personality theory in general,

and in terms of a horizontal class of constructs. Particularly, the

notion of "difficulty" of temperament, at least as that perceived by

parents, has received confirmation of importance in this study.











Measurement of level of ego development during middle childhood by the

SCT has been found more difficult in regard to rater agreement. It is

interesting that ego level, while still showing significant associa-

tions, does not appear to be as confounded by IQ or age in this age

range as in other developmental samples (Loevinger, 1979); these

correlations are shown in Appendix F. More importantly, the relation

of ego development to theoretically relevant constructs, such as

impulsivity or possibly activity level, has been found to be minimal.

The most reasonable explanation for this absence of association is

that the SCT does not carry sufficient variance in this admittedly

restricted ego level range within middle childhood.

This research has involved normal males in middle childhood in

the study of normal personality constructs. However, there are impli-

cations to behavior disorders of childhood. Recent research on the

classification of child psychopathology has established the importance

of two broad and opposing classes of behavior disorders called exter-

nalizing and internalizing dimensions (Achenbach and Edelbrock, 1978).

The externalizing group of disorders includes conduct disorder, aggres-

sion, and acting out problems; the internalizing dimension involves

anxiety and personality disorder problems. Underlying this broad

distinction may be basic temperament differences such as sociability,

activity level, and impulsivity. Previous temperament research has

suggested such an implication also (Thomas and Chess, 1977; Terestman,

1981; Carey, McDevitt, and Baker, 1980). For example, Thomas and

Chess (1977) have found that "difficult" temperament children have











a higher incidence of clinical disorders upon longitudinal followup.

Carey, McDevitt, and Baker (1980) found that temperament dimensions

of adaptibility, intensity, activity, and distractibility are more

frequent in clinical samples of child behavior disorders, but they

could not reliably discriminate minimal brain dysfunction, learning

disability, and mixed conduct disorder groups on the basis of tempera-

ment. Carey, McDevitt, and Baker (1980) do caution against the diag-

nosis of minimal brain dysfunction from observational behavior data

alone because of the confound with temperament dimensions in the

absence of neurological confirmation. In the present study, self-

control, hence impulsivity, and temperamental activity level dimensions

were found orthogonal. Therefore, Carey, McDevitt, and Baker's (1980)

cautionary message to diagnosticians is reinforced by the independence

of impulsivity and activity level, which in clinical practice are

often lumped into categories such as hyperactivity and conduct dis-

order. It is also interesting to speculate on the nature of the

activity level dimension elucidated in the present study in regard to

its association with lower intellectual performance. It may be a

dimension such as this that resembles more closely an externalizing

form of behavior disorder vis-a-vis joint effects of high activity

level and limited cognitive appraisal abilities. Of course, this

suggestion needs to be examined directly in further research.















REFERENCE NOTES


1. Swensen, C. Invited address, 1978, ego development scoring
workshop. St. Louis: Washington University, Social Science In-
stitute, November, 1978.

2. Redmore, C., Loevinger, J. and Tamashiro, R. Measuring Ego
development: Scoring manual for men and boys (Pre-publication
version). St. Louis: Washington University, Social Science
Institute, August, 1978.
















REFERENCES


Aaronfreed, J., and Reber, A. Internalized behavior suspension
and the timing of social punishment. Journal of Personality
and Social Psychology, 1965, 1, 3-16.

Achenbach, T., and Edelbrock, C. The classification of child psycho-
pathology: a review and analysis of empirical efforts. Psycho-
logical Bulletin, 1978, 85, 1275-13Q1.

Achenbach, T., and Weisz, J.R. Impulsivity-reflectivity and cogni-
tive development in preschoolers: A longitudinal analysis of de-
velopmental and trait variance. Developmental Psychology, 1975,
11, 413-414.

Allport, G.W. Personality: A psychological interpretation. New
York: Holt, 1937.

Bandura, A. Self-efficacy: Toward a unifying theory of behavior
change. Psychological Review, 1977, 84, 191-215.


Block, J., Block,
the Matching
impulsivity.


J.H., and Harrington, D.M.
Familiar Figures Test as a
Developmental Psychology,


Some misgivings about
measure of reflection-
1974, 10, 611-632.


Brinkerhoff, R.S. Ego development in adolescent girls. Unpublished
doctoral dissertation, University of Chicago, 1971.

Buss, A.H., and Plomin, R. A temperament theory of personality de-
velopment. New York: Wiley, 1975.

Carey, W.B., McDevitt, S.C., and Baker, D. Differentiating minimal
brain dysfunction and temperament. In S. Chess and A. Thomas
(Eds.), Annual progress in child psychiatry and child develop-
ment. New York: Brunner/Mazel, 1980.

Cattell, R.B. The scientific uses of factor analysis in behavioral


and life sciences.


New York: Plenum Press, 1978.


Cronbach, L.J. Essentials of psychological testing (3rd ed.), New
York; Harper and Row, 1970.











Davids, A. Ego functions in distrubed and normal children: Aspira-
tion, inhibition, time estimation, and delayed gratification.
Journal of Consulting and Clinical Psychology, 1969, 33, 61-70.

Davison, M.L., King, P.M., Kitchener, K.S., and Parker, C.A. The
stage sequence concept in cognitive and social development.
Developmental Psychology, 1980, 16, 121-131.

Farrell, G.E.P. The relation of ego development to intellectual and
ethical development (Doctoral dissertation, Washington Univer-
sity, 1974). Dissertation Abstracts International, 1975, 35,
4648B-4649B.

Frank, S., and Quinlan, D.M. Ego development and female delinquency:
A cognitive-developmental approach. Journal of Abnormal Psy-
chology, 1976, 85, 505-510.

Freud, S. The interpretation of dreams. London: Hogarth, 1953.

Goldfried, M.R., and Merbaum, M. (Eds.). Behavior change through self-
control. New York: Holt, Rinehart, and Winston, 1973.

Gorsuch, R.L. Factor analysis. Philadelphia: Saunders, 1974.

Grinder, R.E. Parental child-rearing practices, conscience, and re-
sistance to temptation of sixth grade children. Child Develop-
ment. 1962, 33, 803-820.

Groesch, S. Ego development and cognitive complexity: integrative
and discriminative complexity. (Doctoral dissertation, University
of Notre Dame, 1978). Dissertation Abstracts International,
1978, 39, 978B-979B.

Hauser, S.T. Loevinger's model and measure of ego development: A
critical review. Psychological Bulletin, 1976, 83, 928-955.

Holt, R.R. Review of "Measuring ego development, Volumes I and II."
Journal of Nervous and Mental Disease, 1974, 158, 310-316.

Holt, R.R. Loevinger's measure of ego development: Reliability and
national norms for male and female short forms. Journal of
Personality and Social Psychology, 1980, 39, 909-920.

Homatidis, S., and Konstantareas, M.M. Assessment of hyperactivity:
Isolating measures of high discriminant ability. Journal of
Consulting and Clinical Psychology, 1981, 49, 533-541.












Hoppe, C.F., and Loevinger, J. Ego development and conformity: A
construct validity study of the Washington University Sentence
Completion Test. Journal of Personality Assessment, 1977, 41,
497-504.

Karoly, P. Behavioral self-management in children: Concepts, methods,
issues, and directions. In M. Hersen, R.M. Eisler, and P.M.
Miller (Eds,,. Progess in behavior modification, Vol. 5. New
York: Academic Press, 1977.

Kaufman, A.S. Intelligent testing with the WISC-R. New York: Wiley,
1979.

Kendall, P.C., and Finch, A.J., Jr. A cognitive-behavioral treatment
for impulsivity: A group comparison study. Journal of Consulting
and Clinical Psychology, 1978, 46, 110-118.

Kendall, P.C., and Finch, A.J., Jr. Developing non-impulsive behavior
in children: Cognitive-behavioral strategies for self-control.
In P.C. Kendall and S.D. Hollon (Eds.). Cognitive behavioral
intervention: Theory, research, and procedures. New York:
Academic Press, 1979.


Kohlberg, L. Development of moral
M.L. Hoffman and L.W. Hoffman
research (Vol. 1). New York:


character and moral idealogy. In
(Eds.), Review of child development
Russell Sage Foundation, 1964.


Kopp, C.B. Antecedents of self-regulation: A developmental per-
spective. Developmental Psychology, 1982, 18, 199-214.

Lamb, N.J. The relation of role-taking to ego development (Doctoral
dissertation, Washington University, 1971). Dissertation Ab-
stracts International, 1971, 32, 2401B.

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

Leary, T. Interpersonal diagnosis of personality. New York: Ronald
Press, 1957.


Lerner, R.M., Palermo, M., Spiro, A., and Nesselroade, J.
the dimensions of temperamental individuality across
span: The Dimensions of Temperament Survey (DOTS).
velopment, 1982, 53, 149-159.


Assessing
the life
Child De-


Lidz, T. The person. New York: Basic, 1976.

Loevinger, J. Three principles for a psychoanalytic psychology.
Journal of Abnormal Psychology, 1966a, 71, 432-443.









Loevinger, J. The meaning and measurement of ego development. Ameri-
can Psychologist, 1966b, 21, 195-206.

Loevinger, J. Ego development: Conceptions and theories. San
Francisco; Jossey-Bass, 1976.

Loevinger, J. Construct validity of the sentence completion test of
ego development. Applied Psychological Measurement, 1979, 3,
281-311.

Loevinger, J., and Wessler, R. Measuring ego development 1. Construc-
tion and use of a sentence completion test. San Francisco:
Jossey-Bass, 1970.

Loevinger, J., Wessler, R., and Redmore, C. Measuring ego develop-
ment 2. Scoring manual for women and girls. San Francisco:
Jossey-Bass, 1970.

Lorr, M., and Manning, T.T. Measurement of ego development by sen-
tence completion and personality test. Journal of Clinical Psy-
chology, 1978, 34, 354-360.

Lucas, P. Validation of a test of ego development by means of a
standardized interview (Doctoral dissertation, Washington Uni-
versity, 1971). Dissertation Abstracts International, 1971,
32, 2204B.

Maccoby, E.E., Dowley, E.M., Hagen, J.W., and Degerman, R. Activity
level and intellectual functioning in normal preschool children.
Child Development, 1965, 36, 761-770.

McCrae, R.R. and Costa, P.T. Openness to experience and ego level
in Loevinger's sentence completion test: Dispositional contribu-
tions to developmental models of personality. Journal of Person-
ality and Social Psychology, 1980, 39, 1179-1190.

Maslow, A. Motivation and personality. New York: Harper and Row,
1970.

McCammon, E.P. A comparison of oral and written forms of the sentence
completion test for ego development. Unpublished master's thesis,
Washington University, 1979.

Mead, G.H. On social psychology. Chicago: University of Chicago
Press, 1964.

Meichenbaum, D. Teaching children self-control. In B. Lahey and
A. Kazdin (Eds.). Advances in child clinical psychology, Vol. 2.
New York: Plenum, 1979.









Mischel, W. Processes in delay of gr;
(Ed.). Advances in experiment.
New York: Academic Press, 1974.


atification. In L. Berkowitz
l social psychology, Vol. 7.


Nie, N.H., Hull, C.H., Jenkins, J.G., Steinbrenner, K., and Bent, D.H.
Statistical package for the social sciences (2nd ed.). New York:
McGraw-Hill, 1975.

Parsons, T. Social structure and personality. London: Free Press
of Glencoe, 1964.

Paulsen, K., and Johnson, M. Impulsivity: A multidimensional concept
with developmental aspects. Journal of Abnormal Child Psychology,
1980, 8, 269-277.

Pavlov, I.P. Twenty years of study of higher nervous activity (beha-
vior) in animals. Warsaw: PZWL, 1952.

Piaget, J. The moral judgment of the child. New York: Free Press,
1932.

Plomin, R. Childhood temperament. In B. Lahey and A. Kazdin (Eds.),
Advances in clinical child psychology (Vol. 6.). New York: Plenum,
1982.

Redmore, C.D., and Loevinger, J. Ego development in adolescence:
Longitudinal studies. Journal of Youth and Adolescence, 1979,
8, 1-20.

Redmore, C.D., and Waldman, K. Reliablity of the sentence completion
test of ego development. Journal of Personality Assessment,
1975, 39, 236-243.


Rowe, D.C. and Plomin, R. Temperament in early childhood.
of Personality Assessment, 1977, 41, 150-156.


Journal


Rozek, F., Wessman, A.E., and Gorman, B.S. Temporal span and delay
of gratification as a function of age and cognitive development.
Journal of Genetic Psychology, 1977, 131, 37-40.


Scarr, S.
ment,


Genetic factors in activity motivation. Child Develop-
1966, 37, 663-673.


Sharron, S., and Weller, L, Classification patterns in underprivi-
leged children in Israel. Child Development, 1971, 42, 581-594.


Sheldon, W. The varieti
tional differences.


es of temperament:


A psychology of constitu-


New York: Harper, 1942.










Skinner, H,A. Factor analysis and studies on alcoholism: A methodo-
logical review. Journal of Studies on Alcohol, 1981, 42, 1091-
1101.

Sullivan, E., McCullough, G., and Stager, M. A developmental study
of the relationship between conceptual, ego, and moral develop-
ment. Child Development, 1970., 41, 399-411.

Sullivan, H.S. The interpersonal theory of psychiatry. New York:
Norton, 1953.

Terestman, N. Mood quality and intensity in nursery school children
as predictors of behavior disorder. In S. Chess and A. Thomas
(Eds.), Annual progress in child psychiatry and child develop-
ment. New York: Brunner/Mazel, 1981.

Thomas, A., and Chess, S. Temperament and development. New York:
Brunner/Mazel, 1977.

Thomas, A., and Chess, S. The dynamics of psychological development.
New York: Brunner/Mazel, 1980.

Thomas, A., Chess, S., and Birch, H. Temperament and behavior dis-
orders in children. New York: New York University Press, 1968.

Thomas, A., Chess, S., Birch, H., Hertzig, M., and Korn, S. Beha-
vioral individuality in early childhood. New York: New York
University Press, 1963.

Wardell, D., and Royce, J.R. Relationships between cognitive and
temperament traits and the concept of "style". Journal of Mul-
tivariate Experimental Personality and Clinical Psychology,
1975, 1, 244-256.

Wardell, D., and Royce, J.R. Toward a multi-factor theory of styles
and their relationships to cognition and affect. Journal of
Personality, 1978, 46, 474-505.

Waugh, M.H. The verisimilitude of Loevinger's measure of ego develop-
ment in a clinical population. Unpublished master's thesis, Uni-
versity of Florida, 1979.

Waugh, M.H, Reliability of the sentence completion test of ego de-
velopment in a clinical population. Journal of Personality As-
sessment, 1981, 45, 485-487.

Waugh, M.H., and McCaulley, M.H. Relation of level of ego development
to type and severity of psychopathology. Journal of Consulting
and Clinical Psychology, 1981, 49, 295-296.





54




Weisz, J.R., O'Neill, P,, and O'Neill, P.C. Field dependence-inde-
pendence on the Children's Embedded Figures Test: Cognitive
style or cognitive level?. Developmental Psychology, 1975, 11,
539-540.,















APPENDIX A



DRAW-A-,T4NE


START

55













APPENDIX B

SHORT FORM SENTENCE COMPLETION TEST FOR BOYS

i. A man's job

2. Being with other people

3. The thing I like about myself is

4. If my mother

5. If I can't get what I want

6. When I was younger

7. Women are lucky because

8. What gets me into trouble is

9. A good father

10. If I were king

11. I feel sorry

12. A man should always

13. Rules are

14. He felt proud that he

15. Men are lucky because

16. My father and I

17. A man feels good when

18. When I get mad

19. At times he worried about

20. My main problem is

21. Sometimes he wished that

22. If I had more money

23. I just can't stand people who












APPENDIX C


ACTIVITY SCHEDULE

Ask the child if he likes to do each activity (yes or no) at home after
school; repeat for at home on the weekend and at school. Then, ask the
child if he does each activity at home after school; repeat for at home
on the weekend and at school.

ACTIVITY LIKES (in general) DOES
After Wkend. School

jump rope
monkey bars, etc.
read books
watch tv.

hop scotch
tag
football
baseball
basketball
soccer
swim
run

board games, cards
dolls
movies
hide & seek
make models
toy soldiers
bicycle
kickball
play with pet
kick-the-can
draw, paint, color
skateboard
roller skate

pinball

walk

dodge ball






58



APPENDIX C Activity Schedule (Cont.)

ACTIVITY LIKES (in general) Does
After Wkend. School


army, cowboy, spacemen
go shopping
music
hunt
fish












APPENDIX D

PARENT RATING FORM


Name of child Grade

Rater Age of Child

Please rate this child in terms of how characteristic the statements
are of him. The underlined 3 in the center of each row represents
where the average child would fall on this item. Please do not hesi-
tate to use the entire range of possible ratings.


1. Child tends to be nervous in new situations. 1 2 3 4 5

No Yes

2. Child often acts on the spur of the moment. 1 2 3 4 5

No Yes


3. Child gets bored easily


1 2 3 4 5
No Yes


4. Child sometimes does "crazy" things just
to be different.


2 3


4 5
Yes


5. Child tends to be shy.



6. Child has many friends.


1 2 3 4 5
No Yes


2 3


7. Child yells and screams more than most
children his age.


8. Child finds self-control easy to learn.



9. When displeased, child lets people know
right away.


10. Child likes to plan things way ahead of
time.


4 5
Yes


2 3 4 5
Yes


2 3


4 5


Yes


1 2 3


4 5


Yes


2 3


4 5


Yes





60



11- Child has trouble resisting temptation.



12. Child likes to wear himself out with ex-
ertion.


13. There are many things that annoy him.



14. Child has fewer fears than most children
his age.


15. Child often cries.



16. Child is very sociable.


17. Child usually seems to be in a hurry.



18. Most of my child's activities are fast
paced.



19. For relaxation, child likes to slow down
and take things easy.


20. Child often appears to be bursting with
energy.


2 3


4 5


1 2 3 4 5
No Ye&

1 2 3 4 5


2 3


1 2 3
No

1 2 3
No

1 2 3


1 2 3


2 3



2 3


4 5
Ye:

4 5
Ye:

4 5


Yes


4 5


Yes


4 5


Yes


4 5


4 5


Yes


21. Child often feels sluggish.


22. When child does things, he does them
vigorously.


23. Child likes to keep busy all the time.


1 2 3 4 5


1 2 3
No

1 2 3


No Yes


4 5


Yes


4 5






61



24. Child's movements are forceful and emphatic. 1


25. Child will try anything once.



26. Child makes friends very quickly.



27. Child usually prefers to do things alone.



28. Unfinished tasks really bother child.



29. Child often says the first thing that
comes into his head.


30. Child often has trouble making up his mind.



31. Child likes to make detailed plans before
he does something.


32. Child is almost always calm--nothing ever
bothers him.


33. Child frequently gets upset.



34. When child gets scared, he panics.



35. It takes a lot to get him mad.



36. Child is happiest in familiar surroundings.



37. Child has trouble controlling his impulses.


1 2 3 4 5


2 3


1 2 3
No

1 2 3


1 2 3


2 3


2 3


1 2 3
No


2 3



2 3



2 3



2 3



2 3


2 3


4 5


Yes


Yes

4 5


Yes


4 5


Yes


4 5
Yes


4 5
Yes

4 5
Yes


4 5
Yes

4 5


Yes


4 5
Yes

4 5


Yes


4 5
Yes

4 5


Yes


4 5






62



38, Child likes to be off and running as soon
as he wakes up in the morning,


39. Child can stand frustration better than
most.


40. Child often feels insecure.



41. Usually child can't stand waiting.



42. Child is easily frightened.



43. Child is known as hot-blooded and quick
tempered.


44. Child generally likes to see things through
to the end.


45. Child gets excited easily.



46. Once child gets going on something he
hates to stop.


47. Child is somewhat emotional.



48. Child tends to hop from interest to
interest quickly.


49. Child generally seeks new and exciting
experiences and sensations.


2 3


4 5


Yes


1 2 3 4 5


2 3



2 3



2 3


2 3


2 3


1 2 3


2 3



2 3



2 3


4 5


Yes


4 5


Yes


4 5


4 5


4 5
Ye:

4 5


4 5
Ye:

4 5


Yes


4 5
Yes


2 3 4 5
Yes


50. Child tends to give up easily.


4 5


2 3










APPENDIX E


RESISTANCE TO TEMPTATION OBJECTS


,"<./'I \ . ; S
1-f -J ^
^r ?,,\^t~


0









APPENDIX F


COMPLETE CORRELATION MATRIX


2 -.18

3 .26 .29

4 .48 .10 .46

5 -.08 .16 -.06 -.05

6 -.08 .18 -.03 -.09 .12


7 -.27 .02

8 .12 .09


.04 -.13 -.16 -.10

.07 .28 .05 .09


9 .11 .13 .14 .14 -.29 .05


10 .05 .01 -.11 -.03

11 -.04 .04 .00 .08


-.04

.10 -.18


.12 .41 -.08 .13


.03 .01 .02 .22 -.30 -.28


12 .07 .06 .06

13 .01 .05 .03


.08 .04 .05

.09 .04 .03


.01 .26 -.21 -.11 .49

.02 .28 -.30 -.24 .89 .84


14 -.12 .04 -.06 -.13


.11 .26 .07 .06 -.04 .32 .22 .23 .27


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


Variables (N = 78)
1. Emotionality
2. Sociability
3. Activity Level
4. Impulsivity
5. Resistance to Temptation
6. Delay of Gratification
7. Draw Normal
8. Draw Slow Difference
9. Activity Schedule
10. Age
11. Block Design
12. Vocabulary
13. I.Q. 64
14. Ego Level









APPENDIX G

DESCRIPTIVE STATISTICS


Standard
Variable Mean Deviation Minimum Maximum


Resistance to Temptation .76 .43 0 1

Delay of Gratification .68 .47 0 1

Draw Normal 9.14 5.46 1.8 26

Draw Slow Difference 21.76 17.86 .50 90.9

Age 8.45 1.39 6.09 11.25

Block Design 10.86 3.07 3 18

Vocabulary 11.71 2.59 4 18

IQ 22.56 4.88 10 35

Impulsivity 59.99 9.53 42 92

Activity Level 33.91 9.55 12 78

Activity Schedule 45.26 14.81 14 75

Ego Level 87.83 8.19 69 107













BIOGRAPHICAL SKETCH


Mark H. Waugh was born in Augusta, Georgia, and grew up in Lexington,

Kentucky, and Greenville, North Carolina. He majored in psychology at

the University of North Carolina at Chapel Hill. He attended graduate

school at the University of Florida and is completing a clinical intern-

ship at the University of Texas Health Science Center at Dallas. He

is married to Marcia Snowden Waugh and they have one son, Robert Mark

Waugh. He will take an assistant professor of psychology position at

Roosevelt University in Chicago, Illinois, after graduation.










I certify that I have read this study and that in my opinion it
conforms to acceptable standards of scholarly presentation and is
fully adequate, in scope and quality, as a dissertation for the degree
of Doctor of Philosophy.


High C. vis, Jr., Chairma
Professor of Clinical Psychology

I certify that I have read this study and that in my opinion it
conforms to acceptable standards of scholarly presentation and is
fully adequate, in scope and quality, as a dissertation for the degree
of Doctor of Philosophy.


acquflin R. Goldan
Professor of Clinical Psychology

I certify that I have read this study and that in my opinion it
conforms to acceptable standards of scholarly presentation and is
fully adequate, in scope and quality, as a dissertation for the degree
of Doctor of Philosophy.


Wer K. Blashfield -
Associate Professor of Clinical Psychology

I certify that I have read this study and that in my opinion it
conforms to acceptable standards of scholarly presentation and is
fully adequate, in scope and quality, as a dissertation for the degree
of Doctor of Philosophy.


Mary If/McCaulley -
Assistant Professor of Clinical Pchology

I certify that I have read this study and that in my opinion it
conforms to acceptable standards of scholarly presentation and is
fully adequate, in scope and quality, as a dissertation for the degree
of Doctor of Philosophy.


aundra Damico
Professor of Foundations of Education










This dissertation was submitted to the Graduate Faculty of the
College of Health Related Professions and to the Graduate Council, and
was accepted as partial fulfillment of the requirements for the degree
of Doctor of Philosophy.


Dean, College of Health Related Professions



Dean for Graduate Studies and Research

August, 1982




Full Text

PAGE 1

TEMPERAMENTAL AND DEVELOPMENTAL DIMENSIONS OF SELF-CONTROL IN MIDDLE CHILDHOOD BY MARK HOWARD WAUGH A DISSERTATION PRESENTED TO THE GRADUATE COUNCIL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 1982

PAGE 2

---------COPYRIGHT 1982 BY MARK HOWARD WAUGH

PAGE 3

ACKNOWLEDGMENTS Acknowledgments I wish to e~press must be located within my educational experiences in clinical psychology. First of all, for first indicating a sense of the excitement of psychology, I thank George Welsh, Ph .D., who taught and inspired me as an undergraduate at the University of North Carolina, For providing lasting influence, I express deep gratitude to the committee members, teachers, supervi sors, role models, colleagues, and good people who are Drs. Hugh Davis, Jacquelin Goldman, Roger Blashfield, and Mary Mccaulley. Dr. Saundra Damico has assisted well in my labors in the dissertation process. I also must thank my friends and colleagues R. Peter Fortney and Leslie Morey, Ph.D., for their friendship and intellectual stimu lation. And most importantly, to my family, especially my wife and son, Marcia S. Waugh and Robert M. Waugh, my heartfelt thanks for years of encouragement and tolerance. iii

PAGE 4

TABLE OF CONTENTS ACKNOWLEDGMENTS LIST OF TABLES ABSTRACT CHAPTER I II III INTRODUCTION AND FRAMEWORK A Proposed Model of Personality Self-Control. Temperament . Ego Development METHODOLOGY Subjects. Measures. Activity Level Self-Control Independent Measures .... Procedure Statistical Analyses ... Factor Analysis of Dependent Measures .. Multiple Regression of Independent and Dependent Variables RESULTS. . Preliminary Analyses .. Interrater Reliability Assessment of the SCT iii vi vii 1 2 3 6 10 17 17 18 18 19 21 22 23 24 25 26 26 Ratings . . . . 26 Psychometric Assessment of the Activity Schedule . 27 Dimensionality of the Set of Dependent Variables. 27 Relations of Age, IQ, and Ego Level to Factor Dimensions ..... Post Host Factor Analyses iv 34 34

PAGE 5

IY DIS CUSSIQN . . Principal Findings. Eyaluatian of Implications REFERENCE NOTES REFERENCES APPENDICES. A DRA,W-A-LINE. Methodology of the Stud y . B SENTENCE COMPLETION TEST FOR BOYS. C ACTIVITY SCHEDULE. D PARENT RATING FORM E RESISTANCE TO TEMPTATION OBJECTS F. COMPLETE CORRELATION MATRIX. G. DESCRIPTIVE STATISTICS . ' . . . 4 1 4 1 4 2 4 4 4 7 4 8 55 55 56 5 7 59 63 64 6 5 BIOGRAPHICAL SKETCH . . . . . . . . . . 116 V

PAGE 6

LIST OF TABLES Table 1. Major Systems of Temperament Dimensions . 2. Some Milestones of Ego Development .... 8 11 3. Principal Axes Factor Analysis of Dependent Variables with Varimax Rotation. . . . . . . . . 29 4. Principal Axes Factor Analysis of EASI-III Scales with Varimax Rotation. . . 32 5. Summary Statistics of Regressions of Age, IQ, and ego level on Factor Scores. . . ....... 35 6. Principal Axes Factor Analysis of EASI-III Scales and Dependent Variables with Varimax Rotation. . . 37 7. Principal Axes Factor Analysis of EASI-III Scales, Dependent, and Independent Variables with Varimax Rotation. . . . . . . . . . . 39 vi

PAGE 7

Abstract of Dissertation Presented to the Graduate Council of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy TEMPERAMENTAL AND DEVELOPMENTAL DIMENSIONS OF SELF-CONTROL IN MIDDLE CHILDHOOD by MARK HOWARD WAUGH August 1982 Chairman: Hugh C. Davis, Jr., Ph.D. Major Department: Clinical Psychology Psychological study of self-control was reviewed in terms of a proposed model of personality. This model of personality considers develo pm ental and dispositional constructs to be orthogonal and super ordinate sectors of personality. Precedents to this model in the literature include similar models and the distinction between developmental and trait variance. When applied to the phenomena of self control, this model suggested that the temperamental dimension of activity level would be dispositional and that other measures of impulsivity would be developmental. The major developmental dimension of self-control was hypothesized to be ego development. Previous research on self-control has indicated a multidimensional character, but the earlier research lacks a comprehensive conceptual base and suffers methodological shortcomings. The present study measured seven variables thought to reflect either activity level or developmental vii

PAGE 8

self-control on sevent y -eight normal male school children of middle childhood (ages six to twelve). A principal axes factor anal y sis resulted in three dimensions. The first f actor was interpreted as either a "difficulty" of parentall y perceived temperament or a method specific dimension. The second factor was seen as an activity level dimension. The third factor was interpreted as a self-control dimen sion. Independent variables of age, intellectual abilit y and level of ego development were used to predict factor scores on the factor dimensions to test for their developmental status. Factor one was independent of age, IQ, and ego level. Factor two was associated with IQ, but post hoc analysis suggested IQ was a correlate as opposed to an intrinsic component of the activity level dimension. Factor three was related to both age and level of ego development, but ego develop ment contributed minimal independent variance. Factor three was seen to be a developmental component of self-control. Results of the study were discussed in terms of the importance of the developmental and dispositional distinction, the importance of the proposed model of personality sectors, and implications to personality factors in child psychopathology. viii

PAGE 9

C~TER I INTRODUCTION AND FRA11-E.WORK The aim of this investigation is to explore the construct of s elf-control within a proposed model of personality. Self-control is taken to be a ref lexi-ve property of the self (Mead, 1964) which involves volitional regulation of impulse or wish, usually in the service of goal attainment. In recent years, self-control and im pulsivity have become rapidly expanding topics of research (e.g., Goldfried and Merbaum, 19]3; Meichenbaum, 1979; Bandura, 1977; Karoly, 197]; Kendall and Finch, 19]8; i9J9). This explosion of research is characterized primarily by a cognitive-behavioral orientation. Langer (1969) has grouped theories of behavioral development in terms of their allegiance to three general models: the psychoanalytic, the mechanical mirror (pehavioral), and the organic lamp (cognitive developmental). Although isolated research efforts have been made into self-control from the psychoanalytic perspective (e.g., Davids, i969) and the organic lamp model (~.g., Rozek, Wessman, and Gorman, 19]7)., it is clear that the area of self-control or impulsivity re search is dominated by the mechanical mirror image of man. One pur pose 9f the present study i s tq further the marriage of diverse litera tures. a,nd orientations within the study of self-control. 1

PAGE 10

2 Loevinger's (l966a; 1966b; 19761 theory of ego development is a theory of self and self-development which is explicitly psychoanalytic and cognitive-developmental in character. Loevinger (1976) has articu lated and synthesized much of the common ground explored by Freud (1953), Sullivan (1953), Piaget (1932), and Kohlberg (1964). Loevinger (1976) incorporates self-control as one of a number of central dimen sions of ego development. Therefore, in considering that ego develop ment includes self-control, that ego development preserves the most central principles of psychoanalysis (Loevinger, 1966b), and that ego development possesses the breadth of scope intrinsic to cognitive social developmental models (Davison, King, Kitchner, and Parker, 1980), it is reasonable that the theory of ego development holds high promise of providing an understanding of self-control as an aspect of an integrated theory of personality development. A Proposed Model of Personality The conceptual framework for this study of self-control consists of a proposed model of personality. This model posits two major and orthogonal dimensions of personality. The horizontal dimension or axis includes such constructs as temperament, interpersonal style, and cognitive style. The vertical dimensional set consists of primary de velopmental constructs such as ego and cognitive development. This model is not entirely original; there are precedents. For example, Swensen (1980, Note 1) invokes a similar notion in his personality scheme consisting of ego development and Leary's (1957) interpersonal

PAGE 11

style system. Additionally, Royce and colleagues seem to be utilizing a similar model in their conceptual and empirical work on individuality theory, which involves integration of self and cognitive style theory (see Wardell and Royce, 1975; 1978). Another precedent to this pro posed model is implicit in Weisz, O'Neill and O'Neill (1975) and Achenbach and Weisz (1975). These investigators have drawn the dis tinction between cognitive ability and cognitive style. The former is considered a developmental variable and the latter a trait variable. They have shown independent developmental and trait components of variance within the domains of reflectivity-impulsivity and field de pendence-independence. This distinction between development and trait is identical to that made between horizontal and vertical sectors of personality. This proposed model of personality is not finely articulated nor has the model been subject to programmatic research. In fact, given the relative absence of scientific confirmation, the model's value is best seen as heuristic. Nevertheless, some preliminary support for the model exists in other empirical studies. For example, Lorr and Manning (1978) found very few significant correlations between a large set of interpersonal style variables and the sentence completion test of ego development. Furthermore, Macrae and Costa (1980) found ego development unrelated to neuroticism and extraversion (measured by the Eysenck Personality Inventory), and also unrelated to the temperament dimensions of emotionality, sociability, and activity level (using Buss and Plomin's (1975) EASI-III temperament survey). Macrae and

PAGE 12

4 Costa (_1980) ,also found ego deyeloJ_Jment and the inhibi tion of i.m~uls.e s ub.scale of the. EASJ;-l U impuls.ivity scale to be independent, although they do not provide the resu],ts for the full impulsiyity scale. Accordingly, insofar as interperson~l styles and temperament dimensions appear orthogonal to ego development, the proposed model gathers credibility. Fleshing out the full structure and ramifications of this proposed model will require programmatic research beyond the scope of the research project herein. A portion of this model, how ever, is used as the conceptual framework to guide this study. That is, the independence of ego development, as a vertical dimension of personality, and temperament dimensions, as a horizontal axis, is accepted. Sel.J;-Control The terms self-control and impulsivity are taken to be inter changeable. The psychological literature on self-control is quite diverse (Karoly, 1977), and often this literature is segregated con ceptually and empirically into foci such as child versus adult or treatment versus assessment. The present study seeks to establish self-control as an aspect of ego deyelopment and thereby making arti ficial such distinctions as child versus adult approaches to the study of self-control. Self-control is cons idered an as pect of the theory q~ ego deveJ9P,ment and is thus hrpothesized to be orthogonal to tempera ment dimensions. Self-control, then, is ytewed as qualitatively dif ferent from the temperamental variable of activity level.

PAGE 13

5 Self-control or impulsivity has often been regarded as multidi mensional and implicated in accounts of hyperactivity, conduct dis order, and learning disability (Block, Block, and Harrington, 1974; Homatidis and Konstantareas, 1981; Paulsen and Johnson, 1980). The proposed model of personality described previously would suggest that previous investigators have confounded measures of temperament (activity level) with ego development (self-control). Confounding activity level with self-control assures a spurious multidimensional character. Paulsen and Johnson (1980) studied the multidimensionality of a set of impulsivity measures using a sample of 55 preschool children. Scales such as delay of gratification choices, the Matching Familiar Figures Test (_MFFT), teacher ratings, and the Porteus Maze test were given and factor analyzed by a principal components procedure. The authors found evidence for factors of sex (defined by activity level), age (.delay of gratification and MFFT errors), and IQ (Porteus Maze). Notice that activity level and age-related measures (probable ego developmental aspects) emerged as separate factors, a finding con gruent with the reasoning that self-control and activity level are artificially confounded. However, Paulsen and Johnson's (1980) study is weakened in import because their small sample size seriously limits the reliability of their factor analysis (_i.e., 10 variables were measured on 55 subjects).

PAGE 14

6 Temperament Reerences have been made to notions of temperament. Before pro ceeding, some of the ideas about and research on temperament are pre s ented. In psychology, there is a long history of interest on this topic, and interest has intensified recently. Historical approaches include theories of bodily humors, Pavlov's (J952} studies of nervous system types, and Sheldon's (1942) theory of correspondences between somato type and personality. The single greatest impetus in the recent ex pansion of interest in temperament is the work of Thomas, Chess, and Birch (Thomas, Chess, and Birch, 1968; Thomas, Chess, Birch, Hertzig, and Korn, 1963; Thomas and Chess, 1977; Thomas and Chess, 1980). How ever, before presenting the Thomas and Chess approach, it is helpful to consider definitions of temperament. Allport (1937) makes the dis tinction between temperament and other aspects of personality as fol lows: Temperament refers to the characteristic phenomena of an individual's emotional nature, including his susceptibility to emotional stimulation, his cus tomary strength and speed of response, the quality of his prevailing mood, and all peculiarities of fluctuation and intensity in mood, these phenomena being regarded as dependent on constitutional make up, and therefore largely hereditary in origin. (Allport, 1937, p. 54) Allport's (1937) remains the core from which contemporary defini tions are derived. For example, at the 1980 Temperament Research Sym posium in New Haven, the following definition was proffered:

PAGE 15

7 Tem-perament invol y es those dimensions of personalit y that are largely genetic or constituional in origin, exist in most ages and in most socities, show some consistenc y across situations, and are relativel y stable, at least within major developmental areas. (flomin, 1982, p. 269) The work of Thomas and Chess represents the major force in con temporary temperament research. Using their clinical experience and parental interviews of a subsample of 22 infants from their sample of 133 infants in the New York Longitudinal Stud y which now has been followed into young adulthood, Thomas and Chess induced 9 basic cate gories of temperament. The y are presented in Table 1. Buss and Plomin (_1975) represent another significant current in temperament study. They utilized criteria of heritability and stability to opera tionally define personality dimensions as candidates for the status of temperament dimensions. Through factor analytic studies, Buss and Plomin (1975) ascertained 4 core dimensions of temperament, which are listed in Table 1 also. Rowe and Plomin (1977), in a factor analytic study of items written from both the Thomas and Chess (1977) and Buss and Plomin (1975) approaches and administered to parents of 91 twins of ages 5 months to 9 years (total of 182 children), found that 4 factors accounted for 93 % of the common variance and 53 % of the total variance in the item ratings. Rowe and Plomin (_1977) called these 4 factors emotionalit y activity level, sociability, and impulsivity. These factors involved Thomas and Chess (1977) and Buss and Plomin (1975) temperament categories in the expected pattern. Therefore, there is considerable overlap between these two temperament systems as

PAGE 16

TABLE 1. Major Systems of Temperament Dimensions, Activity Level Rythmicity Approach/Withdrawal Adaptibility Intensity of Reaction Threshold of Responsivity Quality of Mood Distractibility Attention Span/ Persistence Emotionality Activity Level Sociability Impulsivity Thomas and Chess level, tempo, frequency of motoric behavior regularity of biological functions initial reaction to novel stimulus ease or difficulty of change of behavior in response to a different situation degree of energy in responding magnitude of stimulation required before response pleasant/friendly versus unpleasant/unfriendly tone ease with which extraneous stimulation interferes with ongoing behavior maintenance of a behavior in face of obstructing circumstance Buss and Plomin intensity of reaction tempo and vigor of response affiliativeness, responsiveness to others quick versus inhibition of response 00

PAGE 17

9 s.hown in empirical results, the ma,jor difference being that Thomas and Chess (J9JJ) temperament categories are more specific and numerous. One major problem in temperament research has been a lack of sophisticated measuring instruments (]?lomin, 1982). Until very re cently, upon the publication of Lerner, Palermo, Spiro, and Nesselroade's (J982) psychometrically excellent Dimensions of Temperament Survey, only Buss and Plomin (1975) have published a systematically constructed and validated survey measure of temperament. As the Buss and Plomin (_19]5) EASI-III Temperament Survey has been used in the present study, their instruments are reviewed. Buss and Plomin (19]5) have developed 3 instruments which can be modified slightly, by minor word substitutions, for child or adult ratings by either self report or observer methods of rating. These scales are termed EASI-I, -II, and -III (emotionality, activity level, sociability, impulsivity). The scales have been shown to satisfy standard psychometric criteria of internal consistency, factorial homogeneity, and test re-test reliability (Buss and Plomin, 1975); across a number of studies using children and adults, reliability coefficients always exceed .7, usually substantially higher. Their major sample of subjects involved a group of 137 families with twins in the 2 to 6 year age range (~n addition to college student samples). They have demonstrated a moderately s.ized heritability component to the temperament dimensions of emotionality, activity level, and soci abili ty in their twin study (Buss and )?lomin, 19]5). However, the impulsivity dimension did not show heritability, but the authors consider impulsivity a temperament dimension on theoretical grounds.

PAGE 18

10 As noted in the above section on self-control, this view is challenged; that impulsivity is properly considered a developmental construct is consistent with Kopp's (1982) theoretical review of the development of self-regulation in children. More specifically, impulsivity is hypothe sized to be a developmental construct within the purview of the major developmental dimension of ego development or self development (Loevinger, 1976), and therefore fails the temperament criteria of stability--and heritability (Buss and Plomin, 19751, Ego Development Loevinger's (1966a; 1976) theory of ego development is a high level abstraction along which developmental characterologies are ordered. Ego development is separated conceptually from other de velopmental traces such as physical, psychosexual, and intellectual growth. A particular level of ego development, according to Loevinger (_1976), determines, or more precisely, is the characteristic manner of making sense of and imparting order to reality. The flavor and sub stance of Loevinger's (J976) conception is seen in the descriptions of the various levels of ego development. The first period of growth includes the presocial and symbiotic stages ($ee Table 21. Ego is considered absent at birth, the infant's task being to differentiate self from the non-self of the physical surround. Following elementary differentiation of self, the mothering figure becomes the center of the infant's relationship to the external world in the symbiotic phase. Language development is thought to con solidate emergence from symbiosis. Hence, these early stages of ego

PAGE 19

11 TABLE 2, Some Milestones of Ego Development Sta,ge Pres ocial Symb.:Lotic ImI_Julsive Self-?rotect:Lye Ccmformi st Consci entious Conformist Conscientious I.ndividualis ti.c Autono-qious I.ntegrate,d Code 1-1 I-2 I-3 I-3/4 I-4 I-4/'5 I-5 I-6 Impulse Control Character Development Impulsive, fea~ 9f Retali.ation Fear of bei.ng caught, externalizing b.lame, opportunisti c Conformity to external rules? sha,me, gilt for b re,qking rules Differentiation of norms, goals Self-eyc1-luated stan dards, self-criticism, guilt for consequences, long-term goals and ideals. Add: Respect for i_ndi yiduals Add; CQping with con --flicting inner needs toJ.e+at:L:on Add: Reconciling inner confli~ts, renuncia, ti.qn 9f unattainable Interpersonal Strle Autisti c Symbiotic Receiving, dependent, exploitative Wary, manipulative, exploitative Belonging, superfi cial niceness Aware of self in re lation to group, helping Intensive, responsi sible, mutual, con cern for communica tion Add: Dependence as an emotional problem Add: Respect for au tonomy, interdepen dence Add: Cherishing of individuality Nqte; "Ad d'' mean~ in a,dd :;i:_ ti.'i)n. t.9 tbe de.s cril)t:lon applying to the previ_qus level. ~ou;-ce; Loevi'_nge r ().9J6 pp 24-25)

PAGE 20

Conscious Preoccupations Self vs. non-self Bodily feelings, especially sexual and aggressiye 12 Sel{-protection, trouble, wishes, things adyantage, control Appearance, social acceptability, banal feelings, behavior Adjustment, problems, reasons, opportu nities (yague) Differentiated feelings, motives for behavior, self-respect, achievements traits, expression Add: Development, social problems, differ entiation of inner life from outer Vividly conveyed feelings, integration of physiological and psychological, psycho logical causation of behavior, role con ce.i:ition? self-fulfiJ,lment, self i n so.. cia.l context Add: Identity Cognitive Style Stereotyping, conceptual confusion Conceptual simplicity, stereotypes, cliches Multiplicity Conceptual complexity, idea of patterning Add: Distinction of proce ss and outcome Increased conceptual complex patterns, toleration for ambigu ity, broad scope, pbjectivey

PAGE 21

13 deyelo~ment a;-e inaccessible to stud by verbal methods such as Loevinger' s (=19161 sentence completion test of ego development. In the impulsive stage the world is v iewed with full egocentricit y Others are seen in terms of good or bad, often construed as "good to me'' and "bad to me." Conscious preoccupations are physical and, when age appropriate, sexual or aggressive. Action is valued solel y in terms of reward and punishment. Incommoded states are attributed to places or situations rather than within oneself. The self-protective stage is marked by opportunistic hedonism. First steps toward impulse control occur, but delay of gratification is minimal. The concept of rules is recognized but self-interest and expediency override. When responsibility for wrong is acknowledged, it tends to be disclaimed through blame on body or other minimizing aspect. In the conformist stage rules are accepted and obeyed precisely because the y are rules. Compliance with group consensus rather than the consequences of action characterizes the moralit y of the stage. Social approval is a powerful motivator. In the transition to the conscientious from the conformist stage, a dawning capacity for introspection and relativism marks the self aware level. This appears to be the modal level for adults in Ameri can societ y O: .. ,oevinger, .. 19 J 6); Holt ( J. 980) has confirmed that the sel~-a~~re level is the modal ego level in our societ y in his study of 3,522 ~ ubjects in a national probabilit y sample. Rudimentar y self criticism, understanding of ps y chological causation, and appreciation of multiplicity are experienced at this ego level.

PAGE 22

14 In the conscientious stage long-term goals and achievement for its own sake become salient. Choice, as opposed to fate, is a pre occupation. Inner life is differentiated; others are seen in terms of traits, motives, and patterns of behavior. Time perspective increases and often events are perceived in a broader social context. In the individualistic level there occurs a concern for depen dence as an emotional necessity despite the hallmark theme of indi vidualit y Awareness of inner conflict grows, but here the conflict is seen as only partially internal as opposed to the strong sense of personal determination at the conscientious stage. Instead of re ducing contradictions to polarities, toleration of paradox begins to emerge. Interpersonal relations are often seen as in conflict with drives for achievement and responsibility. The autonomous stage of ego development is conceptuall y complex. Inner conflict is readily acknowledged and coped with. Poignancy, existential humor, and paradox are often found on test protocols of subjects at this stage. The limitations of autonomy are recognized and this ushers in a sense of emotional interdependence. Self-ful fillment is salient, taking priority over achievement. The integrated stage appears only rarel y M aslow's (1970) de scription of the self-actualizing individual is isomorphic w ith Loevinger's (19]6) integrated level of ego development. The standard method for assessing level of ego development is a sentence completion test (SGT) of 36 items. Loevinger and Wessler (1970), using several hundred SGT protocols, have worked out scoring procedures and manuals for the instrument. There have been several

PAGE 23

15 reviews of the SCT of ego development (Holt, 19]4; Hauser, 1976; Waugh, 19]9; Loevinger, 1979); therefore, only portions of research bearing on the construct validity of the SCT are presented. Reliability study of the SCT includes interrater and internal consistency analyses. Across diverse samples and raters, Loevinger and Wessler (1970) report high indices of reliability. Typical fig ures are 60% or greater for total protocol rating agreements and 95% for agreement within a half step. Median interrater correlations are .85. These levels of reliability have been matched by independent re searchers (e.g., Frank and Quinlan, 1976; Sullivan e~ al., 1970; Red.more and Waldman, 1975; Waugh, 1981). Internal consistency study repeatedly has shown coefficient alphas in the .85 to .90 range, and principal component analysis has demonstrated a single dimension (Loevinger and Wessler, 1970). Convergent validity of the SCT with similar measures of constructs is also favorable. For example, Frank and Quinlan (1976) found that lower levels of ego development were highly associated with delin quency in female teenagers. Sullivan et al. (1970) found significant relationships with moral and conceptual level development. Groesch (1978) showed convergence with cognitive complexity, as did Lamb (19]1} with role taking perspective. Hoppe and Loevinger (1977) demonstrated the theoretically predicted curvilinear relation of conformity with ego development in a sophisticated study. Waugh and Mccaulley (J981) demonstrated discriminant validity with broad types of psychopathology. These studies are representative of the many reviewed in the major review papers referenced previously.

PAGE 24

16 Loeyip.ger 's (J. 9]6) a~sum!)ti,on of seq uential, hierarchical develop ment of stages has seen only min:i.,mal empirical study, the segment of contruct validation most lacking regarding the SCT. There is, however, one important longitudinal study in the literature. Redmore and Loevinger (J9J9). report on several longitudinal samples. Briefly, for their 8 subsamples, there was a mean rise in ego level and each test retest correlation was positive (_10 of 14 significant). Loevinger (J979) reasons that these results approach the maximum strength of as sociation possible given the reliability of the SCT and problems in re peated measurement. However, more molecular longitudinal study of ego development within homogenous groups and comparisons with cognitive de velopment are possible fruitful ways to studying the sequentiality assumption further. In summary, the construct validity of the SCT of ego development is favorable to the conception and methodology. The solid theoretical and empirical bases of Loevinger's (1976) measure provide a powerful frame work within which to study the interface of self-control, temperamental activity level, and psychological development.

PAGE 25

CHAPTER II METHODOLOGY It is hypothesized that self-control is a function of ego de velopment independent of the temperament dimension of activity level. This set of relations is believed to hold across all age levels. How ever, in this investigation, the age range is restricted to make the study practical. Accordingly, the age range chosen is that of middle childhood (Lidz, 1976); that is, the years within which children begin to broaden their social world beyond the family is called middle childhood. This socialization process largely takes place in elemen tary schools and the peer group becomes highly significant to the child (Parsons, 1964; Sullivan, 1953). This study follows a two-pronged procedure. First, a set of im pulsivity measures is examined for evidence of multidimensionality, in the expectation that a self-control and activity level dimension will emerge. Second, ego development is compared to each dimension. It is hypothesized that ego level will be related to self-control and not related to activity level. These relations are expected to hold in dependent of the chronological age and intellectual level of the in diyiduals. Subjects Male school children are chosen for study in the interest of controlling for the variable of sex because problems in self-control 17

PAGE 26

18 are more common in boys than girls. Subjects were recruited from the P.K. Yonge Laboratory school of the University of Florida in Gaines ville, Florida. The P.K. Yonge Laborator y school is especiall y ap propriate for this study because the composition of the student body is balanced with respect to race and socio-economic status, making the students roughly representative of children in the population of the Gainesville, Florida, community in terms of demographic variables. The subjects were recruited by sending a packet of material to the parent(s) of all the male students in grades 1 through 5. The packet included an informed consent letter and a parent rating form of the EASI-III. This material (see appendix) was sent to the parent(s) of 141 children. Those potential subjects whose parents did not return the material were recruited two additional times over a period of weeks. The final sample contained 78 subjects, representing a 55 % recruitment rate. The final sample for study includes 20 (25 % ) first graders, 17 (22%) second graders, 21 (27 % ) third graders, 12 (15 % ) fourth graders, and 8 (10 % ) fifth graders. Measures Seven measures of dependent variables and 3 measures of inde pendent variables are used. There are 3 dependent variables of activit y level and 4 dependent variables of self-control. Each is described below. Actiyity Level Sharron and Weller's (1971) Draw-A-Line test provides a measure of motoric activit y level. The time taken for a subject to draw a

PAGE 27

19 line between two horizontal lines (~ distance of 6 inches or 15.5 cen timeters) is considered an index of fine motor activity level. This measure is the first measure of activity level. The second measure is derived from a structured interview used by Scarr (1966). Scarr (1966) used a structured interview covering the kinds of activities children engage in to demonstrate that activity level possesses age netic component and thereby qualifies as a temperament dimension. Scarr (1966) found that the total number of activities endorsed as en gaged in was a function of heritability. In this research project an Activity Schedule is tailored from Scarr's (1966) structured inter view. The Activity Schedule (see appendix) consists of a list of 33 activities (e.g., jump rope, football, pinball, reading, bicycling) which is read to the subject. The subject is to state for each activity whether he engages in it. Each activity is considered in three situa tions: at home after school, at school, and on the weekends. The score on this measure is the sum of endorsements of all activities for all situations. The third measure of activity level is the activity level scale of Buss and Plomin's (1975) EASI-III temperament survey. Self-Control Sharron and Weller's (19712 Draw-A-Line test directl y yields a measure of motoric self-control. After the subject draws a line be tween two horizontal marks, the time taken to draw another line as slowly as possible is recorded. The difference between the two times is an index of fine motor inhibition. This procedure has also been

PAGE 28

20 used by Mac co by, J)owle:y, Hagen, and Degerman (1965)_ and Homatidis and Konstantareas (1981). The second measure of self-control is the im pulsivity scale of the EASI-III (Buss and Plomin, 1975). The third measure is a resistance to temptation experimental procedure. Resis tance to temptation situations has a long history of use in studies of self-control (e.g., Aronfreed and Reber, 1965; Grinder, 1962), but no standardized measure is available; investigators typically construct one appropriate to their study. In this study, the following items are used as stimulus objects: a high quality metaliic, moving-part toy steam roller truck; two attractive animal-shaped pencil sharpeners; three working miniature tools (wrench, hammer, pliers); and a furry, touchable toy animal. See appendix E for illustration of the items. These objects are placed in front of the subject on a desk and he is asked to remain seated and not to touch the toys as the examiner leaves the room for a few minutes. The examiner observes through a one-way window; if the subject does not touch the objects within five minutes, he has passed the procedure. The fourth measure of self control is a delay of gratification procedure. Delay of gratification experimental methods have been the topic of scores of studies in psychology (see Mischel, 1974). The most typical delay of gratifica tion situation has the subject choose between receiving a smaller reward immediately or a larger reward in the future. Selection of the larger reward is considered an index of self-control. In this study, subjects are presented the choice of either one balloon or colored pencil at the conclusion of the testing session, or if they wait for one week, a choice between either three colored pencils or three

PAGE 29

21 larger balloon~. The delayed reward choices are obviously more at tractive and valuable. Independent Measures Three independent variables are measured in this study. They are chronological age, recorded to the nearest month, a short-form intel lectual ability measure, and a short-form test of level of ego de velopment. Intellectual ability is estimated from the sum of the scale scores of the Vocabulary and Block Design subtests of the Wechsler Intelligence Scale for Children-Revised. Kaufman (1979) recommends this sort of short-form from the WISC-R. Level of ego development is typically assessed by a well-studied sentence completion test (Loevinger and Wessler, 1970). In this study, however, subjects are younger than subjects with whom the instrument has heretofore been used. Therefore, the SCT requires extension downward in age. As the theory of ego development covers life-span development and evaluations of the SCT are favorable, there is reason to believe such an extension is feasible. Moreover, the investigations of Lucas (1971), Farrell (1975), and Brinkerhoff (1971) have shown that ego development can be judged from interview material similar to that from the SCT. Thus, the construct appears general izable across methods of measurement. In addition, McCammon (1979) found that for sixth grade boy subjects, oral and written forms of the SCT were equivalent in ego level scores and correlation with IQ. Given these precedents, it is reasonable to adopt an oral interview form of the SCT in this research study.

PAGE 30

22 Each subject is giyen a st:i;-uc;tui;ed interview in which 23 items from the lQ-68 form of the SCT are stimulus items. Because of time constraints. and difficulty level of some of the items in the full 36 item SCT, a short form has been used. Holt (J980l has developed a 12 item SCT which retains excellent psychometric properties. Therefore, the 23 item short form should not compromise the psychometric quality of the SCT. Applying Cronbach' s (19701 procedure for estimating ef fects of changes in test length on coefficient alpha to the SCT shows that a 23 item SCT should reduce the alpha value from .9.1 to .87. This sacrifice of internal consistency reliability is considered mini mal relative to the gains in decreasing difficulty level of the items and ease of administration. Procedure Pilot study with the above described measures has shown that administration time involves about 40 minutes per subject. Therefore, administration has been conducted in one assessment session. The following order of administration has been used: (ll Draw-A-Line (normal and slow as possible conditions); (2) oral 23 item SCT inter view; (,31 activity schedule; (4) Vocabulary and B lock Design subtests of the WISC-R; (51 resistance to temptation experimental procedure; (61 delay of gratification choice. Subjects' parents completed the parent rating form of the EASI-III. Subjects were tested individually in a private room near the library of P,K. Yonge Laboratory School during the s chool day during s pring 1981. The large window in the door to the room was equipped witn one-way viewing plastic to permit qbservation of the subject during the resistance to temptation

PAGE 31

procedure, Ten subjects (5 second and 5 fourth graders) were re-tested with the activity schedule 7 to 10 days subsequent to testing to provide data for test re-test reliability assessment of the measure. Statistical Analyses Preliminary statistical analyses involve test re-test reliability assessment of the activity schedule and interrater reliability assess ment of the SCT. Total scores of the activity schedule for 10 sub jects re-tested are compared with a product moment correlation. The five dimensions of the activity schedule (like, afternoon, school, weekend, and total) are subjected to a principal components factor analysis to ascertain the dimensionality of the modified instrument (n = 78). The SCT protocols were typed onto separate lists for each stem of the SCT. This permitted raters to evaluate the responses blind from other responses of each subject. Two raters were used, myself and another graduate student at the University of Florida. The first rater (myself) is self-trained following Loevinger and Wessler's (1970) procedures, trained at the 1978 ego development scoring work shop at Washington University in St. Louis, and has demonstrated acceptable levels of interrater reliability in other samples (Waugh, 1981). The second rater is self-trained and trained with the first rater on other SCT data two years ago. The first rater scored the entire SCT data set; the second rater scored all of the responses for one half of the subjects. A product moment correlation was computed between the raters' total scores of 40 subjects to permit evaluation of interrater reliability. Total scores were formed by assigning the following numerical values to each item rating and summing the values

PAGE 32

24 of the 23 items 9f each protocol: 1-2 = l, 1-2/ t:. = 2; I-t:. = 3, 1t:./ 3 4; 1-3 = 5, 1-3/4 = 6; 1-4= 7; 1-4/5 = 8, 1-5 = 9, 1-6 = 10. SCT rating was accomplished using the manuals of Loevinger, Wessler, and Redmore (J9]0) and Redmore, Loevinger, and Tamashiro (1978, note 2). Factor Analysis of Dependent Measures The dimensionality of the set of 3 putative activity level and 4 self-control measures is evaluated factor analytically. The sample size of 78 subjects meets the minimum requirement of 10 subjects per variable (Skinner, 1931). It is hypothesized that 2 factor dimensions exist in the data; thus the factor analysis is a hypothesis testing procedure, As the number of subjects is relatively small, the number of variables is relatively small, and the purpose is hypothesis testing, a principal factor method is considered the method of choice (Gorsuch, 1974). The iterated principal factor method with rotation to varimax solution of the Statistical Package for the Social Sciences was used (Nie, Hull, Jenkins, Steinbrenner, and Bent, 1975). Two criteria were used to evaluate the number of factors to retain. First, the typical criterion of all eigenvalues greater than 1 was used to establish the number of factors. This number was confirmed by Cattell's (1978) scree test. The scree test involves plotting the eigenvalues of all variables from a principal component analysis against the number of variables and drawing a line between the points. The point at which the curve veers sharply upward is associated with the number of factors to retain. Following Gorsuch (1974), because the factor study is relatively mitigated by a minimum number of subjects, variables

PAGE 33

25 loading greater than or equal to the absolute value of .35 on retained factors are interpreted. This rather atringent interpretational criterion increases the likelihood that results are reliable. Follow ing the iactor analysis of the 7 variable set of dependent measures, factor scores of each subject on each retained factor dimension were computed. These factor score computations permit evaluation of rela tions with the set of independent variables. Multiple Regression of Independent and Dependent Variables It is predicted that ego level is associated with self-control and not activity level, independent of the effects of age and intel lectual ability level. These relations are tested by multiple regres sions of age, IQ, and item sum SCT scores on each of the sets of fac tor scores for the retained factor dimensions. The SPSS multiple re gression statistical procedure was used.

PAGE 34

CHAPTER III RESULTS Preliminary Analyses Preliminary analyses include interrater reliability assessment of SCT ratings, test-retest reliabilit y assessment of the activity sched ule, and factor anlaytic assessment of the activity schedule. Interrater Reliability Assessment of the SCT Ratings A Pearson correlation coefficient was computed between the two raters' sets of SCT item sum scores for 40 subjects. The resulting correlation coefficient of .73 is considered of sufficient magnitude for research purposes but lower than values typically obtained with the SCT (Loevinger and Wessler, 1970). The most obvious explanation for this relatively lower level of agreement is that the rating manuals were developed from samples that did not include subjects of the age range in this study. Therefore, specific verbal content character istic of children in middle childhood is not represented in the manuals as aids to rating, and raters must make discriminations more on the basis of theor y and extrapolation from the language st y le of older but low ego level subjects, Thus, use of the SCT with subjects in middle childhood seems to entail a sacrifice, although not severe, in degree of rater agreement. 26

PAGE 35

27 Psychometric Assessment of the Activity Schedule Test-retest reliablity and factor anal y tic study of the activity schedule were conducted to ascertain whether the minor modifications of Scarr's (1966) procedure were deleterious to the psychometric char acteristics of the structured interview. The test-retest Pearson cor relation coefficient between the initial and 7 to 10 day retest activity schedule total scores for 10 subjects gave a value of .73. This is within acceptable ps y chometric limits. The principal components factor analysis of the 5 variables in the activit y schedule (like, afternoon, school, weekend, and total) yielded 1 factor with an eigen value of 3.5 accounting for 70% of the total variance. The following factor loadings were obtained: like= .56; afternoon= .92; school= .72; weekend= .92; total= .99. Since the total score is a composite of the afternoon, school, and weekend scores, it is not independent and tends to inflate the degree of intercorrelation. However, it is obvious that a single source of variance underlies the activity schedule and the best expression of this variance is the total score variable. These analyses confirm that the minor modifications of Scarr's (1966) procedure in developing the activit y schedule have not compromised its psychometric credibilit y Dimensionality of the Set of Dependent Variables An iterated principal factor method of factor analysis was ap plied to the set of 7 dependent variables measured on 78 subjects. This analysis required 18 iterations and y ielded a 3 factor solution

PAGE 36

28 accounti_ng for 59.3% of the total variance. Both the eigenvalue greater than or equal to -1 and Cattell' s 0978) scree test criteria indicated that a 3 factor solution, as opposed to the hypothesized 2 factor solution, was appropriate (?ee Table 3 and Figure l). Table 3 shows the loadings of the variables on the 3 factors subsequent to Varimax rotation. Using the criterion of interpreting only those loadings greater than .35, it is seen that the first factor (Yl) is comprised of high and positive loadings for the EASI-111 impulsivity and activity level scales. The second factor dimension (F2) has a high and positive loading for the activity level total score and a moderate and negative loading for the resistance to temptation measure. The third factor (FJ) is defined only by a high and positive loading for the delay of gratification measure. These results are contrary to the predicted number of factors and pattern of loadings. Although exact predictions were not supported, the meanings of the results are consistent with the hypotheses, as discussed below. Interpretation of the Fl dimension is problematic. 1bere are two ways in which to understand this dimension. First of all, because only EASI-111 variables define this dimension, and both variables share a common method of measurement (parental ratings), it is reason able to assume that this dimension simply is a method-specific factor for parental rating. However, a second possibility is that Fl repre sents a "difficult temperament" dimension. The literature on tempera ment supports the idea that a small set of second order factors subsume

PAGE 37

29 TABLE 3. Principal Axes Factor Anal y sis of Dependent Variables with Varimax Rotation. Variables Communalit y Eigenvalue Factor 1 Factor 2 Factor 3 resistance to .196 1.628 -.054 -.363 .248 temptation delay of grati.275 1.464 -.060 .015 .521 fication normal draw rate .094 1.057 -.092 177 -.233 draw slow differ.140 .948 2 65 -.238 .116 erence score activity schedule .655 .853 .124 .797 .071 EASI-III impul.991 .624 .995 .000 -.012 sivity EASI-III activit y .219 .426 4 52 .115 -.043 level

PAGE 38

.65 1.60 .55 .50 .45 .40 .35 .30 .25 .20 .15 .10 .05 1.00 .95 .90 .35 .80 .75 70 .65 .60 .55 .50 .45 .40 1 30 2 3 4 5 6 7 Number of Factors FIGURE 1 Scree Test Indicating 3 Factor Solution

PAGE 39

31 th~ ThqII,laS. an d Ches ) ,, .. (J.9J7) cons.tellati_on of teI!lperament dimens ions (J'lo_m:j_p 1 19.82)_. Thomas and Chess (1968) __ have delinec;1ted three groups of children def ine.d b .. f different pat terns of temperament. The y are termed the "difficult," "easy,'' and "slow to warm up." The "diffi cult" pattern is. characterized b y low rhymicit y negative mood, low approach, low adaptibility, and high intensity ('.Thomas and Chess, 1977) In order to investigate the possibilit y that the Fl dimension reflects the EASI-III version of a "difficult y second order factor, the full EASI-III was subjected to a principal factor anal y sis on a post hoc basis. After 15 iterations, 2 factors with eigenvalues greater than 1 emerged and accounted for 76.1% of the total variance. The first EASI-III factor is defined by emotionality, impulsivity, and activity level (?ee Table 4); the second factor is composed of socia bility and, to a lesser extent, activity level. This first factor appropriately can be called "difficult y and the second factor appears to be an "easy" dimension. There.fore, it would be reasonable to re gard the Fl dimension, with its impulsivity and activity level scale composition, as an EASI-III "difficult y dimension. Unfortunately, in the present study, there is no independent means of ascertaining which--if either--interpreation of Fl is justified. The most cautious interpretation is to consider Fl to be a method specific rating factor. The second factor ( f ll is defined b y the activit y schedule total score and the resistance to temptation score (high scores on this di mension are related to touching the to y s when alone. after being asked not to}. The activit y schedule is thought to assess a heritible

PAGE 40

32 TABLE 4. Principal Axes Factor Analysis of EASI-III Scales w:ith Yarimax Rotation. Variables Eigenvalue Factor 2 Factor 2 Emotionality l.d31 .729 -.243 Activity Level 1.212 5-14 .469 Impulsivity .519 728 .191 Sociability .438 -.036 .647

PAGE 41

33 temperP,II1ent of actiyity level (~carr, 1966), and this variable most saliently defines F2. Although the resistance to temptation measure was hypothesized to reflect self-control, in view of its pairing with activity level, it is more properly seen as an index of amount of motoric activity rather than ability to inhibit motoric response. Therefore, the F2 dimension appears to be an activity level factor. The third factor (F3) is defined by the delay of gratification experimental procedure. Persons scoring high on F3 demonstrate the ability to wait for a preferred reward as opposed to accepting a less preferred one immediately. This dimension most closely reflects the hypothesized self-control factor. In summary, the factor analytic results show that a three factor solution is appropriate to explain the interrelations in the set of dependent variables. The first factor is ambiguous and open to in terpretation as either a method-specific or "difficulty" dimension; in the interest of caution, the method-specific interpretation is chosen. The second factor is considered a temperamental dimension of activity level. The third factor, although defined only by delay of gratification, appears to be a self-control dimension. In the next section, these dimensions are related to age, intellectual level, and level of ego development to determine whether Fl, F2, and F3 are developmental dimensions, specifically if they are ego developmental dimensions independent of age or level of intellectual ability.

PAGE 42

34 Relations of Age, IQ, and Ego Level to Factor Dimensions Relationships of age, IQ, and level of ego development to the three factor dimensions were examined b y computing multiple regressions of the independent variables on the factor scores of Fl, F2, FJ. Table 5 shows summary statistics. The Fl dimension is not related to age, IQ, or ego level singl y or in combination. Therefore, Fl can not qualify as a developmental factor. Dimension F2, called activity level, is predicted mostly b y IQ; the contributions of age and ego level are minimal. F2 seems to be associated with lower intellectual performance. In view of the fact tht IQ tests, such as the WISC-R, assess scholastic ability, it is not surprising that individuals less scholastically able, and by inference less scholastically-minded and reflective, tend to be more action-oriented. The F3 dimension is sig nificantly predicted by both age and ego level, but the independent contribution of ego level is negligible. Therefore, this delay of gratification/self-control dimension qualifies as a developmental dimension. Although level of ego development is associated with this developmental dimension of self-control, ego development is less important than simple chronological age. Post Hoc Factor Analyses Previous hypothesis testing methods have yielded results which support the hypotheses, in general; however, ambiguit y was found in the interpretation of the first factor in the set of dependent vari ables. In addition, the second factor, called activity level, was found to be related to intellectual ability. The problem with this

PAGE 43

35 TABL E 5, Summar y Statistics of Regression of A ge, IQ, and Ego Level on Factor Scores. Dependent Variable Factor l Factor 1 Factor 1 Factor 2 Factor 2 Factor 2 Factor 3 Factor 3 Factor 3 Independent Variable (s) Age Age+IQ Age+IQ+Ego Level Age Age+IQ Age+IQ+Ego Level Age Age+IQ Age+IQ+Ego Level Multiple R .023 .096 .193 .035 .330 .342 .408 .420 .425 Simple R -.023 .096 -.122 -.035 -.310 -.045 .408 .001 2 23 Note. = p < .025; ** = p < Ol; *** = p < .001 F test of Model .038 .351 .957 .093 4.567 ** 3.263 15 .130 *** 8.013 *** 5.429 ***

PAGE 44

36 interpretation is thq,t other research ('.Thomas and Chess, 1977; Plomin, 1982) finds no association between I Q and temperament, and the present putative activit y level dimension thus becomes open to question. In order to provide further means of responding to these ambiguities, two factor analyses are presented. Both analyses are tentative in impli cation because they are post hoc and severely strain their reliabilit y because the ratio of measures to subjects is so large. The first method uses a principal axes factor analysis of the set of 7 dependent variables combined with the two temperament dimensions of the EASI-III not included in the hypothesis-testing portion of this study. If the first factor obtained in the original factor analysis indeed is a difficulty factor, then this factor should include the variable of emotionalit y (_intensity of response). If this result fails to occur, or all of the EASI-III variables themselves define the factor, then the interpretation of it as a method-specific dimension gains credibilit y Table 6 presents the results from this analysis, which required 25 iterations and accounted for 66.1 % of the total variance among the 9 variables. The number of factors to retain was determined by the eigenvalue greater than or equal to 1 criterion. Four factors were retained. As can be seen, the first factor in this analysis appears to be a "difficult y dimension defined b y impulsivity, activity, and emotionalit y Therefore, although this result is tenta tive, more weight is added to the argument that this dimension is not simply a method-speci f ic rating factor. The second factor is nearl y identical to the previous dimension called activit y level. The third factor is defined onl y b y sociabilit y and probabl y can be called an

PAGE 45

37 TABLE 6. Principal Axes Factor Anal y sis of E A SI-111 Scales and Dependent Variables with Varimax Rotation. Variable Factor 1 Factor 2 Factor 3 Factor 4 Emotionality .494 .117 -.287 .467 Activity Level .590 .108 .115 -.093 Sociability .259 .017 758 -.196 Impulsivity .821 .046 -.090 158 Normal Draw Rate -.042 .121 -.103 -.609 Draw Slow Diference Score .278 -.236 .045 .056 Delay of Gratification -.075 -.011 .314 .121 Resistance to Temptation -.063 355 .276 .149 Activity Schedule -.098 .842 .109 -.005

PAGE 46

38 "eas y" tempe rcnent dimension, The fourth factor is comprised of the Draw-A-Line activity level measure and emotionalit y A tentatiye interpretation is that this represents an intensit y of response, as the Draw-A-Line task requires subjects to draw without instructions, and those who are temperamentall y intense responders would be expected to execute the task quickl y ; this is in keeping with the strong loading of emotionality. It is interesting that delay of gratification failed to emerge in the definition of a factor. M ost likely, the develop mental variance in this set of variables is minimal and thereby swamped, preventing it from emerging as a dimension. This reinforces the notion that developmental variance is not strong in temperament, activity level, and putative self-control measures. The second post hoc method of analysis is even weaker statistically than the first because 12 variables measured on 78 subjects are used. Nevertheless, tentative influences of the independent variables (age, IQ, ego level) are observable directl y in their contributions to the factor structure, and can help in evaluating the importance of de velopmental variance and intellectual ability variance, especiall y re garding the putative activit y level dimension. The principal axes factor analysis required 25 iterations and accounted for 67 % of the total variance Table 7 shows the results. Factor one again is the "difficulty" of temperament dimension. Factor two is a solid develop mental dimension made up of age, dela y of gratification, and ego development, respectively. Thus, insofar as self-control represents a developmental phenomenon, onl y dela y of gratification appears to qualify as a developmental/self-control dimension. Factor three is

PAGE 47

39 TABLE 7. Principal Axes Factor Analysis of EASI-111 Scales, Dependent, and Independent Variables with Varimax Rotation. Variable Factor 1 Factor 2 Factor 3 Factor 4 Factor 5 Emotionality .709 -.032 -.038 -.209 -.003 Activity Level .444 -.096 .035 .379 .151 Sociability -.014 .107 .022 799 -.145 Impulsivity .786 -.086 .083 .180 .051 Normal Draw Rate -.240 -.049 .076 .081 .320 Draw Slow Difference Score .261 .140 .288 .063 -.164 Delay of Gratification -.045 .528 .029 .140 -.068 Resistance to Temptation -.077 .126 .026 .150 -.513 Activity Schedule .133 .063 -.305 .153 .554 Age .081 .814 -.193 -.100 -.117 IQ .034 -.043 .912 .025 -.065 Ego Level -.146 .468 .299 .004 .037

PAGE 48

40 defined only by IQ, Factor four is. the "easy" temperament dimension. Factor fiye is the putative activity level dimension, consisting of resistance to temptation and activity schedule variables. It is important to note that, in this analysis when IQ was allowed to covary among the complete set of variables, IQ did not contribute to the activity level factor. Thus, although IQ predicts the factor dimen sion of activity level in multiple regression analysis, IQ cannot be considered an intrinsic component of the dimension. This finding, weak and unreliable as it is, serves to dampen the criticism of the interpretation of the dimension as activity level because of its correlation with IQ, which after all, amounts to only about 9 % of the variance in activity level.

PAGE 49

CHAPTER IV DISCUSSION Principal Findings This study began from the perspective that two major classes of personality constructs, called horizontal and vertical dimensions, can offer conceptual power in understanding personality. Specific use of this framework was made in studying the hypothesized independence of self-control and activity level temperament dimensions through factor analytic procedures. It was found that common measures of activity level and self-control could not be separated into two major dimensions analogous to a developmental and trait distinction; rather, three dimensions were necessary in this particular set of measures. However, two of these dimensions appeared to conform to the developmental and trait distinction. The other factor was viewed as either a method specific or "difficulty" of temperament factor. Relations of these dimensions to predictors of age, IQ, and level of ego development revealed both expected and unanticipated results. The first factor, probably a "difficulty" of temperament factor, but cautiously labelled a method-specific factor, was not predicted by age, IQ, or ego development; this is consonant with theory. The second factor, considered an activity level dimension, was unexpectedly predicted by intellectual ability. This is discrepant with previous temperament research. A 41

PAGE 50

42 post hoc f~ctor analysis suggested that IQ is not an intrinsic compo nent of this activit y level factor but a correlate. This correlation is understood as a function of the temperamentally active child's action orientation as opposed to a more reflective, scholastic style. However, this interpretation is not established in the present study because of the lack of external confirmation. The third factor was related to both age and level of ego development, but the contribu tion of ego level was insignificant compared to simple chronological age. This third factor is considered a self-control dimension which is solidly developmental in nature, an interpretation partially but tressed by the tentative post hoc analysis. Evaluation of the Methodology Methodologically, this study possesses relative strengths and weaknesses. Factor analytic procedures played an important role in the study. The major factor analysis used 7 variables measured on 78 subjects. This ratio approaches the limits of providing results that can be considered statistically reliable (Skinner, 1981). On the other hand, care was exercised in extracting an appropriate number of factors through the use of two reputable techniques and factor loadings were interpreted b y a stringent criterion of loadings greater than .35. These safeguards enhance the reliability of the factor results even though no cross-validation was possible. Moreover, an approp riate factor model was employed, the principal axes model. This is a poor comment on the state of multivariate study of self-control,

PAGE 51

43 impulsivity, and activity level constructs to claim as an advantage the appropriate use of factor models, but most previous research (e.g., Paulsen and Johnson, 1980; Homatidis and Konstantareas, 1981) has used an inappropriate model (principal components), which serves to inflate spuriously the magnitude of the relations. In considering the methodo logical strengths and shortcomings of the present study, the balance indicates that this study represents a significant advance in methodo logical sophistication. Another possible criticism of this research concerns the specific measures. The Buss and Plomin (1975) EASI-III temperament survey is apt to be replaced by the Lerner, Palermo, Spiro, and Nesselroade (1982) Dimensions of Temperament Survey in terms of providing a stan dardized temperament measure applicable to a broad age range. Thus, the present findings may not be directly comparable to future research employing the more sophisticated Dimensions of Temperament Survey; however, this instrument was not available when the present study was conducted. More damaging, however, is the use of the Draw-A-Line procedure. This measure was not found to relate to any other; thus it is likely that the Draw-A-Line technique is a poor choice in studies of impulsivity and activity level (_the Draw-A-Line activity level score did relate to an intensity temperament dimension in a very tentative post hoc analysis). It is possible that this absence of effect results from the fact that the Draw-A-Line test consisted of 1 "item'' and therefore is an underestimate of a possible "true score" which might manifest upon multiple item assessment. At any rate, the

PAGE 52

. 44 delay of gratification and resistance to temptation variables showed solid inter-variable relations even though they too are 1 "item" measurements (~nd scored on a dichotomous basis). Thus, these experi mental procedures should be considered superior construct measurements for use in further research, which could be enhanced by multiple "item" measurement. Implications of the Study This factor analytic study of temperamental and developmental di mensions of self-control in middle childhood has many implications for study of self-control, temperament, ego development, and by inference, behavior disorders of middle childhood. These implications are indi cated below. Prominent constructs in personality theory, particularly within the realm of developmental personality theory, include self-control, temperament, and ego development. In this study, it has been seen that the proposed horizontal and vertical dimensions distinction is teneable. This idea strengthens the trait and developmental distinc tion of Achenbach and Weisz (1975) and Swensen (1980, Note 1). An important implication is that a host of personality variables hereto fore considered traits may be due evaluation in terms of horizontal and vertical dimensional status. In addition, the importance of temperament theory is evident both in personality theory in general, and in terms of a horizontal class of constructs. Particularly, the notion of "difficulty" of temperament, at least as that perceived by parents, has received confirmation of importance in this study.

PAGE 53

45 Measu1;ement of level of ego development during middle childhood by the SCT has been found more difficult in regard to rater agreement. It is interesting that ego level, while still showing significant associa tions, does not appear to be as confounded by IQ or age in this age range as in other developmental samples (Loevinger, 1979); these correlations are shown in Appendix F. More importantly, the relation of ego development to theoretically relevant constructs, such as impulsivity or possibly activity level, has been found to be minimal. The most reasonable explanation for this absence of association is that the SCT does not carry sufficient variance in this admittedly restricted ego level range within middle childhood. This research has involved normal males in middle childhood in the study of normal personality constructs. However, there are impli cations to behavior disorders of childhood. Recent research on the classification of child psychopathology has established the importance of two broad and opposing classes of behavior disorders called exter nalizing and internalizing dimensions (Achenbach and Edelbrock, 1978). The externalizing group of disorders includes conduct disorder, aggres sion, and acting out problems; the internalizing dimension involves anxiety and personalit y disorder problems. Underl y ing this broad distinction may be basic temperament differences such as sociability, activity level, and impulsivit y Previous temperament research has suggested such an implication also (Thomas and Chess, 1977; Terestman, 1~81; Carey, McDevitt, and Baker, 1980). For example, Thomas and Chess (1917)_ have found that "difficult" temperament children have

PAGE 54

46 a higher incidence of clinical disorders upon longitudinal followup. Carey, McDevitt, and Baker (1980) found that temperament dimensions of adaptibility, intensity, activity, and distractibility are more frequent in clinical samples of child behavior disorders, but they could not reliably discriminate minimal brain dysfunction, learning disability, and mixed conduct disorder groups on the basis of tempera ment. Carey, McDevitt, and Baker (1980) do caution against the diag nosis of minimal brain dysfunction from observational behavior data alone because of the confound with temperament dimensions in the absence of neurological confirmation. In the present study, self control, hence impulsivity, and temperamental activity level dimensions were found orthogonal. Therefore, Carey, McDevitt, and Baker's (1980) cautionary message to diagnosticians is reinforced by the independence of impulsivity and activity level, which in clinical practice are often lumped into categories such as hyperactivity and conduct dis order. It is also interesting to speculate on the nature of the activity level dimension elucidated in the present study in regard to its association with lower intellectual performance. It may be a dimension such as this that resembles more closely an externalizing form of behavior disorder vis-a-vis joint effects of high activity level and limited cognitive appraisal abilities. Of course, this suggestion needs to be examined directly in further research.

PAGE 55

REFERENCE N OTES l. Swensen, C. Invited address, 19J8, ego development scoring workshop. St. Louis: Washington Universit y Social Science In stitute, N ovember, 1978. 2. Redmore, C., Loevinger, J. and Tamashiro, R. Measuring Ego development: Scoring manual for men and boys (Pre-publication version}. St. Louis: Washington Universit y Social Science Institute, August, 1978. 47

PAGE 56

REFERENCES Aaronfreed, J., and Reber, A. Internalized behavior suspension and the timing of social punishment. Journal of Personality and Social Psychology, 1965, .!._, 3-16. Achenbach, T., and Edelbrock, C. The classification of child psycho pathology: a review and analysis of empirical efforts. Psycho logical Bulletin, 1978, ~. J275-13Ql. Achenbach, T., and Weisz, J.R. Impulsivity-reflectivity and cogni tive development in preschoolers; A longitudinal analysis of de velopmental and trait variance. Developmental Psychology, 1975, 11, 413-414. Allport, G.W. Personality: A psychological interpretation. New York: Holt, 1937. Bandura, A. change. Self-efficacy: Toward a unifying theory of behavior Psychological Review, 1977, 84, 191-215. Block, J., Block, the Matching impulsivity. J.H., and Harrington, D.M. Familiar Figures Test as a Developmental Psychology, Some misgivings about measure of reflection1974, 1:Q_, 611-632. Brinkerhoff, R.S. Ego development in adolescent girls. Unpublished doctoral dissertation, University of Chicago, 1971. B.uss, A.H., and Plomin, R. A temperament theory of personality de velopment, New York: Wiley, 1975. Carey, W.B., McDevitt, S.C., and Baker, D. Differentiating minimal brain dysfunction and temperament. In S. Chess and A. Thomas (~ds.), Annual progress in child psychiatry and child develop ment. New York: Brunner(Mazel, 1980. Cattell, R.B. The scientific uses of factor analysis in behavioral and life sciences. New York: Plenum Press, 19J8. Cronbach, L.J. Essentials of psychological testing (3rd ed.). New York; Harper and Row, 1970. 48

PAGE 57

49 Davids, A. Ego functions in distrubed and normal children: Aspira tion, inhibition, time estimation, and delayed gratification. Journal of Consulting and Clinical Psychology, 1969, l.2, 61-70. Davison, M.L., King, P.M., Kitchener, K.S,, and Parker, C.A. The stage sequence concept in cognitive and social development. Developmental Psychology, 1980, .:!_, 121-131. Farrell, G.E.P. The relation of ego development to intellectual and ethical development (Doctoral dissertation, Washington Univer sity, 1~74). Dissertation Abstracts International, 1975, 35, 4648B-4649B. Frank, S., and Quinlan, D.M. Ego development and female delinquency: A cognitive-developmental approach. Journal of Abnormal Psy chology, 1976, ~' 505-510. Freud, S. The interpretation of dreams. London: Hogarth, 1953. Goldfried, M.R., and Merbaum, M. (Eds.).. Behavior change through self control. New York: Holt, Rinehart, and Winston, 1973. Gorsuch, R.L. Factor analysis. Philadelphia: Saunders, 1974. Grinder, R.E. Parental child-rearing practices, conscience, and re sistance to temptation of sixth grade children. Child Develop ment. 1962, l.2, 803-820. Groesch, S. Ego development and cognitive complexity: integrative and discriminative complexity. (Doctoral dissertation, University of Notre Dame, 1978). Dissertation Abstracts International, 1978, 12_, 978B-979B. Hauser, S.T. Loevinger's model and measure of ego development: A critical review. Psychological Bulletin, 1976, 83, 928-955. Holt, R.R. Review of "Measuring ego development, Volumes I and IL" Journal of Nervous and Mental Disease, 1974, 158, 310-316. Holt, R.R. Loevinger's measure of ego development: Reliability and national norms for male and female short forms. Journal of Personality and Social Psychology, 1980, 1.2_, 909-920. Homatidis, S., and Konstantareas, M.M. Assessment of hyperactivity: Isolating measures of high discriminant ability. Journal of Consulting and Clinical Psychology, 1981, ~. 533-541.

PAGE 58

50 Hoppe, C. F. a,nd Loevinger, J. Ego development and conformity: A construct ya,lidity study of the W ashington University Sentence Completion Tes.t. Journal of Pe_rs ona,lity Assessment, 19]7, ~, 49-7-504. Karoly, P. Behavioral self-management in children: Concepts, methods, issues, a,nd directions, In M. Hersen, R.M. Eisler, and P.M. Miller (~ds,l. Progess in behavior modification, Vol. 5. New York: Academic Press, 19]7, Kaufman, A.S. Intelligent testing with the WISC-R. New York: Wiley, 19-79. Kendall, P.C., and Finch, A.J., Jr. A cognitive-behavioral treatment for impulsivity: A group comparison study. Journal of Consulting and Clinical Psychology, 1978, ~. 110-118. Kendall, P.C., and Finch, A.J., Jr. Developing non-impulsive behavior in children: Cognitive-behavioral strategies for self-control. In P. C. Kendall and S. D. Hollon (Eds. 2 Cognitive behavioral intervention: Theory, research, and procedures. New York: Academic Press, 1979. Kohlberg, L. Development of moral character and moral idealogy. In M.L. Hoffman and L.W. Hoffman (Eds.) Review of child development research (Vol. 1). New York: Russell Sage Foundation, 1964. Kopp, C.B. Antecedents of self-regulation: A developmental per spective. Developmental Psychology, 1982, ~' 199-214. Lamb, N.J. The relation of role-taking to ego development (Doctoral dissertation, Washington University, 1971). Dissertation Ab stracts International, 1971, ~, 2401B. Langer, J. Theories of development. New York: Holt, Rinehart and Winston, 1969. Leary, T. Interpersonal diagnosis of personality. New York: Ronald Press, 1957. Lerner, R.M., Palermo, M., Spiro, A., and Nesselroade, J. the dimensions of temperamental individuality across span: The Dimensions of Temperament Survey (DOTS). velopment, 1982, 53, 149-159. Li dz, T. The person. New York: Basic, 19]6. Assessing the life Child DeLoevinger, J. Three principles for a psychoanalytic psychology. Journal of Abnormal Psychology, 1966a, l...!_, 432-443.

PAGE 59

51 Loevinger, J. The meaning and measurement of ego development. Ameri can Psychologist, 1966b, Q, 195-206. Loevinger, J. Ego development: Conceptions and theories. San Francisco: Jossey-Bass, 1976. Loevinger, J. Construct validity of the sentence completion test of ego development. Applied Psychological Measurement, 1979, 1_, 281-311. Loevinger, J., and Wessler, R. Measuring ego development 1. Construc _t_i_o_n_a_n_d_u_s_e_o_f_a_s_e_n_t_e_n_c_e_c_o_m_.p_l_e_t_i_ o_n __ t_e_s_t. San Francisco: Jossey-Bass, 1970. Loevinger, J., Wessler, R., and Redmore, C. Measuring ego develop ment 2. Scoring manual for women and girls. San Francisco: Jossey-Bass, 1970. Lorr, M., and Manning, T. T. Measurement of tence completion and personality test. chology, 1978, 34, 354-360. ego development by sen Journal of Clinical PsyLucas, P. Validation of a test of ego development by means of a standardized interview (Doctoral dissertation, Washington Uni versity, 1971). Dissertation Abstracts International, 1971, & 2204B. Maccoby, E.E., Dawley, E.M., Hagen, J.W., and Degerman, R. Activity level and intellectual functioning in normal preschool children. Child Development, 1965, 1._, 761-770. Mccrae, R.R. and Costa, P.T. Openness to experience and ego level in Loevinger's sentence completion test: Dispositional contribu tions to developmental models of personality. Journal of Person ality and Social Psychology, 1980, 12_, 1179-1190. Maslow, A. Motivation and personality. New York: Harper and Row, 1970. McCammon, E.P. A comparison of oral and written forms of the sentence completion test for ego development. Unpublished master's thesis, Washington University, 1979. Mead, G.H. On social psychology. Chicago: University of Chicago Press, 1964. Meichenbaum, D. Teaching children self-control. In B. Lahey and A. Kazdin (~ds.). Advances in child clinical psychology, Vol. 2. New York: Plenum, 1979.

PAGE 60

52 Mischel, W.. Processes in delay of gratification. In L. Berkowitz (Ed.) Advances in experimental social psychology, Vol. 7. New York : Academic Press, 19-74. Nie, N.H., Hull, C.H., Jenkins, J.G., Steinbrenner, K., and Bent, D.H. Statistical package for the social sciences (2nd ed.). New York: McGraw-Hill, 1975. Parsons, T. Social structure and personality. London: Free Press of Glencoe, 1964. Paulsen, K., and Johnson, M. Impulsivity: A multidimensional concept with developmental aspects. Journal of Abnormal Child Psychology, 1980, ~. 269-277. Pavlov, I.P. Twenty years of study of higher nervous activity (beha vior) in animals. Warsaw: PZWL, 1952. Piaget, J. The moral judgment of the child. New York: Free Press, 1932. Plomin, R. Childhood temperament. In B. Lahey and A. Kazdin (Eds.), Advances in clinical child psychology (Vol. 6.). New York: Plenum, 1982. Redmore, C.D., and Loevinger, J. Ego development in adolescence: Longitudinal studies. Journal of Youth and Adolescence, 1979, ~. 1-20. Redmore, C.D., and Waldman, K. test of ego development. 1975, 12_, 236-243. Reliablity of the sentence completion Journal of Personality Assessment, Rowe, D.C. and Plomin, R. Temperament in early childhood. Journal of Personality Assessment, 1977, Q, 150-156. Rozek, F., Wessman, A.E., and Gorman, B.S. Temporal span and delay of gratification as a function of age and cognitive development. Journal of Genetic Psychology, 1977, 131, 37-40. Scarr, S. Genetic factors in activity motivation. Child Develop ment, 1966, 37, 663-673. Sharron 1 S., and Weller, L, Classification patterns in underprivi leged ch~ldren in Israel. Child Development, 1971, ~, 581-594. Sheldon, W. The varieties of temperament: A psychology of constitu tional differences. New York: Harper, 1942.

PAGE 61

, 53 Skinner, H,A. Factor analysis and studies on alcoholism: A methodo logical review. Journal of Studies on Alcoh ol, 1981, 42, 109J.1101. Sullivan, E., McCullough, G,, and Stager, M. A developmental study of the relationship between conceptual, ego, and moral develop ment. Child Development, 19]0 ~, 399-411. Sullivan, H.S. The interpersonal theory of psychiatry. New York: Norton, 1953. Terestman, N. Mood quality and intensity in nursery school children as predictors of behavior disorder. In S. Chess and A. Thomas (Eds.), Annual progress in child psychiatry and child develop ment. New York: Brunner/Maze!, 1981. Thomas, A., and Chess, S. Temperament and development. New York: Brunner/Maze!, 1977. Thomas, A., and Chess, S. The dynamics of psychological development. New York: Brunner/Maze!, 1980. Thomas, A., Chess, S., and Birch, H. Temperament and behavior dis orders in children. New York: New York University Press, 1968. Thomas, A., Chess, S., Birch, H., Hertzig, M., and Korn, S. Beha vioral individuality in early childhood. New York: New York University Press, 1963. Wardell, D., and Royce, J.R. Relationships between cognitive and temperament traits and the concept of "style". Journal of Mul tivariate Experimental Personality and Clinical Psychology, 1975, .!., 244-256. Wardell, D., and Royce, J.R. Toward a multi-factor theory of styles and their relationships to cognition and affect. Journal of Personality, 1978, 46, 474-505. Waugh, M.H. The verisimilitude of Loevinger's measure of ego develop ment in a clinical population. Unpublished master's thesis, Uni versity of Florida, 19]~. Waugh, M.H, Reliability of the sentence completion test of ego de velo?ment in a clinical population. Journal of Personality As sessment, 1981, ~. 485-487. Waugh, M.H., and McCaulley, M.H. Relation of level of ego development to type and severity of ps y chopathology. Journal of Consulting and Clinical Psychology, 1981, ~. 295-296.

PAGE 62

54 Weisz, J.R., O'Neill, P,, and O'Neill, P.C. Field dependence-inde pendence on the Children's Embedded Figures Test: Cognitive style or cognitive level?. Developmental Psychology, 1975, l.!_, 539-540 .

PAGE 63

APPENDIX A START 55

PAGE 64

APPENDIX B SHORT FORM SENTENCE COHPLETION TEST FOR BOYS l A man s j ob 2. Being with other people 3. The thing I like about myself is 4. If my mother 5. If I can't get what I want 6. When I was younger 7. Women are lucky because 8. What gets me into trouble is 9. A good father 10. If I were king 11. I feel sorry 12. A man should always 13. Rules are 14. He felt proud that he 15. Men are lucky because 16. My father and I 17. A man feels good when 18. When I get mad 19. At times he worried about 20. My main problem is 21. Sometimes he wished that 22. If I had more mone y 23. I just can't stand people who 56

PAGE 65

APPENDIX C ACTIVITY SCHEDULE Ask the child if he likes to do each activity (yes or no) at home after school; repeat for at home on the weekend and at school. Then, ask the child if he does each activity at home after school; repeat for at home on the weekend and at school. ACTIVITY LIKES (_in general) jump rope monkey bars, etc. read books watch tv. hop scotch tag football baseball basketball soccer swim run board games, cards dolls movies hide & seek make models toy soldiers bicycle kickball play with pet kick-the-can draw, paint, color skateboard roller skate pinball walk dodge ball DOES After 57 Wk.end. School

PAGE 66

58 APPENDIX C Actiyity Schedule (~ont.2 ACTIVITY LIKES (jn general) army, cowboy, spacemen go shopping music hunt fish Does After Wkend, School

PAGE 67

APPENDIX D PARENT RATING FORM Name of child Grade -------------------------Rater Age of Child --------------------------Please rate this child in terms of how characteristic the statements are of him. The underlined l in the center of each row represents where the average child would fall on thls item. Please do not hesi tate to use the entire range of possible ratings. l. Child tends to be nervous in new situations. 1 No 2. Child often acts on the spur of the moment. 3. 4. 5. 6. ]. 8. 9. Child gets bored easily Child sometimes does ''crazy" things just to be different. Child tends to be shy. Child has many friends. Child yells and screams more than most children his age. Child finds self-control eas y to learn. When dis.pleased, child lets. ~eol'le know it right away, 10. Child likes to plan things wa y ahead of time. 59 1 N o l No l No l No 1 No l No 1 N o 1 N o 1 N o 2 3 2 3 2 3 2 3 2 3 2 3 2 3 2 3 2 3 2 3 4 4 4 4 4 4 4 4 4 4 5 Yes 5 Yes 5 Yes 5 Yes 5 Yes 5 Yes 5 Yes 5 Yes 5 Yes 5 Yes

PAGE 68

60 li... Chi.ld ha.s trouble resisting temptation. 12. Child likes to wear himself out with ex ertion. 13. There are many things that anno y him 14. Child has fewer fears than most children his age. 15. Child often cries. 16. Child is very sociable. 17. Child usually seems to be in a hurry. 18. Most of my child's activities are fast paced. 19. For relaxation, child likes to slow down and take things easy. 20. Child often appears to be bursting with energy. 21. Child often feels sluggish. 22. When child does things, he does them vigorousl y 23. Child likes to keep bus y all the time. 1 N o 1 N o 1 N o 1 N o 1 N o 1 N o 1 No l No l No 1 No l No l N o l N o 2 3 2 3 2 3 2 3 2 3 2 3 2 3 2 3 2 3 2 3 2 3 2 3 2 3 4 4 4 4 4 4 4 4 4 4 4 4 4 5 Yes 5 Yes 5 Yes 5 Yes 5 Yes 5 Yes 5 Yes 5 Yes 5 Yes 5 Yes 5 Yes 5 Yes 5 Yes

PAGE 69

61 24. Child's movements are forcefu l, and emphatic. l 2 3 4 5 N o Y es 25. Child will tr y an y thing once. l 2 3 4 5 N o Yes 26. Child makes friends ver y quickl y l 2 3 4 5 N o Y es 27. Child usuall y prefers to do things alone l 2 3 4 5 N o Y es 28. Unfinished tasks really bother child. l 2 3 4 5 N o Y es 29. Child often says the first thing that l 2 3 4 5 comes into his head. N o Yes 30. Child often has trouble making up his mind. 1 2 3 4 5 No Yes 31. Child likes to make detailed plans before l 2 3 4 5 he does something. No Yes 32. Child is almost always calm--nothing ever 1 2 3 4 5 bothers him. No Yes 33. Child frequentl y gets upset. 1 2 3 4 5 No Yes 34. When child gets scared, he panics. l 2 3 4 5 No Yes 35. It takes a lot to get him mad. l 2 3 4 5 N o Yes 36. Child is happiest in familiar surroundings. l 2 3 4 5 N o Yes 37. Child has trouble controlling his impulses. 1 2 3 4 5 N o Y es

PAGE 70

62 38, Child likes to be off and running as soon 1 2 3 4 5 as he wa k es up in the mornin g N o Y es 39 Child can stand frustration better than 1 2 3 4 5 most N o Y es 40. Child often feels insecure. l 2 3 4 5 N o Y es 41. Usuall y child can't stand w aiting. 1 2 3 4 5 N o Yes 4 2 Child is easil y fri g htened. 1 2 3 4 5 N o Y es 4 3. Child is known as hot-blooded and quic k 1 2 3 4 5 tempered. N o Y es 4 4. Child generall y likes to see things through 1 2 3 4 5 to the end. N o Y es 45 Child gets e x cited easil y 1 2 3 4 5 No Y es 4 6. Once ch i ld gets going on something he 1 2 3 4 5 hates to stop. No Yes 47 Child is somewhat emotional. l 2 3 4 5 No Yes 48. Child tends to hop from interest to l 2 3 4 5 interest quic k l y N o Yes 4 9 C h ild generall y seeks new and ex cit i ng l 2 3 4 5 e x r e riences and sens.ation s. N o Y es 50 Child tends to giv e up e asil y l 2 3 4 5 N o Y es

PAGE 71

APPE N DI X E RESIST AN CE TO TEMPT A TIO N OBJECTS 0 63

PAGE 72

APPENDIX F COMPLETE CORRELATION MATRIX 2 -.18 3 .26 .29 4 .48 .10 .46 5 -.08 .16 -.06 -.05 6 -.08 .18 -.03 -.09 .12 7 -.27 .02 .04 -.13 -.16 -.10 8 .12 .09 .07 .28 .05 .09 -.04 9 .11 .13 .14 .14 -.29 .05 .10 -.18 10 .05 .01 -.11 -.03 .12 .41 -.08 .13 .01 11 -.04 .04 .00 .08 .03 .01 .02 .22 -.30 -.28 12 .07 .06 .06 .08 .04 .05 .01 .26 -.21 -.11 .49 13 0 1 .05 .03 .09 .04 .03 .02 .28 -.30 -.24 .89 .84 14 -.12 .04 -.06 -.13 .11 .26 .07 .06 -.04 .32 .22 .23 .27 1 2 3 4 5 6 7 8 9 10 11 12 13 Variables (N = 78) 1. Emotionality 2. Sociability 3. Activity Level 4. Impulsivity 5. Resistance to Temptation 6. Delay of Gratification 7. Draw Normal 8. Draw Slow Difference 9. Activity Schedule 10. Age 11. Block Design 12. Vocabulary 13. I.Q. 64 14. Ego Level

PAGE 73

APPENDIX G DESCRIPTIVE STATISTICS Standard Variable Mean Deviation Minimum Maximum Resistance to Temptation .76 .43 0 1 Delay of Gratification .68 .47 0 1 Draw Normal 9.14 5.46 1. 8 26 Draw Slow Difference 21.76 17.86 .so 90.9 Age 8.45 1.39 6.09 11.25 Block Design 10.86 3.07 3 18 Vocabulary 11. 71 2.59 4 18 IQ 22.56 4.88 10 35 Impulsivity 59.99 9.53 42 92 Activity Level 33.91 9.55 12 78 Activity Schedule 45.26 14.81 14 75 Ego Level 87.83 8.19 69 107 65

PAGE 74

BIOGRAPHIC~ SKETCH Mark H. Waugh was born in Augusta 1 Georgia, and grew up in Le x ington, Kentucky, and Greenville, North Carolina. He majored in ps y chology at the University of North Carolina at Chapel Hill. He attended graduate school at the University of Florida and is completing a clinical intern ship at the Universit y of Texas Health Science Center at Dallas. He is married to Marcia Snowden Waugh and the y have one son, Robert Mark Waugh. He will take an assistant professor of ps y cholog y position at Roosevelt University in Chicago, Illinois, after graduation. 66

PAGE 75

I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. y --.<---'---~L-4~ -W~Ck.a:--d HJ~is, Jr., Chairmai Professor of Clinical Psychology I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. _,.L ~i Q ~--r-~4& -=: ~in R.Golan Professor of Clinical Psychology I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. eriC Blashfield Associate Professor of Clinical Psychology I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. Mary Mc Caulley Assistant Professor of Clinical P I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. (' _/h1.fA .u t)~ ./ aundra Damico Professor of Foundations of Education

PAGE 76

This dissertation was submitted to the Graduate Facult y of the College of Health Related Professions and to the Graduate Council, and was accepted as partial fulfillment of the requirements for the degree of Doctor of Philosophy. Dean, College of Health Related Professions Dean for Graduate Studies and Research August, 1982

PAGE 77

UNIVERSITY O F FLORIDA II I II IIIIII Ill Ill lllll lllll II IIIIII IIII II IIII IIIIII IIII II IIIII I I 3 1262 08553 6018

PAGE 78

UNIVERSITY O F FLORIDA II I II IIIIII Ill Ill lllll lllll II IIIIII IIII II IIII IIIIII IIII II IIIII I I 3 1262 08553 6018