32mm”
'11:} “4?; “a

  
  
  
   

 

{"""I '3‘}

3 3
E."$3'3;"~3!'}"'. -3\'L"’ 3- y'.9‘"31 3 '3'
_- 3333333331,, 33'“ 3333,3333: 33333:?3

33"; 333333333 33" 3 3333 333

»

-7
9 ~m

   
 

 

‘T'W “u— __,
fl-:.'.l_‘ . _.
ff“: .3
:.;..:

~-— I. r '-
-v's' 4-239» ,
‘ .-. I "‘
4.- __

4—.34 --—v.-
,-

     

.o‘
4

:«Jr' .

5

.v'
' 4 .
u . ' '—'. I
.n-s. vi '_
u

. '.
~371'f‘

   

..
0..
poison
¢~..

    
   
   
 

‘Ibv
r
n . .
.._ _'
1. . '3
. —4- -3 .A.
— c- 3 _33
—'
n
‘
1.4.
.1

    
 
  
  

 
 
  

  
  
   

-- n
-.
v
30‘
>3

of“: .
-_ q.-.

 
 

w

-‘.T'.‘ '
o
r -

    
   
 

.

-o

" -.

v v—V‘.‘
-- y.

:T'v't-z-L. _ 3 .3 .‘.-_ 3 3 _

«‘1‘: .43, « . 3. 3 3,. 3“
4-41313 .. _m531333
i” A u" . 4‘ - . . ‘. .
. ‘ _ . “ 4 .:-.;‘:. érh" '5'. Y} .
h ' ' ‘ ‘ 3. 0 - '

      
      
   
  
 

. ..
3'"
2' ,3 3... 333 3
3. .fu 3 3 . ‘ 3
'b'. I ‘3: .3 '
23:: . 333 3 3
0 J}.$3:'
4 V
‘ '.' £3.5-
' ' ‘ .‘J‘
T ‘ ‘ .7;
h . “I.
.L 3

    
 

    

 

3 .M 33 3333333 33 333333333 3 7333. 3:33 3
33 :33'33'3 3.333333 . 3 .33.3.333
lg; “3' W" " '3“. I" 3;; 1":3::' I33 3' [3.3. :' ,: 33,!z':"3|3“3 l!!i:“| . ' 370,333

.3 33333333333333 333333 33333333: 3
""3333" " 33' 3 ' 33'3' 3"‘33‘

.‘llll:"n\ttl'.' 312" . :333'anu... "15913“ 433631

33 33 -3 333,33

:33'333} $3113"!
9333.33

v _y_
A .. 'b‘
‘i':: " I» _ 3
‘ A - A
-‘ A. V ' " - h. . L
3.3 . .. . , 3 :23? -
_ 3. - V _ _3 3 33 u ‘3
r.—-‘ ‘ ‘ . .. ' --.... ‘ .3 fK- _
A ‘ _ ‘ ‘A ‘ ~ 1.. ~ 3
A .3—- ' ‘ —3 "‘."' —. ‘3‘ ; . -..
- _ 3 3 _ 3 3

 

   
    

  

LIBR/ " Y
Michigan to

UlliVC 1'51 .. Y

333‘} ‘1 1‘» INN“ “m!

  

This is to certify that the

thesis entitled

STYLES OF PARENTj-CHILD INTERACTION AS A MEDIATING
FACTOR IN CHILDREN'S LEARNING FROM
ANTISOCIAL TELEVISION PORTRAYALS

presented by
Felipe Korzenny
has been accepted towards fulfillment

of the requirements for

Ph 0 D. degree in Communication

th

Ma professor U

 

DateSeptember 16, 1977

 

0-7639

 

 

 

STYLES OF PARENT-CHILD INTERACTION AS A MEDIATING
FACTOR IN CHILDREN'S LEARNING FROM
ANTISOCIAL TELEVISION PORTRAYALS

BY

Felipe Korzenny

A DISSERTATION

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

DOCTOR OF PHILOSOPHY
Department of Communication

1977

Accepted by the faculty of the Department of
Communication, College of Communication Arts, Michigan
State University, in partial fulfillment of the require-

ments for the Doctor of PhiloSOphy degree.

M .

irec of Dis er a ion

 

Guidance Committee: Chairman

 

 

 

 

ABSTRACT

STYLES OF PARENT-CHILD INTERACTION AS A MEDIATING FACTOR IN
CHILDREN'S LEARNING FROM ANTISOCIAL TELEVISION PORTRAYALS

BY

Felipe Korzenny

The present study dealt with the role of different modes
of parent—child interaction as mediators of children's model-
ing of antisocial television portrayals, i.e., physical and
verbal aggression, theft, and deceit.

More specifically, two main orientations of parent-child
interactions with regard to social situations have been iden-
tified, namely, an internal and an external orientation. The
main difference between the two types of orientations is that
internally oriented parents are said to provide children with
the necessary cognitive structure for evaluating their social
behaviors. Externally oriented parents do not provide cogni-
tive resources upon which the child can rely when social
decisions have to be made.

Six hypotheses were tested in this study. Three hypoth-
eses dealt with the main effects expected from the internal
and external orientations of parents and children and exposure
to antisocial television portrayals. The fourth hypothesis
predicted the relative magnitude of antisocial predispositions
in children for different subsamples defined by the internal
and external orientations and antisocial television exposure.
The last two hypotheses were concerned with the contingent

relationships between exposure to antisocial television exposure

Felipe Korzenny

and the child's antisocial predispositions at different inter-

sections of the internal and external orientations of parents

and children.

Two hundred and twenty-seven mothers and their children

in the fifth, seventh and ninth grades were tested.

The results of this study were as follows:
Exposure to antisocial television portrayals had a signif-
icant direct independent effect on the children's anti-
social predispositions, as reported by both mothers and
children.
The external orientation of parents had a significant
direct independent effect on physical and verbal aggres-
sion and deceit, based on the mothers' report of the
child's behavior. However, with the children's data the
external orientation of children had a negligible effect
on the children's antisocial predispositions.
The children's internal orientation was found to be the
most powerful negative predictor of antisocial predispo-
sitions. The mothers' data, on the other hand, did not
show an appreciable relationship between the parental in-
ternal orientation and the children's negative predisposi-
tions.
With both sets of data, from parents and children, it was
consistently found that highly internally oriented chil-
dren, who watch small amounts of antisocial television,
and whose parents and themselves are low in their external

orientation displayed the least amount of antisocial

Felipe Korzenny
predispositions. Also, the children highest in antisocial
predispositions were those lowest in the internal orienta-
tion and high in antisocial television exposure and the
external orientation.

5. Those highly internally oriented parents who were low in
their external orientation had children for whom the cor—
relation between exposure to antisocial behaviors on tele-
vision and their display of antisocial predispositions was
the lowest when compared with all the children and with
other subsamples. The children who showed the highest cor-
relation between exposure and predispositions were those
whose parents were highly externally oriented and low in
their internal orientation. With the children's orienta-
tions the pattern of findings was the opposite to the hy-
pothesized expectations. Few of these contingent correla-
tion comparisons were found to be statistically significant.

One of the main conclusions of this study was that tele-
vision remains a contributing source of socialization independ-
ently of parental practices and the children's internalization
of moral values. The results were discussed taking in consider-
ation the research literature dealing with the internalization

of moral standards in children.

To Sandy, for the third time.

iii

ACKNOWLEDGMENTS

I know I won't be able to thank all the peOple that have
contributed to my graduate education and to this dissertation.
However, I will try to do justice to the most notable individ-
uals and organizations.

I would like to thank Dr. Bradley S. Greenberg for his
continued guidance and help. He has been a model for me and
has taught me many direct and indirect lessons. From a schol-
arly point of view, I was lucky to have him as my advisor and
as the chairman of my committee. Thank you, Dr. Greenberg.

Dr. R. Vincent Farace has been very helpful in guiding
me in the direction of the study of communication and change.
His participation in my committee was distinct, from my point
of view, in teaching me about the relativity of our discipline.

Dr. Stephen L. Yelon introduced me to the area of moral
development and took me by the hand in understanding critical
issues regarding the role of parental practices on children's
behavior.

Dr. Charles K. Atkin was always helpful in encouraging
me to be flexible regarding analytic methods and in the con-
sideration of alternatives.

All four, Drs. Greenberg, Farace, Yelon and Atkin, be-
sides being members of my doctoral committee are good friends

that I hope to preserve.
iv

Several other members of the faculty in the Department
of Communication gave me assistance in understanding issues
related to the present dissertation. Dr. Edward L. Fink was
one of the most notorious.

My family played a crucial role throughout my career:
My wife, my parents, my grandparents, and my parents-in-law
gave me always unusual support that I will never forget.

The data for this dissertation were collected as part
of a larger project (CASTLE) sponsored by the Office of Child
Development of the United States Department of Health, Educa-
tion and Welfare. A grant from the National Association of
Broadcasters facilitated the dissertation work. The National
Council of Science and Technology of Mexico (CONACYT)
partially sponsored me during my graduate education.

Ruth Langenbacher efficiently typed the manuscript.

Thank you all

Chapter

I

II

TABLE OF CONTENTS

STYLES OF PARENT-CHILD.INTERACTION AS

A MEDIATING FACTOR IN CHILDREN'S

LEARNING FROM ANTISOCIAL TELEVISION

PORTRAYALS . . . . .

INTRODUCTION . . . . .

Parental Controls over Viewing
Parental Interpretation of Tele-
vision Content . .

Parent-Child Interaction Modes with

Regard to Social Situations .
Family Interaction Perspective
Hypotheses . . . . .
Summary . . . . . .

METHODS . . . . . . . .

PETEST O O O O 0

Preliminary Methodology . .
Instrument Design . . .
Questionnaire Administration .

FINAL STUDY . . . . .
Sample Description . . .
Measures and Descriptive

StatiStics O O O O 0

Internal and External Orientations

Mothers . . . . . .

Children . .

Television Exposure to Antisocial
Behaviors . .

Children's Measures of Antisocial
Predispositions . . .

Validity and Reliability . .
Parent and Child Internal and

External Orientations . .
Exposure to Antisocial Tele-
vision Portrayals . . .

vi

Page

A1 to +4 H

18
24
29

31
31
31
36
44
44
45
45
46
47
54

56
62

64

73

Chapter

Chapter II (cont'd.)

III RESULTS

Child's Antisocial Predispositions
Data Replicating the Emergence of
the Internal and External Orien-

tations .

Analytical Methods .

Summary .

Analysis of Variance and Regression:
and HC

a. Physical Aggression
b. Verbal Aggression

Ha, Hb,

c. Theft .
d. Deceit .
T-tests: H

a. Physical Aggression
b. Verbal Aggression

C. Deceit .

Contingent Correlations

a. Physical Aggression
b. Verbal Aggression

c. Deceit .
Summary .

IV SUMMARY AND DISCUSSION

REFERENCES

Summary .
Discussion .

Dimensions of Parent-Child

action Modes

Independent and Mediational

Effects .

and
e
Inter-

Suggestions for Future Research

Limitations
Conclusion .

vii

Page

74

78
86
87

89

90

91
95
96
103
107
109
110
111
112

113
116
118
121

123

123
130

130
133
140

142
144

146

LI ST OF TABLES

Table Page

1. Means, Standard Deviations, and Number of
Cases for Each Index of Parental or Child
Responses to Social Situations . . . . 39

2. Principal Factor Matrix with Varimax
Rotation, Ordered According to Hypothesized
Indexes Loadings . . . . . . . 41

3. Means, Standard Deviations, and Number of
Cases for Each Index of Parental Responses
to Social Situations (Index range 0-8) . . 51

4. Means, Standard Deviations, and Number of
Cases for Each Index of Child Responses
to Social Situations (Index range 0-8) . . 52

5. Means, Standard Deviations, Number of
Cases and Index Range of Mother and Child
Indexes of Reports of the Child's Anti-
social Behaviors in Hypothetical Situations . 63

6. Pearson Correlation Coefficients Between
Mother and Child Items Concerning the
Parent's Reactions to a Positive and a
Negative Situatian . . . . . . . 66

 

7. Pearson Correlation Coefficients Between
Mother and Child Items Concerning the Child's
Reactions to a Positive and a Negative
Situation . . . . . . . . . 68

 

8. Internal Consistency a Coefficients for the
Indexes of Parental Responses Reflecting
the InternaIIandvfixternaIIOrientations for
Positive and Negative Social Situations . . 69

 

9. Internal Consistency a Coefficients for the
Indexes of Children's Responses Reflecting
the Internal and External Orientations for
Positive and Negative Social Situations . . 71

 

viii

Table Page

10. a. Test-retest Reliability for Children's
Reactions . . . . . . . . . 72

 

b. Test-retest Reliability for Parent's
Reactions . . . . . . . . . 72

ll. Correlations Between the Indexes of Exposure
to Antisocial Behaviors in the First and
Second Years . . . . . . . . . 74

12. Correlations Between Mother and Child
Indexes of Reports of the Child's Anti-
social Behaviors in Hypothetical Situations . 76

13. Test-retest Reliability for the Indexes
of the Children's Antisocial Predispositions
in Hypothetical Situations. Correlations
Between One Year Lag Indexes . . . . . 77

14. Internal Consistency a Coefficients for the
Indexes of the Mother's Report and the
Child's Own Report of Antisocial Behaviors
in Hypothetical Situations. Year 2 . . . 77

15. Principal Factor Matrix with Varimax
Rotation, Ordered According to Hypothe--
sized Indexes Loadings. Parental Reactions . 79

16. Principal Factor Matrix with varimax
Rotation, Ordered According to Hypothe-
sized Indexes Loadings. Children's
Reactions . . . . . . . . . . 81

 

17. Means, Medians, Standard Deviations, Number
of Cases, Index Range and Internal Con-
sistency a Coefficients for the Internal
and External Indexes of Parent-Child Modes
of Interaction . . . . . . . . 83

18. Parental Orientations, Child's Exposure on the
Mothers1 Report of the Child's Physically

A ressive Predispositions: Regression and
AnaIysIs of Variance Results . . . . . 92

 

 

19. Children's Orientations, Exposure on the Self-
report of—Physically Aggressive Predisposi-
tions: Regression and Analys1s of Variance . 94

 

ix

Table Page

20. Parental Orientations, Child's Exposure on the
Mothers' Report of the Child's Verbally Ag-
gressive Predispositions: Regression and
Analysis of Variance Results . . . . . 96

 

 

 

21. Children's Orientations, Exposure on the Self-
report of—Verbally Aggressive Predispositions:
Regression and Analysis of’Variance . . . 98

 

 

22. Parental Orientations, Child's Exposure on the
Motherrs Report of the Child's Predispositions
Towards Theft: Regression and Analysis of
variance Results . . . . . . . . 99

 

23. Children's Orientations, Exposure on the Self-
report of Predispositions Towards Theft:
Regression and Analysis of Variance Results . 101

 

24. Parental Orientations, Child's Exposure on the
Mother's Report of the Child's Predisposi-
tions Towards Deceit: Regression and
Analysis of Variance Results . . . . . 104

 

25. Children's Orientations, Exposure 0n the Self-
report of’Predispositions Towards Deceit:
Regression and Analysis of variance Results . 105

 

26. Contingent Correlations Between Exposure
and the Child's Ph sical A ression as
Reported by Mother and Child at Different
Intersections of the Internal and External
Orientations of Parents and Their Children . 114

27. Contingent Correlations Between Exposure and
the Child's verbal A ression as Reported by
Mother and Child, at Different Intersections
of the Internal and External Orientations of
Parents and Their Children. . . . . . 117

 

28. Contingent Correlations Between Exposure and
the Child's Dispositions Towards Deceit as
Reported by Mother and Child, at Different
Intersections of the Internal and External
Orientations of Parents and Their Children. . 120

CHAPTER I
STYLES OF PARENT-CHILD INTERACTION AS A MEDIATING
FACTOR IN CHILDREN'S LEARNING FROM
ANTISOCIAL TELEVISION PORTRAYALS

INTRODUCTION

During the past decade, a great deal of concern has
been expressed by public and private sectors of society with
regard to the effects that negative television content may
have on children. A large amount of research has been gen-
erated by this concern and the results point to a positive
association between the child's exposure to undesirable tele-
vision content and antisocial attitudes and behavior on the
part of the child (The Surgeon General's Advisory Committee
on Television and Social Behavior, 1972). However, few re-
searchers have focused their attention on the role of other
socialization agencies in mediating television's effects
(Chaffee, 1976).

Parents constitute the first source of socialization
to which children are exposed. This primacy factor makes
parental practices a central point of inquiry when one raises
questions about the social behavior of children. The main
point of the present investigation is to see if contrasting
parental styles of discipline and interaction substanially
mediate the relationship between watching antisocial portray-

als on television and the child's own negative social.behaviors.
l

2
Some recent research has been concerned with differ-
ent activities parents carry out with regard to television
exposure by their children. This research can be categor-
ized into two tendencies: parental control of television

exposure, and parental interpretation of television content.

 

The innovative contribution of this dissertation is the ex-
ploration of the role of parent-child interaction modes on
the relationship between exposure to antisocial portrayals
on television and the children's antisocial predispositions.

In this first chapter, the research on parental con-
trol and interpretation of television will be summarized as
the most immediate antecedent for this line of inquiry.
Next, the role of parent-child interactions with regard to
social situations in general,* as mediators of television
modeling of antisocial behaviors will be explored. Here, also,
the relevant literature on family interaction will be reviewed.
Lastly, the derived empirical hypotheses that constitute the

nucleus of this dissertation will be presented.

Parental Controls over Viewing

 

Hanneman g£_al. (1976) indicate that "research on con-
trol of children's television behavior by parents indicates
that parents generally do not control their children's TV
behavior: (p. 13). Atkin (1972) reported data that suggest

"that the parents are not always aware of the television fare

 

*
Without reference to the TV viewing situation.

3
that their children are exposing themselves to" (p. 1).
Atkin also mentions that parental censorship, when it occurs,
tends to be with regard to sex content rather than violence.

Sears 33 21. (1957) found that some mothers use tele-
vision as a soporific to keep their children "quiet and out
of mischief" (p. 289). They found that 58% of the mothers of
kindergartners interviewed reported from moderate to severe
restrictions on watching: "Children may look only during
specific times, or at specific programs" (p. 291).

Barcus (1969) reported that formal rules were stated
about what the child should not watch, with or without power
exertion, once the infant had started exposure to undesirable
shows.‘ Formal rules use was the most widespread type of
control. Greenberg, Ericson and Vlahos (1972) interviewed
100 fourth and fifth graders and their mothers.

Four items concerned parental regulation of
the child's viewing habits. One item asked
if there were any rules about how late the
child could watch television. Another asked
if there were some shows the child was not
allowed to watch. Two items dealt with the
loss of viewing privileges as punishment and
extended viewing privileges as reward for
something special the child did (p. 398).
They found that the agreement between the mother and her
child was low (r=.19) with respect to the perceived nature
of television rules in the household. However they found
that "all of the parents and 90 percent of the youngsters

said there were rules about how late television could be

watched; two-thirds of each group said there were some

4
forbidden shows; one-third said there was punishment in the
form of not being allowed to watch television; one-third of
the children and one-fifth of the parents said more tele-
vision watching was used as a reward for good behavior"
(p. 403).

Lyle and Hoffman (1972) in a study with first, sixth
and tenth graders found that "although the majority of the
first grade mothers interviewed stated that they did try to
guide program selection for their young children, few indi-
cated that they attempted to restrict amount of viewing.
About one-third of the students themselves said that their
parents tried to control their viewing, either 'now' or
'when they were younger'" (p. 134).

In the studies reviewed here one finds that parental
control of their children's television exposure is not very
widespread. This may be due to the pervasiveness of the

medium. Television has taken roots in the family setting.

Parental Interpretation of Television Content

 

In a large summarizing effort with regard to instruc-
tional television, Chu and Schramm (1967) documented that
discussion of television content promotes learning. Ball
and Bogatz (1972), in their evaluation of the effects of
Sesame Street, found that children who rehearsed the concepts
they watched on Sesame Street with their mothers, learned

more than children who watched by themselves.

5

Atkin and Gantz (1974) discovered that parental expla-
nation and amplification of news program content tends to in-
crease the child's acquisition of current events knowledge.
McLeod, Atkin and Chaffee (1972) found that parents can miti-
gate some of the harmful influences of television violence by
interpreting the violent acts as they are portrayed. Tolley
(1973) found some evidence that parental communication, re-
garding TV news reports, leads to a greater adoption of paren-
tal attitudes about issues such as war.

Hicks (1968) in an experiment with seven year-old
children found that positive and negative sanctions by an
adult co-viewer produced corresponding increases and decreases
in imdtative aggressive behavior when the adult co-viewer re-
mained with the children who had viewed a film of aggressive
behavior. These effects disappeared when the co-viewer did
not remain with the children in the observation period.
Feshbach (1972) found that adult labeling of a violent act
as fantasy or reality tended to decrease or increase, respec-
tively, the level of aggression displayed by his young
subjects. These adults were not the children's parents.

Walling (1976) conducted an experiment with first
graders. He assigned them to three groups at random: Inter-
action with parents, no interaction, and a control group.
After one week, the children were interviewed in order to
assess their learning of problem resolution techniques from

three television shows. He found that "children who

6

interacted with their parents were significantly better able
to complete problems which paralled those" (pp. 21-22) they
had viewed on television than children who did not interact
with parents during viewing and children who did not watch
television at all.

More recently, Atkin and Greenberg (1977) have re-

ported that "the most important factor is joint viewing:

 

the relationship between exposure and aggression is half as
strong for children who frequently view police-detective
shows with their parents than for children who seldom view
with parents" (p. 4).

In general, it seems to be that those parents who
take the time and exert the effort to explain television
content, "and teaching children discernment and discrimination
in regard to television” (Leifer 22,31. 1974) may effectively
innoculate them against possible negative influences. This
may generalize to those occasions in which the child watches
alone or with peers.

When one considers parental interpretation, it seems
that parent-child interaction with regard to television
offerings has the potential for modifying the medium's
impact. In a related vein, it is to be expected that parent-
child interactions in everyday situations, without regard to
television, but with regard to the social behaviors of the
child, may constitute important determinants of what the
child acquires from television portrayals, given the cogni-

tive structure provided by parental tuition.

7

Parent-Child Interaction Modes with Regard to Social
Situations

 

Now the research on parent—child interaction modes
will be considered. In the past this has not been related
to what the child acquires from television, with the excep-
tion of the family interaction paradigm, that will also be
reviewed below.

Here the evidence available with regard to parent-
child interactions will be reviewed, and testable hypotheses
that predict the influence of those interaction modes on chil-
dren's social learning from television will be derived.
Generally, the writer interprets parent-child interaction
during TV time as a special case of the more general phenom-
enon consisting of enduring and ongoing interaction and re-
sponse modes across the wide variety of social situations in
which parents have the opportunity for shaping their children's
response patterns.

The ways in which parents handle the development of
moral behavior in their children should be especially rele-
vant to the examination of the child's acceptance or rejec-
tion of television portrayals of antisocial behavior.
Children may acquire what they watch without necessarily
translating their acquisitions into practice. ”Social learn-
ing theory distinguishes between acquisition and performance
because pe0ple do not enact everything they learn" (Bandura,

1977, p. 28).

8

Children whose parents have enabled them to differ-

entiate between "good" and "evil" should be able themselves

to approve or disapprove of television behaviors to which

they are exposed. Bandura explains:

Parents cannot always be present to guide
their children's behavior. Successful
socialization requires gradual substitution
of symbolic and internal controls for ex-
ternal sanctions and demands. After moral
standards of conduct are established by
tuition and modeling, self—evaluative con-
sequences serve as deterrents to trans-
gressive acts (Ibid, p. 43).

Moral training has been a major concern with child

psychologists. Internalization of moral values, identified

as conscience or superego, should be the result of certain

parental practices.

Internalization has been defined in a variety of ways,

for example:

By 'internalization' I mean that they have a
set of cognitions that they employ under ap-
propriate conditions and thus guide their own
behavior, and that they experience self-rein-
forcement and self-punishment as a function
of their behavior (Staub, 1975, p. 118).

Generally to those qualities of children's
behavior represented by the control of impulse
and the reactions to loss of such control . . .
resistance to temptation and feelings of fear,
shame, or guilt concerning deviation (or antici-
pated deviation) from right and proper conduct

. . . it is a part of the mind that controls
other parts, directing behavior in ways that

are mainly inhibiting and self-punishing (Sears
EE.El-r 1965, p. 199).

Is the predisposition to subordinate one's hedo-
nistic needs in favor of the social and moral
requirements of a situation, without regard for
external sanctions . . . it is reasonable to

A11

9

assume that the most important socialization
experiences for the development of such moti-
vation are those in which the child faces the
conflict between hedonistic needs and moral
demands, as communicated by the parent early
in life (Hoffman, 1975, p. 234).

The tendency of society to satisfy itself as
cheaply as possible results in appeals to 'good
conscience,‘ through which the individual pays
to himself the wages for his righteousness,
which otherwise would probably have to be as-
sured to him in some way through law or custom
(according to Simmel, Hoffman, 1970, p. 262).

The concept of internalization is often used to
refer to the child's adOption of social norms
or roles as its own, and to the resulting con-
trol of its behavior by the most complex media-
tional functions of cognitive and verbal
processes (Aronfreed, 1969, pp. 264-265).

these definitions of internalization have at least

three elements in common. First, internalization requires

the adopti

on of social norms; second, these social norms give

 

internal direction to the behavior of the child; and thirdly,

 

such internal direction is achieved by self-generated conse-

 

quences* for the social behavior of the child.

The most highlighted point of the concept of inter-

nalization is precisely the idea of inner-directedness. The

child who

has been brought up to be independent of external

stimulation for the control of his/her behavior should also

be more re
counter to
one aspect
resistance

counter to

sistant to the modeling of behaviors that run

the child's values. Hoffman (1970) mentions that
of internalization is precisely "the amount of
offered to pressures and temptations to behave

the standard" (pp. 286-287).

 

*
The

child's internal punitive or rewarding reactions.

10

The child who is inner directed should then be able
to behave according to what he/she considers apprOpriate,
and then, when exposed to direct or vicarious models that do
not fit his/her norms, discount and criticize their be-
haviors.

Conscience, also called internalization or supergo,
has been Operationalized in a variety of ways. Some commonly
accepted indicators of conscience are: (l) "maintenance of
control, in the face of temptation, when there is no one
present to insist, and when there is little danger of being
caught" (Sears gplal., 1957, p. 366); (2) occurrence of
guilt feelings when temptation is not overcome (Ibid.); (3)
admitting transgression when asked (Grinder, 1962); (4)
emotional upset and confession (Sears pp 31., 1965, p. 240);
(5) moral judgments about others which are based on internal
rather than external considerations, e.g., confession, accept
responsibility, consideration for others (Hoffman and
Saltzstein, 1967, p. 45); (6) "moral stage advance" (Lickona,
1976, p. 25); (7) self-condemnation (Hoffman, 1970, p. 264);
etc.

A low level of moral internalization has been identi-
fied in the literature as "behaving prOperly through fear of
external sanctions and evaluating moral action on the basis
of the likelihood of external punishment" (Hoffman, 1975, p.
232), or as external resolution (or externally oriented

initiation and performance) indicated in story completion

11
tasks (Aronfreed, 1961, p. 227).

How is moral internalization achieved? The most
likely source of influence to greatly affect moral develop-
ment are parental practices and interactions with regard to
social behavior. The importance of parental practices is
emphasized because parents are the initial source of influ-
ence on their children (Hoffman, 1975, p. 232).

Numerous studies have shown that there is a relation-
ship between parental disciplinary practices and the intern-
alized control of their children's social behavior. Among
them one can identify the following: Allinsmith (1960);
Aronfreed (1961); Bandura and Walters (1959); Burton, Maccoby
and Allinsmith (1961); Hoffman and Saltzstein (1967); and
Sears gp‘gl. (1957).

Of these studies, perhaps the most salient are the ones
by Aronfreed (1961), and by Hoffman and Saltzstein
(1967). Aronfreed utilized a projective story completion
technique to elicit response tendencies in 122 white children
from the sixth grade classes of two public schools. In the
stories, the central child committed an act of aggression
directed against parents, a friendly neighbor, or a close
companion. The mothers of the children were interviewed at
home to find out "about the different things which mothers
do to help their children grow up" (p. 227). The mother was
asked 12 questions, each describing a particular form of ag-

gressive behavior and then asking the mother to report how

12
she usually responded when her child behaved in that way.
In general, the findings indicated that middle-class children
were more likely to give internalized moral responses, while
lower socioeconomic class children tended to be more con-
cerned with the external aspects of the transgressions. In
a congruent fashion, middle-class mothers were more likely to
report the use of love oriented practices, such as reasoning
and explanation. Lower-class mothers showed a preference
for externally oriented, or power assertive modes of disci-
pline. Also, there was a clear tendency for love oriented.
inductive, mothers to have children who focus their attention
on internal aspects of the wrongdoing, while power assertive,
or externally oriented mothers were found to be more likely
to have children who focus their attention on the external
aspects of transgressions.

Hoffman and Saltzstein assessed the degree of internal-
ization of morality of a sample of 444, seventh grade children
by means of projective story completion and hypothetical trans-
gression items. Parents, teachers and peers were also asked
to rate the children on their level of moral internalization.
Children reported their parents' disciplinary practices, and
the parents were interviewed to obtain a report of their own
disciplinary techniques.

Each respondent (the child or parent) was
asked to imagine four concrete situations.
. . . Following each situation was a list
of from 10 to 14 practices. The respondent

was asked to look over the list, then rate
the absolute frequency of each (p. 42).

13
The authors found that parental power assertion tech-
niques were consistently associated with weak moral intern-

alization in the children.

. . . in sum it is a pattern of infrequent
use of power assertion and frequent use of
induction by middle-claSs mothers which

generally appears to facilitate the facets
of morality included in this study (p. 49).

No parental data was obtained from lower-class parents,
and the results for this subsample were mixed.

Two main global categories of disciplinary types of
interactions among parents and their children have been
identified by Aronfreed (1976) and Hoffman (1975). One of
these disciplinary types is called induction, and the other
sensitization.

Induction types of discipline have in common
. . . that they tend to make the child's con-
trol of its behavior independent of external
contingencies. In contrast, disciplinary
habits of direct physical and verbal attack
may be characterized as sensitization because
they tend merely to sensitize Ehe child to

the anticipation of punishment (Aronfreed,
1969, PP. 309-310).

 

 

In general, induction techniques include the use of
reasoning and explanation, e.g., to point out the require-
ments of a situation, or the consequences of the behavior
for the child or others. Also, appeals to the child's pride
and achievement are considered to be inductive techniques
(Hoffman, 1970, p. 286).

Sensitization, sometimes called power assertion,

"includes physical punishment, deprivation of material objects

14

or privileges, the direct applications of force, or the threat
of any of these" (Ibid., p. 285). Although less documented in
the literature, the use of external rewards is also expected
to belong to the class of parental practices that have been
labeled sensitizing techniques. External rewards "reduce
intrinsic motivation by creating the impression that one's be-
havior is externally prompted and by weakening feelings of
competence and self-determination" (Bandura, 1977, p. 107).

The main difference between these two modes of parental
discipline is that induction is communication oriented and
sensitization is based on the exercise of actual and implied
power. Induction can be said to provide the cognitive struc-
ture with which the child will be able to categorize his
social experiences: "Cognitive structure facilitates intern-
alized suppression by serving as an intrinsic mediator of
anxiety which can intercede before the commission of a
punished act" (Aronfreed, 1969, p. 276).

Sensitization, by requiring continuous survellance,
and by being limited to the disciplinary event, reduces the
length of time during which the child experiences anxiety for
transgressions (Ibid., p. 313).

According to Hoffman (1970), dissonance theory would
also account for the effectiveness of induction in forming an
independent moral orientation. Induction techniques exert
little external power over the child and if she/he refrains

from a negative behavior, she/he will be more likely to

15
reduce dissonance by thinking that she/he refrained volun—
tarily. However, sensitizing techniques are too obvious and
the child may just think that she/he refrained due to extern-
al demands (p. 284). Furthermore, because inductive tech-
niques point out the consequences of the child's behavior for
others, they may develop the child's capacity for empathy.
So if transgression occurs, the child who learned by induc-
tion strategies will be more likely to experience emotional
discomfort or guilt (p. 328).

Actually, inductive techniques may be more severe than
sensitizing techniques, due to their property of conservation
of anxiety, e.g., guilt feelings may be more long lasting
than the physical pain derived from a spanking.

Induction and sensitization are not mutually exclusive.
The parent may balance both techniques in a variety of ways.
It has been suggested that a maximizing situation for promot-
ing internalization is a minimal amount of sensitization and
some larger proportion of inductive parental acts (Hoffman
and Saltzstein, 1967). "Apparently, a spoonful of medicine
helps the reasoning go down" (Lickona, 1976, p. 25).

An affective and love oriented family seems to be a pre-
requisite for inductive techniques to be effective. "Being
well loved provides the emotional security and feeling of
confidence in the essential good of the world . . . neces-
sary for considering the needs of others" (Hoffman, 1970,

p. 286). Sears g£_al. (1957) in their extensive study

16
of children's acquisition of social behaviors, found that
accepting mothers who love their children and use inductive
more than sensitizing techniques produce children with more
moral internalization than other mothers. Henry (1971),
found that in the homes of mentally disordered children "a
person is viewed as an adversary and treated like one. 'The
child as enemy' is a fairly constant theme in these families."
This stands as evidence of the possible negative effects of
lack of love. In general, a fairly constant positive cor-
relation has been found between nurturance of parents and
internalization (Aronfreed, 1969, p. 305).

Love withdrawal in the form of ignoring, refusing to
speak, explicitly stating dislike, and isolating the child,
are considered as inductive techniques by Aronfreed (1968,
p. 314). However, this has not been found to be as con-
sistently related to the internalization of morality as the
other inductive techniques detailed above (Saltzstein, 1976,
p. 254; Hoffman, 1975, p. 233). The reasons cited for this
lack of consistent relationship is that other-oriented in-
duction capitalizes on the child's capacity for empathy,
while love withdrawal only emphasizes egoistic concerns,
more in accord with sensitizing child rearing practices.

To summarize, one can say that a love oriented family
will tend to have children who are less dependent upon ex-
ternal stimulation for proper behavior to take place. A
love oriented family is that in which parents use power as-

sertion (sensitization) only when absolutely necessary, but

l7

tend to guide their children's social behavior on the basis
of considering the consequences of the child's behavior for
others, and use reasoning and explanation as the mode of
problem resolution. Loss of love as a disciplinary measure
is not necessarily effective in promoting moral internaliza-
tion.

It is illuminating to call attention to a parallel
that Saltzstein (1976) has drawn between the set of techniques
described above and Kelman's types of attitude change in re-
sponse to social influence: "Power assertion goes with com-
pliance; love withdrawal and sometimes parent-oriented, and
even peer-oriented, induction with identification; and the
reasoning component of induction with internalization (pp.
261-262).

A child who has internalized moral standards can be
expected to be in a better position to reject portrayals of
negative behaviors offered on the television screen. The
core of the matter resides in the differentiation between
"good" and "evil" based on internal conviction and not on
external considerations. A child who does not have an in-
ternal pilot for social behavior will model those negative
portrayals which he/she has witnessed when the likelihood of
discovery and external punishment is minimal. A strong
moral conscience can be thought of as a servant who follows
the child everywhere and prevents him/her from undesirable

behavior. This servant is inside the child and its

18
punishments and rewards may be more powerful than all the
spankings or prizes of a parent. Children may learn, from
many televised instances, that by resorting to force they can
obtain certain satisfactions. However, only those with in-
ternalized moral values will say "no, it's wrong." Some also

will say "no," but because "I may be punished."

Family Interaction Perspective

 

Another trend of related research has evolved from the
work of a team of social scientists at the University of Wis-
consin (McLeod, Chaffee, and Eswara, 1966). This trend of in-
quiry emerged from an interest in studying "the possibilities
of changing personality through the change in the person's
interaction patterns" (p. 4) mainly within the family. These
researchers "felt that families were likely to differ in the
degree to which they emphasize harmony in interpersonal rela-
tions and impose restrictions on the expression of anger"

(p. S).

MCLeod and Chaffee (1972, p. 83) report that they have
consistently found two uncorrelated dimensions of communica-
tion structure in families. The first of these dimensions is

called concept-orientation, and the second one socio-

 

orientation.

 

In concgpt-oriented families, "the child is stimulated

to express his ideas and to challenge others' beliefs. He

MM

is frequehfiyfiexpased to both sides of an issue, and takes
part in controversial discussions" (McLeod, Chaffee and

Wackman, 1967, p. 2).

19

In families which are socio-oriented, "the child is

 

encouraged to maintain harmonious personal relations with
his parents and others. He may be advised to give in on
arguments, avoid controversy, repress anger, and generally
keep away from trouble" (Ibid., p. 1).

These two dimensions interact to render a fourfold

family style typology: (l) ppotective families emphasize

 

harmonious and hierarchical socio relations but do not stress
exposure to controversy or expression of ideas; (2) consen-
sual types emphasize both harmonious, hierarchical relations

and controversy or expression of ideas; (3) laissez faire

 

families, where neither socio, nor concept orientation is

emphasized; and (4) pluralistic families which stress con-

 

cept orientation and not harmonious, hierarchical social
relations.

This typology was first used to predict public affairs
knowledge and participation. It was expected that children
from pluralistic families would be the most likely to acquire
and rehearse political knowledge. The pluralistically raised
child should feel free of constraints which might inhibit
expression of his/her own opinions, and the exposure to dif-
ferent points of view should foster diverse interests. The
results of at least two studies supported the prediction:
"Pluralistic parents and children appear to be more politic-
ally informed and active, and more often use the media for

information" (Ibid., p. 6), while the children from the other

20
three types of family differed little among themselves.

The offsprings of pluralistic families have been
shown to be more independent in the judgment of social situ-
ations. They should also be expected to be more resistant to
persuasive messages. In two experiments with 9th grade
children Eswara (1968) and McLeod, Chaffee and Wackman
(1967) found that students from concept-oriented homes were
less persuasible and more sensitive to the amount of infor-
mation substantiating arguments in a persuasive message.
Persons from a socio-oriented background indicated they
would be more likely to abandon a plan if a neighbor friend
said he thought it was a poor idea.

In an experiment with adolescents, Stone and Chaffee
(1970) replicated those findings. Youngsters from highly
socio-oriented homes were more persuaded than others to
change their opinion with respect to a social truism (p.
243-244), when attacked by a highly credible source.

Pluralistic families may in fact immunize their chil-
dren against many sources of influence in the child's en-
vironment, as bases for decision making, and provide him/her
with a tool for judging evidence in order to arrive at sat-
isfactory solutions. "The socio-oriented person . . . asks:
Who is involved? Do I like them? Do they like me? The con-
cept-oriented person asks: What is involved? Is it a good
idea? How does it compare to what I know?" (McLeod and

Chaffee, 1972, p. 87).

21
Further studies using the family typology approach
have found suggestive evidence on the role of family inter-
action patterns in mediating the relationship between TV
violence viewing and aggressive behavior. Protective fam-
ilies have children who are the heaviest television viewers,
and are quite high in violence viewing. Children from plu-
ralistic families have been found to be very low consumers
of television content, and specifically violent television
fare. Protective families tend to have highly aggressive
children, while pluralistic families raise quite pacific
children. The consensual and laissez faire types fall some-
where in between. This set of findings by McLeod, Atkin and
Chaffee (1972) and others, have rendered suggestive evidence
of the possible indirect role of parental practices in re-
ducing the relationship between television violence viewing
and the child's aggressive behavior.
For example, Atkin (1972) reported that:
The relationship between violence viewing
and aggressive behavior in homes where the
parent tried to teach the child not to act
aggressively was compared to homes where a
more laissez-faire attitude was implemented.
The relationship between violence viewing
and aggressive behavior was much stronger
in the half of the . . . samples where no
emphasis was placed on nonviolent behavior--
while only a slight positive relationship was
found where the parents did emphasize non-
violence (p. 2).
Reiss (1971) reports that families of delinquents have

been found to experience the acceptance of "suggestions, ob-

servations, or ideas of others as a sign of his own weakness"

22
(p. 23) rather than in a rational frame of mind. One would
expect that a child who is positively oriented towards the
expression of ideas and their consideration and who is less
likely to be manipulated by external influences, would more
likely discount television violence as an appropriate means
of problem resolution. This type of child may consider many
options, including reasoning, before resorting to alternatives
that may go counter to his/her empathy concerns. This child
can be expected to be more likely to try to understand the
ideas and needs of others, without necessarily adopting them.

Two main theoretical and research tendencies have been
outlined with regard to parent-child interaction modes and
their effects on social behavior. Some implications of both
perspectives have been indicated for the relationship between
anti-social television viewing and negative behaviors on the
part of the child. A synthesis of these two perspectives is
now required.

From our review, one can observe the similarities
between the moral internalization perspective and the family
interaction orientation. Children from pluralistic families
can be expected to be children of inductive parents, and
furthermore, to be those children who are relatively inde-
pendent of external pressures in order to decide about appro-
priate social behavior. These children may watch less anti-
social television content, but even if they watch the same
amount as their counterparts, they will be less influenced by

it.

23

The parental styles of both research trends are con-
gruent. Concept orientation goes with reasoning, explanation
and an informational orientation in general. Power assertion
or sensitization fits the social power orientation where ex-
ternal pressures towards conformity are the rule.

At this point two general labels are proposed in order
to characterize the two general categories that have been
outlined. We can think of parental practices and child char-
acteristics as belonging to either external or internal types
of orientation (Aronfreed, 1968b, p. 34).

The externally oriented child is one who looks for the
possible external consequences of his/her behavior in a sen-
sitizing or socio-oriented parent. The internally oriented
child will focus on his internalized resources for judging
the behavior that might be appropriate under certain circumr
stances, and will use the information supplied by the family
in an atmosphere of love, for furthering his/her understand-
ing of the world.

An externally oriented child is expected to be morally
heteronomous or dependent, and an internally oriented child
is expected to be morally autonomous. The first will repro-
duce undesirable behaviors if they are expected to go un-
punished or undetected; the latter will refrain from such
behaviors when it is known that the welfare of others is in

jeopardy.

 

24

Hypotheses

The combination of a high regimen of internal guide-
lines and scarce externally oriented behavioral tendencies
should be the optimum mix to permit the child to evaluate
social situations and self-responses in terms of internal
considerations. Internally oriented children will be in a
better position to reSpond to the behaviors they have intern-
alized as prOper or inappropriate. On the other hand, when
parents generally resort to power assertion and less to
reasoning and explanation, their children should be external-
ly oriented and be less inhibited from performing the antisocial
behaviors witnessed on television. This is specifically so
if they feel that they can get away with it in order to re-
solve conflict or attain satisfaction of their needs.

The basic model underlying the present research is:

EXPOSURE TO ANTISOCIAL TV
BEHAVIORS
PARENT-CHILD INTERACTION
A
ORIENTATIONS i7
CHILDREN'S MODELING OF TV

ANTISOCIAL BEHAVIORS

Parent-child orientations should mediate the relation-
ship between TV exposure to antisocial portrayals and the

child's performance of such behaviors.

25

Before examining the complex contingent relationships
that properly test the mediation of parent-child interaction
orientations on the relationship between exposure and anti-
social predispositions, attention should first be focused on
the main effects and specific cell contrasts derived from
the theoretical statements presented above.

It is clearly the case that an internalized approach
or orientation to social situations should inhibit antisocial
attitudes and behaviors on the part of the child. Externally
oriented parents and children should, on the other hand,
facilitate antisocial attitudes and behaviors. Exposure to
antisocial portrayals on television has been repeatedly shown
to be associated with antisocial predispositions in children.
Consequently the independent effect of the internal and ex-
ternal orientations and exposure should behave in the follow-
ing way:

H : The more exposure to antisocial
portrayals on television, the more
antisocial behavioral predisposi-
tions will be displayed by children.
H : The more externally oriented the
parents and the children, the more
antisocial behavioral predisposi-
tions will be displayed by children.
H : The more internally oriented the
parents and the children, the less
antisocial behavioral predisposi-
tions will be displayed by children.
The cells resulting from the intersections of the

three factors, internal and external orientations and ex-

posure, should result in specific contrasts that can be

26
logically derived.

Let §l€§8 be the means in each of the following cells

INTERNAL ORIENTATION

 

 

LO HI
EXTERNAL ORIENTATION EXTERNAL ORIENTATION
LO HI LO HI
E
’1: L0 3E1 322 is x6
0
S HI x3 x4 x7 x8
U
R
E

 

 

 

 

 

 

and given Ha' Hb, and Hc, we expect that i4, which represents
the mean of the children's antisocial predispositions at a
high level of the external orientation and exposure and a low
level of the internal orientation, to be the highest mean in
the table.

On the other hand is, which is the mean of the chil-
dren's antisocial predispositions at a high level of the
internal orientation and a low level of exposure and the ex-
ternal orientation, should be the lowest mean in the table,
since the only factor impinging upon those children in any
substantial manner is the internal orientation.

According to Hypotheses a, b and c, it is expected that
the low internal orientation cells should be higher than the
high internal orientation cells: that the high exposure cells

should be higher than the low exposure cells; and that the

27
high external orientation cells should be higher than.the low
external orientation cells. Following these expectations,

one can posit the rank order of the cells in Hypothesis d.

Hd: 3E4>§3 = §2>§E1>328>3€7 = §6>§ES

Since there is no theoretical reason to expect that
exposure to antisocial TV portrayals would be more powerful
in affecting antisocial predispositions than the external
orientation, or vice versa, it is indicated that i3 should
be equal to i2 and that I7 should be equal to E6.

Finally, the relationship between exposure to tele-
vision antisocial portrayals and the children's antisocial
predispositions should be mediated by the external and in-
ternal orientations. Highly internally oriented and low ex-
ternally oriented parents and children should result in the
lowest correlation between exposure and antisocial predis-
positions in the children. Highly externally oriented and
low internally oriented parents and children should result
in the highest correlation between exposure and antisocial
predispositions. Also, the overall correlation between ex-
posure and antisocial predispositions, and the correlation
at the high-high and low-low intersections of the internal
and external orientations should all be localized in between
the two extremes specified above.

To illustrate, let's visualize the combinations that
are expected from both the external and the internal parent-

child orientations:

28

INTERNAL ORIENTATION

 

 

LOW HIGH
.33 E
g: .4 r1 r2
9
2% E3
xva H r3 r4
nag :n

 

 

 

 

Let rA be the overall correlation between exposure to anti-

social portrayals on television and antisocial predisposi-

tions in the child.

r1 = correlation between exposure to antisocial portrayals

r2

r3

r4

on television and the antisocial predispositions of
the child, at a low level of the internal orientation
and a low level of the external orientation.
correlation between the same two variables among chil-
dren with a low external orientation and a high
internal orientation.

correlation between the same two variables among chil-
dren with a high external orientation and a low
internal orientation.

correlation between the same two variables among chil-
dren with a high internal orientation and also a high
external orientation.

Considering the theoretical framework presented above

one would expect cell r2 to yield the lowest correlation in

the table, cell r3 to be the highest, and rA and cells r1

and r4 to have a coefficient lying in between cells r2 and

r3.

In other words,

29

r2 <rA<r3 and

53:11

r25rlir3

r2_<_r4_§r3

It should be noted that the expectations with regard
to Hf are logically derived but less documented. In the case
of r1, the child may obtain influential socialization experi-
ences mainly outside of the home, and the precise prediction
of the effect of those socializing forces lies outside of
the realm of this research. The children in r4 receive con-
tradictory messages from their parents, and the prediction
of the effects of such inconsistencies is not within the
domain of this study either, although it has been found that
a hypocrite model can enhance negative behaviors at the ex-

pense of pro-social ones (Bryan, 1975).

Summagy

Two main orientations, internal and external, concern—
ing parent-child types of responses to social situations have
been discussed. The main expectation resulting from a liter-
ature review and a theoretical effort, is that the internal
orientation should have a hindering effect on the child's
modeling of antisocial television portrayals, and that the
external orientation should have a facilitative effect on the
same phenomenon.

It has been hypothesized that exposure to antisocial

television portrayals and the two parent-child orientations

30
should have an independent effect on the child's display of
antisocial predispositions. The specific contrasts that de-
rive from the independent effects of exposure and the parent—
child orientations have in turn been hypothesized.

Finally, a set of hypotheses were derived that deal
with the potential role of those two dimensions in mediating
the relationship between exposure to antisocial portrayals
on television and the child's display of attitudes or be-

haviors modeled after such examples.

CHAPTER II

METHODS

In this chapter several sets of data will be dealt
with. First the results of some pretest data will be pre-
sented and discussed. Second, the sample of the present
study will be described, and the measures and descriptive
statistics pertaining to the data collected in this project
will be presented. Third, the reader will find the validity
and reliability coefficients for the measures used in this
study. Fourth, the data replicating the emergence of the
two dimensions, internal and external will be presented and
discussed. Lastly, the analytical methods to be utilized
in order to test the hypotheses proposed in Chapter I will

be introduced.

PRETEST

Preliminary Methodology

A first attempt at creating an instrument for measur-
ing the internal and external orientations will be reported
here. At the same time, the hypothesized emergence of the

two external and internal dimensions will be tested.

31

32

Instrument Design

 

The instrument designed was based on the one utilized
by Hoffman and Saltzstein (1967) in which "each respondent
(the child or parent) was asked to imagine four concrete
situations: one in which the child delayed complying with a
parental request to do something, a second in which the
child was careless and destroyed something of value, a third
in which he talked back to the parent, a fourth situation in
which he had not done well in school" (p. 48).

In response to the hypothetical situations it was
desired to obtain parental reactions towards the child's
behavior, and the child's own responses to social situations.
Although the greater bulk of research on moral internaliza-
tion, especially in field surveys, has been concerned with
responses and practices in relation to transgression (Hoffman,
1970, pp. 291, 348), it was felt that socially approved be-
havioral situations should be included. Consequently, posi-
tive social situations were also designed.

The rationale underlying the construction of positive
as well as negative situations is that children who are in-
ternally oriented with regard to transgressions should also
be internally oriented with regard to "good” doing. Extern-
ally oriented children should also reflect their preference
consistently across both types of situations. Likewise,
parents who are power assertive in negative situations should

also be externally oriented in positive situations by the

33
use of external rewards. Parents who are more inclined to-
wards the use of inductive techniques should also be more
likely to show their approval consistently in the form of
internal types of rewards.

The negative situations constructed were: Suppose
"you get mad and yell at your parents"; "your parents asked
you to do something for them, and you didn't do it"; "you
lie to your parents and they find out that you lied to them";
you tell another kid that you are going to hurt him"; "you
hit a kid in the neighborhood because of an argument."

The positive situations designed were: Suppose "you
do something really nice for someone in your family"; "you
help a kid in the neighborhood with some hard work"; "you
do something bad and your parents punish you. The next day
you tell your parents you are really sorry for what you
did": you do something nice for your parents to show them
that you love them."

The response items were designed for the child to
indicate "what would your parents do?" and what their own
reaction "would" be. Positive and negative situations were
repeated twice in different parts of the questionnaire, one
instance was for the parental reaction items and the second
for the child's own reactions.

The response items were designed with the internal

and external orientations in mind, based on past research.

34

Parental responses as perceived by the child were
designed for negative and positive situations. The paren-
tal orientation practices in the case of negative situations
were:

"Keep you from watching TV"

"Don't let you go out for a while"

"Hit or slap you"

"Yell at you"

"Don't talk to you for a while"

"Say that they are disappointed in you"

"Tell you to make up for what you did"

The first five items attempted to measure the extern-
al orientation, and the last two to reflect the internal one.

The external items include one with regard to tele-
vision exposure control as a disciplinary practice. The
item "don't talk to you for a while" was included as a
measure of love withdrawal, which has been found to be more
externally than internally oriented: "both love withdrawal
and power assertion direct the child to the consequences of
his behavior . . . for the child himself, and to the external
agent producing these consequences" (Hoffman and Saltzstein,
1967, pp. 54-55).

Responses like "don't let you out for a while," "hit
or slap you," and "yell at you" have been consistently found
to be externally sensitizing (Adkins gg_§l., 1974, p. 118).
The two internal items "say that they are disappointed in
you" and "tell you to make up for what you did” have been

reported to lead to moral internalization, as Opposed to the

external items (Aronfreed, 1969, p. 309).

35

The parental response items in the case of positive
situations were intended to reflect the same predispositions
as the ones for the negative situations. The reSponse items
were:

"Let you watch extra TV"

"Give you something special"

"Pay extra attention to you"

"Kiss or hug or pat you on the back"

"Tell you they are proud of you"

"Tell you to feel good about what you did"

The first three items were intended to reflect the
external orientation. "Kiss or hug or pat you on the back"
was designed to tap general affection. "Being well loved pro-
vides the emotional security and feeling of confidence in
the essential goodness of the world . . . necessary for con-
sidering the needs of others" (Hoffman, 1970, p. 286). Also,
"there is some evidence . . . that the learning processes
which underly internalization may be dependent on a certain
minimum of nurturance (warmth, affection, etc.) in the general
climate of child rearing" (Aronfreed, 1969, p. 304). The
last two positive items were intended to be the counterparts
to the internal negative ones.

The child items for negative situations were:

"WOuld you feel guilty?"

"WOuld you try to make up for what you did?"

"WOuld you be afraid your parents (or your teacher

or other people) may find out?"
"Would you be afraid you may be punished?"
The first two items which include predispositions to

guilt feelings and towards reparatiOn were designed to tap

the internal dimension (Hoffman and Saltzstein, 1967, p. 228).

36
The last two items were intended to tap external considera-
tions for the commitment of a negative action (Ibid).

The child items for positive situations were:

"Would you feel happy with yourself?"

"Would you feel that you deserve a prize?"

"Would you want people to know about it?"

The first item should correspond to its guilt counter-
part in terms of the internal orientation. Reparation
("would you try to make up for what you did?") could not be
issued a mirror image in the positive situations. The last
two items were intended to measure external expectations of
reward and recognition.

All of the items were to be answered with a "yes" a
"maybe" or "no." A "no" would be coded with a 0 (zero), a
"maybe" with a 1, and a "yes" with a 2. The higher the like-

lihood of each parental or child item, the higher the nu-

merical value that would be assigned to it.

Qgestionnaire Administration

In the Spring of 1976, the above instrument was ad-
ministered to 202 children in two schools of the Lansing,
Michigan area. Half the children were fourth graders and
the other half were sixth graders, about equally divided by
sex. The self administered instruments were given to the
children in their classrooms, and were guided through it by
one adult. The instrument contained other items for a larger
research project. No items presented special difficulties

for the children, except for a minority of fourth graders

37
with reading problems, in which case all situations and re-
sponse items were read to them. The time of administration
was 15 minutes in the average. The children were assured

anonymity, and were asked to tell us "how you really feel."

Index Construction

For both parental and child items, all responses were
summed across identical items* to form an index for that
item. The parental response indexes for positive situations
could range from 0 to 8, and for the negative situations
from 0 to 10. The child's response indexes for positive
situations could go from O to 8, and for the negative situ—
ations they could range from 0 to 10.

An example may help in clarifying this procedure.
If a child reported that his/her parents would yell ("yes")
if he/she committed the actions described in all five of
the negative situations, then the index for ”yell at you"
would have a score of 10. However, a child who said that
his/her parents would never yell at them, would receive a
score of 0. Table 1 presents the means, standard deviations,
and number of cases for each index.

The addition of-scores across identical items was
reasoned to be a more valid measure of enduring parental or
child response modes than each separate, situation specific,

'response.

 

*
Recall that each item could be answered with a "yes"
(2), a "maybe" (1) or a "no" (0).

‘4

be.

   

. . . . v
-w .- .. .. n-.. --. . q..— —-. .
. — — - , e .41 ., N
d. | 1 1' . . ‘ . .’ )‘1 - JJ . .4- l
.. e
r , - V ..- r 1 ; - . -: 1 v1." "7
. I I 1 . o I - J . ' 4 A I.
’ ,
v- . 1:, - 7- pr w ' , ‘ “a".-
,. (r . __ '- J x -1 l _ . .— A .A- I: .. ' ,1 ,. l .. . .. Q I:- '
" " \ —' a. 7 1' ‘ /\ ~( ,\ -
‘ _ '- . \n 2 >1 ) .‘ _' .l J . I. .. .J 1.
u
. . . . . ‘ .‘ i . .. . \ 3 : -. f . ,
. l ' '\ a I ‘ A h‘ .4 '.‘ a I . 3' s
I ' '4 I
( . ' ’ - 'f “I
‘ , . t 1 4 x,_ -~ . ‘. .J
.‘ , .
q a . . . .
V s f I \ '. . I r‘ (a
. I 2 g }\ ‘1 \ _} i ’1‘ I ~. ( I
. ‘ I
9 . i 1 l- w - ‘
t I n I ' ,4
L - o } .' ‘ A : I‘L -‘ I

.,
lb . DO _. ' “. ‘5. , i o ‘ .' ’Q’ ~ '
I' h.
'5 ,. ' , I ‘ u ) F
.4. 4' ‘ \ I '
.- ‘ . ~5f _ J _. ‘ -
"cult , s . a <~:.H'~ 1. r}'
u . 0 1 J ~ .4 D“ 1'. to}. I- o
v
' r
’ I. ‘ -' \I l."" -v, ’ 'n l< .
. a l' r s , -
' T '4 {I J.. a ‘ _. ~) v ' I'. ‘ 6" 1..) ~ 0 ‘.
_

b

:11—"36‘": -?l‘{.f€ E; I (9;.531‘ci ,' r nl‘l ‘V . t) :Lk.

rarer. ~. -- . r. . a . .
1 ~. ,*_,':€.' 5 If. '4’ ." e- s :- " ..' /.
‘ -' T. .)1\ "~ .4” ~v.:: ;" r I ‘ L . ‘

-I.L K -4 r...‘ J' i _. ~1'.‘ . \' . x L. ,t it

‘ Q
I
.b
‘
x
\
g.
In
H
r'.
J
"N
‘4

38

In Table 1 all indexes have some degree of vari-
ability. Even in the case with the smallest standard devia-
tion ("WOuld you feel happy with yourself?"), the average
variation was 1.8 units in a scale of 0 to 8.

In the case of parental responses to negative situa-
tions, parents were mostly expected to yell and express dis-
appointment in the average given the means of 5.5 and 5.4,
respectively. Love withdrawal ("Don't talk to you for a
while"), on the other hand, resulted in a mean score of 1.9
in a scale of 0 - 10, being this the least likely parental
disciplinary technique as perceived by the children.

The parental responses to positive situations indicate
that in the average children expect their parents to express
pride more often than anything else. However, the item ”pay
extra attention to you" showed the lowest mean.

In hypothetical negative situations, children were
most likely in the average to indicate that they would try
to repair their undesirable actions. They didn't show, how-
ever, that much concern for being caught or punished.

In positive situations, the highest mean score was
for "feel happy with yourself," and the lowest for feeling
that they "deserve a prize."

One way of deciding which items should be reused
would be to select items frequently endorsed in the average,
and with a minimally acceptable variation. This procedure,

however, may not be acceptable in the present case. The

39

Table 1. Means, Standard Deviations, and Number of Cases
for Each Index of Parental or Child Responses to
Social Situations.

 

Y s N

 

Parental Responses

Negative Situations (Scale 0 - 10)

 

1. Keep you from watching TV . . . . . . 2.8 2.7 195
2. Don't let you go out for a while . . 4.1 3.0 194
3. Hit or slap you . . . . . . . . . . . 2.7 2.8 194
4. Yell at you . . . . . . . . . . . . . 5.5 2.9 190
5. Don't talk to you for a while . . . . 1.9 2.4 190
6. Say that they are disappointed in you 5.4 3.0 190
7. Tell you to make up for what you did 3.8 3.1 194
Positive Situations (Scale 0 - 8)
8. Let you watch extra TV . . . . . . . 2.6 2.3 195
9. Give you something special . . . . . 2.7 2.0 193
10. Pay extra attention to you . . . . 2.1 2.0 189
11. Kiss or hug or pat you on the back . 4.0 2.6 194
12. Tell you they are proud of you . . . 5.1 2.2 193
13. Tell you to feel good about what you
did 0 O O O O O O O O O O C O O O O 4.4 2.6 197
Child Responses
Negative Situations (Scale 0 - 10)
14. Would you feel guilty? . . . . . 5.6 2.9 193

15. WOuld you try to make up for what

you did? . . . . . . . . . . . 6.0 3.0 194
16. Would you be afraid your parents(or

teacher or other people) may find

out? . . . . . . . . . . . . . . 3.8 3.1 193
17. Would you be afraid you may be

punished? . . . . . . . . . . . . . 3.8 3.2 191

Positive Situations (Scale 0 - 8)

18. Would you feel happy with yourself? . 6.3 1.8 198
19. Would you feel that you deserve a

prize? . . . . . . . . . . . . . . 1.6 2.2 195
20. WOuld you want people to know about

it? . . . . . . . . . . . . . . . . 2.7 2.1 196

 

40

theory presented in Chapter I predicted the emergence of
two dimensions, internal and external. Those indexes of
response clearly associated with each dimension should be
the ones to be retained. Furthermore, the indexes should
be closely associated with the dimension of which they were
predicted to be a part. It was considered that the most
appropriate technique to use would be confirmatory factor
analysis (Rummel, 1970, p. 22). A certain type of response
may not be very widely utilized and still constitute an
important mediator of what children model from antisocial
television portrayals.

Two questions are to be answered in this section:
1. Do an internal and an external orientation emerge as
hypothesized?; and 2. Which are the items that better dif-
ferentiate one dimension from the other?

To answer the above questions, the indexes for paren-

tal and child responses were submitted to a principal axis

 

factor analysis routine with varimax rotation. Since two

 

dimensions were hypothesized, the solution was limited to
the extraction of two factors. According to the numbering
of indexes in Table l, the items predicted to belong in the
internal orientation dimension were: 6, 7, ll, 12, 13, 14,
15, and 18. The items hypothesized to load on the external
orientation dimension were: 1, 2, 3, 4, 5, 8, 9, 10, 16,

17, 19, and 20.

41
Table 2 presents the results of the factor analysis
procedure. A loading of .35 is considered to be the cutoff
point for including an index in a certain factor, since
”loadings exceeding .35 have been found to be stable and
replicable" (Overall and Klett, 1972, p. 109).

Table 2. Principal Factor Matrix with Varimax Rotation,
Ordered According to Hypothesized Indexes Loadings.

 

Item Factor 1 Factor 2

 

Internal Orientation

 

6. Say that they are disappointed in you .13 .58

7. Tell you to make up for what you did .41 .47

11. Kiss or hug or pat you on the back .11 .53

12. Tell you they are proud of you .12 .62
13. Tell you to feel good about what

you did .20 .64

14. WOuld you feel guilty? .08 .60
15. WOuld you try to make up for what

you did? .02 .68

18. Would you feel happy with yourself? .02 .50

External Orientation

1. Keep you from watching TV .81 .08
2. Don't let you go out for a while .57 .25
3. Hit or slap you .46 -.05
4. Yell at you .36 .09
5. Don't talk to you for a while .53 .07
8. Let you watch extra TV .62 .14
9. Give you something special .48 .22
10. Pay extra attention to you .47 .26

16. Would you be afraid your parents (or
teacher or other people) might

find out? .36 .36
17. WOuld you be afraid you may be

punished? .43 .34
19. WOuld you feel that you deserve a

prize? .42 -.06

20. WOuld you want people to know about
it? .44 .16

 

42

Factor loadings can be interpreted to represent

"correlations between original measurements and the factors"

 

(Ibid., p. 91). Looking at Table 2 one can say that in
general, the two hypothesized dimensions contain the pre-
dicted indexes. Factor 2 clearly represents the internal
orientation dimension. All hypothesized internal indexes
load highly and consistently on Factor 2, and in general
show low or almost zero loadings on Factor 1. The one ex-
ception is response index number 7, which does not seem to
differentiate between the two factors.

Factor 1 consistently represents the external orien-
tation dimension. Two response indexes, numbers 16 and 17
do not discriminate between the two factors.

The results support the expectation that postulated
the existence of two parental and child orientations, one
external and one internal. Second, although the majority of
the indexes pretested to tap both dimensions in fact seem to
underly them, three different items correlate about equally
with both factors. This last finding deserves further
consideration.

Index number 7 “Tell you to make up for what you did"
was originally hypothesized to be part of the internal orien-
tation dimension. This index does show a higher loading on
the predicted dimension, however, its loading on the external
factor is above the .35 criterion, and the difference between

both loadings is very small, .06. One may conclude that

43

parental requests for reparation accompany both a spanking

in one case, and an expression of disappointment in another,
or that at least children expect their parents to act in this
manner.

One course of action would be to simply delete such
response item from a future instrument. However, it is not
known whether the actual parental response would behave in
the same fashion. This writer considered collecting the
parents own responses in the final study, and did.

Two more indexes did not discriminate between the two
types of orientation: 16. "WOuld you be afraid your parents
(or your teacher or other peOple) may find out?" and 17.
"WOuld you be afraid you may be punished?" These two indexes
were originally hypothesized to belong in the external orien-
tation. Index 16 loads equally on both factors, and index 17
loads higher on the external dimension but the difference
between the two loadings is small (.09). Besides, index 17
is close to reaching the .35 criterion. Both indexes reflect
the children's own responses as reported by themselves. The
first explanation that comes to mind is that perhaps these
responses, as they are formulated, are too vague. The chil-
dren who report other reactions also report fear of punish-
ment and fear of detection. Some children might have
internalized moral values, but if they do transgress then
they are afraid of the consequences. In the case of detec-

tion or finding out, it seems to be best to delete

44
the response item from future consideration.

In the case of being afraid of punishment, the item
could be rewritten in two separate ones, for example: "WOuld
you be afraid your parents may spank you?" and "Would you
worry about disappointing your parents?" With these two new
items, one could identify children who are afraid of the
painful physical consequences for themselves, and children
who wouldn't want to hurt their parents. In the first case
we would have an external orientation, and an internal in the

second.

F INAL STUDY

Sample Description

Three hundred mothers and their children who were
part of a panel sample contacted a year before (Atkin and
Greenberg, 1977) provided the data base for this study. The
mothers were interviewed at home by trained interviewers,
and the children were administered questionnaires in their
school classrooms. No more than 5% of the children were
administered the questionnaires in their homes, since they
could not be reached in the schools. The data were collected
in two comparable cities of the United States, Haslett,
Michigan, and Verona, Wisconsin, from middle class respon-
dents. The children were from the fifth, seventh and ninth

grades.

45

The actual data return was from 227 respondent pairs
from the original sample of 300. This was due mainly to
geographical mobility and refusals.

The composition of the final sample was as follows:

Location:

Haslett, Michigan: 130
Verona, Wisconsin: 97

Grade in school:

Fifth: 74

Seventh: 81

Ninth: 72
Sex:

Males: 112

Females: 115
The average interviewing time for the mothers was
approximately 45 minutes. The children took about one hour

in the average to complete their questionnaires.

Measures and Descriptive Statistics

Several sets of data will be presented in this sec-
tion. In the first place the reader will find the measures
utilized in order to tap the internal and external dimensions
of parent-child interactions. Next, the measures of tele-
vision exposure used in.this project will be presented.

Finally, the childs antisocial measures will be introduced.

Internal and External Orientations
The measures used in order to tap the internal and

external orientations of parent-child interactions do not

46

differ substantially from the measures utilized in the pre-
test, and the results of the preliminary study were taken
into consideration in order to improve the instrument.

Four sets of items were assembled to tap the two
dimensions. The mothers responded to the parental practices
items, and the children responded to the child response items.
Each, mother and child, was presented with eight hypothet-

ical situations, four positive and four negative, as follows:

Mothers:
Positive situations

Suppose (NAME OF CHILD) does something really nice
for you (and your husband) to show that (HE/SHE)
loves you. What would you (and your husband) do?

Suppose (NAME OF CHILD) helps a friend in the
neighborhood with some hard work, and you (and your
husband) hear about it. What would you (and your
husband) do?

Suppose (NAME OF CHILD) does something really nice
for someone in your family. What would you (and
your husband) do?

Suppose (NAME OF CHILD) apologizes and tells you

(and your husband) (HE/SHE) is really sorry for some-
thing bad (HE/SHE) did to you. What would you (and
your husband) do?

W

Suppose (NAME OF CHILD) hits a kid in the neighbor-
hood after an argument, and you (and your husband)
find out. What would you (and your husband) do?

Suppose you (or your husband) asked (NAME OF CHILD)
to do something for you, and (HE/SHE) doesn't do it.
What would you (and your husband) do?

Suppose (NAME OF CHILD) lied to you (and your husband)
and you find out. What would you (and your husband)
do?

47

Suppose (NAME OF CHILD) gets mad and yells at you
(or your husband). What would you (and your
husband) do?

Children:

Positive situations

 

Suppose you do something really nice for your
parents to show that you love them.

Suppose you help a friend in the neighborhood
with some hard work.

Suppose you do something really nice for some-
one in your family.

Suppose you apologize and tell your parents you
are really sorry for something bad you did to them.

Negative situations

Suppose you hit a kid in the neighborhood after an
argument.

Suppose you lie to your parents.

Suppose your parents ask you to do something for
them, and you don't do it.

Suppose you get mad and yell at your parents.

The results of the pretest data presented above were
taken into consideration for the generation of the response
items used in the final study. The rationale for including
the different response items was presented in the preceding
section dealing with the pretest, and it will not be repeated
here. However, two major types of items were added. For
both parental and children orientations response items were
included to tap parental emphasis about the consequences of

the children's behaviors on others, and the children's

48
consideration for the welfare of others including the
parents. These types of responses have been found to be
important components of the internal orientation of both
parents and children (Aronfreed, 1968b and 1976; and Staub
and Feinberg, in press).

The final set of items also included reasoning and
explanation as a parental technique since it has been found
to promote the internalization of moral values in children
(Sears EE.2l-r 1957; Aronfreed, 1969; Aronfreed, 1976; and
Saltzstein, 1976).

The number of items used to tap the internal and ex-
ternal orientations in children and parents was determined
by the antecedents present in the literature reviewed, the
pretest results, and by a set of ten parental and children
interviews conducted in order to check for those response
modes that take place in the home setting.* Consequently,
the reader will find that the number of items for the intern-
al and external orientations and for both parents and chil-
dren is not necessarily the same.

i The following parental responses were generated for
the positive situations:

Internal:

Say you are proud of (HIM/HER).

 

*

These were informal interviews in the home setting,
where parents and children were asked for criticisms and
elaboration. Also, a team of researchers highly familiar
with this type of research provided inputs for the question-
naire formation.

tions

tions

49
Tell (HIM/HER) to feel good about what (HE/SHE)
did.
Kiss or hug or pat (HIM/HER) on the back.
Explain why it was a good thing to do.

Say that you appreciate the good things
(HE/SHE) does.

Tell (HIM/HER) reasons why (SHE/HE) should
keep doing these things.

External:

Let (HIM/HER) do something (HE/SHE) wanted to
do very badly.

Let HIM/HER) watch extra TV.

Give (HIM/HER) something special.

The parental response items for the negative situa-
were:
Internal:

Say you are disappointed in (HIM/HER).

Explain why (HE/SHE) shouldn't behave that way.
Say (HIS/HER) behavior makes you feel bad.

Tell (HIM/HER) another way to solve (HIS/HER)
problem.

External:

Don't talk to (HIM/HER) for a while.
Yell at (HIM/HER).

Hit, spank or shake (HIM/HER).

Keep (HIM/HER) from watching TV.

Don't let (HIM/HER) go out for a while.

The response items for the children's positive situa-
were:
Internal:

Would you feel good about it?
Would you be proud of yourself?
Would you feel happy for them?

were:

a yes,

50

Would you feel happy with yourself?

External:

Would you feel you deserve a treat?

Would you want people to know about it?

Would you think you should get something special?

The items for the children's negative situations

Internal:

Would you feel guilty?

WOuld you worry about how they feel?

Would you worry about disappointing your parents?
Would you try to make up for what you did?
External:

Would you worry that your parents may hit,
spank or shake you?

Would you worry that your parents may yell
at you?

Wbuld you worry that your parents may take
something away from you?

Would you worry that your parents may not
let you watch your favorite TV show?

The response categories for all the above items was
maybe or a no, coded as 2, l, and 0, respectively.

All responses to identical items were summated to

create indexes of response, across varied situations for both

parents and children, and for positive and negative situa-

tions.

The descriptive statistics of these indexes of re-

sponse are presented in Tables 3 and 4.

51

 

mNN
MNN
mNN

«NN
mmm
mmm
mmm
mmm
own

wNN
mNN
QNN
mNN
wNN

mmm
mmm
ANN
mam

NON HON o o o o o o o
m.~ e.e . . . . . . .
m.~ e.m . . . . manta
e.~ m.e motes» mama»
~.H e.e . . . . .
HON mtw O O O O O O O
e.~ H.e . . . . . . .
HON mow O O O 0 O I O
”OH ”ON. 0 O O o 0 0 0

amoeba)
O
ONNNN

MV'NV'M
OMANV‘
o
o
o
o

mmha
O
u>h\0h-
O
O
O
O

m .m

. . . . . . . . . . Hmwoomm maesuofiou Ammm\zH=V w>ww

. . . . . . . . . . . . >9 tuuxo nouns lemmszms hog
sum> on on cause: 1m=m\mms scenuosom on lemmxzem. awn

AGZmWme

mcaoo mmox pasonm Amm\mmmv was accuses Ammm\szv Haws
moon Amm\mmmv moses» ooom one macaooummm no» nos» mum

. . . . . on on mean» ooom n was us has camamxm
. . . . . . xomn on» so .mmm\szv use no man so when
. can .mmm\mms .umee uaonm poem Heme op lemmsze. Heme
. . . .mmm\szv mo ozone mum sch hem

AdzmmBZH

.mH
.hH
.mH

.mH
.vH
.MH
.NH
.HH
.OH

 

acoHumsuwm m>wuwmom

. . . . . . . . Tawn3 o now use omammm\szv you u.:oa
. . . . . . . . . . . >9 ma~noum3.&ouu Ammm\szv moon
0 o c o o o o o o o o .- AmmmgHmv mxm-m HO xcmmm ~U...Hm~
o o o o o o o o o o o o o o o o o AmmmEHmv U.“ HHwM
. o o o o o . . wflflnz M HON AmmmgHmv CU. u—HMU U. C8

Adzmmfixm

amenoum Ammm\mHss meson 0» so; hmeuoem Ammmxzems Heme

. . . own Home 50% nexus H0a>mnon Ammm\mmmv mum
. . . >m3 page m>m£0n u. coaaonm Ammm\mmv has sandman
. . . . . . . . Ammm\=Hmv a“ omucwommmmwo one so» mom

AdzmmBZH

 

maceumsuwm m>wummmz

 

.Amlo mucus xmoch ncoHumsuem Heeoom on noncommom

Hmucmumm no xwocH even you

mommo mo umnfisz one .m:0eum«>oa tumocmum .mcmoz .m manna

52

 

vNN
mNN
mNN

wNN
mNN
bNN
bNN

oNN

MNN
emm
omm

mNN
MNN
mNN
«NN

m.H m.H
h.H m.m
m.H m.H
m.H m.m
m.H o.m
o.N o.w
v.H 0.5
m.N m.H
».N m.~
m.N H.v
¢.N ¢.N
H.N m.m
m.N v.m
H.N w.m
N.N m.m

m _m

H50» noum3 5o» umH uoc
~50» EOHH »m3m mcHnuosom 0xmu
O O O O O O O O ~50“ pm Hflmh
~50» mecca H0 Hemem .uHs

NHMHomem mcHnumeom umm ©H5ocm 50» stnu
muH u50nm zocx ou 0He00e ucm3

. mumeu m 0>H0mmo 50» Hmmm

. «HH0EH50» suH3 »eemn Homm

»ME
»ME
»08
»m&

. «Beau How »eemn Home
. mmHmmH5o» Ho o50He on
. wuH u50n0 poom Hmwm

so» eHnoz .mH
50» oH503 .vH
so» eHsos .mH

Aézmmaxm
so» pesos .NH
50» UH503 .HH
no» ounce .oH

50» 0H503 .m

AdzmmBZH

 

mGOHu05uHm 0>HuHmoe

musmHme
muamee
mucmHme
mucmHme

H50» umnu »HH03
H50» umnu »HH03
H50» umuu »HH03
H50» umnu »HH03

mpHp 50» uses How e5 0x08 ou »Hu
mmusmHme H50» msHucHoeemmHt u50nm »HH03

«Homm »mnu 30: u5onm »HH03

m»uHH5m mem

N30£m >9 OuflHO>MH

50» ©H503 .
50» 0H503
50» UH503 .
50» UH503

IDKDI‘Q

{Adzmmfixm

50» 6H503
50» oHsoz
5o» 6H503 .
50» UH503

O
HNMQ‘

AdszBZH

 

mcoHum5uHm 0>Hummoz

 

OHHfiU HO mecH comm

.Amlo TOGMH xmocHV NGOHum5uHm HMHoom 0u momsoemmm

How mommo m0 HonE5z can .mcoHuMH>oe onocmum .mcmoz

.v OHQMB

53

The indexes in all cases could range from 0 to 8. In
Table 3 one can observe that, in general, the means for the
response items belonging to the positive situations are high-
er than for the negative ones, but more interesting is to
notice that across both positive and negative situations,
the internal items have higher means than the external ones.
Seemingly, parents tend to be more likely to explain and show
affection than to withdraw it or physically punish their chil-
ren, or at least these are the things that they are most
willing to report.

Congruent with the pretest results, item 5 shows the
lowest mean in the table and the lowest variability as indi-
cated by the standard deviation. Item 5 is at this point a
good candidate for exclusion from the analysis. All other
items in Table 3, at this point seem to be acceptable, since
the discriminatory value for the internal and external orien-
tations of an item with a low mean can be of importance in
selecting those cases that are extremes in either orientation.

In Table 4 one finds a similar pattern of means for
children reactions to that of the parental practices in Table
3. The means for the internal are consistently higher than
the means for the external orientation items. At this point,
all of the items seem to be appr0priate for further analysis,
since all the means evidence the existence of the behaviors

of interest to an acceptable degree.

54

Television Exposure
to Antisocial Behaviors

The children in this study were asked to rate a list
of 29 shows as to whether they watched each of the shows
"every week," "most weeks," "some weeks," or "never." The
scale points were coded as 5, 4, 2, or 1 respectively. The
exclusion of the value 3 was done because it was considered
that the interval between watching "most weeks” and ”some
weeks" was about twice as large the distance between "every
week" and "most weeks," or the distance between "some weeks"
and "never."

The 29 shows were selected on the bases of high view—
ership, distribution across the days of the week and the be-
haviors characteristically portrayed in those shows. The
selection of high viewership levels of each show and the
variability of portrayal of antisocial behaviors in each
show was carried out according to the results of a recent
viewership and content analysis set of studies (Greenberg,
Atkin, Edison and Korzenny, 1977 and forthcoming).

The shows included in the questionnaire were: The
Six Million Dollar Man, Rhoda, The Jeffersons, Little House
on the Brairie, Happy Days, Laverne and Shirley, Baa Baa
Black Sheep, Police Woman, Kojak, Good Times, the Bionic
Woman, Charlie's Angels, The Waltons, Welcome Back Kotter,
Hawaii Five-O, Sanford and Son, Chico and the Man, The Rock-

ford Files, Emergencyl, The Mary Tyler Moore show, Bob Newhart,

55

Starsky and Hutch, All in the Family, Jabber Jaw, The Rink
Panther, The Bugs Bunny/Roadrunner hour, Shazam, Isis, and
Fat Albert and the Cosby Kids.

The viewership rating of each show was weighted by
the frequency of occurrence of antisocial behaviors in each
show, as determined by the content analytic studies cited
above. In those studies it was found that the most prevalent
antisocial behaviors available on the current television

fare are, in order of prevalence: verbal aggression, physi-

 

cal aggression, deceit and theft. Based on those results it

 

was decided to limit the study of antisocial behaviors to
those more available for modeling on television programming.

Four indexes of exposure to antisocial behaviors were
then created, one for each type of antisocial behavior. In
each index, each amount of exposure to a show in the ques-
tionnaire list was multiplied by the frequency of each of
the four types of antisocial behaviors in that show. So, if
a show did not contain any physical aggression, the product
of exposure by frequency of portrayal would be zero.

The advantage of such indexes of exposure is that they
are intended to reflect more accurately than other methods
used in the past the degree to which individuals are actually
exposed to specific television portrayals. The stability of
television portrayals by the methods of content analysis used
by Greenberg pp 31. (Ibid) has been demonstrated in ongoing

research across three episodes of a subsample of programs.

56

Below are the means, Standard deviations and

number of cases for each index of exposure to antisocial
behaviors on TV.
Index of Exposure to: Y S N
TV Physical Aggression

(range 365-1502) 848.63 249.84 213

TV Verbal Aggression

(range 534-1857) 1020.31 248.33 212
TV Theft

(range 32-116) 62.92 17.77 215
TV Deceit

(range 136-518) 287.60 73.93 213

Children's Measures of
Antisocial Predispositions

In order to tap the children's proclivity to utilize
antisocial modes of conflict resolution, the mothers and
their children were asked to report the likelihood that the
child may respond antisocially across varied social situa-
tions. This method has been widely utilized in the research
concerning children's modeling of antisocial behaviors,
especially physical aggression (e.g. Leifer and Roberts,
1972).

The mothers interviewed were presented with the fol-

lowing social situations and response items:

57

For instance, what if someone out in front of (HIM/
HER) in a long line. What would (HE/SHE) do? . . .

Would (HE/SHE) push them out . . . yes, maybe,
or no?1

Would éHE/SHE) yell at them . . . yes, maybe,
or no?

Would (HE/SHE) tell them politely to leave?

The cashier at a store gives (NAME OF CHILD) one
dollar more in change than (HE/SHE) is supposed to
get. What would (HE/SHE) do?

Give the extra money back

Keep the money3

Suppose (HIS/HER) friends are all going to the movies,
but you order (HIM/HER) to stay home and finish (HIS/
HER) homework for school.

Would (HE/SHE) argue with you?2

Would (HE/SHE) finish (HIS/HER) homework like you
tell (HIM/HER)?

Would (HE/SHE) lie and say (HE/SHE) already
finished it?4

What if a kid is using (NAME OF CHILD)'s bike without
asking (HIM/HER). (HE/SHE) asks to ride it, but the
kid won't give it back. What would (HE/SHE) do?

Scream at the kid?2

Wait awhile until the kid gives it back?

Hit the kid?1

 

1These items constituted the index of physical
aggression.

2These items constituted the index of verbal
aggression.

3These items constituted the measure of theft.

4These items constituted the index of deceit.

58

Suppose your child is playing a game. No one is
watching when (HE/SHE) takes (HIS/HER) turn, and
(HE/SHE) doesn't do well. What would (NAME OF
CHILD) do?

Say (HE/SHE) did well4

Say (HE/SHE) hasn't had (HIS/HER) turn yet,
and takes another turn

Tell the truth

All response items could be answered with a yes, may-
be or no, coded respectively as 2, l, and 0. Items indicat-
ing the same type of response, e.g. physical aggression,
verbal aggression, theft or deceit were summed to form in-
dexes for that type of response.

Table 5a contains the means, standard deviations, the
number of cases and the range for each of the indexes of the
mothers' report regarding their children's antisocial pre-
dispositions.

The children's questionnaires contained the following

social situations and response items.

What if someone cut in front of you in a long
line. What would you do?

Would you push them outl. . .

Would you yell at themz. . .

 

1These items constituted the index of physical
aggression.

These items constituted the index of verbal
aggression.
3

4

These items constituted the measure of theft.
These items constituted the index of deceit.

59

Suppose you are playing a ball game and one kid
throws the ball at your head while you aren't
looking, and it hits you. What would you do?

Hit the kid with the balll. . .

Kick the kidl. . .

Shout bad names at the kidz. . .

The cashier at a store gives you one dollar more in
change than you are supposed to get. What would you
do?

Give the extra money back . . .

Keep the money3. . .

A kid is using your bike without asking you. You
ask to ride it, but the kid won't give it back.
What would you do?

Scream at the kidz. . .

Hit the kidl. . .

Knock the kid off your bikel. . .

What if a friend is feeling bad because they keep
making stupid mistakes while trying to play a new
game. What would you do?

Help them to learn the rules . . .

Say a mean thing to themz. . .

Say you still like them anyway . . .

 

1These items constituted the index of physical

aggression.

These items constituted the index of verbal

aggression.

3These items constituted the index of theft.

4These items constituted the index of deceit.

60

Suppose you are playing a game. No one is watching
when you take your turn, and you don't do well.
What would you do?

Say you did well4. . .

Say you haven't had your turn yet, and take
another turn4. . .

Swear at the kid who tells on youz. . .

Your friends are all going to the movies, but your
mother orders you to stay home and finish your home-
work. What would you do?

Argue with your motherz. . .

Lie and say you already finished it4.

You are playing a game with another kid and you are
winning. The other kid decides to quit in the middle.
You ask them to stay, but the kid says no. What
would you do?

Say a mean thing to the kidz. . .

Grab the kidl. . .

What if you find somebody's pen on the floor at
school. What would you do?

Try to find the owner . . .

Keep the pen3. . .

You work really hard to finish your homework for
school. Some kid takes your papers and rips them
up. What would you do?

Shout at the kidz. . .

 

1These items constituted the index of physical
aggression.

These items constituted the index of verbal ag-
gression.
3

4

These items constituted the index of theft.
These items constituted the index of deceit.

61

Beat up the kidl. . .

C0py someone else's homework4. . .

You are walking down the street and dr0p a dollar
bill. Another kid grabs the money and won't give
it back. What would you do?

Yell at the kidz. . .
Start a fight with the kidl. . .

Throw something at the kidl. . .

What if your friends are playing a game and really
need one more person to play, but you don't like
that game. What would you do?

Join in so they can play it . . .

Help them find another player . . .

Lie and say you have to go home4. . .

All response items could be answered with a yes, may-
be or no, coded as 2, 1, and 0 respectively. Items indicat-
ing a similar type of antisocial behavior were added to form
and index for that type of behavior.

Table 5b contains the means, standard deviations, the
number of cases and the range for each of the indexes of the

child's own report of antisocial predispositions.

 

1These items constituted the index of physical
aggression.

2These items constituted the index of verbal
aggression.

3These items constituted the index of theft.
4These items constituted the index of deceit.

62

Within the range for the mother and child indexes,
one finds that the means indicate that all behaviors were
reported to exist to some extent, and that none of them is
not within the potential behavioral repertoire of the child.
The standard deviations indicate that, within the limits of
the index, all the behaviors are reported to varying extents

by different mothers and children.

Validity and Reliability

 

Validity addresses the question of whether we are
measuring the phenomenon that we intended to measure or not.
Reliability addresses the question of whether measurements
render consistent results over time or internally, within a
measurement instrument. This last reliability estimate is
better known as internal consistency. In the instruments
there were three main sets of variable measurements: 1.
Parent and child responses to social situations in terms of
internal and external orientations; 2. Exposure to anti-
social television portrayals, in the areas of physical ag-
gression, verbal aggression, theft and deceit; and 3.
Antisocial predispositions on the part of the child for re-
solving social conflict situations. The questions of valid-
ity and reliability will be addressed for each set of

measurements separately.

63

Table 5. Means, Standard Deviations, Number of Cases

and Index Range of Mother and Child Indexes
of Reports of the Child's Antisocial Be-
haviors in Hypothetical-Situations.

 

Indexes of the

 

Mother's Report Y S N Index Range
Physical aggression 1.2 1.1 224 0-4
Verbal aggression 3.6 1.7 225 0-6
Theft 0.4 0.7 223 0-2
Deceit 1.2 1.5 223 0-6

Indexes of the
Child's Report

 

Physical aggression 7.4 5.1 220 0-18
verbal aggression 8.3 3.9 222 0-18
Theft 1.9 1.4 225 0-4

Deceit 2.4 2.2 221 0-10

 

64

Parent and Child Internal and External Orientations

 

a. Validity: As documented in the first chapter, several
studies in the past have found both an internal and an
external orientation. Through a factor analytic confirmr
atory solution, with the pretest data presented above, it
was found that the two hypothesized factors did appear with-
in the same type of instrument described for use in the
present study. The confirmatory factor analysis solution
is an indication of "factor validity" which can be inter-
preted as evidence of construct validity. Those items de-
signed to measure the constructs clustered together, accord—
ing to the indications of past research, some of which is
based on direct observations of behavior. Nunally (1967)
explains: "With content validity, factor analysis mainly is
important in suggesting ways to revise instruments for the
better. With construct validity, factor analysis provides
some of the tools that are most useful for determining
internal structures and cross-structures for sets of vari-
ables" (p. 101).

The ideal and most definitive validation of these
instruments would consist of correlating actual observations

of behavior with responses to the questionnaire items. However,

65
it was not possible to conduct such concurrent test valida-
tion given the resources available.

Convergent validity estimates were derived from items
administered independently to the mother and the child. The
mothers were asked to respond to one positive and one neg-
ative child situations, and the child was asked to respond
to one positive and one negative parental situations.

The mothers were asked to report how the child "would
feel or react if (HE/SHE) did any of these things," and the
child was asked to report "what your parents would do to you
if you did any of these things."

The two situations with the appr0priate wording for
mothers and children were the ones in which the child helps
"a friend in the neighborhood with some hard work," and the
one where the child gets mad and yells at his/her parents.
Tables 6 and 7 contain the situations and the response items
used for validating parental and child responses respectively.
These tables also contain the zero order correlation coef-
ficient between the mother and the child responses, the
number of cases, and the significance of the coefficient.

In Table 6 one can observe that when children are
asked to report about parental practices the covariation be-
tween the mother's responses and the child reports is gen-
erally very low. Only one statistically significant
correlation is found in the case of the positive situation,

and two in the case of the negative situation. The average

66

Table 6. Pearson Correlation Coefficients*Between Mother
and Child Items Concerning the Parent's Reactions
to a.Positive and a Negative Situation.

 

 

Suppose (NAME OF CHILD) helps a
friend in the neighborhood with
some hard work, and you (and your
husband) hear about it. What
would you (and your husband) do?

Say you are proud of (HIM/HER) .

Kiss or hug or pat (HIM/HER) on
the back . . . . . . . . . . .

Explain why it was a good thing
to do . . . . . . . . . . . .

Let (HIM/HER) watch extra TV . .

Give (HIM/HER) something special.

Suppose (NAME OF CHILD) gets mad
and yells at you (or your
husband). What would you (and
your husband) do?

Say you are disappointed in (HIM/
HER) o e o o o o e e o e e e

Yell at (HIM/HER) . . . . . .

Explain why (HE/SHE) shouldn't
behave that way . . . . . . . .

Hit, spank or shake (HIM/HER) .

Keep (HIM/HER) from watching TV .

._£_ .;E_
.04 227
.18 227
-.02 226
.01 226
.08 227

Average r

.01 226
.02 226
-.04 226
.27 226
.29 225

Average r

Signif-
icance

 

*
Zero-order

67
correlation for the positive situation was .07, and .12 for
the negative one.

In Table 7 the reader finds that the mothers seem to
perceive their children's reactions quite differently from
themselves. The average correlation for the positive situa-
tion is .05, and .15 for the negative situation.

The correlations, in general, are very low, despite
the statistically significant coefficients, and at this
point the only validation available is that forthcoming from
the factor validation replicating the existence of the in-
ternal and external dimensions. A separate section, below,
will deal with factor analytic results replicating the

emergence of the two dimensions.

b. Reliability: A measure of internal consistency for the
newly created indexes was obtained for all response items
corresponding to parental and child responses, and to the
positive and negative situations. Cronbach alpha as an ex-
tension of the Richardson Kuder formula 20 was utilized for
this purpose. Cronbach alpha is a summary measure of the
average correlation between all possible random halves of
the components of the scale.

Table 8 contains the alpha coefficients for the in-
dexes of parental responses to social situations. The
reader should recall that these items were only presented to
the mothers. With one exception, all the indexes seem to be

internally consistent to an acceptable degree. The only

68

HOUHOIOHON

.4.

 

mH.
Hoo.

.mofi

woo.

hoo.
.m.G
mo.
mo.

Hoo.

mo.

.m.c
.m.z
.m.:
.m.:
.m.c
.m.c
mo.

v
TOGMOHHHcmHm

H 0m0H0><

wmm

oNN
mmm

mmm
MNN
omm
mmm
oNN

mm. . . . . . . . . . . . . . m30nm >9 ouHHo>mm H50» noumz
50» umH uo: »mE mucmHme H50» umnu »HH03 50» UH503
no. . . . . . . . moHo 5o» umnz How e5 0x08 ou »Hu 50» 6H503
NH. . o o o o o o o o o o o o o o o o o o o «0.50% EOHW xmzm
mcHnuTEom mxmu »mE mucone H50» umnu »HH03 5o» oH5oz
0H. . . . mmucmHme H50» msHucHoeemmHo u5onm »HH03 5o» oH503
mo. . . . ~50» um HHo» »ME mucmHme H50» umsu »HH03 5o» oH503
NH. . . . . . . . . . . «Homm »0su 30: u50um »HH03 5o» oH503
HH. . o o o o o o o o o o o o o o o o ngflfigm Hmmm 50W Uflsoz
mN. ..oooooooocooococa-oooNSOHAmvfimgm

H 0m0H0><

hNN
mNN
»NN
ham
QNN
NNN
hNN

2

mo. . . .
oo. . . .
vo.u . . .
Ho. . . .
mo. . . .
mo. . . .
MH. . . .
IMI

H0 xcmem .an »mE mucone H50» umsu »HH03

50» oH503

.mucmHme H50» um HHw» can one umm 5o» mwoee5m

. . . . . . . NHHmmH5o» nuHB »eemn Hmmm

NHMHomem moHanEOm umm UH50nm 50» xCHnu
. . . . . . . . . . NETS“ HON %QQM£ Hmmm
. . . . . . . . . mHHmmH5o» H0 @50He on
. . . . . muH u5onm Socx ou oHeome ucm3
. . . . . . . . . . muH u5onm ooom Hmom

. . . . . . . mumoHu m 0>Homoo 50» Hoom

5o» ©H503
5o» oH503
5o» oH503
5o» UH503
50» 0H503
5o» oH5oz
50» oH503

.xHo3 onn 080m zqu coonHonanoc
ecu :H ocmHHm m eHmn 5o» omoee5m

 

coHum5uHm 0>Hummmz m can 0>HuHmoe m ou mcoHuommm m.oHHnU ecu
moHcHoocoo mEmuH oHHno one Honuoz c003u0mtmuc0H0Hmmoou coHuMHoHHou comHmme

 

.5 0H959

69

Table 8. Internal Consistency a Coefficients for the
Indexes of Parental Responses Reflecting the
Internal and External Orientations for Positive
and Negative Social Situations.

 

 

Item Index a

 

Positive Situations
Say you are proud of (HIM/HER) . . . . . . . . . .70

Tell (HIM/HER) to feel good about what (HE/SHE)
did 0 O O O O O O I O O O O O O O O O O O O O O .76

Kiss or hug or pat (HIM/HER) on the back . . . . .86
Explain why it was a good thing to do . . . . . . .78

Let (HIM/HER) do something (HE/SHE) wanted to
do very badly . . . . . . . . . . . . . . . . . .78

Let (HIM/HER) watch extra TV . . . . . . . . . . .86

Say that people appreciate the good things
(HE/SHE) does . . . . . . . . . . . . . . . . . .80

Give (HIM/HER) something special . . . . . . . . .84
Tell (HIM/HER) reasons why (HE/SHE) should

keep doing these things . . . . . . . . . . . . .85

Negative Situations

Don't talk to (HIM/HER) for a while . . . . . . . .00
Say you are disappointed in (HIM/HER) . . . . . . .66
Yell at (HIM/HER) . . . . . . . . . . . . . . . . .81
Explain why (HE/SHE) shouldn't behave that way . .67
Hit, spank, or shake (HIM/HER) . . . . . . . . . .84
Say (HIS/HER) behavior makes you feel bad . . . . .81
Keep (HIM/HER) from watching TV . . . . . . . . . .90
Don't let (HIM/HER) go out for a while . . . . . .87

Tell (HIM/HER) another way to solve
(HIS/HER) prOblem O O C O O C O I O O O O O O O O 71

 

70
index for which there is no internal consistency is the first
one for the negative situations (don't talk to him/her for a
while). That one item has been pointed out as a candidate
for exclusion from the final analysis, as indicated by its
descriptive statistics.

Table 9 presents the alpha coefficients for the in-
dexes of child responses to social situations for the intern-
al and external orientation dimensions. All indexes seem to
be internally consistent to varying degrees. The lowest co-
efficient in the table is .61 for "would you feel good about
it," and the highest is .88 for "would you worry that your
parents may not let you watch your favorite TV show?."

Measures of test-retest reliability were also obtain—
ed for both parental and child responses. All items dealing
with the internal and external orientations were included at
the beginning of the questionnaire or interview, and one of
the situations containing a subset of the items was repeated
at the end of the questionnaire or interview in order to
obtain an estimate of test-retest reliability.

Table 10, parts a and b contain the correlation co-
efficients between duplicate response items in the mother's
interviews and in the children's questionnaires. The social
situation in both cases is that in which the parents ask the
child to do something for them, and the child doesn't do it.
All the test-retest correlation coefficients are above .40,

and all are statistically significant. The average

Table 9.

71

Internal Consistency o Coefficients for the
Indexes of Children's Responses Reflecting
the Internal and External Orientations for
Positive and Negative Social Situations.

 

 

Item Indexes

 

Positive Situations
Would you feel you deserve a treat? . . .

WOuld you feel good about it? . . . . . .
W0uld you want pe0ple to know about it? .
W0uld you be proud of yourself? . . . . .
Would you feel happy for them? . . . . . .

Would you think you should get something
speCial? I O O O O O O O O O O O O O I 0

Would you feel happy with yourself? . . .

Negative Situations

Would you worry that your parents may hit,
spank or shake you? . . . . . . . . . .

W0uld you feel guilty? . . . . . . . . . .
Would you worry about how they feel? . . .

Would you worry that your parents may yell
at you? O O I O O O O O O O O O 0 O O I

Would you worry about disappointing your
parentS? O O O O O O O O O O O O O I O Q

Would you worry that your parents may take
something away from you? . . . . . . . .

Would you try to make up for what you did?

W0uld you worry that your parents may not
let you watch your favorite TV show? . .

.76
.61
.64
.78
.67

.76
.83

.81
.73
.76

.79

.82

.86
.82

.88

 

72
Table 10. a. Test-retest Reliability for Children's

 

 

 

Reactions.

Suppose your parents asked you to Signif-
do something for them, and you icance
don't do it. r* N p <
Would you worry that your parents

may hit, spank or shake you? . . . .65 226 .001
Would you feel guilty? . . . . . . . .57 225 .001
Would you worry about how they feel? .60 225 .001

Would you worry that your parents
may yell at you? . . . . . . . . . .49 225 .001

Would you try to make up for what
you did? . . . . . . . . . . . . . .66 224 .001

Average r = .60

b. Test-retest Reliability for Parent's
Reactions.

 

Suppose you (or your husband) asked
(NAME OF CHILD) to do something
for you, and (HE/SHE) doesn't do
it. What would you (and your
husband) do?

Don't talk to (HIM/HER) for a while . .41 225 .001
Yell at (HIM/HER) . . . . . . . . . . .60 223 .001
Explain why (HE/SHE) shouldn't

behave that way . . . . . . . . . . .52 222 .001
Hit, spank or shake (HIM/HER) . . . . .69 225 .001

Say (HIS/HER) behavior makes you
feel bad . . . . . . . . . . . . . .68 224 .001

Average r = .58

 

* O O O
Zero-order Pearson correlat1on coeff1c1ents.

73

test-retest correlation coefficient is .60 for both parental
and child responses. These measures were considered to be

reliable enough for further analysis.

Exposure to Antisocial Television Portrayals

 

a. Validity: Direct validation of the exposure measure
that was used has not been conducted. Bechtel, Achelpohl,
and Akers (1972) attempted such a task with direct observa-
tion of viewing behaviors of 20 families and a diary record
kept by the participants. Respondents in that study were
found to overestimate their exposure time.

No attempt was made to obtain exposure measures from
both mothers and children since children are the best source
of information with regard to the television programs they
watch. Besides, the mothers' interviewing schedule was al-

ready lengthy.

b. Reliability: A measure of test-retest reliability was
obtained for the indexes of exposure for each of the four
areas of concern in this research, namely, physical aggres-
sion, verbal aggression, theft and deceit. Identical indexes
to the ones created in this study were available from the
survey and content analysis conducted one year earlier.

Table 11 contains the correlation between the first
and the second year indexes of exposure to antisocial be-
haviors. All the correlations are statistically significant

(p j .001) and all of them are of the magnitude of about .60.

74

Table 11. Correlations*Between the Indexes of Exposure to
Antisocial Behaviors in the First and Second

 

 

Years.
Significance
Indexes of Exposure r N p 1
Physical Aggression .65 188 .001
Verbal Aggression .62 186 .001
Theft .56 199 .001
Deceit .62 190 .001

 

.i
Zero-order

Given the time lag of one year between one index and the
other, and given the fact that there are measures for only
two points in time, it is impossible to ascertain what part
of the variance left unexplained is due to unreliability
and what part is due to true change. However, the coeffi-
cients obtained are conservative in any case.

Given the composition of the indexes, the reliabil-
ity coefficients reported do not only refer to viewing re-
liability, but to show content stability as well. Concept-
ually, one can interpret the coefficients to mean the amount
of stability of exposure to the four types of behaviors of

interest.

Child's Antisocial Predispositions

a. Validity: Leifer and Roberts (l972)obtained a correla-
tion between similar items to the ones used here and a
teacher's rating of r=.33 and r=.49. A similar coefficient
was obtained by this writer with Latin American children.

In this study, the mothers and the children were asked to

75

report the antisocial behavioral tendencies of the child,
and similar measures obtained from the mother and the child
were correlated in order to obtain an estimate of convergent
validity. The indexes detailed in the section above were
correlated for each one of the subcategories of antisocial
behaviors that we are concerned with. The results of that
analysis are reported in Table 12. The correlations for
physical aggression and for theft were found to be statist-
ically significant and not very high but in line with the
coefficients typically encountered with this type of data.
The correlations for~verbal aggression and deceit were
found to be unexpectedly low. Given the low degree of
agreement between mother and child about the child's pre-
dispositions, it was decided that the data for mother and
child should be separately analyzed with regard to these

measures.

b. Reliability: Leifer and Roberts (1972 ) found a test-
retest reliability coefficient of r=.72 with measures similar
to this writerTL In order to obtain a test-retest reliabil-
ity coefficient for the child's antisocial predispositions

as reported by the mother and the child, indexes similar to
those created for the second year were created for the

first year data. Theft was absent from the first year data,
consequently no coefficients could be calculated for that

type of behavior.

76

Table 12. Correlations*Between Mother and Child Indexes of
Reports of the Child's Antisocial Behaviors in
Hypothetical Situations.

 

 

Significance
Index r N p 3
Physical Aggression .30 217 .001
Verbal Aggression .10 220 .067
Theft .23 221 .001
Deceit .10 217 .074

 

*Zero-order

Table 13, parts a and b, presents the correlations be-
tween the indexes of antisocial behaviors for the first
year with the ones for the second year. The reliability
coefficients for the children's data are generally higher
than for the mothers' report, with the exception of deceit.
For both mothers' and children's reports the correlations
are statistically significant, although not as substantial
as one might have expected. Here again, one does not know
how much of the unexplained variance is due to true change,
and what part is due to unreliability.

Table 14 presents the internal consistency coefficients
for the mothers' and children's reports of antisocial pre-
dispositions of the child. Since a minimum of three items
are necessary in order to compute any measure of internal
consistency, physical aggression and theft for the mothers'
report are excluded from the table. It can be seen that the

internal consistency for the children's indexes is generally

Table 13.

77

Test-retest Reliability for the Indexes of the

Children's Antisocial Predispositions in Hy-
Correlations*Between

pothetical Situations.
One Year Lag Indexes.

 

 

Significance
r N p i
a. Mothers' Report
Physical Aggression .52 223 .001
Verbal Aggression .51 221 .001
Deceit .43 221 .001
b. Children's Self-report
Physical Aggression .68 214 .001
Verbal Aggression .57 217 .001
Deceit .37 217 .001

 

1r
Zero—order

 

 

 

Table 14. Internal Consistency a Coefficients for the
Indexes of the Mother's Report and the Child's
Own Report of Antisocial Behaviors in Hypo-
thetical Situations. Year 2.
Index a
Mother
Verbal Aggression .50
Deceit .58
Child
Physical Aggression . .88
Verbal Aggression .81
Theft .61
Deceit .72
Note: Physical aggression and theft coefficients for the

mother's report were noncomputable due to the small

number of variables.

78
higher than for the mothers'. This difference, however,
may be due to the larger number of items available for the
children's indexes. All coefficients in Table 14 are of
the magnitude of .50 and over.

Qata Replicating the Emergence of the
Internal and External Orientations

 

 

As indicated earlier, mothers were asked to report
parental practices, and the children responded to the child
reaction items dealing with the internal and external dimen-
sions. Both parental and child items were submitted to a
principal axis factor analysis routine with varimax rotation,
limiting the extraction of factors to the number of two.

This was done in cider to try to replicate the findings of
the pretest and to further validate the existence of the
internal and external dimensions. The results of this fac-
tor analysis were discouraging. Two clear dimensions emerged,
but contrary to expectations, the children's responses
clustered on one factor and the mother's responses on the
other. Given this initial finding, the two sets of data for
mothers and children were factor-analyzed separately. Since
these were confirmatory factor analytic solutions, the ex-
traction of factors was limited to two with the two sets of
data.

Table 15 contains the results of the ppincipal axis fac-

 

tor-analytic solution for parental practices. The indexes were

 

ordered according to their position in the internal or in the

79

 

em. mH. . . . . . . . . . . . . . . . . . HmHowem OGH£HQEOm Ammm\ZH=v 0>H0

v9 0 HH 0 o 0 o e e 0 e o e 0 0 0 o e e o 0 0 0 a ”Huxm sou“; Ammm\2Hmv “.3
Ho. mm. . . . HHooo Hum> on on counts Ammm\mmv mcHeuoaom oo 1mmm\zems nod
mm 0 mH o o o e 0 o 0 0 e 0 0 0 o 0 o mHflFfi; M “OH “:0 cm Amm=\szv umfl U.- Goo
”m o N N o o o o 0 0 0 o 0 0 0 o I 0 O O 0 >8 @Gflnoums duo-Hm Ammm\zH=v “”0”
mm 0 0o 0 I. o 0 0 o o o 0 0 0 0 O I o 0 O O O I mmm=\=Hmv wxmgm Ho §xcmmm § “H”
mm 0 mo 0 II o o o o o e o 0 o o o o 0 o o o o o e o 0 sammm\szv “m HHmM
5H 0 v0 0 0 I o 0 0 o O O 0 O O O 0 O o mHnoflgz a HON Ammm\2Hmv O“ xnfim“ “a “on

 

:oHumucoHHo HmcHouxm

«H. om. modes» moon» mduoo dome oHsono 1mmm\mms Hes woodman 1mmm\zumc HHoe
HH. mm. . . . . . noon mumm\mmv mmoHnu ooom Tau oumHooHeem 5o» umnu »mm
no. me. . . . . . . . . . . on ou mcHnu 000m 0 nmz uH »33 :HmHexm
«H. mm. . . . . . . . . . . . . . xomn onu so Ammm\szv ume Ho 55: Ho mme
mo. em. . . . . . . . oHo Ammm\mmv umsz u5oum oooo Home ou Ammm\szv HHoa
$0. Nmo o o o o o o o o o o o o o o o o Ammm\szv Ho QQOHQ mHm 50% “mm
Ho.l me. . . . . . . .EanOHe Ammm\mHmv 0>H0m ou »03 Hosuocm Ammm\szv HHoa
hN. we. . . . . . . . . . own Home 50» nexus H0H>m£0n .mmm\mHmv »mm
no. he. . . . . . . . . . . »03 umnu 0>mn0n u. :oH5ozm Ammm\mmv »n3 :HmHexm
wm. we. . . . . . . . . . . . . o . .Ammm\SHmv :fl Uflflflflommmmflfl OHM 90% %mm

 

coHumuomHHo HmcHouoH

 

N Houoce H Houome souH

 

 

.mcoHuommm Hmucone .mocHomoq moxmocH oTNHm
Ionuoe»m ou mnHoH000¢ oonoHo .coHumuom xMEHHm> guHS NHHumz Houome HmeHocHHe .mH oHnma

80

external orientation. As can be seen the two expected di-
mensions appeared. The reader should be reminded that a
loading of .35 was considered to be the cut-off point.
Factor 1 underlies the internal orientation, and factor 2
comprises the external orientation of parental practices.
Only one item loaded about equally and very low on both fac-
tors, namely, "don't talk to him/her for a while." It al-
ready.has been noted that the mean and standard deviation
for that index was the lowest, and also that this index

had the lowest internal consistency coefficient. Apparently,
the withdrawal of attention is not a pervasive parental
technique neither does it accompany any consistent parental
strategy. This one index also had a very low mean in the
pretest. However it clearly loaded in the external dimen-
sion as hypothesized. In the pretest, the children provided
the information regarding both parental and child reactions,
and perhaps for children this technique of withdrawal of
attention or love, is perceived to be symptomatic of ex-
ternally oriented practices by their parents. However, when
the parents themselves are questioned, the technique is not
a part of the externally oriented syndrome.

In the factor-analytic solution, the internal dimen-
sion accounted for 28% of the total variance, and the ex-
ternal orientation accounted for 12%.

Table 16 presents the results of the factor analysis

of the child responses indexes. Here, once again, the

81

 

he. Hm.u . . . . . . . . mHmHoQO mcHnumfiom pom cHnonm so» xcHsu Dom oHsoz

MH. mo. . . . . . . . . . . . . . muH usonm zocx cu meomQ pan: :0» cHsoz
om. m~.n . . . . . . . . . . . . . . . mason» m m>ummmo so» Hmom so» oHsos
H“. OH. 0 o o O O o o o o o o o o o o O I O o O 0 O O “zonm >8 mufiuo>mw

 

use» nouns so» uwH no: hue mucmumm Mao» umsu muuoz so» cHsos

up. NH. msom Eoum muss manuoEom mxwu was mucwumm mac» uwnu >HH03 so» oHsoz
om. mm. . . . . . . . . «so» no HHmm hue mucmumm use» was» huuo3 so» cHsos
Hp. vH. . . mac» mxmnm Ho xcmmm .HHn hue mucmumm use» was» muuoz no» 6H503
COHumucwHuo Hmcumuxm

no. om. . . . . . . . . . . . . . . . NMHmmuaoa suHs mamas Hmmm no» oHsoz
mo. 00. . . . . o . . . . . . . . . . . . N505“ HON kmmfln wam 90% ©H903
0H. mm. . . . . . . . . . . . . . . . . . NMHGmHSO% HO USOHQ 0Q 90% OHSOS
No.l HF. . . . . . . . . . . . . . . . . . . Nfifl “Sonm @000 H60“ 50% DHDOS
MH. on. . . . . . . . . . . . . mcHn so» ums3 HON mu mme cu »Hu :0» UHsoz
so. om. . . . . . . . . mmucwnmm know mcHusHomech usonm muuo3 90> cHsoz
mo. Nm. . . . . . . . . . . . . . . . NHGOH %0£# 303 HDO£M hHHO3 50% OHDO3
No. cm. 0 o o o o o o o o o o o o o o o o o o o o o Nhuflflgm Hmmm 5.0% Mudflaoz

 

coHumucwHuo HmcuwucH

 

N Houomm H Houomm

ETUH

 

 

.mcoHuommm m.cmH©HH£U
Imnuommm ou mcHUHooo¢ commune .coHumuom meHum> nuHs

.mmchmoq mmxoccH cmNHm
xHuumz Houomm HmmHocHum .mH mHnma

82

clear emergence of the two dimensions is found. Factor 1
underlies the internal orientation and factor 2 reflects
the external one. One item loads about equally low on both
dimensions, "would you want peeple to know about it." This
finding is surprising since the same index did discriminate
between factors in the pretest, and in both the pretest and
in this study the children were the respondents for this
item. The children in this study were older than the chil-
dren in the pre-test. However, if age was the explanatory
factor for the discrepancy, other indexes should reflect the
discrepancies as well. At this point this inconsistency
escapes explanation.

The children's internal factor accounted for 33% of
the total variance, and the external dimension accounted
for 19%.

At this point the existence of the two internal and
external dimensions for both parental practices and children's
responses to social situations have been successfully repli-
cated. However, the replication was done for parental and
child indexes separately, since the solution including both
sets of data renders a parental set of practices and a child's
cluster of responses without discrimination between the two
orientations. The low convergent validity coefficients re-
ported below may explain to some extent the overriding co-
variation within the two separate sets of data. It is ap-

parent that mothers and children see things quite differently.

83

To represent the internal and external orientations
for parental practices and children's responses, four in-
dexes were created by summating those indexes that loaded
together in different factors. The two indexes that did
not clearly load in any one dimension were excluded.
Parental internal and external indexes were created as well
as children's internal and external indexes.

Table 17 contains the means, medians, standard devia-
tions, number of cases, index ranges and the internal con-
sistency alpha coefficients for the four indexes. From
examining this table it is clear that the external orienta-
tion is consistently lower than the internal. The variation
within indexes is proportionally higher, however, for the
external dimensions.

Table 17. Means, Medians, Standard Deviations, Number of
Cases, Index Range and Internal Consistency a

Coefficients for the Internal and External
Indexes of Parent-Child Modes of Interaction.

 

 

 

Index i' Median S N Index Range a
Child

Internal 47.6 50.5 11.6 216 0-64 .89
External 14.1 12.17 9.7 218 0-48 .79
Parents

Internal 68.6 72.92 11.8 217 0-80 .85

External 18.3 17.45 10.2 219 0-56 .72

 

84
The alpha coefficients of internal consistency indi-
cate that the indexes are sufficiently internally reliable
given that the lowest of them all was .72.
The four indexes were intercorrelated among them-

selves and the results are as follows:

Child Internal x Child External r=.l3 p<.05
Child Internal x Parent Internal r=.ll p<.06
Child External x Parent Internal r=.08 n.s.
Child External x Parent External r=.27 p<.001
Child Internal x Parent External r=.03 n.s.

Parent Internal x Parent External r=.36 p<.001

According to the literature reviewed in the first chap-
ter, besides expecting to find one factor for both parental
practices and child reactions for each of the internal and ex-
ternal orientations, it was to be expected that the internal
parental orientation should be associated with the internal
child orientation, and that the parental external orientation
should be associated with the external child orientation.

The correlation obtained between the two internal indexes was
low (r=.11, p=.053) and fell short of significance. However,

the correlation found between two external orientations was

85

.27 and statistically significant. These results seem to
indicate that either the perceptions of parents and children
differ substantially, or that internally oriented parents do
not necessarily contribute to the internalization of moral
values in their children. However, the more obvious types
of discipline encompassed by the external orientation do
seem to contribute to the child's focus on external factors
of social situations.

The correlations obtained between the internal and
external orientations for both the parental practices and
the children's reactions were .36 and .13 respectively, both
of them statistically significant. These findings seem to
indicate that to the extent that parents and children are
more internally oriented, they tend to also be more con-
cerned with external considerations. In the children's case,
the correlation, although statistically significant, is not
substantial in magnitude. In the parents case the coeffi-
cient is more substantial and it may indicate that parents
that are more concerned about their.chi1dren tend to do more
of everything, although it is noted that generally there is
stronger tendency towards being internally oriented.

Given the discrepancies obtained from the mothers'
and the children's data, it was decided that the analysis to
test the hypotheses presented in Chapter I should be done

separately for the mothers' and the children's data.

86

Analytical Methods

 

There were six main hypotheses to be tested, Ha - Hf,
and different analytical tools seemed most apprOpriate in
each case.

Hypotheses Ha - HC were tested by means of multiple
regression analysis. Multiple regression analysis renders
regression coefficients that, when standardized, indicate
the magnitude of the unique contribution of one standard
deviation of one of the independent variables on the depend-
ent variable when holding all other factors constant. This
form of analysis will also supply us with a multiple correla-
tion coefficient that, when squared, indicates the amount of
variability in the dependent variable that is accounted for
by the set of independent variables. The interactions among
variables can also be ascertained by this method when enter-
ing the interaction terms as independent variables in the
regression equation.

To complement the regression analysis, a multiple
analysis of variance was conducted to obtain contingency
tables where the magnitude of individual cell means could be
visually inspected. The results from both forms of analysis
were expected to be very similar since both forms are de-
rived from one general analytic model (Kerlinger and Pedhazur,
1973, pp. 6-9).

Hd, was tested by means of Student t-tests.

87

Finally, Hypotheses He and Hf were tested by means
of contingent correlations. A contingent correlation is a
conditional statement that separates those cases that fall
in certain categories for which a Pearson Product Moment
correlation is computed.

All analyses were conducted in a CDC 6500 computer
system, with programs of the Statistical Package for the
Social Sciences, at the computer facilities of Michigan

State University.

Summary

First, the results of a preliminary study pretesting
a set of instruments intended to tap two dimensions of
parent and child modes of response were introduced.

The sample utilized in this study was described and
the descriptive statistics of the measures used in this re-
search project were presented. Three main sets of variables
were considered: a. Internal and external orientations of
parents and children; b. Exposure to antisocial television
portrayals; and c. Antisocial behavioral predispositions of
children. Validity and reliability estimates for the meas-
ures utilized were obtained and discussed.

The existence of an internal and an external orienta-
tion of parental and children's reactions to social situa-
tions was replicated, and it was found that althought the
two dimensions emerged for parents and children separately,

they did not appear when the data for parents and children

88
were jointly analyzed. It was decided that the hypotheses
of this study would be tested for the parental and the
children's data separately.
Finally, the analytical tools utilized for testing

the hypotheses of this research were presented and

described.

CHAPTER III

RESULTS

The results of this study will be presented in three
general sections, according to the order in which the hy-
potheses were originally introduced. The first section pre-
sents the results dealing with the main effects of the
internal and external orientations, and exposure to anti-
social television portrayals on the child's antisocial pre-
dispositions. Here, the interactions are explored.
Secondly, the hypothesized contrasts among different cells
resulting from the intersections of exposure to television
antisocial portrayals and the internal and external orien-
tations on the child's antisocial predispositions, are pre-
sented. The last section comprises the results dealing with
contingent correlations between exposure to antisocial por-
trayals on television and the child's antisocial predispo-
sitions at different intersections of the internal and
external dimensions of parent-child orientations.

Within each section the internal and external parental
and child modes of interaction are separately considered.
Also, the four different categories of antisocial portrayals
and children's predispositions are separately analyzed,
namely, physical aggression, verbal aggression, theft, and

deceit.
89

90

Analysis of Variance and Regression: Ha, Hb, and Hc

 

In this section, Hypotheses a, b, and c are tested
by means of regression analysis and analysis of variance.
The hypotheses were worded as follows:

H : The more exposure to antisocial
portrayals on television, the more

antisocial behavioral predisposi-
tions will be displayed by children.

The more externally oriented the
parents and the children, the more
antisocial behavioral predisposi—
tions will be displayed by children.
H : The more internally oriented the
parents and the children, the less
antisocial behavioral predisposi-
tions will be displayed by children.

All three hypotheses will be simultaneously tested by
both analytic methods, regression and analysis of variance,
for each of the sets of data we have, the mothers and chil-
dren, and for each category of antisocial behaviors of
interest in this study: physical aggression, verbal aggres-
sion, theft and deceit.

Since there are three main effects that are being hy-
pothesized, there are three possible two—way interactions,
and one possible three-way interaction. In the analyses of
variance the interactions are dealt with by the traditionally
accepted methods (Overall and Klett, 1972, pp. 448-449). In
the regression analyses, however, the interaction terms are

constructed as multiplicative terms of the different combina-

tions of the independent variables. These multiplicative

91

terms, by including the independent variables in themselves,
are highly correlated with the independent variables. This
collinearity presents a problem in determining the magnitude
of regression coefficients.* In order to avoid collinearity
the interaction or multiplicative terms are standardized, and
through such standardization procedure the correlations be-
tween the independent variables and the interaction terms are
attenuated (Pink and Noell, 1972). The resulting regression
coefficients for the interactions are directly interpretable.

Now, the test of Hypotheses a, b, and c will be presented.

a. Physical Aggression

 

As the reader will recall, the parental internal and ex-
ternal orientations were reported by the mother, and the child's
internal and external orientations were reported by the child.

Table 18a indicates that exposure to physical aggres-
sion on television and the external orientation of the parents
have a relatively strong effect on the children's aggressive
predispositions, when the interaction terms and the internal
orientation are held constant, as evidenced by the partial
standardized regression coefficients (B).

The same trend is observed when one looks at the zero
order correlation coefficients. The interaction term of ex-

posure by the external dimension also shows a statistically

 

*

If extreme collinearity exists, it may be impossible
to invert the correlation matrix of the independent vari-
ables, and if regression coefficients are obtainable they
tend to be unstable from sample to sample. See Kerlinger
and Pedhazur (1973).

Table 18.

92

Parental Orientations and Child's Exposure on the
Mothers' Report of the Child's Physically
Aggressive Predispositions: Regression and
Analysis of Variance Results-

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

a. Regression B r
Exposure .29** .31**
External .20* .28**
Exposure x Internal x External -.15 .01
Exposure x External .14 .16*
Exposure x Internal -.13 .03
External x Internal .07 .09
Internal -.02 .01
R = .41** R2 = .17
b. Analysis of Variance INTERNAL
LO HI
External External
LO HI LO HI
LO §1=.83 §2=1.os §5=.54 ‘ §6=l.23
Exposure N=35 N= 17 N=24 N=22
HI §3=1.50 '§4=2.00 §7=.77 §8=l.70
N=26 N=21 N=l7 N=34
OVERALL MEANS
‘Tfiange0;1)
EXPOSURE** EXTERNAL** INTERNAL
HI LO HI LO HI LO
1.55 0.89 1.54 0.92 1.14 1.29
*
p 5 .05
**<
P _ .001
Note: No interactions were significant

93
significant correlation coefficient which is not relevant
since that correlation is still the result of the attenuated
original correlation between the external orientation and
exposure.

Table 18b which presents cell means for the dichot-
omized independent variables and significance of the analy-
sis of variance test, provides identical results.

The multiple correlation in Table 18a when all the
variables are correlated with the dependent variable is
statistically significant, and the coefficient of determina-
tion (R2) indicates that all the regressed variables jointly
contribute 17% to the variation of the child's physically
aggressive predispositions.

Hypotheses a and b are supported and Hypothesis c
is rejected in favor of the null, since the internal parent-
al orientation does not appear to diminish the children's
favorable predispositions towards physical aggression. None
of the interaction terms exhibit a significant regression co-
efficient, nor were the interactions found to be significant
in the analysis of variance.

When we turn to the children's data in Table 19a and
b we find that both forms of analyses render identical results.
The internal dimension is overwhelmingly the most critical
factor in predicting the children's aggressive predisposi-
tions. Exposure and the external child orientation are also

found to be significant contributors, and none of the

94

Table 19. Children's Orientations and Exposure on the Self-
report of Physically Aggressive Predispositions:
Regression and Analysis of Variance.

 

 

 

 

 

 

a. Regression B r
Internal -.49** -.45**
Exposure .19* .23**
External .15* .14*
External x Internal -.10 .05
External x Internal x Exposure -.03 -.13
Exposure x External .02 .09
Exposure x Internal .01 .09

R=.53** R = .29

 

 

 

b. Analysis of Variance INTERNAL
LO HI
External External
LO HI LO HI
LO §1=7.55 §2=10.73 §s=3.79 §6=5.22
Exposure N=30 N=11 N=28 N=27
HI ‘§3=9.06 '§9=11.63 '§7=5.7o ‘§8=6.17
N=l8 N=32 N=23 N=24

 

 

 

 

 

 

OVERALL MEANS
(Range 0-18)

 

 

EXPOSURE* EXTERNAL* INTERNAL**
HI LO HI LO HI LO
8.39 6.16 8.29 6.32 5.16 9.66
*
p': .05
**
p i .001

Note: No interactions were significant

95

interactions are statistically significant. The contribu-
tions of all three variables are in the expected direction,
and it was decided to reject the null hypotheses in

all three cases, for Ha, Hb, and Hc‘ ‘It is observed that
all three independent variables and interaction terms con-
tribute 29% of the variation in the dependent variable, the
principal contributors being the main effects.

It is interesting to note that while the external
dimension was significant and the largest predictor with the
mother's data, the internal orientation was a substantial
and significant predictor in the child's case. With the
mother's data the internal orientation was unimportant, and
with the Child's data the external orientation contributed
a smaller amount to the child's physically aggressive pre-
dispositions. Exposure to television physical aggression

was about equally important in both sets of data.

b. verbal Aggression

 

Table 20a and b present the results regarding the
parental orientations and the mother's report of the child's
verbally aggressive predispositions. Again, one finds that
both forms of analyses give the same results, and that these
results are very similar to those for physical aggression.
The external orientation and exposure to TV verbal aggression
are the most important predictors of the child's verbally
aggressive predispositions. Neither the interaction terms

nor the internal orientation were found to be significant or

96

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Table 20. Parental Orientations and Child's Exposure on the
Mothers' Report of the Child's Verbally Ag-
ressive Predispositions: Regression and
AnaIysis of Variance Results.

a. Regression B r
External .20* .24**
Exposure .17* .23**
Exposure x Internal -.06 -.07
Internal -.08 .06
Internal x External -.07 -.08
Exposure x External .04 .05
Exposure x External x Internal -.01 .10

R = .31* R2 = .09
b. Analysxs of Variance INTERNAL
LO HI
External External
LO HI LO HI
LO '§1=3.17 ‘§2=3.81 ‘§5=3.12 §6=3.58
Exposure N=35 N=16 N=26 N=24
HI x3=3.81 x4=4.26 x7=3.13 x8=4.29
N=26 N=23 N=15 N=31
OVERALL MEANS
(Range’046)

EXPOSURE* EXTERNAL* INTERNAL

HI L0 HI LO HI LO

3.97 3.36 4.02 3.31 3.62 3.69

*
p 5 .05
**
p 5 .001
Note: No interactions were significant

97
substantial. Nine percent of the variance in the dependent
variable was accounted for by total set of predictors.

Hypotheses a and b were supported by these data, and
Hypothesis c was rejected in favor of the null.

The results of the analyses that tested the three
hypotheses with the children's data are presented in Table
21a and b. The children's internal orientation shows again
to be the most potent negative predictor of the child's pre-
dispositions, verbal aggression in this case. Exposure is
again found to be important, and the external orientation
has a contribution of zero. None of the interactions turned
out to be significant or important in any sense.

With the children's data it was decided to reject the
null hypotheses in the cases of Ha and Hc, but it wasn't
possible to reject the null in the case of Hb' All the pre-
dictors in the system accounted for 12% of the variance in

the children's verbally aggressive predispositions.

c. Theft

 

Table 22a and b presents the results for the mothers'
data regarding theft, for Hypotheses a, b, and c.

The analysis of variance results indicate that none
of the main effects or interactions were statistically sig-
nificant. The regression analysis results are almost iden-
tical except for the regression coefficient for the interaction
of exposure with the external orientation. This discrepancy

between both forms of analyses can be readily explained due

 

98

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Table 21. Children's Orientations and EXPOSUFe on the Self-
repOrt'of—Verbally Aggressive Predisp051tions:
Regression and Analysis of Variance.

a. Regression B r
Internal —.26** -.29**
Exposure .19* .18*
Exposure x Internal .05 .14*
Exposure x Internal x External -.06 -.l4*
Internal x External -.05 .06
Exposure x External .04 .07
External -.01 .01

a = .35** R2 = .12
b. AnalySis of Variance INTERNAL
LO HI
External External
LO HI LO HI
LO §l=8.70 §2=8.77 §5=6.77 §6=6.96
Exposure N=27 N=13 N=35 N=23
HI §3=9.33 SE4=9.9 SE7=9.41 328:7.68
N=21 N=30 N=l7 N=28
OVERALL MEANS
(Range 0-18)

EXPOSURE* EXTERNAL INTERNAL“

HI LO HI LO HI LO

9.05 7.61 8.37 8.28 7.50 9.26

t
p S. .05
**
Note: No interactions were significant

99

Table 22. Parental Orientations and Child's Exposure on the
HathngE Report of the Child's Predispositions
Towards Theft: Regression and Analysis of
Variance Results.

 

 

 

a. Regression B r
Exposure x External .17* .17*
External .14 .14*
Internal -.13 -.10
Exposure x Internal -.14 -.01
Exposure .07 .09
Internal x External .05 .10
Exposure x External x Internal -.05 -.01

2

R = .27 R = .08

b. Analysis of Variance

 

 

 

 

 

 

 

 

 

 

INTERNAL
LO HI
External External
LO HI LO HI
L0 351=.51 352:.47 355:.28 §6=.44
Exposure N=39 N=15 N=25 N=18
HI §3=.44 §4=.54 §7=.20 §8=.43
N=23 N=24 N=15 N=37
OVERALL MEANS
(Range 0-2)
EXPOSURE EXTERNAL INTERNAL
HI LO HI LO HI LO
0.42 0.43 0.47 0.39 0.36 0.50

 

p‘: .05
'k

p i .001
ote: No interactions were significant with ANOVA

*

N

100
to the elimination of variation in the independent variables
resulting from their dichotomization into high and low. The
significant interaction means that although none of the com-
ponents, exposure or the external orientation, contribute to
favorable predispositions towards theft in children, the
joint action of both variables does seem to have an impact.
Originally the writer did not hypothesize this two way
interaction effect, and will not elaborate further on it,
except to recognize its interest as a post hoc finding.

In this instance, Hypotheses a, b, and c are rejected
in favor of the null. The inconsistency of this set of find-
ings with regard to the pattern that has been identified with
physical and verbal aggression can be explained in terms of
the restricted variability of the dependent variable. The
mothers were only presented with one item regarding their
children's predispositions towards theft, and the range of
the scale was extremely limited as can be seen in Table 22b.

The children's data in Table 23a and b are more con-
sistent with the patterns of results identified above. The
internal orientation on the part of the child is the most
powerful negative predictor of the child's predispositions
towards theft. Neither the external orientation or the
measure of exposure were found to predict the dependent vari-
able to any extent. In the regression analysis, it was

found that there was a small significant effect of the

101

Table 23. Children's Orientations and Exposure on the Self-
report of Predispositions Towards Theft:
Regression and Analysis of Variance Results.

 

 

 

 

 

a. Regression B r
Internal -.33** -.33**
Internal x External -.l7* -.07
Exposure x External x Internal -.09 -.l6*
External -.08 -.13*
Exposure .03 -.01
External x Exposure .01 .01
Internal x Exposure .01 .03

R = .38** 32 = .15

 

 

 

b. Analysis of Variance INTERNAL
LO HI
External External
LO HI LO HI
LO §1=2.07 §2=2.39 §5=1.59 ‘§6=1.23
Exposure N=29 N=13 N=32 N=22
HI '§3=2.oo ‘§4=2.16 '§7=1.81 ‘§8=1.3§
I N=20 N=32 . N=21 N=30

 

 

 

 

 

 

OVERALL MEANS

 

(Range 0—4)
EXPOSURE EXTERNAL INTERNAL**
HI LO HI LO HI LO
1.82 1.76 1.72 1.85 1.44 2.13

 

.3
-<_ .001
: No interactions were significant with ANOVA

102

interaction of the internal and the external orientations,
and this constitutes another post hoc finding that is pre-
sented at face value. This interaction term indicates that
the unique effect due to the joint action of both orienta-
tions depresses the likelihood of the child's favorable
attitudes towards theft. The larger range of this scale
makes the results to be trusted a little more than the ones
for the mother's data, however they must be treated with
caution.

Hypothesis c is supported and Hypotheses a and b are
rejected in favor of the null.

Fifteen percent of the variance in the dependent vari-
able is accounted for by all the independent variables,
being the internal dimension the most important of all of
them.

At this point it is important to emphasize that no
correlation between exposure to television theft and the
children's predispositions towards theft was found either
with the mothers' or the children's data. Further exams
ination of the data dealing with theft would be useless
since the original focus of this study is the analysis of
parental mediation of children's learning from antisocial
television portrayals. Since no original correlation was
found no mediation can be investigated. Consequently the

rest of the hypotheses to be tested will not include theft.

103

d. Deceit

 

The results regarding deceit for both the mothers'
and children's data are in Tables 24 and 25.

Both regression and analysis of variance have render-
ed identical results in this case, with one exception. The
mothers' data in Table 24a and b show that exposure and the
external orientation are significant predictors of the
child's favorable predispositions towards deceit. The in-
ternal dimension was found to be a significant predictor
with the regression analysis, but not so with the analysis
of variance. The reason for the discrepancy can one more
time be explained by the restriction of variability due to
the dichotomization of the independent variables in the
analysis of variance approach. Given this restriction of
variability we can confidently Opt for placing more weight
on the regression analysis results and decide to reject the
null hypotheses in favor of the research Hypotheses a, b,
and c with the mothers' data. The complete set of independ-
ent variables accounted for 18 percent of the variance in
the dependent variable, and no interactions were found to
be significant.

When one considers the results for the children's
data, we find that one more time the internal dimension is
a very powerful negative predictor by itself, and that no
other main effects or interactions are statistically signif-

icant. These results are in Table 25a and b. Taking into

Table 24.

104

Parental Orientations and Child's Exposure on the
Mother's Report of the Child's Predispositions

: Regression and Analysis of
Variance Results.

Towards

Deceit

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

a. Regression B r
External .30** .28**
Internal -.29** -.08
Exposure .21* .28**
External x Internal -.12 -.02
Exposure x External .07 .09
Exposure x Internal -.06 -.04
Exposure x Internal x External .01 .08

R = .42** r2 =.18
b. Analysis of Variance INTERNAL
LO HI
External External
LO HI LO HI
LO §1=.81 §2=1.31 §5=.64 §6=1.09
Exposure N=37 N=16 N=25 N=22
HI x3=l.39 x4=l.65 x781.00 x8=l.61
N=23 N=23 N=16 N=33
OVERALL MEANS
(Range 0-6)

EXPOSURE* EXTERNAL* INTERNAL

HI 1 L0 HI LO HI LO

1.46 0.91 1.45 0.93 1.14 1.22

*
p i .05

**p < .001

Note? No interactions were significant with ANOVA

Table 25.

105

Children's Orientations and Exposure on the Self-

 

report of Predispositions Towards Deceit:
Regression and Analysis of Variance Results.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

a. Regression B r
Internal -.40** -.36**
Exposure .12 .15?
Internal x External -.12 .00
External .09 .06
Exposure x External x Internal -.02 —.11
Exposure x Internal .02 .08
Exposure x External -.00 .05

R = .41** R2 = .17
b. Analysis of Variance INTERNAL
LO HI
External External
L0 HI LO HI
LO §1=2.41 §2=2.54 §s=1.53 §6=1.91
Exposure N=29 N=13 N=32 N=23
HI x3=3.00 x4=3.97 x7=l.90 x8=1.79
N=l9 N=29 N=;9 N=29
OVERALL MEANS
(Range 0-10)

EXPOSURE EXTERNAL INTERNAL**

HI LO HI LO HI LO

2.71 2.02 2.60 2.14 1.76 3.06

*
p §_.05
** <
p‘_ .001
Note: No interactions were significant

106

consideration all the independent variables, the amount of
variance explained in the child's favorable predispositions
towards deceit is 17%. Hypothesis c is confirmed and Hy-
potheses a and b are rejected by the results obtained.

To summarize, two main patterns of results have been
found, one with the mothers' and another with the children's
data. In the mothers' case it was consistently found that
television exposure to antisocial behaviors and the parent's
external orientation towards the child's behavior are the
most consistent predictors of the child's antisocial pre-
dispositions.

In the children's case, it was found that exposure
to antisocial behaviors in general is also a consistent pre-
dictor of the child's antisocial predispositions, but not
as powerful as the internal orientation of the child with
regard to social situations.

The internal orientation on the parents part was not
found to always significantly predict the children's anti-
social predispositions, in the sense of reducing them. With
the children's data the external orientation of the child
was not found to always enhance the antisocial predisposi-
tions of the child.

The analyses of variance have given us the opportun-
ity to inspect the means of the dependent variable at differ-

ent intersections of the dichotomized independent variables.

107
In Hypothesis d specific predictions were made regard-
ing the differences or similarities among the cells in those
tables, and will now proceed to present the results of those

hypothesized comparisons.

T-tests: Hd

 

In order to check for the specific effects of differ-
ent values of the independent variables on the dependent
variable,a set of one-tailed t-tests was conducted. The
reference for the specific means being tested are the tables
in the above section that deal with the analyses of variance
results. A special notation will be utilized for indicating
the statistical significance of the comparisons hypothesized
and this notation will be exemplified with the original

hypothesis. It was stated and expected that:
§4>SE3 = §2>§1>§8>§7 = SE6>SES

If the hypothesis were completely supported with a set
of data it would look exactly as the original statement of
the hypothesis except that it would have the actual means
between the symbols larger than (>) or equal (=). Whenever
a symbol ">" is placed between two means the difference
between them is at a probability level of less than .05.

In the cases of hypothesized equalities, in order to avoid
a type I error, the probability level is established at .5

two-tailed, and only in those occasions in which the

108
statistical significance level of "t" is larger than .5 it
is said that the equality is confirmed. Both symbols "> "
or "==" will only be used for hypothesized differences or
equalities.

At this point the reader should be reminded that
according to the theory in_the first chapter, there are
reasonstxnexpect that i4 should be the highest mean and E5
should be the lowest. Clearly, those children who themselves
or their parents are highly externally oriented and who are
highly exposed to television antisocial portrayals, and who
themselves or their parents are low in their internal orien-
tation should be the children most likely to be antisocially
predisposed. However, those children who are not highly
exposed to television antisocial behaviors, who themselves
or their parents are highly internally oriented, and low in
their external orientation, should be the children least
likely to be antisocially predisposed.

The above relation is expressed in Hypothesis d, and
the rest of the comparisons are derived from that expectation
giving equal weights to exposure and the external and in-
ternal orientations, the external orientation being con-
sidered a facilitator and the internal an inhibitor of anti-
social predispositions. With these considerations in mind
we proceed to present the results of comparisons between

specific means.

109
a. Physical Aggression
Below are the results for Hypothesis d regarding
physical aggression for the mothers' and the children's

data.

Hd: §4>§3=§2>§1>§8>§7=§6>§5

Mothers: 2.0 1.5 1.6 0.83 1.7 > 0.77 1.23> 0.54
Children: 11.63> 9.06 10.73> 7.53 6.17 5.7 = 5.22 3.79

The pattern of the means for the mothers' data con-
firms the expectations for the comparisons comprising §4-§1,
but not so for the comparisons for the second half of the
hypothesized order of means. There are no statistically
significant differences between the pairs from E4 to El, and
there are two significant comparisons, as postulated, in the
second part of the expression.

The comparison between £4 and 55 was found to be sta-
tistically significant (t=4.55, df=30.25, p < .001 with a
separate variance estimate) in the expected direction, and
in general we can conclude that Hd is partially supported
with the mothers' data.

The general pattern of the means is consistent with
the results presented in Table 18a and b. In that table it
was found that exposure and the external orientation had
significant main effects but not so the internal orientation,
that is why we find that i8 is larger than £1 (t=-3.38, p <

.001) in the direction Opposite to that hypothesized.

110
The means corresponding to the children's data are all
in the expected direction. Three of the comparisons were
statistically significant. The comparison between §4 and i5

is also significant (t=8.08, df=58, p < .001) in the predicted

direction. With the children's data Hypothesis d can be said

to be partially supported as was the case with the mothers' data.

b. Verbal Aggression

 

The results pertaining to the mothers' and the chil-
dren's data with regard to verbal aggression are presented

below:

H §4>§3=§2>§1>§3>§7=§6>§5

d:
Mothers: 4.26 3.81 = 3.81 3.17 4.29> 3.13 3.58 3.12
Children: 9.9 9.33 = 8.77 8.70 7.68>’9.91 6.96 6.77

With the mothers' data the means are generally in the
direction predicted. The equality between i3 and i2 is sig-
nificant below the probability level established of .5, and
the other hypothesized equality was not statistically signif-
icant. Only one of the hypothesized comparisons was signifi-
cant, and one post hoc comparison was found to be significant
in the opposite direction to that hypothesized (t= -2.86,
df=64, p i .006, two-tailed test) for 3'61 and £8. The differ-
ence between E4 and x5 is one more time found to be signifi-
cant (t=2.43, df=47, p:<.02). Hypothesis d regarding verbal
aggression, with the mothers' data can be said to be partially

supported.

111

The means for the children's data are all in the pre-
dicted direction with one exception, that is, there was a
post hoc significant difference between E7 and i6, where an
equality was initially expected (t=2.13, df=38, p < .05,
two-tailed test). The equality between i3 and i2 was found
to be statistically significant, and the hypothesized dif-
ference between R8 and i7 was significant in the predicted
direction. The difference between i4 and is, the two ex-
tremes of Hypothesis d was significant (t=3.77, df=63,
p ‘<.001), and Hypothesis d is supported in part one more

time.

c. Deceit
The results dealing with pred18positions towards

deceit are:

Hd: §4>§3=§2>§1>§8>§7=§6>§5

Mothers: 1.65 1.39 = 1.31 0.81 1.61 1.0 = 1.09 0.64
Children: 3.97 3.0 = 2.54 2.41 1.79 1.9 8 1.91 1.53

The mothers' data follow the same pattern found for
physical and verbal aggression. One more time i8 was found
to be larger than 321 (t= -2.55, df=68, p < .02, two-tailed
test) in a post hoc fashion. The two hypothesized equalities
were corroborated, and the difference between i4 and i5 was
found to be significant (t=2.59, df=36.05, p < .02, with a
separate variance estimate). These results one more time
follow the main effects encountered in Table 24a and b, that
is, no main effect was found for the parental internal orien-

tation. We can say that Hypothesis d received partial support

112

with the mothers' data in the case of deceit.

The children's data exhibit the hypothesized pattern
of means, but only the equalities found statistical support.
The difference between x4 and §5 is statistically significant
in the hypothesized direction (t=4.9l, df=46.21, p <.001 ,
with a separate variance estimate). In this case we can say
that Hypothesis d follows the pattern identified for physical

aggression and verbal aggression, and is partially supported.

Contingent Correlations: He and Hf

 

The internal and external dimensions of parent-child
orientations were divided at the median* in order to obtain
the following four cells, each of which contains a zero order

correlation coefficient.
INTERNAL ORIENTATION

6 LOW HIGH
g r1 r2
0

 

 

r3 r4

 

 

 

HIGH LOW

 

The correlation coefficient in each of the cells is be-
tween one of the types of exposure to antisocial behaviors on
television, and the same type of antisocial predisposition in

the child.

 

*

Three different breakdowns also were considered: the
dimensions were subdivided into three equal segments and then
two extreme combinations were produced. Whatever the break-
downs the results rendered were similar. Consequently the
partition at the median was chosen as the least arbitrary for
presentation. ‘

113

As indicated in the first chapter, if rA is the over-
all correlation between a type of specific television expo-
sure and the same type of antisocial predisposition in the

child, it was then hypothesized:

He: r2<rA<r3 and
Hf: r2_<_rl:r3

r2:r4:r3

Z tests for difference between independent correla-
tions were used in order to assess the statistical signifi-
cance of the differences between pairs of correlations

(Bruning and Kintz, 1968, pp. 191-192).

a. Physical Aggression

 

First of all, the overall zero order correlations between
exposure to TV physical aggression and the child's and mother's
reports of the child's physically aggressive predispositions

were:
Child's Physically Aggressive
Predispositions

I ' I
Exposure to TV Mother 8 Report Child 8 Repgrt

Physical Aggression r = .31 r = .23
N = 210 N = 206
p 3.001 p 1.001

The reader should be reminded that the measure of ex-
posure to television was obtained only from the child.

Table 26 presents the contingent zero order correlations

 

obtained at different intersections of the internal and external

114

Table 26. Contingent Correlations Between Exposure and

the Child's Physical Aggression as

 

Reported by

Mother and Child at fiifferent Intersections*
of the Internal and External Orientations of

Parents and Their Children.

 

a. Internal and external parental orientations.

 

 

 

 

 

 

 

 

Correlations of exposure with the mother's report
—of the child's physical aggression.
INTERNAL

LOW HIGH
3”; r = .27 r = .19
T LOW N = 57 N = 45

S = .02 S = .12
E
g r = .44 r = .23
A HIGH N = 36 N = 58

S = .004 S = .04
L
b. Internal and external child orientations.

Correlations of exposure with the child's own
report of physical aggression.
INTERNAL

LOW HIGH
:3 r = .05 r = .29
T LOW N = 48 N = 51

S = .37 S = .02
E
S r = .19 r = .03
A HIGH N = 44 N = 50

 

 

 

 

 

*
When the two dimensions are divided at the median

115

parental orientations. Each of the dimensions was parti-
tioned at the median. The correlations reported include the
measures of the child's predispositions that correspond to
the orientations utilized, be them the parents or the child's.
The portion a_of the table always refers to parental orien-
tations and the child's predispositions as reported by the
mother. The portion h_of the table uses the child's internal
and external orientations and his/her own report of predis-
positions.

Table 26a, at first glance seems to corroborate the
original expectations, as stated in Hypotheses e and f. The
correlation at the low external, high internal intersection
is lower than the overall correlation and the one at the
high external, low internal cell. Also, the correlations
at the high-high and low-low cells are in between the other
two correlations.

The Z value for p‘: .05 is 1.64 in the case of one-
tailed tests. When comparing the 2 value of the difference
between pairs of correlations in Table 26a, and between the
correlations in the table and the overall, with 1.64 it was
found that none of them was higher, and consequently none of
the comparisons were found to be statistically significant
at the probability level of .05.

Although for the individuals in the sample Hypotheses
e and f are not rejected, inferentially they are. All the

differences among correlations were found to be in the expected

direction in the sample, but they were not statistically

116
significant.

When considering the children's data in Table 26b one
finds that none of the expectations are fulfilled. The
pattern of correlations is reversed, and the inferential
significance tests are all insignificant. Hypotheses e and
f are rejected in favor of the null both for the sample cases

and for the population these cases represent.

b. Verbal Aggression

 

The overall zero order correlations between exposure to
Tvrand verbal aggression and theschild's predispositions to be

verbally aggressive were:

Child's verbally Aggressive Predispositions

 

 

Mother's Report Child's Report
Exposure to TV
verbal Aggession r = .23 r = .18
N = 210 N = 207
p 5.001 p 3.005

Table 27 presents the correlations at different inter-
sections of the internal and external dimensions for both
parents and children.

For the parental orientations and the mother's report
of the children's verbally aggressive predispositions, Table
27a contains the correlation coefficients obtained.

The patterns of correlations in Table 29a is close to
the expectations, except that neither the overall correlation,
neither the correlations at the high-high or low-low inter-

sections fall between the high internal-low external and the

117

Table 27. Contingent Correlations Between Exposure and
the Child's Verbal Ag ression as Reported by
Mother and Child, at Different Intersections*
of the Internal and External Orientations of
Parents and Their Children.

 

 

a. Internal and external parental orientations.
Correlations of exposure with the mother's
report of the child's verbal aggression.

 

 

 

 

 

 

 

 

INTERNAL
LOW HIGH
i r = .26 r = -.09
T LOW N = 57 N = 45
S = .03 S = .27
E
E r = .15 r = .26
A HIGH N = 37 N = 57
L S = .20 S = .03
b. Internal and external child orientations.
Correlations of exposure with the child's
own report of verbal aggression.
INTERNAL
LOW HIGH
f: r = -.09 r = .36
T LOW N = 48 N = 52
E S = .27 S = .004
S r = .14 r = .13
A HIGH N = 44 N = 50
L S = .19 S = .18

 

 

 

 

 

*
When the two dimensions are divided at the median

118
high external-low internal coefficients. However, as expect-
ed, the lowest correlation in the table is that at the high
internal-low external cell. All comparisons among the cor-
relation coefficients are nonsignificant, except for the com-
parison between the overall coefficient and the one located
at the high internal-low external cell, and it is in the pre-
dicted direction (Z=l.92, p <.05). In this sample the
original Hypotheses e and f are partially confirmed, and
inferentially, for the statistically significant finding,
Hypothesis e is also partially confirmed.

When we turn to Table 27b one finds a pattern of cor-
relations contrary to our expectations. None of the com-
parisons among correlation coefficients are significant,
except for one post hoc finding. The correlation in the low-
low cell is significantly lower than the coefficient in the
high internal-low external cell, according to a two-tailed 2
test of significance (z=2.26, p <.05). This statistically
significant finding is contrary to these expectations. In
general with the children's data it was decided to reject

Hypotheses e and f.

c. Deceit

 

The overall zero order correlations between exposure
to deceit on television.and the child's favorable predispo~

sitions towards deceit were:

119

Child's Favorable Predispositions
Towards Deceit
Mother's Report Child's Report

 

 

Exposure to TV

Deceit r = .28 r = .15
N = 209 N = 207
p 1.001 p i .02

Table 28 presents the correlations obtained between
exposure to TV deceit and the child's favorable predisposi-
tions towards this type of behavior at different intersec-
tions of the internal and external orientations of both the
child and the parents.

For the mothers' data, Table 28a indicates that at
first glance the expectations are disconfirmed. The low in-
ternal-high external cell shows the lowest correlation in
the table. This is in the exact opposite direction of the
hypothesized relationships. Z tests of significance indicate
that no comparisons are statistically significant.

Part b of Table 28 shows that for the children's data,
the pattern of correlations is contrary to Hypotheses e and f.
No inferential statistical tests were significant for any of
the comparisons among correlation coefficients.

To summarize, it can be said that in general, the
patterns of correlations for the mothers' data support the
expectations expressed in Hypotheses e and f. However, when
inferential statistical tests are used no significant differ-
ences among correlation coefficients appear. If a larger

sample had been utilized, and if the coefficients had not

120

Table 28. Contingent Correlations Between Exposure and
the Child's Dispositions Towards Deceit as
Reported by Mother and Child, at Different
Intersections* of the Internal and External
Orientations of Parents and Their Children.

 

a. Internal and external parental orientations.
Correlations of exposure with the mother's
report of the child's dispositions towards

 

 

 

 

 

 

 

 

deceit.
INTERNAL
LOW HIGH
)E‘ r= .38 r= .20
T LOW N = 56 N = 45
S = .002 S = .10
E
If} r = .12 r = .28
A HIGH N = 37 N = 57
L S = .24 S = .02
b. Internal and external child orientations.
Correlations of exposure with the child's
own report of'his/her dispositions towards
deceit.
INTERNAL
LOW HIGH
i r= -.04 r= .19
T LOW N = 48 N = 51
E S = .39 S = .09
If} r= .15 r= -.03
A HIGH N = 43 N = 51
L S = .16 S = .41

 

 

 

 

 

*
When the two dimensions are divided at the median

121

changed in magnitude, Hypotheses e and f would have been
mostly supported, especially in the cases of physical ag-
gression and verbal aggression.

With the children's data, one encounters a general
pattern that not only disconfirms the hypotheses but
that'is reversed. The tests of statistical significance used
showed that the correlations in each of the cells tended not
to significantly differ from each other, but the pattern is

surprising.

Summary
Chapter III has dealt with the results of the tests

of the hypotheses of this study. The results were divided
into three main sections: 1. regression and analysis of
variance; 2. t-tests for comparisons between pairs of means;
and 3. contingent correlations. Each method utilized dealt
with a set of hypotheses.

When analysis of variance and regression were used to
test Hypotheses a, b, and c, it was generally found that
with the mothers' data exposure and the external orientation
were significant predictors of the childrens' antisocial pre-
dispositions in general. In the children's case, the in-
ternal orientation was found to be a potent and significant
negative contributor to the child's antisocial predisposi-
tions, and to a lesser extent exposure in some occasions and

the external orientation in others.

122

The hypothesized differences among subgroup means
were tested by t-tests. These results were generally in
agreement with the regression and analysis of variance re-
sults, and rendered partial support to Hypothesis d in most
comparisons.

It was generally found that with the contingent cor-
relation analyses the data :support Hypotheses e and f in
the mothers' case, although the results were not statisti-
cally significant in general.

Hypotheses e and f were disconfirmed by the observed
results in the children's case, and they were not statisti-
cally significant. The patterns of correlations observed
was contrary to that expected.

The results above represent an interesting set of
complementary data. The inconsistencies require that ex-
planations be attempted, and this will be done in the dis-

cussion chapter.

CHAPTER IV

SUMMARY AND DISCUSSION

Summary
The present study dealt with the role of different

modes of parent-child interaction as mediators of children's
modeling of television antisocial portrayals. This type of
research is innovative in that it attempts to explain the
mechanisms by which children model the behaviors they witness
on the television screen. Past research has been mainly con-
cerned with the modeling process per se without further con-
sideration for the social conditions that facilitate or
inhibit it.

Some research as been done that deals with parental
and adult mediation of children's television learning, but
these research efforts have mainly been concerned with media-
tion during television exposure or with regard to it. This

study argues that there are enduring patterns of parental be-

haviors that in turn affect the children's internalization

of values or behavioral norms. Both parental and children's
orientations have been shown to be related, and in this study
it has been argued that parental and children's orientations
should act as mediators of the children's display of predis-
positions modeled after negative television examples.

123

124

More specifically, two main orientations of parent-
child interactions with regard to social situations have been
identified, namely, an internal and an external orientation.

The parental internal orientation is composed of a
set of behaviors that tend to focus the attention of the
child on the consequences of his/her behaviors on others.
This is an inductive form of child rearing where affection
is the norm and is accompanied by reasoning, explanation,
and appeals to self evaluation on the part of the child.
Internally oriented parents have been expected to lead their
children towards the internalization of moral values which
comprises consideration for the welfare of others, self
evaluation,reparation of wrongdoings, and guilt feelings
after transgressions occur.

The parental external orientation is constituted of
power assertive techniques which emphasize the external con-
sequences of the behavior for the child with little consider-
ation for others. External punishments or rewards have been
found to teach the child to expect extrinsic consequences
for his/her behaviors, be they positive or negative. An
externally oriented child may behave in socially desirable
ways when there is the possibility of some external rule en-
forcement. External rewards sensitize the child to behave
in socially desirable ways for the sake of social recognition
or material rewards but little internal satisfaction is

achieved.

125

The main difference between the two types of orienta-
tions was said to be that the internal orientation in parents
provides the children with the necessary cognitive structure
for evaluating his/her social behaviors. Externally oriented
parents do not provide cognitive resources upon which the
child can rely when social decisions have to be made.

The theoretical contribution of this study has been
to extrapolate the results of parent-child interaction re-
search to the instance of television modeling of four anti-
social behaviors--physical aggression, verbal aggression,
theft and deceit. The child who has been provided with cog-
nitive resources for evaluating his/her social behaviors
should be able to discount antisocial television portrayals
to a larger extent than a child who is externally oriented.

The internally oriented child should be less influenced
by negative television examples than the child who bases his/
her social behaviors on external considerations.

The first task attempted was to construct a pilot
instrument to tap the internal and the external dimensions
or parent-child interaction modes. A pretest instrument ad-
ministered to children rendered two clear separate dimensions,
one external and one internal, when the items were factor
analyzed. In the design of a final study the parental items
were administered to the mothers and the children's orienta-
tion items were presented to the children. When all parental

and child items were factor analyzed the two dimensions that

126

emerged discriminated between parents and children but not
between the internal and the external orientations. When
factor analyzed separately, the parental and the children's
data showed in each case to be decomposable into the two ex-
pected internal and external orientations. Given this find-
ing, the pr0posed hypotheses were separately tested for the
parental and the children's data.

Six hypotheses were tested in this study. Three hypoth-
eses dealt with the main effects expected from the internal
and external orientations and exposure to antisocial tele-
vision portrayals. The next hypothesis predicted the relative
magnitude of antisocial predispositions in children for dif-
ferent subsamples defined by the internal and external orien-
tations and antisocial television exposure. The last two hy-
potheses were concerned with the contingent relationships be-
tween exposure to antisocial television exposure and the child's
antisocial predispositions at different intersections of the
internal and external orientations of parents and children.
Below, each of the hypotheses tested will be reintroduced and
the evidence encountered for each one will be summarized.

Ha: The more exposure to antisocial port
trayals on television, the more anti-
social behavioral predispositions will
be displayed by children.

Exposure to antisocial television portrayals had a
significant independent effect on the children's antisocial

predispositions with the mothers' and the children's reports.

Specifically, with the mothers' data, exposure had an

127
independent effect on physical and verbal aggression and
deceit. With the children's data, antisocial television ex-
posure had a significant contribution on physical and verbal
aggression. All statistically significant coefficients were
in the predicted direction.
Hb: The more externally oriented the

parents and the children, the more

antisocial behavioral predispositions

will be displayed by children.

The external orientation of parents had a significant
independent effect on physical and verbal aggression and
deceit as reported by the mother; however, with the children's
data, the external orientation of children only had a sig-
nificant contribution on the children's physically aggressive
predispositions. All significant coefficients were in the
expected direction.

He: The more internally oriented the parents
and the children, the less antisoc1a1
behavioral predispositions will be dis-
played by children.

The children's internal orientation was found to be the
most powerful negative predictor of antisocial predispositions.
The children's internal orientation showed high negative
partial regression coefficients with antisocial predisposi-
tions towards physical and verbal aggression, theft, and
deceit. The mothers' data, on the other hand, only rendered
one statistically significant negative coefficient with deceit.

Since exposure to television theft did not correlate

or predict theft predispositions, it was deleted from further

analysis, since no mediation effects were possible where there

128

was no initial television modeling.

Hypothesis a was said to be supported with both the
mothers' and the children's data. Hypothesis b was supported
in general with the mothers' data but not with the children's.
Hypothesis c was supported with the children's data but not
with the mothers'. None of the two and three-way interactions
among the independent variables were found to be statistically
significant.

Hd: i4 > £3 = $2 > i1 > £8 > £7 = R6 > is

Where i1 - E8 were the means in each Of the following

 

 

 

 

cells:
INTERNAL ORIENTATION

LO HI
E EXTERNAL ORIENTATION EXTERNAL ORIENTATION
§ LO f HI - LO p HI
OLO 21 l 352 L 325 J Re ]
s
U HI ‘E3 1 ‘i4 1 ‘i? J ‘iaj
R , , ,
E

The cell means were the average amount Of each Of the
three antisocial predispositions--physical aggression, verbal
aggression and deceit.

With both sets Of data, parents and children, it was
consistently found that highly internally oriented children,
who watch small amounts Of antisocial television, and whose
parents and themselves are low in their external orientation

display the least amount Of antisocial predispositions when

129
compared to other children. Also the children highest in
antisocial predispositions were those low in the internal
orientation and high in antisocial television exposure and
the external orientation. Other subsample comparisons were
generally found to follow the predicted patterns but just a
few were statistically significant. Hypothesis d was said

to be partially supported by the data.

He: r2 < rA < r3 and
Hf: r2 3 r1 fi’r3
r2 5’r4 §_r3

where rA was the overall correlation between exposure to
antisocial portrayals on television and antisocial predis-
positions in the child. r1 - r4 represented contingent cor-
relations at the following intersections Of the internal and

the external orientations of both parents and children:

INTERNAL ORIENTATION

 

 

LOW HIGH
EXTERNAL LOW ‘1 f ‘2 l
ORIENTATION HIGH r3 1 r4 ]

 

 

Those highly internally oriented parents who were low
in their external orientation had children for whom the cor-
relation between exposure tO antisocial behaviors on tele-
vision and their display Of antisocial predispositions was
the lowest when compared with all the children and with other

subsamples. The children who showed the highest correlation

130
between exposure and predispositions were those whose parents
were highly externally oriented and low in their internal ori-
entation. With the children's orientations, the pattern Of
findings was Opposite tO the hypothesized expectations. Almost
none Of these contingent correlation comparisons were found to
be statistically significant. This finding was replicated when
no three-way interactions were found to be statistically sig-
nificant with regression analyses or analyses of variance. Hy-

potheses e and f were inferentially rejected in favor Of the null.

Discussion

 

The discussion of the results Of this study will first
focus on the dimensionality Of parent-child interaction modes
and then on the independent and mediational effects Of parent-
al and child orientations. To finalize, some suggestions for
future research, limitations Of this study and some concluding

remarks will be presented.

Dimensions Of Parent-Child Interaction Modes

 

The pretest data Obtained from children regarding their
parents' and their own orientations showed that when data are
collected from the same respondents the expectations regarding
the emergence of two overall dimensions of internal and ex-
ternal parent-child orientations were fully corroborated.

This finding encourages the most recent theoretical formula-
tions in the literature dealing with the internalization Of

:moral principles (Lickona, 1976; and DePalma and Foley, 1975).

131

On the other hand, when data were collected from
parents and children separately parent-child dimensions
failed to appear with both sets Of data; however, the intern-
al and external dimensions did emerge for parents and child-
ren separately. These results suggest that either the
mothers and children perceive things differently or that for
other reasons they prefer to report dissimilar perspectives
about a common phenomenon.

That different peOple perceive Objects or events in
dissimilar forms is not a new Observation. "Lewin argued
that the phenomena to which the psychologist should direct
his attention are what the individual subjectively perceives,
not what the Observer perceives as the 'Objective reality'"
(Shepherd, 1964, p. 24). This notion is similar to Weick's
(1969) enactment process: "The human creates the environment
to which the system adapts. The human actor does not react
to an environment, he enacts it" (p. 64). If human subjec-
tivity is crucial for the understanding of psychological
phenomena, it is not surprising to find that parents and
children differ in their reports about the same phenomenon,
or about phenomena that are supposed to be related.

Future research on parent-child interactions may well
benefit from collecting all relevant information from the
subjects of direct interest, be they parents or children.
Children who perceive their parents to be oriented in a

certain direction do perceive themselves to be oriented in

 

132

a similar direction. However, the association between
parental perceptions Of their orientation and children's
perception of their reactions is not as consistent.

The research literature does not specify the degree
of association between the two internal and external dimen-
sions. This research has shown that the external and intern-
al orientations do not lie on a continuum, they tend to be
somewhat positively correlated with parental orientation
reports (r=.36, pp: .001), and mildly associated with the
children's report Of their orientations (r=.13, p i .05).
What these associations suggest is that parents that are con-
cerned aboutwhat their children do, do more Of everything in order
to guide their social behaviors, and that children who are
socially active tend to be oriented in both directions to
some extent.

Another contribution of this study to the parent-
child interaction literature regarding the internalization
of moral guidelines is that parental and children's reactions
to positive social situations are consistent with the orien-
tations derived from responses to negative situations. Past
research has been mainly concerned with reactions to trans-
gressions. In this study, based on scarce literature, posi-
tive social situations were presented to mothers and children
and their reactions were assessed. Parents who internally
respond to their children's transgression do also internally

respond to their children's prosocial or positive behaviors.

133

Externally oriented parents with regard to transgressions
were found to at the same time be externally oriented re-
garding positive acts. The children's data showed the same
consistency regarding their internal and external orienta-
tions across positive and negative situations. Future re-
search may well benefit from incorporating positive situations
in order to more fully account for the internalization Of

standards in children.

Independent and Mediational Effects

The role of parent-child interaction modes in mediat-
ing the children's modeling of antisocial television por-
trayals was found to be negligible in this study. However,
some indications of a possible inhibiting effect due to the
internal orientation and some enhancing effect on the part
Of the external orientation Of parents was suggested by the
data. Future research utilizing larger samples may more
definitely unveil these patterns.

It was of utmost importance to Observe that television
antisocial exposure has a moderate and very consistent effect
on the children's antisocial predispositions regardless of
parental or children orientations. Television as a pervasive
agent of socialization showed to be as important as extern-
ally oriented parental practices in directly contributing
to antisocial predispositions in children.

Given the demonstrated independent effect Of tele-

vision exposure it can be said that among all the sources Of

134
socialization of children in modern society television is
one more "significant other" (WOelfel and Hernandez, 1972)
for the child. Television along with other media, parents,
siblings, teachers and peers may account for the complete
socialization of children. Different interactions of the
child with significant forces in his/her environment may do
little to alleviate the pervasive influence of the tele-
vision medium. The mediation impact Of other socialization
forces besides parents still has to be demonstrated.
Parents were considered here due to their constant social
proximity to the child and due to the potential that parents
represent for modifying the social environment of the child.

Parental and children orientations were demonstrated
to contribute to different extents to the children's anti-
social predispositions. However, regardless of whether or
not parents use inductive or power assertive socialization
techniques, or whether children have internalized moral guide-
lines Or not, television and its fare of socially undesirable
portrayals teach the seemingly unintended lesson that the way
to solve problems is through aggression and deceitful
behavior.

Future research should clearly look for other modifi-
able aspects of the children's environment for attenuating
negative television exposure effects. It might be that the
only answer to the problem of antisocial television modeling

consists of constant surveilance and company at the time

135

that antisocial television programs are aired (Atkin and
Greenberg, 1977). It might also be necessary for broadcasters
to reconsider their position in view of these findings. If
there is little that parents can do to alleviate negative
television effects, broadcasters and producers may find
themselves undertaking a major responsibility regarding the
examples they provide to the youth.

Parents, on the other hand are not exempt of responsi-
bility since this study showed that parents do contribute to
their children antisocial predispositions if they are extern-
ally oriented in their child's rearing practices. It is not
television alone that is the cause of social illnesses. If
parents provide an example of power assertion when dealing
with their children they will learn the lesson independently
of what television does.

Children who by some means have internalized moral
values are less likely to behave antisocially but the effect
of such internalization was found to be independent Of tele-
vision exposure. SO, when controlling for the moral develop-
ment of the child, television exposure to antisocial portrayals
enhances the children's antisocial predispositions. If there
had been a significant interaction between the children's
internal orientation and exposure, one might have concluded
that the more the internalization of moral values by children,
the less the contributing effect of television exposure. No
such interaction was encountered in our analyses. Children

benefit from internalized guidelines but if they watch negative

136

television examples, they seem to be willing to cOpy them.

The question is now raised with regard to the differ-
ential effects found for parental and children orientations.
Why should the external orientation in parents predict the
antisocial predispositions in children and not so the intern-
al orientation? In the first place, the external orientation
in parents consists Of behaviors similar to those expressed
in the children's antisocial attitudes. Parents who hit and
yell provide a direct example of antisocial behavior (Bandura,
1973). On the other hand, inductive or internally oriented
parents provide an example Of reasoning and explanation that
might be reflected in the prosocial attitudes of children,
but prosocial predispositions were not studied here. The
parental internal orientation was expected to provide a cog-
nitive structure for the children's evaluation of their social
behaviors, but it was Observed that the relationship between
the internal orientation reactions of parents and their children
was low and fell short of statistical significance (r=.ll, p=.053).
Future research that includes prosocial attitudes or behaviors
as the dependent variable may reveal that internally oriented
parents also serve as models for their children's positive be-
haviors.

Second, it may be that verbal exhortations by parents
are modeled as verbal endorsements by children, and that motor
behaviors by parents are modeled in the same behavioral cate-

gory. Children who hear some preaching from their parents may

137

model their verbal behavior and motorically imitate what the
parents do. Since here it was found that the internal and
external orientations of parents are moderately related (r=.36,
p i .001), it wouldn't be farfetched to say that since parents
that do more of one thing also do more Of the other, their
children match their behaviors both ways. Overtly, the chil-
dren express antisocial predispositions for problem resolution
and they may verbally reason and explain as their parents do.
Bryan (1975) reports that "children who had witnessed a non-
giving model would preach charity but, in fact, practice self-
ishness. In effect, then, hypocritical children had been
produced" (p. 104). In this research it was not investigated
whether children model their parents' expressions of concern
or reasoning and explanation. At this point the evidence pro-
vided by this study suggests that it may be the case that
children say what they hear and do what they observe.

Why should children be influenced by their own internal
orientation and not by their externalized expectations? The
components of the children's internal orientation are behav-
iors that indicate guilt reactions, consideration for others,
reparation and self evaluations. A child who is likely to
experience anxiety after transgressions was expected to be
less antisocial than a child who does not experience inner
conflict. This expectation was confirmed and it can be argued
that according to the theory (Aronfreed, 1968b) internally

produced anxiety is more long lasting and potent in inducing

138
the child to refrain from antisocial behavior than external
considerations.

Not only should the children's external orientation be
less effective than internalized values in reducing antisocial
predispositions. An externally oriented child should have
been found to be more willing to behave antisocially since
worries regarding external consequences should be less en-
during (Aronfreed, 1969, p. 313) than the anxiety produced by

internalized principles. It was found that among the anti-

 

social behaviors investigated here, only physical aggression 1
was at all predicted by the child's focus on external consider-
ations. Children may in general worry about external conse-
quences regardless of their internalization of moral values.
The correlation found between the external and internal
orientation of the children was indeed low (r=.13, 9.1 .05).
When transgressions do occur, or in its case, when positive
behaviors take place the fear of discovery might be a function
of the specific contingencies of the situation, e.g., the
actual possibility of discovery in a certain circumstance.

At this point the theoretical perspective that guided
this investigation should be reconsidered to incorporate the
results of this study and to guide future investigations.
Although parental practices are related with the children's
orientations, they do not underlie common factors. .Parental
practices contribute to the children's socialization, but

other influences in the social environment of the child should

139

be considered. Television exposure might not only be an in-
fluence on the children's antisocial predispositions, but it
may in turn serve as a source for the children's acquisition
of internal controls. In this sense, multiple influences
might be considered as determinants Of the children's acqui-
sitions Of internal controls.

The present research suggested that parental extern-
alized practices and television exposure are two contributors
to negative attitudes. On the other hand, the children's
degree of internalization of moral values was shown to inde-
pendently reduce those antisocial predispositions.

With external parental practices and television ex-
posure a modeling effect has been supported. With the chil-
dren's internalization of moral standards a more cognitively
mediated type of effect has been shown. What needs to be in-
vestigated now is the nature of the antecedents of internal-
ization in children. Clearly, the relationship between
parental practices and the children's internalization of
moral standards has not been found to be substantial enough
in this study or in the past (Hoffman, 1970, p. 291).

This investigation was initiated under the basic prem-
ise that children who have internalized moral standards should
be less likely to model antisocial television examples. Child
psychologists may want to consider the findings of this re-
search in order to reevaluate the theory behind the socializa-

tion of moral values in children.

140

Research regarding television effects may benefit from
having encountered that television has an independent effect
on the children's display of negative behavioral predispo-
sitions. Parents will be well advised to consider that tele-
vision is not the only source of negative behaviors in their
children. Broadcasters may want to take into consideration
that parental socialization practices do not seem to change

the likelihood of television effects on children.

Suggestions for Future Research

 

Future research should consider the possibility that
parental or child orientations may enhance television model-
ing of positive or desirable television examples. It might
well be that parent-child interaction modes effectively inter-
act with television exposure in such a way that low doses of
external considerations and high internally oriented parents
and children enhance modeling of socially acceptable tele-
vision portrayals. Parents who guide their children to be
internally oriented may cue their children to selectively
focus on the virtues of prosocial television content. In
certain television contexts, children may find models who are
suggestive of opportunities for giving form to their intern-
alized moral values. Prosocial television examples may
channel the internalized orientations of children. The child
may find the examples of prosocial television to be a model

for expressing their internalized convictions.

141

For television research purposes it might be profit-
able to continue exploring the conditions under which parent
child orientations underlie common factors. If data are col-
lected from parents regarding their practices and their
children's reactions, and if data are collected from children
regarding their parents' practices and their own reactions
it may be found that when all data are collected from the per-
spective of one of the participants in parent-child inter-
actions two clear dimensions appear. If this is found to be
the case in future replications, the results of this research
may be altered and more definitive evidence for the role of
parent-child interaction modes in mediating children's learn-
ing from antisocial television portrayals may be obtained.

Further attempts at validating parent-child interaction
data should be carried out. Siblings might be asked to re-
port on parental practices regarding the child of interest in
the family. Also siblings can be asked about the response
orientations and the social behaviors of the child of interest.
Perhaps of more importance would be the collection of data
from mothers and fathers independently. In this study mothers
were asked to report for themselves and their husbands and
were asked to generalize. This type of measurement might not
have reflected what we call parental practices. Obtaining
both sets of data can provide for different analyses which

might untangle the effect of both parental sources of influence.

142

On the other hand, children might be asked to report
on their mothers' and fathers' practices separately. In this
study the child was asked to report about what both his/her
parents in combination might do.

If all the data suggested above are collected, the
analyses would become more complex but may render clearer
insights into the process under study by separating sources
of influence and at the same time providing more valid obser-

vations.

Limitations

 

The data for this study were collected from interviews
and questionnaires. verbal reports of behavior may not come
pletely reflect the interactions that take place in the home
or family environment. The reader should be cautioned that
since no Observational data were collected and since the con-
vergent validation of a key portion of the instrument was
generally low, the research results may deviate from the
actual phenomena that the research attempted to analyze. On
the other hand, the reader should also be aware that Observa-
tional data are not free of validity problems. Most methods
for collecting observational data are Obstrusive to some ex-
tent and the behavior of parents and children can also be
influenced by the presence of an observer or a recording
instrument.

The results of this study are also limited by its asso-

ciational nature. We Obtained indications of the degree of

143

predictability and relationship between and among measures
but in no sense can it be said that the evidence supports a
causal relationship. The theoretical statements presented do
give indications about causality. However, the analysis was
done with data collected at one point in time. In order to
establish causality the temporal order of the variables under
consideration has to be demonstrated and this was not done
here. Longitudinal data collected from parents and children
may help in assessing the causal flow indicated by the theory.

The socioeconomic status of parents and children in this
study was not utilized. Past research indicates that parental
orientations or disciplinary practices are related to their
socioeconomic status (Aronfreed, 1961). The relative contri-
bution of the socioeconomic class of the respondents should
be considered in order to assess the degree of independence
between parental practices and class. If their contributions
are relatively independent both sets of antecedents may be
considered as separate predictors. If their contribution is
shared, the social class conditions that promote different
disciplinary practices should be studied.

The main focus of this study was to investigate the
potential role of parental practices and children's responses
to social situations in mediating negative television effects;
the roles of peers, siblings, teachers and other socializa-
tion agents were ignored. The deletion of alternate sociali-

zation agents limits the conclusions that can be derived from

144

this study. As children grow older different socialization
agents increase in impact on their social behavior and moral
considerations.

The distributions of reports of disciplinary practices
by parents and the children's reactions to social behaviors
were skewed. Although the analytic methods used in this re-
search are robust, alternate methods for diminishing the
skewness of the distributions should be considered. One way
of achieving this is by expanding the range of the scales
utilized, e.g., "Out of ten times that John lied to you, how
Often would you consider spanking him?" Less skewed distri-
butions should allow the researcher to place more confidence
on parametric statistical analyses like those conducted here.

Another limitation of this study is the implicit bias
in calling a set of behaviors "antisocial." Although he-
haviors such as yelling, hitting, stealing and cheating are
considered antisocial by a large segment of society, some of
those behaviors are encouraged and found positive in some
situations, e.g., "if somebody hits you, you should hit back."
The value judgment implied in the selection of "antisocial"
behaviors may affect the nature of the analysis and conclu-

sions that can be drawn from this research.

Conclusions

 

Although some indications were found that parent-child
orientations may partially account for what the children model

from television portrayals, the overall results of this study

145

indicated that television exposure is in itself a consistent
and independent contributor to the children's favorable atti-
tudes towards the negative behaviors they watch on television.

The task of further clarifying the determinants and the
nature of children's internalization of moral standards be-
longs to child psychologists. However, this task cannot be
disassociated from research dealing with television effects
if scientists, policy makers, broadcasters and the public are
to understand the broader context in which television effects
take place.

Studies of this type are unlikely to give definitive
answers to complex questions in an initial attempt. The

bases upon which to build have been explored here.

REFERENCES

REFERENCES

Adkins, Dorothy C., Frank C. Payne and J. Michael O'Malley

1974

Moral DevelOpment. In F.M. Kerlinger and J.B.
Carroll, editors, Review of Research in Educa-
tion, Vol. 2. Itasca, 111.: F.E. Peacock'
Publishing, Inc.

 

r Allinsmith, W.

1960

Aronfreed,
1961

1968

1968b

1969

1976

The Learning of Moral Standards. In D.R. Miller
and G.E. Swanson, editors, Inner Conflict and
Defense. New York: Holt, Rinehart and Winston.

 

J.

The Nature, Variety, and Social Patterning of
Moral Responses to Transgression. Journal of
Abnormal and Social Psychology, Vol. 63, No. 2,
pp. 223-240.

 

 

Aversive Control of Socialization. In William J.
Arnold, editor, Nebraska Sym osium on Mgtivation.
Lincoln, Neb.: University 0 Nebraska Press.

Conduct and Cpnscience: The Socialization of
Internaliged Control over Behavior. New York:
Academic Press.

The Concept of Internalization. In D.A. Goslin,
editor, Handbook Of§gcializationrgheory and
Research. Chicago: Rand McNally College
Publishing Company.

Moral Development from the Standpoint of a General
Psychological Theory. In T. Lickona, editor,
Moral Development and Behavior Theory; Research
and Social Issues. New York: Holt, Rinehart and
WInston.

 

Atkin, Charles K.

1972

Statement at Hearings before the Subcommittee on
Communications of The United States Senate,
March 24.

Atkin, C.K. and W. Gantz

1974

Children's Response to Broadcast News: Exposure
Evaluation and Learning. Paper presented at the
meeting of the Association for Education in
Journalism, San Diego, Ca., August.

146

147

Atkin, C.K. and B.S. Greenberg
1977 Parental Mediation of Children's Social Behavior
Learning from Television. Paper presented at the
annual convention of the Association for Educa-
tion in Journalism, Madison, Wisconsin, August.

Ball, S. and G.A. Bogatz
1972 Summative Research of Sesame Street: Implications
for the Study of Preschool Children. In A.D. Pick,
editor, Minnesota Symposia on Child Psychology,
Vol. 6. Minneapolis, Mihn.: University of
Minnesota Press.

 

Bandura, Albert
1973 Aggression: A Social Learning Analysis. Engle-
wood Cliffs, N.J.: Prentice Hall.

 

1977 Social Learninngheory. Englewood Cliffs, N.J.:
Prentice Hall.

 

Bandura, A. and R.H. Walters
1959 Adolescent Aggression. New York: Ronald Press.

 

Barcus, F.E.
1969 Parental Influence on Children's T.V. Viewing.
Television Quarterly, 8: 63-73.

 

Bechtel, R.B., C. Achelpohl and R. Akers.

1972 Correlates Between Observed Behavior and Question-
naire Responses on Television Viewing. In E.A.
Rubinstein, G. A. Comstock, and J. P. Murray,
editors, Television and Social Behavior, VOl. Iv:
Television in Day-to-DayL Life: Patterns of Use.
Washington, D. C.: U. S. Government Pr1nting Office,
pp. 274-344.

 

 

Bruning, J. L. and B. L. Kintz
1968 Computational Handbook of Statistics. Glenview,
111.: Scott, Foresman and CO.

 

Bryan, James H.
1975 You Will be Well Advised to Watch What We Do
Instead of What We Say. In D. J. DePalma and
J. M. Foley, editors, Moral Development: Current
Theory and Research. Hillsdale, N. J.: Lawrence
Erlbaum Associates.

 

Burton, R.V., Maccoby, E.E. and W. Allinsmith
1961 Antecedents of Resistance to Temptation in Four-
Year-Old Children. Child Development, 32:
689-710.

 

148

Chaffee, Steven H.
1976 Comparing Television to other Agencies of Social-
ization. Memorandum to the Ford Foundation,
September.

Chu, G.C. and W. Schramm
1967 Learning fromTelevision: What the Research Says.
Washington, D.C.: National Association of Educa-
tional Broadcasters.

 

DePalma, D.J. and J.M. Foley (Eds.)
1975 Moral Development: Current Theory and Research.
Hillsdale, N.J.: Lawrence Erlbaum Associates.

 

Eswara, Harogadde S.
1968 An Interpersonal Approach to the Study of Social
Influence: Family Communication Patterns and
Attitude Change. Unpublished Ph.D. dissertation,
University of Wisconsin.

Feshbach, S.

1972 Reality and Fantasy in Filmed Violence. In J.P.
Murray, E.A. Rubinstein and G.A. Comstock, editors,
Television and Social Behavior, Vol. 2.

TeleviSion and Social Learning. Washington, D.C.:
Government Printing Office.

 

 

Fink, Edward L. and James J. Noell
1972 On the Representation of Interaction Effects in
Multiple Regression: A Note on Portes' 'On Class
Consciousness.‘ Unpublished manuscript, Department
of Communication, Michigan State University,
East Lansing, Michigan.

Greenberg, B.S., C.K. Atkin, N.G. Edison and F. Korzenny
1977 Pro-Social and Anti-Social Behaviors on Commercial
TElevisiorin 1975976. Department of Communication,
Michigan State University. Report to the Office
of Child Development, HEW, Washington, D.C.,
February.

 

 

Ero-Social and Anti-Social Behaviors on Commercial
TelevisiOn in 1976-77. Department ofLCOmmunication,
Michigan State University. Report to the Office

of Child Development, HEW, Washington, D.C.,
forthcoming.

 

Greenberg, B.S., P.M. Ericson, and M. Vlahos
1972 Children's Television Behavior as Perceived by
Mother and Child. In E.A. Rubinstein, C.A. Com-
stock and J.P. Murray, editors, Television and
Social Behavior, Vol 4. TelevisiOn in Dayeto-Day
Life: Patterns of Use. Washington, D.C.: Govern-
ment Printing Office, pp. 359-409.

 

 

 

 

149

Grinder, R.E.
1962 Parental Childrearing Practices, Conscience and
Resistance to Temptation of Sixth-Grade Children.
Child Development, 33: 803-820.

Hannenman, Gerhard J. and Herbert S. Dordick 22' El
1975 The Family Hour: A Solution in Search of a
Problem. A Research Study of Los Angeles TV
Viewers' Response to the Family Hour. Summary.
Center for Communications Policy Research of the
Annenberg School of Communications. Los Angeles:
University of Southern California.

Henry, Jules
1971 Pathways to Madness. New York: Random House.

 

Hicks, D.J.
1968 Effects of Co-Observer's Sanctions and Adult
Presence on Imitative Aggression. Child Develop-
ment, 38: 303-309.

 

Hoffman, Martin L.
1970 Moral Development. In Paul H. Mussen, editor,
Carmichael's Manual of Child Psychology. New
YorK: John Wiley and Sons, Inc.

 

1975 Moral Internalization, Parental Power, and the
Nature of Parent-Child Interaction. Developmental
Psychology: 11(2): 228-239.

 

Hoffman, M.S. and H.D. Saltzstein
1967 Parent Discipline and the Child's Moral Develop-
ment. Journal of Personality_and Social Psy-
chology, V01. 5, No. 1, pp. 45-57.

Kerlinger, Fred N. and Elazar J. Pedhazur
1973 Multiple Regressign in Behavioral Research.
New York: Holt, Rinehart and Winston.

Leifer, A.D. and D.F. Roberts
1972 Children's Responses to Television Violence. In
J.P. Murray, E.A. Rubenstein, and G.A. Comstock,
editors, Televigion and Social Behavior, Vol. 2:
Television and Social Learning. Washington, D.C.:
Government Printing Office, pp. 43-180.

 

Leifer, A.D., N.J. Gordon, and S.B. Graves
1974 Children's Television: More Than Mere Entertain-
ment. Harvard Educational Review, 44: 213-245.

150

Lickona, Thomas
1976 Critical Issues in the Study of Moral Develop-
ment and Behavior. In Thomas Lickona, editor,
Moral Develgpment and Behavior Theory, Research,
aid Social Issues. New York: Hoit, Rinehart
aHHLWinston.

Lyle, J. and H.R. Hoffman
1972 Children's Use of Television and Other Media.
In E.A. Rubinstein, G.A. Comstock, and J.P.
Murray, editors, Television and Social Behavior,
Vol 4. Television in Day-to-Day LifezIratterns
of Use. Washington, D.C.: Government Printing
Office, pp. 129-256.

 

McLeod, J.M., C.K. Atkin, and S.H. Chaffee
1972 Adolescents, Parents and TV Use: Adolescent
Self-Support Measures from Maryland and Wisconsin
Samples. In G.A. Comstock and E.A. Rubinstein,
editors, Television and Social Behavior. Vol. 3,
Television ass Adolescegt Aggressiveness. Washing-
ton, D.C.: Government Printing Office, pp. 173-238.

McLeod, J.M. and S.R. Chaffee
1972 The Construction of Social Reality. In James T.
Tedeschi, editor, The Social Influence Processes.
Chicago: Aldine-Atherton.

McLeod, J.M., S.H. Chaffee and H.S. Eswara
1966 Family Communication Patterns and Communication
Research. Paper presented to the Theory and
Methodology Division, Association for Education
in Journalism Convention. Iowa City, Iowa, August.

McLeod, J.M., S.H. Chaffee, and Daniel Wackman
1967 Family Communication: An Updated Report. Paper
presented to the Theory and Methodology Division,
Association for Education in Journalism. Boulder,
Colorado.

Nunnally, Jum C.
1967 Psychometric Theory. New York: McGraw Hill.

 

Overall, J.B. and C.J. Klett
1972 Applied Multivariate Analysis. New York:
McGraw Hill.

 

Reiss, David
1971 Varieties of Consensual Experience. 1. A Theory
for Relating Family Interaction to Individual
Thinking. Family Process, March, Vol. 10, No. 1:
1-27.

151

Rummel, R.J.
1970 Applied Factor Analysis. Evanston, 111.:
Northwestern University Press.

 

Saltzstein, Herbert D.
1976 Social Influence and Moral Development: A
Perspective on the Role of Parents and Peers.
In Thomas Lickona, editor, Moral Development
and Behavior Theory, Research and Social Issues.
New York: Holt, Rinehart and Winston.

 

 

Sears, R.R., E.E. Maccoby, and H. Levin
1957 Patterns of Child Rearing. Evanston, 111.:
Peterson and Co.

 

Sears, R.R., L. Rau and R. Alpert
1965 Identification and Childrearing. Stanford:
Stanford University Press.

 

Shepherd, C.R.
1964 §mallGroups= Some Sociological Perspectives.
San Francisco, CalifOrnia: Chandler Publishing
Company.

Staub, Ervin
1975 To Rear a Prosocial Child: Reasoning, Learning
by Doing and Learning by Teaching Others. In
David J. DePalms and J.M. Foley, editors,
Moral DevelOpment: Current_Theory and Research.
Hillsdale, N.J.: Lawrence Eribaum Associates,
Publishers.

 

Staub, E. and H.R. Feinberg
Personality, Socialization, and the Development
of Prosocial Behavior in Children. In D.R. Smith
and J. Macauley, editors, Voluntary Action
Research, Lexington Press, in press.

 

Stone, vernon A. and Steven H. Chaffee
1970 Family Communication Patterns and Source -
Message Orientation. Journalism:guarterly,
47: 239-246.

The Surgeon General's Scientific Advisory Committee on
Television and Social Behavior
1972 Television and Growin Up: The Impacrof Tele-
Vised'Violence. Mary and: NatiOnal Institute
of Mental Health.

 

 

Tolley, H.
1973 ghildren and War: Political Socialization to
International ggnflict. New York: Teachers
COllege Press, COlumbia University.

 

 

152

Walling, James I.

1976 The Effect of Parental Interaction on Learning
from T.V. Communication Education, V01. 25.

Weick, R.E.

1969 The Soggal Psychology of Organizing. Menlo
Park, CalifOrnia: Addison-Wesley.

Woelfel, J. and D. Hernandez

1972 Media and Interpersonal Effects on Attitude

Formation and Behavior. Unpublished manuscript,
University of Illinois at Urbana.

PF A--Burks 9

character. In a stable environment, under what circumstances will an evolu-
tionary process approach a limit? With respect to what properties and to
what extent does :volution maximize? What kind and degree of randomness is
needed to make evol~ ion operate? What is the r 1e of levels of organization

(hierarchies) in evol ion? It has been argue that there has not been time

 
  
 

for evolution to produce its results by chanc , and countered that evolution

proceeds hierarchically, p-oducing small co ounds from.atoms, large compounds
from small compounds, and so on up to org and organisms. Such questions
should be investigated by sim lation and so formulated quantitatively, at
least in ideal form.

These are all questions tv be 3 died by computer scientists and bi-
ologists, rather than philosophers. B the answers should give rise to con-
ceptual and foundational problems in t e philosophy of science. Some tele-
ological or maximizing theories are eo‘ivalent to mechanical explanations,
for example, Snell's law and the law of .east time. Does this equivalence
hold in biology, for example, for th: evol.tion of natural species? What is
the relation of maximization in evol tion to »aximization in utility theory,
game theory, and economics? What k ad or kind- of relevant quantitative
measures Of complexity can be assigted to organi~us and computers and their
abilities, so that the information processing powe of different organisms
and computers can be meaningfully Compared and so thz development in evolu-
tion, learning, and computational a-ility can be measu ed quantitatively?

I'll close my remarks by con acting this last su;gestion to the ra-
tionalist-empiricist controversy. S ppose we have an autou:ton-1ike account
of evolution beginning with physics, yroceeding through chemistry, biochem-
istry, the origin of life, and biology and ending with mode men and their
genetic programs of, say, 3000 years ag-. Suffix to this thz subsequent his-
tory of science up to today. The whole p ocess has produced modern science.
How much of science is attributable to evol~tion up to 1000 :.C., and how
much to subsequent history? How much of man's\present complexity is due to
evolution, and how much to learning from the environmenrg///

The rationalist would hold that in fact innate ideas and principles
were produced by evolution and existed in man's genetic program 3000 years
ago. Mere strongly, he might maintain that innate ideas are necessary for
science in the sense that there has not been time for modern science to de-
velOp in 3000 years without such a head start. In contrast, the empiricist

would hold that evolution only produced a very general learning program, and

emit“.-

d!”

:‘l‘

 

"- , "I"
D ' “
11! ’.~..\‘
V7“ '1

- um -
, - n

..r.o aw:

ILNEC’EQ
"c. or”
hit. ‘68

‘ .0.)

mt

«1 I mean.
"l{'dl
w u E-Iu

.\.a L”‘:
lit-ml,

..; .713a1

. 1; Juana.

, . n. no no:

2" than!

.- dial 7: m

l :4. .nnlaolavi

n ."I.K.-;" sd'. .
>21me
',.A-to oval .III
_,-..,..-. . “oo- “1“”

"Q Vh'

2'; \ wit!“ W. M w

’1‘ nlqwaw‘.

  
 

 

"WITHIN)11113