=
.—
=
=
=
—
—
—
—
—
_——
—
—
—
—

 

(Hit-fig

       

__.., . 3129310

HER A R Y “‘1?
Michigan State
University

This is to certify that the

thesis entitled

CHILDREN'S IMPRESSIONS OF THE VALUES

OF THEIR PEERS AND PARENTS:
PRESCRIPTIVE 0R PROSCRIPTIVE?

presented by
William M. Bukowski, Jr.

has been accepted towards fulfillment
of the requirements for

MA degree in PSYCHOLOGY

 

 

 

0-7639

 

      
 

  

'-
‘ o"

(xi.

.; ' ~::\i1r,. i '

OVERDUE FINES:
\
25¢ per day per item

RETUMIB LIBRARY MATERIALS:
Place 1

n book return to ve
roan circulation records

     
  
    
  

  

charge f

_ I.‘.'-)L;9; [I 'L_ '

i SEW?
' ,» MAGEC2

‘. ‘r

.\

SEP 012003
n 4 26 03]

 

CHILDREN'S IMPRESSIONS OF THE VALUES
OF THEIR PEERS AND PARENTS:
PRESCRIPTIVE OR PROSCRIPTIVE?

BY

William M. Bukowski, Jr.

A THESIS

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

MASTER OF ARTS

Department of Psychology

1980

ABSTRACT
CHILDREN'S IMPRESSIONS OF THE VALUES

OF THEIR PEERS AND PARENTS:
PRESCRIPTIVE 0R PROSCRIPTIVE?

'By
William M. Bukowski, Jr.

Sixty-two students in grades five, eight, and eleven were
asked to finish incomplete sentences which were designed to
examine their impressions of why they are rewarded and
punished by their parents and peers. Their responses were
scored according to whether they emphasized "right doing"
(i.e., prescriptive values) or "wrong doing" (i.e., pro-
scriptive values). The responses to the peer stems were
compared with the responses to the parent stems in order

to test the hypothesis that children and adolescents would
perceive their peers as having a stronger prescriptive
values orientation than the values orientation of their
parents. Although the subjects used more prescriptive
responses to complete the peer stems than the parent stems,
the differences between the responses to these two groups
of stems were not significant. It was suggested that this
lack of significance may have been due, in part, to the poor

reliability of the measurements used by the experimenter.

ACKNOWLEDGEMENTS

Several persons generously assisted me on various
aspects of this project. The faculty, administration and
students of the Mason School District in Michigan partic-
ipated in and helped me with the data collection. The
Department of Psychology of Michigan State University pro-
vided me with materials, supplies and secretarial services.
Several students, including Steven Truhon, Debbie Hotch
and Arthur Pomerantz read and made comments about earlier
drafts of this paper. DeWayne Moore was especially helpful;
he had many valuable and instructive suggestions for nearly
every aspect of this study, particularly the data analysis.
I have greatly appreciated and benefited from his patient
help. The members of my committee (Drs. E. A. Strommen,

R. Levine and J. P. McKinley) also generously provided me
with their time and their help. Particularly I am indebted
to John McKinney for his friendship and encouragement.

Finally, I would like to express my gratitude to my
parents for their financial and moral support. This work

is dedicated to them.

ii

TABLE OF CONTENTS

Introduction . . . . .
Method . . . . . .

Subjects . . . .

Procedures . . .
Results . . . . . .
Discussion . . . . .

Appendix A: Parental Permission

Appendix B: Incomplete Sentence
Instructions . . . . .

References . . . . .

iii

Letter

Stems and

16
16
17
20
25
36

38

41

LIST OF TABLES

Table Page

1 Frequency of Proscriptive Responses to Peer
and Parent Punishment Stems By Age . . . 20

2 Summary of Analysis of Variance of Subjects'
Responses to Peer and Parent Stems . . . 22
3 Mean Sureness of Subjects' Responses to Peer
and Parent Stems By Grade . . . . . 22
4 Summary of Analysis of Variance for Subjects'
Sureness Ratings . . . . . . . . 24

iv

LIST OF FIGURES

Distribution of mean peer ratings. The
number following each line is the frequency
of that response. . . . . . .

Distribution of mean parent ratings. . .

INTRODUCTION

In the past ten years researchers have been attrac-
ted to two topics: 1) children's impressions of the factors
which influence social interaction, and 2) the ways that
children acquire prosocial behaviors. Previously a large
amount of attention had been paid to the ways children
learned not to perform behaviors which are socially pro-
hibited (e.g., not to lie, not to cheat, etc.) and to
children's impressions of why an act was good or bad (e.g.,
Pieget's work on moral judgment). The interest in children's
impressions of regulations has continued; only recently,
however, have persons investigated how children develop
tendencies to perform behaviors which are socially admirable
or acceptable. Very little of this research has been con-
cerned with the ways that peer interaction influences these
processes. In this paper I would seek to discuss the ways
the ways that parents and peers affect the ways children
learn what is permitted and prohibited by society. Essen-
tially I will be discussing the development of values.

The topic of values has been discussed by many
persons and a clear definition has not emerged. The term
"values" has been used to describe impressions, opinions,
priorities, goals, etc. in fields as diverse as psychology,
economics, aesthetics and philOSOphy. Rokeach (1973) has

1

2
examined the notion of values and has suggested that
behavioral values can be divided into two groups:
instrumental or means values and terminal or ends values,r”fl
Instrumental or means values refer to behavior or modes
of conduct. Rokeach explains that having an instrumental
value for a particular type of behavior means that one
finds this behavior to be socially or personally preferable
to alternative forms of behavior. Anexample of an instrumental
value is honesty. Terminal or ends values are concerned
with "end states of existence." That is, a person with a
given terminal value considers that particular end state
to be desirable and "worth striving for" (p. 160). An
example of what a terminal value might be is peace or
justice.

When I speak of values in this paper I am referring //
to the internal individual standards used by persons to
guide and judge their own behavior and the behavior of
other persons. By "internal" I mean that a person follows
these values even in the absence of external enforcement.

By "individual" I mean that the values held by one person
may be quite different from the values held by another
person. It is important to remember that there are different
types of values: economic, aesthetic, behavioral, etc. The
major emphasis of this paper will be on behavioral values.

Interest in values can be seen among at least three
groups of psychologists: (1) the psychoanalysts; (2) the

cognitive structuralists; and (3) the learning theorists.

3
Freud considered values to be a part of the child's
superego. The contents of the superego were acquired during

the child's identification with his/her parents. At this

fl... Mew—H...“ .. w- .. ,4...“ “”"“ '

time the child internalizes the standards of the parents.

a... v‘mw— w—c-u -—-—- -——-.~-.—._.._.
s“ —-.. . 7r-~.-.e_. .4

These standards include notions about which behaviors and

 

thoughts are acceptable (i.e., the ego ideal) and which
thoughts and behaviors are unacceptable (i.e., the con-
science). The ego ideal serves as a standard of personal
excellence and it rewards the child through pride when the
child aspires toward this standard. The conscience guides
the person away from those behaviors or thoughts which are
unacceptable. When persons think or do something which is
unacceptable to the conscience they are punished through
guilt (Freud, 1959).

I The Freudian notion of values implies that they are
a part of the personality. Their acquisition is mostly a
result of parent-child interaction and this process occurs
during childhood and is complete before asolescence (Freud,
1959). More recent accounts of these processes have sug-
gested that although a large part of the superego develops
very early in life it can change over time. It has been
suggested that these processes can be influenced by persons
such as coaches, teachers, etc. Very little, if any, con-
sideration is given to peers. (Brenner, 1972)

‘From an analytical pointmnfwnieug if one would like

to understand the development of values one must consider

the structure of the personality. Another perspective

which persons have used to examine the development of
values has not been the structure of personality but
instead the structure of thought. Piaget, and more
recently Eisenberg-Berg and Kohlberg, have examined the
ways that children cognitively approach moral issues.

Piaget classified the development of morals and
values into two distinct stages: first, a period of con-
straint and second, a period of c00peration. The period
of constraint is characterised by the young child's
belief that rules are rigid and unchanging laws conforming
to some type of moral order enforced by an external agent.
Piaget claims that young children think that "things are
‘as they ought to be." He says that children during this
early stage do not base their conception of good on moral
or ethical principles; instead they decide upon an act's
goodness depending on whether it conforms to rules enforced
by parents and other powerful external agents. That is,
the goodness of an act is a function of its adherence to
rules,not to its appeal to abstract moral principles.
(Piaget, 1948)

The second stage is one of cooPeration wherein the
person is no longer primarily concerned with the strict
obedience to rules. Instead they consider the principles
which underlie rules and try to account for the reasons
why persons behave in certain ways. Piaget notes that

these changes are influenced by the child's peer relations

5
and the child's shift in cognitive level. He asserts that
their ability to consider the intentions and motives of
others depends on their ability to decenter. That is, it
depends on their awareness that their point of view may be
different from the perspective of another person. Piaget
claims that these abilities are enhanced by_children's play
in which they experience the roles of other persons (Piaget,
1948). This notion that play and particularly role playing
are important for the development of moral judgment has
been supported by the research of Robert Selman (1971).

Kohlberg has expanded Piaget's notion into a scheme
. of six stages of "value orientations" (1963, p. 13) which
he divides into three levels of morality. He claims these
stages occur in an invariant sequence. (Kohlberg, 1963)

He describes the first level as a premoral level.
Children at this level base their decisions of whether an
act is good or bad on the act's consequences for the actor.
An act is good if the actor is rewarded (stage 2) and is
bad if the actor is punished (stage 1).

Kohlberg claims this first level is replaced by one
which is characterized by conformity to moral rules. He
says that persons in this stage will make judgments accord-
ing to whether an act conforms to a person's role (stage
three) or to rules drawn up by authorities (stage four).

The third level is reached by persons who hold self-
accepted moral principles. Persons on this level look to

individual rights and standards agreed upon by the whole

6

society (stage five) or ethical principles such as equality
of justice (stage six) when they form their moral judgments.

The child's approach to moral judgment as described
by Kohlberg starts at a point where judgments are based on
egocentric considerations and then changes to a stage based
on roles and interpersonal duties. The end of the course
is reached when persons refer to abstract concepts when
resolving conflict between social norm, laws and personal
desires.

Kohlberg claims that changes in moral judgment are
a function of changes in cognitive ability and role-taking
skills. Both of these notions have received strong support.
(See Tomlinson-Keasey and Keasey, 1974 and Selman, 1971.)
Unfortunately, moral judgment is only slightly associated
with prosocial behavior and it does not address the question
of why a person would, or would not, act in a prosocial
manner.

Kohlberg and Piaget are primarily interested in
children's conceptions of rules and their moral decision

making skills. Recently, however, Eisenberg-Berg has

questioned children about the reasons motivating someone
to act prosocially (Eisenberg-Berg, 1977). Essentially
she has found an age-related trend similar to the one
outlined by Kohlberg. That is, young children referred
to instrumental needs and satisfaction when explaining
why someone should or should not help another person.

Older children referred to social norms or approval (e.g.,’

7
"It's nice to help," "He'd like it if I helped"), empathic
understanding (e.g., "I know how he must feel," "I care
about him") or internalized standards (e.g., "I have a re-
sponsibility to help people in need").

The trend that she has found nicely parallels the
changes observed by Kohlberg. Like Kohlberg, she feels that
these changes are due to greater role taking skills and
more sophisticated cognitive abilities among older children.
Although neither Kohlberg nor Eisenberg-Berg directly
assessed the influence of peer experience, they both suggest
that it plays a part in these processes. There are at least
two drawbacks to Eisenberg-Berg's and Kohlberg's schemes.
First, they deal only with changes in reasoning and do not
examine changes in the behavioral tendencies of children
faced with moral dilemmas. Second, they have noted changes
in the ways children think about the morality of a person's
behavior in a hypothetical situation without explaining the
amount of attention that children pay to the different
features of the stories they use. In other words, they have
noted differences in reasoning and they have concluded that
these changes are due to the acquisition of more sophistica-
ted cognitive interpretations of these stores when in fact
it could be due to a change in the patterns of attention
that children pay to the stories' features at different ages.

A third group of psychologists have proposed another

 

model to explain the development of values. Unlike Freud

8
do not look to structural aspects of thought or personality.
Instead they stress behavior and observation of other
persons. Hill (1960) asserts that very little can be said
about internal processes of values and he suggests that
persons should be concerned with behaviors indicative of
values. A large amount of research has demonstrated that
children will imitate other persons' behaviors and that
observation of models can influence a child's generosity and
propensity toward helping (Midlarsky, Bryan and Brickman,
1973). The effects of observation can be seen even two
months later (Rushton, 1975) and can be more influential
' than verbal encouragement (Yarrow, Scott and Waxler, 1972).

Perhaps learning theory can provide a useful frame-
work to examine the ways children acquire an understanding
of the distinction of those behaviors which are socially
admirable and those which are socially prohibited. Hill
(1960) points out that the notions of punishment and reward
are implicit in Freud's discussion of the conscience and
ego ideal.

McKinney (1971) has proposed another dimension to
complement the reward-punishment dimension. The dimension
described by McKinney involves "right doing" and "wrong
doing." When these two dimensions are combined they form
a four-fold model in which a person could be rewarded for
either "doing" or "not doing" and punished for either
"doing" or "not doing." McKinney outlines two systems:

the prescriptive system and the proscriptive system. In

9
the prescriptive system a person is rewarded for doing and
punished for not doing. For example, a person might be
rewarded for doing well in school and punished for not doing
well in school. In a proscriptive system a person may be
rewarded for not failing in school and_punished for failing
in school. In other words, the proscriptive system stresses
the "thou shalt nets" and the prescriptive system stresses
the "thou shalts." McKinney (1971) conducted a study to
determine the extent of the prescriptive and proscriptive
orientations. He asked 67 college students to complete
sentence stems and then scored them according to their
prescriptive or proscriptive orientation. Almost all the
responses to the sentence stems involving a reward situation
. were prescriptively oriented. That is, nearly all the
pe0ple thought they were rewarded for doing. The stems
involving punishment, however, were completed with both
prescriptive and proscriptive responses. Apparently,
sometimes persons thought that they were punished for
doing and sometimes they were punished for not doing.

The notion of prescriptive and proscriptive values
has not received much attention in the literature. However,
one can find inferences to this dimension in the results of
some studies conducted to examine related topics. For
example, Crandall, Katkovsky and Crandall (1965) reported
the results of a study designed to examine children's locus
of control in academic situations. One of these results

bore a resemblance to the prescriptive/proscriptive

10

dimension discussed by McKinney. They found that "there
is a trend for a young child who wishes to appear socially
acceptable to deny that he is at fault when he fails and
for the older child who seeks social acceptance to claim
credit for his successes" (p. 107).

In other words, the young children they referred to
attributed their failure to external sources whereas the
older children they referred to attributed their successes
to internal causes. The yOunger subjectsof their study
were in grades three, four and five and the older subjects
were in grades six through twelve. From the perspective
of the prescriptive/proscriptive system it seems that the
younger subjects they mentioned were concerned with "wrong
doing" (i.e., denying failure) while the older subjects were
concerned with "right doing" (i.e., claiming credit for
successes).

The results of a study conducted by Ervin Staub are
similar to the results of the Crandall, Katkovsky and Cran—
dall study. Staub (1971) was interested in knowing how
persons would respond to cries of distress. He told his
subjects he was interested in collecting some information
and he asked them to fill out a form containing a series
of questions. Each subject was placed in a room adjoining
another room where they were told another subject was
completing the same form. Actually, no one was in the other

room. Staub divided the subjects into three groups; each

11
group received a different set of instructions. One group
was told they were permitted to go into the other room
while they were completing the form or after they were
finished. A second group was told they were not permitted
to go into the other room in order to prevent interaction
between the subjects. The third group was given no informa-
tion about going into the other room. Within each group
there were two age levels: the younger subjects were seventh-
grade girls and the older subjects were women between the
ages of 18 and 23 years. As the subjects were completing
the forms Staub played a tape recording of a person crying
for help from the adjoining room. Staub recorded how fre—
quently persons in each group responded to the calls of
distress. A response was considered as either going to
the other room or notifying the experimenter that someone
was in trouble.

There were no differences in response according to
age in the permission or prohibition conditions. That is,
in the permission condition both age groups responded with
equal frequency. Again in the prohibition condition both
age groups responded equally as often. Not surprisingly,
the level of response for both age groups was lower in the
prohibition condition than in the permission condition.
There was, however, a difference in the no-information
condition. In this condition the response rate of the
older subjects was similar to the response rate of the

older subjects in the permission condition. The younger

12
subjects in the no-information condition responded at a
rate resembling the prohibition. The older subjects treated
the no-information condition like the permission condition
whereas the younger subjects treated this condition like
the prohibition condition.

In his discussion of these results Staub suggested
that the socialization of young children may overemphasize
the teaching of prohibitions. He claims that when faced
with the possibility of disapproval for not helping or
disapproval for doing something not specifically permitted
the young children prefer the former.

As mentioned earlier there are similarities between
the results obtained by Staub and the findings of Crandall,
Katkovsky and Crandall. It seems that in both studies the
younger subjects were concerned with "not doing" (i.e.,
not accepting failure and not breaking a rule) and the
older subjects were concerned with "doing" (i.e., accepting
credit for successes and helping).

If McKinney's concept of prescriptive and proscrip-
tive values were applied to these results it would appear
that the younger subjects were operating in a proscriptive
system. A possible explanation for this might be found by
considering the social settings of the older and younger
subjects. As children grow older their peers become a more-
salient feature of their environment. Sullivan (1953) has
noted that peer interaction becomes more important as

persons become older. Many studies have noted the

13
importance of peer expectations, especially among young
adolescents (Williams, Iscoe and Harvey, 1963; Hartup,
1970). This does not mean that parental influence is not
important; it merely suggests that peer influence becomes
increasingly important with age. Brittain (1963) demon-
strated that both peer and parental influence are important
and that this importance varies with the characteristics
of the situations.

If there is a relationship between the shift from
parent to peer influence and the shift from a proscriptive
to a prescriptive orientation then there should be a differ-
ence in the ways that the values orientations of peers and
parents are perceived. Peers should be perceived as being
prescriptive and parents should be perceived as being pro-
scriptive.

Data previously collected by McKinney (1971) were
re-analyzed and the results partially supported this pro-
posal. Sixty—seven college students were asked to complete
these four sentence stems: "My parents are upset with me
when I . . . "; "My friends are upset with me when I . . . ";
"My parents are happy with me when I . . . ," and "My friends
are happy with me when I . . . ."

All the reward stems but one were completed with
prescriptive responses.

There were, however, differences among the responses
to the punishment stems. Of the 134 responses to the punish-

ment stems 45 were prescriptive and 84 were proscriptive

14
(five were unscoreable). The first stem (the parent
punishment stem) received 13 prescriptive responses, 51
proscriptive responses and three unscoreable responses.
The second punishment stem (the peer punishment stem)
received 32 prescriptive responses, 33 proscriptive re-
sponses and two unscoreable responses. The raw data
indicate that in regard to punishment peers were perceived
as being equally prescriptive and proscriptive whereas
parents were perceived as being more proscriptive than
prescriptive. This difference between peers and parents
was significant (x2 = 11.9285, p < 0.01).

These results indicate that the subjects had differ-
ent perceptions of their peers' and parents' values orien-
tation in regard to punishment. Parents were regarded as
punishing for proscriptive reasons more often than pre-
scriptive reasons. Peers were regarded as punishing for
prescriptive and proscriptive reasons with equal frequency.
The nature of these data, however, may indicate a stronger
difference than actually exists. Since the responses
were scored as either prescriptive or proscriptive the
scoring of a response as being one of these excluded the
possibility of its being scored as the other. In summarizing
these data it is possible to say that in terms of punishment
orientation parents were considered relatively more pro-
scriptive than prescriptive and relative to peers they are
more proscriptive and less prescriptive.

One difficulty in comparing parents and peers on

15
the dimension of punishment may be that persons are pun-
ished more often by their parents than their peers.
Parents may be regarded as being more proscriptive because
they punish more often than peers do. It would be necessary
to control for this factor when examining the punishment
orientation of parents and peers.

In light of the previously mentioned theories and
studies I would like to propose two hypotheses:

1. Parents will be perceived as being more pro-
scriptive than peers in regard to punishment, even when
the frequency of the punishment is held constant

2. The prescriptive system will be more apparent

among older children

METHOD

Subjects

Sixty-five children (nine boys and ten girls from
grade five, five boys and sixteen girls from grade eight
and seven boys and eighteen girls from grade eleven) par-
ticipated as subjects in this study; they were all students
in the school system of Mason, Michigan.

The experimenter sent letters to each child's
parents; they were asked to permit their child to take
part in this study. (A copy of the letter of consent is
included in Appendix A.) The experimenter met with the
children who obtained this permission in their classrooms.
Each child was encouraged to read the sentence stems care-
fully and to respond to each item honestly; they were
assured there were no right or wrong answers. All the
subjects, but one, completed all the items. Each of them
participated anonymously and they and their parents have
received a letter explaining the study's results and thank-
ing them for their assistance.

The average age of the subjects was 13.7 years;
within each grade the averages were 16.8 years in grade 11,

13.2 years in grade 8, and 10.3 years in grade 5.

16

17

Procedures

 

The subjects were asked to complete thirty sentence
stems and rate, on a scale of one to five, how sure they
were of their responses. Twelve of these stems were
designed to obtain information about how the subjects per-
ceived the values orientations of their peers and parents.
The remaining eighteen stems were included to prevent
response sets.

Half the stems referred to reward and the other

half referred to punishment. The reward stems were:

"My friends are happy with me when I . . .
"My parents are happy with me when I . . .
"My parents are pleased with me when I . . .
"My parents are pleased with me when I . . .
"My mother is pleased with me when I . . ."

"My father is pleased with me when I . . ."
The punishment stems were:

"My friends are upset with me when I . . ."

"My friends are disappointed with me when I . . .

"My parents are upset with me when I . . .

"My parents are disappointed with me when I . . .

 

1This is an accurate list of the stems I used in
this study. Originally I had intended the third stem to
read "My friends are happy with me when I . . . ." Un-
fortunately I failed to notice an error in the questionnaire
form until just before I met with subjects. Since I have
made no hypotheses about the subjects' responses to reward

stems I did not alter the makeup of these stems.

18
"My mother is upset with me when I . . ."

"My father is upset with me when I . . ."

These twelve stems were randomly distributed among
the other eighteen stems (see Appendix B for a copy of the
form used).

Beside each stem a scale numbering from one to five
was provided for the subjects to rate how sure they were of
their responses to that stem. A rating of "one" indicated
they were very sure of their responses whereas a rating of
"five" indicated the subject was unsure of the response.

It took most of the subjects about 30 minutes to
complete the sentence stems and rate their responses.

After they were finished the purposes and the hypotheses of
the study were explained to them. The experimenter answered
any questions they had.

The subjects' responses were scored in this way:
First, each response was scored as being prescriptive or
proscriptive; the responses which indicated that the subject
was rewarded for doing (e.g., "My parents are pleased when
I do well in school") or punished for not doing (e.g.,

"My friends are upset with me when I am not helpful") were
scored as being prescriptive; the responses which indicated
that the subject was rewarded for not doing (e.g., "My
parents are happy with me when I don't do poorly in school")
or punished for doing (e.g., "My friends are upset with me
when I forget about them") were scored as proscriptive.

The values for the ratings were simply the numbers

19
the subjects marked as indicating their sureness of their
responses.
The subjects were also asked to indicate their

age and sex.

RESULTS

In order to analyze these data each subject's
responses were scored as being prescriptive or proscriptive
and each subject was given three scores: two scores were
the number of proscriptive responses when the subject used
to complete the peer punishment stems and the parent punish-
ment stems; the third score was a combination of the first
two scores. The first two scores could be either 0, 1 or 2
and the third score could be either 0, 1, 2, 3, or 4.

An examination of the raw data (illustrated in
Table 1) indicates that the majority of the stems were
completed with proscriptive responses, that proscriptive
responses were seen only slightly more frequently among
parent stems than peer stems and that the number of pro-

scriptive stems is lower among the older subjects than the

TABLE 1

FREQUENCY OF PROSCRIPTIVE RESPONSES TO PEER
AND PARENT PUNISHMENT STEMS BY AGE

 

 

 

Type of Stem Grade 5 Grade 8 Grade 11 Total
Peer 29 28 26 83
Parent 31 29 26 86
Total 60 57 57 169

 

NOTE: N = 57 (19 in each grade level); there were 2
peer stems and 2 parent stems.

20

21
younger subjects.

A three-by-two analysis of variance was performed
using a modified repeated measures design (i.e., an
AX(BXS) described in Keppell, 1973, pp. 433-442) to assess
the differences between the three age groups and the two
types of questions. This design was required because each
subject in each of the three age groups responded to both
the parent and peer stems. Subjects who provided a re-
sponse which was both prescriptive and proscriptive (e.g.,
"My parents are upset with me when I don't pass my courses
and when I do poorly in school") were excluded from the
. data analyses. Also, subjects were randomly withdrawn
from the subject samples of grades eight and eleven to
provide equal sample sizes for each of the three grade
levels. This left nineteen subjects in each grade level.

The results of this analysis of variance indicated
that the differences between the parent and peer questions
and the age groups were not significant (f(l,54) = 0.264,
p < 0.05, f(2,56) = 0.8864, p < 0.05 for peer and parent
differences and age differences respectively). The inter-
action between these two variables was also insignificant.
A summary of these results can be found in Table 2.

The subjects' ratings of their sureness were also
compared to determine if there was a difference between
their sureness of peer and parent questions and between
age groups. The mean ratings (illustrated in Table 3)

indicate more sureness of parent-stem completions than

22

TABLE 2

SUMMARY OF ANALYSIS OF VARIANCE OF SUBJECTS' RESPONSES

TO PEER AND PARENT STEMS

 

 

 

 

Sources SS df MS F P
A 0.89 2 0.445 ' 0.8864 >0.05
S/A 27.11 54 0.502

B 0.1 1 0.1 0.264 >o.05
AxB 0.01 2 0.005 0.01323 >0.05
BxS/A 20.39 54 0.3779

Total 48.5 113

 

NOTE: N = 57, 19 subjects per grade level; A = grade,

B = peer or parent question.

TABLE 3

MEAN SURENESS OF SUBJECTS' REPONSES TO PEER
AND PARENT STEMS BY GRADE

 

 

 

Type of Stem Grade 11 Grade 8 Grade 5 Total
Peer 2.0 2.44 ’ 1.92 2.12
Parent 1.65 1.5 1.57 1.57
Total 3.65 3.94 3.5 3.7

 

NOTE: N = 57, 19 per grade level.

23
peer stem completion. A three-by-two analysis of variance
was used to determine if these differences, the differences
between age groups and the interaction of age and types of
stem were significant. Again, a modified repeated subjects
design was used. The differences between the age groups
were not significant (f(2,54) = 0.6039, p < 0.05), the
differences between the sureness of peer and parents stems
were significant (f(l,54) = 18.58, p < 0.001) and the inter-
action between these two variables approached significance
(f(2,54) = 2.58, 0.05 < p < 0.10). (A summary of these
results is presented in Table 4.) Further analyses of
_ the subjects' sureness ratings were performed to determine
if there were differences within the peer and parent stems
by age and to determine if the differences between peer
and parent stems occurred at all ages. The analysis of
variance was used to compare the three grade levels' ratings
of their sureness of their responses. There were no signif-
icant differences between their ratings of the parent stems
(f(2,54) = 0.638, p < 0.05); the differences between their
sureness of the peer stems was close to significance
(f(2,54) = 2.57, p < 0.10). T-tests were used to compare
peer and parent ratings within each grade level. All the
subjects who participated in this study were included in
these computations. The subjects were more sure of their
parent-stem completions than their peer-stem completions
at all grade levels except grade five (T = 2.14, df = 24,
p = 0.014; T = 2.52, df = 20, p = 0.02; T = 1.62, df = 15,

24

TABLE 4

SUMMARY OF ANALYSIS OF VARIANCE FOR
SUBJECTS' SURENESS RATINGS

 

 

 

Sources SS df MS F P
A 3.9122 2 1.9501 0.6039 >0.05
S/A 174.9 54 3.2388

B 33.7196 1 33.7196 18.5803 <0.001
AXB 9.39 2 4.695 2.58706 <0.01
BXS/A 98 54 1.8148

 

NOTE: N = 57, 19 per grade level; A = grade, B =
peer or parent question.

p = 0.10; for grades 11, 8 and 5 respectively).

DISCUSSION

In this section I would like to discuss three issues:

1) The difficulties of interpreting these results
in light of methodological limitations I A

2) How these results compare with other theories or
notions about parent child or peer interactions

3) Suggestions of how these hypotheses may be
more adequately evaluated

Perhaps the most distressing part of this study is
the low reliability of the measures, particularly the
measures of the peer values orientation and the subjects'
ratings of the sureness of their responses. The reliability
of the values orientations was limited by two factors: a
low number of items used in each measure and an abnormal
distribution of the data. There were only two items used
in each measure. The responses were scored dichotomously
(i.e., as being either prescriptive or proscriptive) and the
distributions were negatively skewed (i.e., in the direction
of the proscriptive items). Having a small number of items
makes it more difficult to account for the total variance
and the skewed distributions limits the power of the
statistical analysis. The reliability of the measures was
determined by computing the correlation between the two
items in each measure. Given the absence of a normal

25

26
distribution one could not find a very high correlation.

Although the data for the subjects' sureness of
their responses weren't dichotomous their reliability was
constrained by the same factors which limited the reliabil-
ity of the values orientation. There were only two items
in each measure and the data were more frequently observed
at the low, or "sure," end of the scale. (See Figures 1
and 2.) The peer ratings were more normally distributed
than the parent ratings.

One interesting aspect of the reliability of these
measures is the higher reliability of the parent items than
~the peer items. The reliability of the peer items was only
0.0876, whereas the reliability of the parent items was
0.3728. The reliability of the values orientation was higher
for the parent items, too. There may be a number of reasons
for this. First, children have a limited number of parents
while they have a large number of friends. It may be easier
to provide reliable information about a limited number of
persons (in this case parents) than a larger number of
persons (in this case friends). That is, when the subjects
were asked about their parents they needed to refer to only
a small group of persons but when they were asked about
their peers they needed to refer to a much larger group of
persons. Perhaps this greater variability among their peers
made it more difficult to make consistent responses about
them.

Another important consideration may be the length of

27

 

 

 

1.5 114

 

 

2 112

 

 

2.5 114

 

 

 

 

Figure 1. Distribution of mean peer ratings. The
number following each line is the frequency of that response.

4.5

28

 

//

 

116

 

18

 

 

Figure 2. Distribution of mean parent ratings.

128

29
time that the subjects have known these two groups of
persons. Clearly children have known their parents for a
longer time than they have known their friends. This longer
period of acquaintance may have helped the subjects to
provide more reliable information about their parents than
their peers.

Perhaps these difficulties also constrained the
subjects' sureness of the peer responses. Since the subjects
may have had more difficulty making general comments about
their peers than their parents they may also be less sure
of their responses to the peer items.

If the measure of the subjects' sureness of their
peer responses had some validity one might expect to observe
a change in this sureness with age. That is, children at
one age may be more sure of their peer responses than
children at other ages. Frequently, early adolescence has
been considered to be a time when children are especially
unsure about their interactions with their peers (Elkind,
1966, 1979; Sullivan, 1953; Dunphy, 1963). These persons
have suggested that new issues have been introduced into the
lives of children at this time and they have not yet become
accustomed to them. One might expect that this unsureness
may be reflected in the subjects' ratings of their sureness
of the peer responses. It is conceivable that the grade-
eight students would be less sure of their peer responses
than the grade-five students (who haven't yet entered ado-

lescence) and the grade-eleven students (who have become

30

accustomed to the new issues of adolescence). Although the
mean rating of the grade-eight students was higher than the
ratings made by the other two grades these differences fell
short of significance. The average rating for grade eight
was 2.44 while the averages for grades five and eleven were
1.91 and 2.0 respectively (F = 2.57, df = 2, p = 0.085).
The power of the analysis of variance, however, was limited
by the absence of normal distributions of the grade-five
and grade-eleven data; both distributions were extremely
skewed. There were no differences on the parent ratings
according to age (F = 0.638, df = 2, p = 0.532).

Another reason for the lower reliability and less-
sure rating of the peer items may be a function of the
quality of peer and parent interaction. Perhaps the con-
sideration of a learning model for the development of
values is more appropriate for parent-child interaction
than peer interaction. Damon (1979), in a recent discussion
of children's social interaction, noted that the parent—
child relationship is frequently thought of as being one of
authority whereas the peer interaction is characterized by
friendship and sharing. In light of this it is doubtful
that the notions of reward and punishment are equally
salient for peer and parent items. Perhaps these issues
are more important in parent-child interaction than in
peer interaction. This may be a reason behind the differences
in reliability and sureness.

It may be useful to consider these differences in

31
salience from the persPective of Kohlberg's ideas of the
stages of moral "values" (1963). He suggests that young
children judge the value of an act according to whether
it is punished or how well it satisfied one's needs or
desires. Later they base these decisions on social
norms and standards; punishment and instrumental satisfaction
have lost their importance. According to this model, and
the age-related trends associated with it, reward and pun-
ishment may be important to young children whereas norms
and social standards become important for older children.
This may explain why this model seemed more reliable for
the parent-child interaction that peer interaction.

It is interesting to note that Freud applied his
notions of reward and punishment in the ego ideal and
conscience (as interpreted by Hill, 1960) to the parent-
child interaction and other interactions where some type of
hierarchy existed (e.g., between children and coaches,
teachers, etc.) but not to peers (Brenner, 1963).

The results of this study are consistent with the
results of one other previous study. The frequencies of the
prescriptive and proscriptive responses in this study re-
sembled the results obtained by McKinney in his original
investigation of these values orientations. The subjects
of this study completed over 70% of the punishment stems
with proscriptive responses, whereas McKinney's subjects
completed 63% of the punishment stems with proscriptive

responses.

32

Although the procedures of this study have yielded
some interesting results which are consistent with other
findings in the literature they were unable to adequately
assess the validity of the hypotheses. Clearly if these
hypotheses are to be properly examined a different strategy
should be used. In light of the difficulties posed by
uneven distributions of the data and low numbers of items
one might hope that increasing the number of items would
be a sufficient solution to these problems. However, it
is doubtful that this type of change would be an adequate
solution. First, the addition of items would probably
increase the possibility of a response set and increase the
probability that the subjects will recognize the purpose of
the study. With only two items some subjects recognized
that some questions were asked more than once. For example,
an eleventh-grade subject who had come to the second parent
stem which read "My parents are disappointed with me when
I . . ." (he had already completed the first parent stem,
which read, "My parents are upset with me when I . . .")
completed the second stem in this way: "My parents are dis-
appointed with me when I answer a question twice." Also,

I spoke with a fifth grader just after he completed the
sentence stems. I asked him if he had enjoyed participating
in this study. He replied that he had but he was a bit
concerned that I had attempted to confuse him. I asked him

what he meant and he said, "You asked some of the questions

33
more than once."

Second, the distribution may not acquire a more-
normal distribution by merely adding more items. In this
study neither parent item had a normal distribution. The
responses which each subject made on each item were
combined to create a new scOre. The distribution of these
data was no less skewed than the distribution of two items
alone.

Perhaps a completely different method would be
needed. One possibility would be to ask a group of children
to rank a number of issues which are relevant to either peer
interaction or parent-child interaction. Each issue would
be provided in prescriptive and proscriptive form. For
example, an issue in parent-child interaction may be "being
neat" or "not being sloppy." The subjects would be asked
to rank them according to their importance within that
type of relationship. The dependent measure would be the
average rank assigned to each item. The experimenter would
examine these ranks to determine if the prescriptive and
proscriptive items were regarded as being equally important.

Another way of investigating this relationship would
be to simply ask children to list ten things that would up-
set their friends and ten things that would upset their
parents. The dependent measure would be the number of pre—
scriptive and proscriptive items mentioned by the subjects.

Although the hypotheses of this study were not

supported it would be difficult to say at this time that

34
they have no validity. The results were insignificant
but in all cases they were in the predicted direction.
This lack of significance may be a function of methodolog-
ical constraints on the data. Or it may be that the learning
model upon which McKinney has based his notion of prescrip-
tive and proscriptive values is salient for parent-child
interactions but lacks validity in peer interaction.

Another issue which must be examined more thor-
oughly is the assumption on which the hypotheses of this
paper are based. It was aesumed that young children seemed
to be Operating within a proscriptive system while older
children were oriented toward a prescriptive system. This
notion was based on inferences from the results of two
studies. These studies involved two select groups of
subjects (one group was high in needs of social approval
and the other included two groups of women who may have
been from widely different populations: one was a group
of seventh graders, and the other was a group of college
students); the results of both studies may not be easily
generalizable to other populations.

In fact, one part of the literature would suggest
that the assumption of this change from a proscriptive to
a prescriptive system may be either too simplistic or even
wrong. Children's tendencies to help may be an indication
of the prescriptiveness of their orientations. If the
assumption of an increase or prescriptiveness is true then

one should observe an age-related increase in helping.

35
Staub (1970), however, found that helping was related
to age in a curvilinear manner. Using a distress situa-
tion similar to the one described earlier in this paper
he noted that children in grade 4 help more than children
in either grade six or grade two. Indeed, the notion that
young children consider their world to be proscriptive
while older children consider it to be prescriptive

should be adequately assessed.

APPENDICES

APPENDIX A

PARENTAL PERMISSION LETTER

Dear Parent or Guardian,

I am a graduate student studying developmental
psychology at Michigan State University. I am currently
studying how children develOp values. In order to
complete this study I would like to have your son/daughter
to complete a questionnaire for me. It should take them
about 30 minutes to do this.

If you child is to participate in this study you
will have to give them written permission to do so. In
order that you can understand what I will ask them to do I
would like to explain to you the procedures of this study.

In this study the participants will be asked to
complete 30 sentence stems and rate how sure they are of

their response. An example of a sentence stem might be:
When I meet a new person I try to . . .

Possible responses might be " . . . to act friendly" or

" . . . not act foolish." Next they will be asked to rate,
on a five-point scale, how sure they are of their answer.

On the scale a child who was very sure of his response
would circle 1 and a child who is very unsure would make
number 5. A child who is some place in-between would circle

36

37
2, 3, or 4.

The questions I would like to ask your children
are very similar to the one in the example. They will be
given clear instructions and asked to be as honest as
possible. Also, they will be assured that there are no
right or wrong answers. All participants will be free to
stop whenever they like. All responses will be treated
with strict confidence and your child will participate
anonymously. After they have completed the questionnaire
I will explain the experiment to them. The full results
of the study will be available upon request.

This study has been approved by the University
Committee on Research Involving Human Subjects and by the
principal of your child's school. I am conducting it
under the supervision of Dr. John McKinney, Department of

Psychology at MSU.

If you have any questions please call me at 353-8418

or 337-8096.
Please sign below if you will permit your child to

participate. Thank you for your help and cooperation.

Sincerely,

William M. Bukowski, Jr.

APPENDIX B
INCOMPLETE SENTENCE STEMS AND

QUESTIONNAIRE INSTRUCTIONS

H c053 we nuHS woman mH Hugues a: munmcam v m N Houam H.m

use m conH>o mu ensues ouch .

n H @000 mnlvlumiimulH A m
H cogs

oE ouHs ooucHommomHo ohm muconom h: whamCSm v m N Houam A.h

so 3 08 m o o moo muons: when .
n H a H m milvnumulmiuH A o
umouum one so we muoofi oHHno manoa m HH ounmcaml v m N Hounm A.m
H cos? GE :uH3 ammo: mH Mosumm >2 ousmcam v m N Houan A.v
H xuo3 HE Honomou as 302m H HH masons Ham A.m

o
muuvuumuumuua
H cons we npws woman who mocoHHm m: mafimGSm v m N Hounm A.N

H 80:3 MHmmse sufis woemmflumm an H munmasm v m m Hausa A.H

.mmcHHoom :30 mac» 0» can»
on on whom om .oocoucom some one» ou Hoofioamm .m Human: may oHouHo
onnmcs no: when HosuHoc who 50% HH .v Hogan: on» oHouHo muons: hhm>

uoc pan nuance mum no> HH “N Hones: may mHoHHo whom >H0> no: pan when

who 90% HH .m “when: on» oHouHo oqucs >uo> mum 50% HH «OHmom may no H
HooEbc may oHoHHo when muo> mum 90% HH .omcommou Hsox no one 90% onnm
so: onom may no game oocoucom on» oouonEoo o>o£ so» Moumm :.Hoonom

ou om ou o>m5 u.coo H omzmoonz no =uon mH “H omsmowag on uanE oocoucom
mHnu Mom mono o3» =.HoEE5m oxHH H: mmz mchchoQ oocoucom onu HH .onme0
Mom .wocoucom ouonEoo m mH “H on pH ochHH con» can hHHsmono mchch
103 oocoucom on“ comm now 0» 50> oxHH o.H puns m.ouom .m on H scum mcHom
onom m mH oco come no acoum :H .moocoucom on no mchchon ecu who SoHom

MH¢EMh mqfiz "xmm mflflmw m0¢

38

39

on oxHH HHHmHoommo H muons: Ham A.hN

w
muneuimuumuuH

H co£3 oE nuH3 common who mucoumm h: OHSmGSm v m N Hanan A.mN
Ho>m o no oGOOEOm no chance » whom .
H H H x H HH mnivinmnuNuuH A mN

ammo so co ensue: ouch .
H n H :3 mlivulmnuanH A vN
c053 oE nuH3 ponds mum muconom Hz ounmcsm v m N HoHSm A.MN

who: m o moon moose oHsmcs when .
H H m mniviumunNunH A NN

a mom: m MHoH o oH muons: whom .
HH H . .mm .Hnfim mlnouimuanuH A HN
H c053 ma nuHS hammn ohm moclom >2 whamcsm v m N Human A.0N
mH Hoonom nacho umwn oxHH H ocHnu one whamcnm v m N HoHSm A.mH
H cogs oE nuH3 uomms mH Hosuom as whamcam c m N Housm A.mH
o m H cabana c ensues mush .
u x H H H mnlvunmllNliH A 5H
H c053 oE :uH3 poucHommomHo mum mocoHHH >2 ohsmccm o m N Human A.mH
Ho Ho: ounmcn ousm .
H o n x H HH milounmilNIIH A mH
c053 Honuo some anon» OHQOoH muons: once A.vH
. miivIIMIINnnH
H c033 oE nuH3 Hagan mum mucoumm Hz muanSm v m N Hoham A.MH
cannon 3o: m 008 cm muons: mush .
H u H :3 mnnviumunmuuH A NH
H c053 oE nuH3 common mum mucwumm Hz oaumczm v m N HoHom A.HH
H c033 oconm onu co mocoHHH HE HHoo H mucosa oHnm A.0H

mnnvuumuumuuH

40

won» cops mmoHono who onoom
H sog3 we :uH3 Home: mH Hosuofi a:

mH kucH3 pecan mcHnu 00H: one

OHHHm g 0H9m .
minvsnmsamnuH A on

NHHHm 5 0am .
munviumnumnuH A mm

mason: mush A.mN

mnivuumuumuuH

REFERENCES

Allen, T. "Consequences when assumptions are not met."
Unpublished lecture notes and handouts, Michigan
State University.

Brenner, C. An elementary textbook of psychoanalysis.
Garden City, N.Y., Anchor Books, 1974.

Brittain, C. V. Adolescent choices and peer parent cross
pressures. American Sociological Review, 1963,
28, 385-391.

 

Crandall, V. C., Katkovsky, W., and Crandall, V. J. Children's
belief in their own control of reinforcements in
intellectual achievement situations. Child Develgp-
ment, 1965, 36, 91-109.

 

Damon, W. The social world of the child. San Francisco,
Jossey-Bass, 1977. '

 

Dunphy, D. C. -The social structure of urban adolescent peer
groups. Sociometry, 1963, 26, 230-246.

 

Elkind, D. Egocentrism in adolescence. Child Development,
1967, 38, 1025-1037.

 

Elkind, D. Imaginary audience behavior in children and ado-
lescence. Developmental Psychology, 1979, 15, 38-44.

 

Freud, S. An outline of psychoanalysis. New York, W. W.
Norton, I969.

 

Guilford, J. P. Psychometric theory. New York, McGraw-Hill,
1954.

 

Hartup, W. W. Peer interaction and social organization. In
Paul Mussen (ed.), Charmichael's Manual of Child Psy-
coloqy (Vol. 2), New York, John Wiley and Sons, 1970.

 

Hill, W. F. Learning theory and the acquisition of values.
Psychological Review, 1960, 61, 317-321.

 

Iscoe, 1., Williams, M., and Harvey, J. Modification of
children's judgments by a simulated group technique:
A normative development study. Child Development,
1963, 34, 963-978.

 

41

42

Kohlberg, W. The development of children's orientations to-
ward a moral order: 1, sequence in the development
of moral thought. Vita Humanae, 1963, 3, 11-33.

 

Lindquist, E. V. Design and analysis of experiments in
psycholoquand education. Boston, Houghton Mifflin,
1953.

 

McKinney, J. P. The development of values: Prescriptive or
proscriptive? Human Development, 1971, 33, 71-80.

 

Midlarsky, E., Bryan, J. P., and Brickman, P. Aversive ap-
proval: Interaction effects of modelling and rein-
forcement on altruistic behavior. Child Development,
1973, 33, 321-328.

Mussen, P., and Eisenberg-Berg, N. The roots of caging and
sharing and helping. San Francisco, W. H. Freeman
Co., 1977.

 

Piaget, J. Moral judgment of the child. New York, Free
Press, 1978.

 

Rokeach, M. The nature of human values. New York, Free
Press, 1973.

Rushton, J. P. Generosity in children: Immediate and long
term effects of modelling, preaching, and moral
judgment. figurnal of Pgrsonality and Social Psy-
chology, 1975, 33, 3, 459-466.

Sekman, R. The relation of role taking to the development
of moral judgment in children. Child Development,
1974, 33, 391-398.

Staub, E., A child in distress: The influence of age and
number of witnesses on children's attempt to help.
Journal of Personality and Social Psychology, 1970,
33, 130-140.

Staub, E. Helping a person in distress: The influence of
implicit and explicit rules of conduct on children
and adults. ‘Journal'of Personality and Sociol Psy-
chology, 1971, 31, 137-144.

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

Tomlinson, Keasy, C., and Keasy, C. B. The mediating role of
cognitive development in moral judgment. Child De-
velopment, 1974, 33, 291-298.

 

43

Yarrow, M. R., Scott, R. M., and Waxler, C. 2. Learning
concern for others. Developmental Psychology,
1973, 3, 240-260.

 

"I'lllllllfllllllllllllI