w
..m‘ .
w. no
.P:
33.3. \P

           

2%
fl

 

 
 

 

      

. ... tun... .7.me , .. 1:44.50 5
:3. wreaqfi: A . r. 4
Ta. P5 Arr ..
s,
.. ”.3 hm
.. 4 7 A
an

 

~r

   

       

‘:

      

A

“9
RM

'rr-.

 

   

i?

       
      

 

r

 
   
 

    

, .
.

ply rll I . mm.

5.... L . 5C1

pl}: W3: 1».

fiafifi.
...

z: _

  

 

.

I

 

:9

.- s
'- USAF

 

‘I

 

Nu.

 

 

1??“

{WEAK}

‘ w

-

'AAAAAD

{8:80
, BAA?
HEARMA

 

.

 

SIAAE

L; .
AAAAA

THE
STUDY ‘0 , 1
1971..

ROOM

R APRRGVA
F'cHAL

"Air the Deg

 

FR? W;

CAHAGIAN
A 3535

l

Ahesi

O TEAC
A

TATUS :IN
T.-

S

 

   

      

  

     

 

4i. ...... A 1.1.2:... ; A. A 1.3,: «at... .5........J..3p9. ...Rultjlw
.51: x .. u. \ ...:a.:au..v...r.d.u . . .. .m'» s k 7 I,
snaking“... “23“ng E..&.ciu.s....=....~..az..§ a. 95% _._wf. ..Aii
I ll. Ev ‘ A it. It I

 

 

21m; w “—7—“

j L my Ii A y

Micfiigrm Scam
Ufitvcrsicy

     
     

This is to certify that the

thesis entitled

STATUS IN THE CLASSROOM AND ITS
RELATIONSHIP TO TEACHER APPROVAL AND DISAPPROVAL '-
A STUDY OF CHILDREN '3 PERCEPTIONS

presented by

Robert W. Herrmann

has been accepted towards fulfillment
of the requirements for

Ph. D. Educational Psychology

degree in

 

 

December 10, 1970
Date

 

0-169

ABSTRACT
STATUS IN THE CLASSROOM AND ITS
RELATIONSHIP TO TEACHER APPROVAL AND DISAPPROVAL -
A STUDY OF CHILDREN'S PERCEPTIONS

by Robert N. Herrmann

A sample of 334 fourth, fifth and sixth grade students (12 classrooms,
h at each grade level) was used in a study of the relationships between
status in the peer group and teacher approval and disapproval. Data was
obtained using a combination sociometric and "guess who" instrument
devised for this purpose. The status variables measured were peer accep-
tance, academic competence and power. Since the instrumentation measured
children's perceptions rather than actual behavior, relationships between
the status and teacher approval and disapproval variables represented the
degree to which a child of a given status was perceived as receiving
teacher approval and teacher disapproval.

The major findings of the study were:

1. Teacher approval showed a moderate correlation with peer accep-
tance for both boys and girls. There was no sex difference in the magni-
tude of the relationship.

2. Teacher disapproval did not show a significant relationship with
acceptance for either girls or boys.

3. Teacher approval showed a high positive correlation (r)0.80) with
competence for both sexes.

4. Teacher approval was more highly related to power for girls than
boys.

5. A significant correlation was obtained between teacher disapproval

and status for boys, but not for girls.

Robert‘w; Herrmann

The findings were interpreted in light of previous studies which
have indicated the extent to which peer group values can influence the
attainment of educational objectives. It was suggested that this study
underscores the irrelevance of the values fostered by the schools in the
awarding of status in the peer group. When such a situation exists, often
the teacher and classroom group work at cross-purposes. The author ad-
vocated the use of democratic procedures in classroom management. It was
suggested that the use of such procedures would bring both the power of
the peer group and the teacher to bear on classroom behavior problems.
It was further suggested that the use of older children to serve as models
for apprOpriate behavior in the school would be helpful in socializing the

young child to the educational environment.

STATUS IN THE CLASSROOM AND ITS
RELATIONSHIP TO TEACHER APPROVAL AND DISAPPROVAL -
A STUDY OF CHILDREN'S PERCEPTIONS

we”
Robert‘w.”herrmann

A THESIS

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

DOCTOR.OF’PHILOSOPHY

College of Education
Department of Educational Psychology

1971

ACKNOWLEDGMENTS

The author wishes to express his sincere thanks for the support and
encouragement given during the entire course of my graduate studies by
my chairman, Dr. Harvey Clarizio. In addition, thanks are due my wife,
Jeanne, for typing and editing the manuscript and also for her limitless

understanding and encouragement during the past 3% years.

11

TABLE OF CONTENTS

T‘b1.0fTab1.3eeeeeeeeeeeeee
T‘b1.0prp.m100’eeeeeeeeeeee

Chap“? I - IntrOduCtion e e e e e e e e e e

Peer Status and Mental Health . . . . . .

Peer Group and Educational Values . . . .

St‘tug O O O O O O O O O O O O O O O O O 0
Purpose and Hypotheses . . . . . . . . . .

Chapter II - Review of the Literature . . .

Definitions of Status Systems . . . . . .
The Sociometric Test . . . . . . . . . . .
The Reliability of Sociometric Tests . . .
Strengths of the Sociometric Technique . .

Limitations of the Sociometric Technique .

Th."G“.83WO"T.3te e e e e e e e e e 0
Factors Related to Sociometric Status . .

Characteristics of those with High Status

Relationships between Acceptance, Competence

The Role of Teacher-Student Interaction on Status

Chapter III - Materials and Methods . . . .
Instrumentation . . . . . . . . . . . . .
subjects . . . . . . . . . . . . . . . . .
Experimenters . . . . . . . . . . . . . .

Pro °.dur. 3 O O I O I O O O 0 O O O O O O 0

iii

and Power

22
2h
25
28
31
31
35
35

39
39

ChapterIV‘ROS‘ultSeeeeeeeeeeeeeeeeeeeeee “2
scoringOfDaueeeeeeeeeeeeeeeeeeeeeee "'2

Tests of Hypotheses Concerning Differences due to Sex and SES
on Status and Teacher Approval/Disapproval Variables . . . . . 43

TCStsorthOHypOthoseseeeeeeeeeeeeeeeeeee “I"

Hypotheses Concerning the Relationships among the Five Major
Variables.............o............ [*7

Relationships among Status SyStems e e e e e e e e e e e e e e 1+8

Relationships between Status Systems and Teacher Approval and
Disappmvaleeeeeeeeeeeeeeeeeeeeeeeee 50

smryoraosfltSeeeeeeeeeeeeeeeeeeeeee 52
heidentalFiNdingseeeeeeeeeeeeeeeeeeeee 53

Relationships between Individual Items and the Three Status
Variables.......................... 53

Correlations with Acceptance . . . . . . . . . . . . . . . . . 57
Correlations with Competence . . . . . . . . . . . . . . . . . 5?
Correlations with Power . . . . . . . . . . . . . . . . . . . 5?
Factor Analyses . . . . . . . . . . . . . . . . . . . . . . . 58
Factor Analysis - Boys . . . . . . . . . . . . . . . . . . . . 59
Factor Analysis - Girls . . . . . . . . . . . . . . . . . . . 62
Adjective Check List . . . . . . . . . . . . . . . . . . . . . 6#
Chapter V - Discussion . . . . . . . . . . . . . . . . . . . . . 67
Representativeness of the Sample . . . . . . . . . . . . . . . 67

Sex Differences on the Mean Teacher Approval and Disapproval

scamseeeeeeeeeeeeeeeeeeeeeeeeeeeo

S

Facmrs Rel‘ted to Status e e e e e e e e e e e e e e e e e e 70
P our e e e e e e e e e e e e e e e e e e e e e e e e e e e e 71
Acceptance e e e e e e e e e e e e e e e e e e e e e e e e e e 72

Comancaeeeeeeeeeeeeeeeeeeeeeeeeee 73

iv

Teacher-Approval/Disapproval . . . . . .
Generality of Status . . . . . . . . . .
The Highly Accepted Child - A Composite
Educational Implications . . . . . . . .

Bibliogr'phyeeeeeeeeeeeeeee

71A
76
78
79
8A»

Table 1
Table 2

TflfleB

Table A

Table 5

Table 6

Table 7

Table 8

Table 9

Table 10

Table 11
Table 12
Table 13

Table In

TABLE OF TABLES

Breakdown of Sample Studied by Grade and Sex . . .

Breakdown of Sample Studied by Parental Occupation
Ratinglndsexeeeeeeeeeeeeeeeeee

Means and Standard Deviations for Each Sex on
the Five Dependent Variables . . . . . . . . . . .

Univariate Analyses of Variance - Sex Main Effect

Relationships among the Five Major variables for
BoySCHdGirISeeeeeeeeeeeeeeeeee

Relationships among the Five Major Variables for
HighandLowSESBoys...........o..

Relationships among the Five Major Variables for
High‘ndLOWSESGiI‘ISeeeeeeeeeeeeee

Names and Brief Content Descriptions of‘Each Item
onthOQuestionnairo.......o.......

Correlations of Each Item with Each of the Five
Major VirifileS for Boys e e e e e e e e e e e e 9

Correlations of Each Item with Each of the Five
143301. V‘riables for Girls e e e e e e e s e e e e

Factor Analysis Boys: Factor'Loadings . . . . . .
Factor Analysis Girls: Factor Loadings . . . . .

Summary of Responses to Adjective Check List for
Boy'deirISeeeeeeeeeeeeeeeeee

Means on the 5 Composite Variables for the 30 Most
and 30 Least Accepted Children of Each Sex . . . .

vi

3?

55

61
63

66

73

Appendix A
Appendix B
Appendix C

Appendix D

Appendix E

Appendix F

Appendix G

Appendix H

TABLE OF APPENDICES

Categorical Listing of Items Used in Scale . . .

McGuire‘White Occupational Level Scale . . . . .

SCflCeeeeeeeeeeeeeee

Correlations of Each Item with Each
Major Variables for High SES Boys .

Correlations of Each Item with Each
Major Variables for'Low'SES Boys .

Correlations of Each Item with Each
Major Variables for High SES Girls

Correlations of Each Item with Each
Major Variables for Low SES Girls .

Mean Scores and Standard Deviations
for Both Boys and Girls . . . . . .

vii

of the Five

of the Five

of the Five

of the Five

on Each Item

89
91
9A

100

101

102

103

104

Al ll‘l lll.‘lltallll'|‘

CHAPTER I -INTRODUCTION

Educators and psychologists have long been concerned with the role
of. the school in promoting the normal development of the child. Some
have argued that next to the home the school has the greatest impact of
any institution on development. However, this impact is not limited to
the cognitive domain, but is felt in the child's social and personality
development as well (Withall and.Lewis, 196“).

From the time a child reaches school age and for a number of years
thereafter, the school provides him with a large share of his interper-
sonal experiences with his age mates. For nine months of the year a
child spends a major portion of his waking hours in school-~much of this
time beirg spent in group activities. By exposure to the in-school peer
group the child comes into contact with a great diversity of viewpoints,
religions. racial and ethnic groups and socioeconomic levels. For some
children-"such as children livirg in isolated or rural areas--the school
msy provide the only Opportunity to interact with children of the same
age (Gronlund. 1959)-

A number of theorists and mental health workers have long advocated
the importance of peer relations in personality development. According
to Harry Stack Sullivan, the peer group serves a dual role:

1. It provides an environment in which the child learns the necessary
social skills.

2. More importantly, by exposure to the wide rage of backgrounds
and skills possessed by other mowers of the peer group the child learns
may of his strengths and weaknesses vie-a-vis other individuals.

2

The end result of this process is that the child attains a more realistic
picture of himself. A consequence of failure to experience normal peer
relations, according to Sullivan, is severe social and emotional maladjust-
ment in adulthood.

Those in the field of mental health have considered peer acceptance
an index of a child's current state of adjustment and a powerful predictor
of future adjustment. They have attempted through the use of sociometric
devices and teacher ratings to identify those children who wen experiencing
difficulties in peer relations. Such a child was considered to have a
high potential for unfavorable adult mental health. Attempts have been
made to use some form of intervention to aid the child in improving peer
relations. Often the intervention strategy involved enlisting the aid of
the teacher (Northway, 1941+).

Is there any evidence to support the notion that peer relations are
important to mental health? Kohlberg, 1970, in a recent review article
on predictors of adult mental health status from childhood behavior.
points out that acceptance by the peer group is one of the best pre-
dictors of future adjustment available. The evidence indicates that
children's sociometric ratings of one another are more predictive than
the rating of adults using clinical judgment and mental health criteria.
Not only have the clinicians been shown to be inaccurate in their per-
oeptions of peer relations, but teachers also are notoriously lacldng
in sensitivity to peer group structure. Such insensitivity has been
demonstrated both through their consents regarding an individual ' s
interpersonal relations in cumulative folders and also by comparisons
of teacher ratings of peer status with sociometric tests (Harnken and

a.

how achievement could take place in a classroom where anarchy reigns.
Thus, the school can be said to stand for a set of values--values which
are often bound to educational objectives.

A number of studies have indicated that peer values are related to
the overall academic effort exerted in a particular educational setting--
even to the extent of influencing the specific subject matter which the
group will learn (Coleman. 1961: Bushnell. 1962: Hughes, Becker and
Greer, 1962).

Colemanwgt‘al. (1966) in the now famous Coleman Report found that
by grade 6 the characteristics of a student body (in terms of background
and educational aspirations) accounted for as much variance in achieve-
ment as did the characteristics of the school itself. Further. for
minority group children these attributes accounted for more variance in
achievement than did either school characteristics or faculty character-
istics (p. 302).

In short, peer group values and characteristics are a potent pre-
dictor of educational outcomes. Where peer group values are supportive
of educational objectives, educational outcomes tend to be favorable.
‘Hhere the peer group does not reward academic effort. educational out-
comes tend to be poor. or at best, mediocre.

To what extent are peer group values consonant with those of educa-
tional institutions? One way of answering such a question might be to
examine the relationships between peer group status and educational
values--as reflected in the behaviors for which students are typically’
rewarded or punished. The notion here is that if educational and peer

group values are consonant. there should be a high relationship between

5
the incidence of those behaviors reflecting educational values and peer
status. If they are not consonant, however, a high relationship between
the two would not be expected.

Status may be defined as one's rank or position in a group. In
research on classroom social structure, one's status is generally
inferred from a score on a rating scale (either peer or teacher ratings)
or through systematic observation of group interaction.

Evidence indicates that status is not unidimensional, but rather
there are several status systems or components to the structure of any
group (Polansky, 1951}; Gold, 1958; Gronlund, 1959). These components
can be separated conceptually. but in reality they seem to be inter-
related (Glidewell gt 5]... 1966). It has been suggested that studies
of classroom social structure must take into account three components:
acceptance. competence and power. Although a large body of research
suggests that both teachers and pupils make distinctions among individuals
along these three dimensions (Shoobs, 19157; Gronlund. 1955a; Evans, 1962:
Glidewell gt a]... 1966) the overwhelming majority of studies have been
concerned only with acceptance (Glidewell 91 51. . 1966) .

Competence, when it has been studied, has generally been defined in
terms of academic achievement. as measured by standardised achievement
tests (Northway, 19M; Bonney. 19M; Grossman and Wrighter, 19%). Very
few studies have dealt with children' s perceptions of their peers' com-
petence and its relation to acceptance. Likewise, power has not been
studied to arv great degree. Most of the studies which have considered
power have used a method for deriving a power score from sociometric
tests which was developed by Moreno (19315). The underlying assumption

of this method is that acceptance is the base of power-an assumption

6

which seems unwarranted (Gardner. 1955: Evans. 1962). As with competence,
very little is known about the rolationship between power and acceptance
(Glidewell 33 g... 1966).

In addition to the general failure to consider components of the
classroom.social structure other than acceptance, there is some question
about the validity of generalizations drawn from these studies due to
the samples used. Many of these studies have used the laboratory schools
of major universities or have used public schools located in university
towns. The result has been an overabundance of studies drawn from
upper middle classcr professional segments of the population. A study
by Pope (1953) indicates, for example, that while direct and open ex-
pression of aggression is negatively related to acceptance for middle
class boys, such is not the case with boys of the lower class. The work-
ing classes apparently'are much more tolerant of and expect a certain
amount of open aggression in boys. This suggests that the bases of
status and the relationships among classroom status systems may show
differences by SE3 of the 3s.

A further weakness of most studies on classroom group structure is
that rarely'if’ever are data presented separataby for'bqys and girls.
There is sufficient evidence to indicate the bases for peer acceptance
show sex differences (Pope. 1953). Kagan and Moss (as quoted by Honsik,
1965) pointed out the necessity of considering data in personality
studies separately for'bqys and girls before assuming that the data from
each sex.may'be pooled. Again, it is probable that the relationships
among classroom.status systems vary for boys and girls.

As was previously indicated, a child's peer group status is a good

7
predictor of future mental health status. If a strong relationship were
established between teacher approval and status, it would suggest that
the status of an isolate might be influenced in a positive direction by

a shift of teacher approval in the direction of the child. Conversely,

if teacher disapproval were shown to be related to low (status, this too
would have obvious implications for action on the teacher' s part.

It was also stated above that the degree to which peer values are
consonant with educational values can influence educational outcomes for
the better or worse. It was suggested that one way of approaching this
question is by examining the relationship between peer status and those
behaviors which are related to academic achievement and the educational
process. If a low relationship beheen the two were found, it would sug-
gest a ”generation gap"--i.e. that the peer group and educators are work-
ing at cross purposes. It would suggest that educators must take into
account the peer group when formulating educational objectives.

As with status, evidence exists that patterns of teacher approval
and disapproval show sex differences. A study by Heyers and Thompson
(1956) indicated that boys are perceived by their classmates as receiving
a significantly larger share of teacher disapproval than girls. There
was no difference in the amount of approval received by boys and girls.

The purpose of the present study is to attempt to clarify some of
‘ the unresolved issues pertaining to the social structure of the classroom.

1. This studyhas attemptedtouseasamplewhichwillbemore
representative of the general population than those used in most previous
studies.

2. This study represents an attempt to explore differences in the

8

relationships among status systems between boys and girls.

3. An attempt was made to examine differences in the relationships
among status systems by social class.

it. An attempt was made to clarify the relationships between teacher
approval, teacher disapproval and status, considering the effects of
both sex and social class on these relationships.

The variables which were considered in this stu® are children's
perceptions of acceptance, competence, power, and patterns of the dis-
tribution of teacher approval and disapproval in their respective class-
rooms. It should be noted that the decision to use children's perceptions
of classroas status and teacher approval and disapproval as opposed to
observational techniques, teacher ratings, etc. was based on the following: I

1. The previously cited evidence pertaining to the use of a socio-
metric measure of acceptance as a predictor of future mental health status;

2. The fact that what seems to be is often more important than what
is. It is a child's perceptions of a peer which guide his behavior towards
this peer. Although 11s perceptions may be to some degree based on objec-
tive criteria, the extent to which this is true is irrelevant. If a child
perceives another as being highly competent or powerful, he will act in an
appropriate manner. Evidence of the relationship between children' s per-
ceptions of their peers and their behavior is illustrated by the so called
ripple effect (Glidewell gt 5].. , 1966) .

The hypotheses which were tested in this study are broken into four
categories:

1. Differences in mean scores between boys and girls on each variable

2. Differences in mean scores between high and low 3155 Se on each

variable

9

3. Relationships among status systems, and
h. Relationships among teacher approval, disapproval and status.
‘Qittaran22§_dus_&9_§sz

1. Girls are perceived as significantly more competent than boys.

2. qus are perceived as significantly more powerful than girls.

3. There is no significant difference in perceived teacher approval

#. qus are perceived as receiving a significantly higher amount of
disapproval than girls.

WW

1. Both high SE8 boys and girls are perceived as significantly
more competent than their lower SES counterparts.

2. Boys and girls of higher 383 are perceived as receiving signi-
ficantly less disapproval than their'lower class counterparts.

3. Boys and girls of higher SES are perceived as receiving signifi-
cantly more approval than.their lower class counterparts.

R a S

l. The relationships between acceptance and competence, acceptance
and power, and competence and power are significantly higher for girls
than for boys.

2. The relationship between acceptance and competence is signifir
cantly higher for high SES boys than for'low'SES boys.

3. The relationship between power and competence is significantly
higher for high SE3 girls than for*low’SES girls.

WWW.
WM:

1. The absolute (unsigned) values of the correlations between

10

teacher disapproval and acceptance and teacher disapproval and competence
are significantly higher for girls than for boys.

2. The relationship between teacher disapproval and power is
significantly higher for boys than for girls.

3. There is no significant difference in the relationship between
teacher approval and acceptance for boys and girls.

1!. Teacher approval is more highly related to power for girls
than for boys.

5. The relationship between teacher approval and acceptance is

significantly higher for high SES boys than low SES boys.

CHAPTER II - REVIEW OF‘THE LITERATURE

o s u s on

As was noted in Chapter I, both teachers and students make Judgments
along three dimensions of classroom social structure--acceptance, come
petence, and power. The first part of this Chapter will be devoted to
looking at previous definitions of these terms. Then the definitions of
these terms as used in this study will be given.

1. Acceptance - Acceptance may be viewed as the degree to which an
individual is liked by or is attractive to others., Most studies of
classroom acceptance have made use of sociometric tests. With such a
test the S is asked to choose the members of the class group which he
would prefer as fellow participants or companions in some activity.

Most such studies have restricted the range of activities to those found
in a school setting--i.e. acceptance in terms of a workmate, playmate, or
someone to sit next to. .A few studies have included sociometric criteria
dealing with out-of-school activities--i.e. companion fer a movie or some-
one to invite to a party, etc. (Gronlund, 1959).

It should be apparent at this point that when talking about acceptance,
one must specify the activity or situation. It has been shown, for
example, that an individual's choices for seating companion, workmate, and
playmate vary considerably with the lowest correlations obtained between
workmate and playmate (Gronlund, 1955a; Shears, as quoted in Evans, 1962,
p. 60).

For the purposes of this paper, acceptance will be defined in terms
of acceptance as a seating companion, as a werkmate and as a playmate on

the playground.

ll

12

2. Eggpgtgggg - Competence refers to the ability component of group
structure--to how well an individual does something in relation to other
group members and in relation to an external standard (French and Raven,
1955). Although a number of definitions of competence have been used,
most studies have not differentiated between interpersonal, work and play
competencies. (Glidewell‘gt‘gl,, 1966). Thus, the specific area of com-
petence must be spelled out for a study dealing with this component of
social structure to have meaning.

Competence has been Operationally defined in a number of ways when
studied in the classroom. Some studies have defined it in terms of
academic achievement as measured by standardised achievement tests
(Northway, 19M; Bonney, 191414; Grossman and Wrighter, 19’48). Other
studies have used sociometric criteria ranging from.the very general (i.e.
‘Who is good at doing what you do in the classroom?) to more specific
ones (i.e.'Who is good at athletich); from.the interpersonal domain (Who
is good at getting along with others?) to specific academic competencies
(Who is good at arithmetic?) (Gronlund, 1959). Competence items of the
"guess who" format have also been used.

In this study, competence will mean classroom work competencies as
Opposed to interpersonal and play competencies.

3. Eggs; - Social power has generally been defined as the potential
ability to influence another or to get another to behave in a certain.way
(Moreno, 193% Gold, 1958; Lippitt and Gold, 1959). Obviously, however,
there are a nmmber of ways in which one can induce another to do some-
thing. Thus, it seems meaningful to distinguish the base of power. The
specific base of power could have a great effect upon the relationship

between power and another component of group structure. For example, it

13

would seem likely that power based upon coercion would have a lower cor-
relation with acceptance than would power based on one's ability to act
as a social reinforcer.

‘Moreno (193“) assumed that acceptance was the underlying base of
power. He inferred the degree of power possessed by an individual from
sociometric data by determining the extent to which this individual was
chosen by high status individuals. According to this definition, the most
pepular child in the class would be the most influential. Such an assump-
tion seems unwarranted (Gardner, 1956; Evans, 1962).

A potentially more useful scheme is that provided by French and
Raven (1955). They defined five types of social power with different bases:

a. Reward power--power based on one's ability to mediate rewards
or lessen punishment for another. I

b. Coercive power-ebased on the ability to mediate punishments for
another:

c. Legitimate power--based on office:

d. Referent power--power which has as its base the identification
of one individual with another, manifested in a desire to be like or imitate
another; and

e. Expert power--a derivative of the perception that another possesses
a high degree of eXpertise in a given area in relation to one's own
abilities in that area, and in relation to external standards.

Originally, an attempt was made to use all five of the above types of
power as defined by French and Raven in this study. However, a number of
difficulties were encountered:

a. Legitimate power would seem to be primarily in the hands of the
teacher in the classroom.

12.

b. While referent power certainly is encountered in the adolescent
group via fads, it was found to be extremely difficult for elementary
children to conceptualize.

0. Expert power could not be differentiated from competence by most
elementary school children.

Power in this study is thus defined four ways:

a. Power in general-“Le. no base is specified;

b. Power as a function of acceptance after Moreno (1931+);

o . Coercive power and

d. Reward power, the last two after French and Raven (1959 .

S c e

One of the most used devices in studying classroom social structure
is the sociometric test and its variants. The sociometric test was
designed to measure the individual's perceptions of the status of other
individuals in a defined group on some specified dimension or criterion.
Since this study deals with children's perceptions of classroom social
structure, it behooves the writer to discuss the sociometric test or
technique in some detail.

The term "sociometry" is generally considered to refer to a number
of techniques devised by Moreno. However some within the field of social
psychology, including Moreno himself, have urged the adaption of the term
to include all techniques designed to measure social behavior (Lindsey
and Borgatta, 1954. p. I405). The following discussion will be limited to
a single of Moreno's techniques-"the sociometric test--and an allied
technique called the "Guess Who?" technique used originally by Hartshorne,
Hay and Maller (1929).

15

The sociometric test can be considered a rating technique in that
each member of a group is asked to rate the other members in terms of
their desirability or attractiveness on some criterion--usually'insolving
the choice of a fellow participant or companion in an activity (Lindsey
and Borgatta, 195”). However, a number of differences between the standard
rating scale and the sociometric test have been noted:

1. No training of raters is required. The criterion upon which other
group members are to be rated is specified. The "rater" then uses whatever
criteria he has at hand to make his choices.

2. The rater, while he is rating the other N-l group members is in
turn rated by the same Nel others. Thus, an individual serves as both
rater and rates on the same test.

3. Interrater reliability is not an important issue with the socio-
metric test. It is assumed s 222.221 that an individual will not be per-
ceived in the same way by each member of the group, and that.rater dif-
ferences on a sociometric test reflect such differences in perception.
Rater differences are treated as true differences--i.e. reflective of the
true group structure--rather than as error.

Moreno (193“) preposed a number of requirements for sociometric test-
ing:

1. The tester should specify the group being rated. ‘Within that
group there should be no restrictions on who can be chosen or rejected.

2. The S should be allowed to choose or reject as many in the group
as he wants.

3. The criterion upon which the group members are to be rated should

be specified. Further, this criterion should be meaningful to the Ss.

16

11-. The results from sociometric tests should be used to restructure
the group.

5. Each individual should be able to make his responses in private
with specific choices and rejections never being revealed to the group.

It has been general practice in research to ignore some of the
requirements specified above--especially those pertaining to restructuring
the group and allowing unlimited choices. The former requires no explana-
tion. In the case of the latter, it has been shown that the reliability
of the test is lowered somewhat by restricting the number. of choices alllowed
(Gronlund, 1959) . However, from a more pragmatic viewpoint, with concern
for rapport with the Ss, time restrictions, and logistics problems in
data analysis, it has become roughly conventional to limit the number of
choices per item to three (Lindsey and Borgatta, 1951+) . One further
desiratum is that some time should have passed between the formation of
the group and testing (Gronlund, 1959) . This allows time for group members
to become acquainted, and form a basis for making Judgments about their
peers. In general, the greater the acquaintance span prior to testing,
the finer the discriminations made by the Ss: hence the greater the informa-
tion about group structure .

WW
One concern in a study of this kind is the degree of stability of the

variables being measured. If a child is high or low in status today, will
he maintain this position over time? A discussion of the reliability of
sociometric tests seems in order.

There is some question as to the applicability of the usual psycho-
metric definitions and methods of determining reliability to sociometric

instruments (Pepinslq, 19’49; Mouton, Blake and Fruchter, 1955: Gronlund,

17
1959). In classical measurement theory, reliability is considered to be
a property of the test instrument itself. However, reliability coefficients
obtained for sociometric instruments using traditional methods are
apparently reflecting a property of the behavior being studied, not a
property of the instrument (Pepinslcy, 19119: Gronlund, 1959) .

While the usual psychological test instrument attempts to measure a
particular behavior by obtaining responses which are related to that
behavior, the sociometric instrument elicits a sample of the actual
behavior being studied-4w. choice behavior (Jennings, 19’4'3) . If we
accept the notion that a sociometric instrument is measuring the actual
behavior being studied, then the problem of reliability becomes one of
determining how stable this behavior is. Thus it would seem impossible
with a sociometric instrument to separate the question of instrument
reliability from that of behavioral stability.

In the absence of the development of special techniques for comput-
ing the reliability of sociometric instruments, investigators have turned
to the usual ones. 01' interest is the fact that the magnitude of the
obtained reliability coefficients depends upon whether one considers the
stability of the actual choices given by each individual, over time, or
the number of choices received by each individual (sociometric status)"-
independent of who gave them. The coefficients obtained using the former
have generally been low. This presumably is because this coefficient
reflects in some degree the fluctuations of interpersonal choice patterns
within the group. However, sociometric status (number of choices received)
seems to reflect a fairly stable individual trait (Gronlund, 1959: Hartup,
Glaser and Charlesworth, 1967) .

18

Studies of the stability of the actual choices given have indicated
that the degree of stability varies inversely with the number of the choice
(i.e. first choice, second choice, etc.) and the length of time between
test and retest.

Criswell (1939) in a study of children in the first.through sixth
grades reported that 38% of the children showed no change in their choices
over a six week period. Forty-two percent showed a single change while
an additional 20% showed two changes. Sixty-nine percent of'the first
choices remained unchanged. The test used had a single criterion with two
choices allowed. Similar results were obtained hy.Austin and Thompson
(19fi8) for sixth graders and for fourth and sixth graders over a four month
period by Gronlund, 1955.

Horrocks and Thompson (19#8) in a sample of sixth through twelfth
grade students found that in general, the choices of girls tended to be
more stable than those of boys. They used a single criterion with three
choices allowed. Seventy percent of the sixth grade girls (versus 50% of
the boys) and 80% of the twelfth grade girls (versus 55% of the boys)
showed no change in their’choices over a two week period.

Singer (1951) reported that 72$ of a group of seventh and eighth
graders showed no change in.their first choice over a period of'1§ years.
The second and third choices showed little stability over this period.

A number of factors have been shown to relate to the degree of
stability of sociometric status (Mouton, Blake, and Fruchter, 1955).

Among these are:
1. The time interval between test and retest:
2. The age of the Ss (the older the Ss, the more stable the results)

3. The amount of time the Ss have known each other;

19

h. The relevancy of the criteria to the activity of the group;

5. The number of choices allowed (the larger the number, the greater
the stability);

6. The size of the group; and

7. The more extreme the score (i.e. stars and isolates), the greater
the stability.

Thompson and Powell (1951) reported five week test-retest coefficients
ranging from.0.85~0.92 for different classrooms of sixth graders. Socio-
metric status was deternined by sunning the number of choices received on
feur criteria-three choices allowed on each. Similar results were reported
by-Witryol and Thompson (1953), using sixth grade classes, the same test-
retest interval, and sociometric test.

Lower coefficients were obtained by Bronfenbrenner (1945) using class-
rooms from.nursery school to sixth grade, and a test-retest interval of
either seven months (nursery school) or five months. Sociometric status
among the nursery school children was considerably less stable than that
among fifth and sixth graders (r = 0.27 versus 0.59).

Hartup, Glazer and Charlesworth (1967) reported a test-retest
reliability coefficient for acceptance. Their sample consisted of 32
nursery school children. The test-retest interval was five months.

'Wertheimer (1957) reported a test-retest coefficient of 0.56 using
high school students as Se and a 1% year test-retest interval. A cor-
ralation coefficient of 0.72 was reported between the initial test and
one given at the end of the first sch061 year (a nine month test-retest
interval).

It is interesting to note that the year-tonyear stability of socio-

metric status appears to be as great as the stability of IQ (Bonney, 19h3;

20

Taylor, 1952) and achievement. Bonney reported one year stability coef-
ficients of 0.67-0.8h for sociometric status, 0.75-0.86 for IQ and 0.60
to 0.83 for achievement (as measured on standardized achievement tests).
The Ss used in this study were followed longitudinally from second to
fifth grade.

W
The sociometric technique has a number of strengths which commend it

to the investigator of peer relations. Some of these strengths have been
mentioned earlier, but it would seem in order to mention them again.

1. W - Unlike questionnaires, rating scales,
observational schemes, etc. sociometric tests are relatively easy to con-
struct. Basically there are two requirements:

a. The criteria used should be relevant to the activities of the
group being tested.

b. The limits of the group to be rated must be specified. Further
contributing to the ease of construct are a number of how-to-do-it books
such as Gronlund (1959).

2. I;gin1ng_gfi_z§tg;§ - Unlike other techniques used for the study
of peer relations, the sociometric technique requires no training of raters.
In addition, unlike the other techniques, interrater reliability is not a
consideration.

3. E§§g_g£_gnmin1§tngtign - The sociometric test can generally be
given to an entire classroom in less than one-half hour. In addition, it
has been stressed in the literature that almost all children find the
experience of taking a sociometric test an enjoyable one (Gronlund, 1959).
The writer observed this first hand while collecting data for the present

Studye

21

4. 3g1igpility_g£_thg_ig§trgmgnt - As was pointed out in the
previous section of this chapter, the sociometric test is a fairly
reliable instrument in terms of test-retest reliability--this in
spite of the relative simplicity of the instrument.

5- WWW - In Chapter I. the use
of sociometric status as a predictor of adult mental health status
was discussed. In this context there is no question but what the
sociometric test has a substantial degree of validity.

6. Nguyen 9f raters psgg - The sociometric test uses a large
number of raters, each of whom has had experience with the criterion--
i.e. acceptance or competence or power, etc. of peers. For any
given individual, the number of raters on a sociometric test is
equal to N-l, where N equals the size of the group.

7. Wm '- The sociometric
test can be considered an unobtrusive measure, a fact which contri-
butes to its validity. Firstly, on many tests it is possible to
gauge the social desirability of particular responses, or to attempt
to answer questions in the manner in which one believes the examiner
desires. It is difficult to see how such factors could be operative
on a sociometric test. Secondly, with many tests involving achievement
or output, the individuals taking the test frequentLy try harder and
score above their typical performance. Again, it is difficult to see how
such would be the case with a sociometric measure. Finally, many tests pro-

voke anxiety in those who take them. Since the results of a sociometric

22

test are not known to any but the individual(s) giving and scoring
it; and since it would be difficult for children to get together and
construct sociometric patterns by comparing reSponses, it would seem
there is little for the subject to fear.

To summarise, the sociometric test is a reliable, valid instru-
ment which has many qualities to commend it both to the classroom
teacher and individuals studying classroom social structure--this in
spite of its seeming simplicity.

WWW

As with any technique the sociometric test has a number of limita-
tions. One can divide these limitations into those inherent in the
test itself and those involving the interpretation of sociometric
data (Lindsey and Borgatta, 1954: Gronlund, 1959).

1. Limitations inherent in the sogiometzic test - It is true
that the sociometric test yields much data about the structure of a
group. It does not, however, reveal the reasons why that particular
group structure came into being. Similarly, it gives information
concerning the status of individuals within the group, but no infor-
mation as to why some group members are highly chosen and some receive
few or no choices. It should be further noted that the data obtained
from a particular sociometric test is based on a particular criterion
or a set of criteria. This places some limitations on the degree to
which one can generalize a child's sociometric status from one
group situation to another. While one's status might be high in one
kind of activity, it might be considerably lower in another.

Hence one should either use fairly general criteria or several

23

criteria in order to provide a basis for generalisation. Although choice
status is relatively stable, the particular choice patterns observed in a
group will fluctuate considerably over a time. Thus any inferences based
on specific choices given and received should be used with.extreme caution.
In some a sociometric test has limitations due to the fact that a particu-
lar group is being rated at a particular time on a particular set of
criteria.

2- WW " Frequently 1’0 has
been the practice to consider a star-that is, an individual who receives
a large number of choices on a sociometric tests-a leader and a person
who is well adjusted. Conversely, an individual who receives few or no
choices or is highly rejected by his peers often has been considered
socially maladjusted or of undesirable character. Although a sociometric
test mty provide a clue as to whom the socially successful and unsuccessful
children in the classroom are, it does not provide information as to why a
child is a star or an isolate. Frequently, one may find that an isolate
is relatively new to the group and thus is not well-known by the other
children. Sometimes an individual may have interests which do not coincide
with those of the group. Further, there are individuals who just don't
tool the need for the close peer relations which are encouraged in this
society. In the case of the Openly rejected individual, the basis of his
rejection mty be for some factor totally extraneous to personality factors.
Thus, in cases where inferences are going to be made about the state of a
child's adjustment, supplementary data is definitely needed.

Sociometric data, because it is based on individual! perceptions of

others may or mty not reflect what is objectively so. Part of the reason

2h
for possible variance from objective reality would be the halo effect.
If a person is perceived as possessing one good quality, it is likely
that he will be perceived as possessing others. The extent to which the
halo effect is operant on sociometric tests has not, to the writer's
knowledge, been investigated.
n n

A technique which is related to the sociometric test is the "guess
who" technique. This technique was originally used in conjunction with a
reputation test by Hartshorne, May and.Maller (1929). It has since been
used in major studies of the personality traits associated with status
(Tryon, 1939: Pope, 1953) and studies of children's perceptions of the
distribution of teacher approval and disapproval among members of class-
room groups (deGroat and.Thompson, 1949: Mayer and Thompson, 1956).

Essentially this technique involves the presentation of a series of
descriptions and having group members identify the individua1(s) who is
best described by each (Gronlund, 1959). Items measuring both positive
and negative characteristics are used. The usual practice is to weight
each mention on a positive characteristic as a plus one and each negative
mention a minus one. The algebraic sum of the positive and negative men-
tions constitutes a "reputation" score. Studies, in which both socio-
metric status and guess who reputation scores were obtained, have indicated
that there is a high correlation between the results obtained with the
two (Gronlund, 1959). In the previously cited studies on teacher approval
and disapproval, the usual procedure of weighting items positively and
negatively was not followed, but rather separate scores for approval and
disapproval were obtained by simply taking the sum of the mentions received

on each kind of item.

25

Stability coefficients obtained for guess who instruments have been
of the same magnitude as those reported for sociometric tests for both
the original reputation test (Tryon, 1939; Pope, 1953) and the teacher
approval-disapproval variant (deGroat and Thompson, 19189) . The restric-
tions as to item content, number of choices allowed for each item and
privacy of choices, which were noted for sociometric tests, also apply
to guess who tests (Gronlund, 1959).

WWW!

In this section research dealing with a number of factors relating
to sociometric status will be discussed. Before proceeding further, how-
ever, it might be helpful to make some cements about sociometric research
in general.

The sociometric test was designed originally not so much for the pur-
pose of research as to serve as an aid in uncovering patterns of friend-
ship, and structuring groups accordingly. During the early history of
sociometric research, the main vehicle for presentation of studies was the
journal W-founded by Moreno. When one reads the early issues of
this journal one immediately gets the impression that its purpose during
this period was to spread the gospel of sociometryu-not present acceptable
studies. Thus one can find numerous articles by school teachers, principals
and admmstrators all extolling the values of the use of the sociometric
test in the classroom. Equal in number were articles by those in the mental
health field praising the use of sociometry in helping them to identify the
isolated, withdrawn, socially maladjusted child.

These articles were long on praise, but short on facts. There
apparently was no standard format for articles. It is not unusual to find

an article which fails to specify anything about the sample other than the

26
fact that it was a group of elementary school children--no grade level,
no number of Se, no number of classrooms used, etc. Further, rarely if
ever was there an acceptable presentation of procedures or statistical
analyses. This period lasted until the middle to late l9#0's. From this
time until the middle 1950's a number of good studies were done, mostly
by researchers interested in group dynamics. Their interest was shifted
to the dynamics of the small group in business and industry during the
late 1950's. During the 1960's there were a small number of studies of
sociometric nature. Most of'these were of reasonable quality.

The writer has attempted, in the following discussion, to be selective
as to the studies presented. An attempt was made to use the most-often-
cited studies. Even these were screened to see if they met reasonable
standards of scientific rigor; and if they did not, they were omitted.
Where possible, a recent study(ies) was included. The writer feels that
what remains represents the "best possible data" that we have at the
present time.

1. IQ_3nd_ggademig_aghigygmgnt_- A number of studies have attempted
to assess the relationship between IQ and acceptance. Bonney (l94h) used
a sample of third, fourth and fifth grade children. He found a moderate
correlation (.hO) between IQ and sociometric status. This finding was
replicated by'Northway (l9hh) and Shoobs (l9h7). Both used samples of
elementary school children. However, neither specified the grade level or SE3
of the children. According to Grossman and wrighter (1948) and Haber
(1956) the relationship between IQ and sociometric status is not linear.
In general, children with lower IQ's tend to receive low scores on

sociometric tests of acceptance and those with high IQ's tend to score

27

high. However, beyond a certain value (IQ of 115) an increase in IQ
does not yield a corresponding increase in sociometric status.

Roff and Belle (1965) compared the relationships between intelligence
and sociometric status in groups of different sociometric backgrounds.
They reported that the difference between groups with high sociometric
status and those with low status differed on IQ from 11.5 to 22.1 IQ
points. These findings were consistent with all socioeconomic levels
and for both sexes. The results obtained for academic achievement are
similar to those reported for IQ. Academic achievement shows low to
moderate correlations with acceptance. However, the reported coefficients
are lower than those obtained between IQ and acceptance (Northway, 19M;
Bonney, 19156: Shoobs, 191W). Unlike IQ, however, there is a difference
in the relationship between academic achievement and status across social
classes. Pope (1953) reported that among lower class boys high achieve-
ment is predictive of low acceptance rather than high acceptance as is the
case for middle class boys.

2. W - Socioeconomic status has been reported to correlate
moderately with acceptance. Lower class children tend to receive lower
acceptance scores than middle class children (Grossman and Wrighter, 1948:
Neugarten, 1952) . The greatest choice tendency is for children to choose
others of the same social class, and then a tendency to select children
of a higher class. Rarely are children of a lower class chosen. However,
if the discrepancy between classes is too great (i.e. upper-upper and
lower-lower) the tendency is to reject one another (Neugarten, 1952) .
Other studies have not supported this relationship between SES and acceptance

(toung and Cooper, 19“” Davis and also Dalke as quoted in Gronlund, 1959,

28

p. 209) . Gronlund (1959) suggests that the degree of relationship
between SES and acceptance may be a function of the degree of social
stratification of the comnity.

3- W " A number of
studies have reported a significant relationship between athletic skill
and acceptance for boys (Bretsch, 1952: Feinberg, 1953) and in at least one
study, for girls (Polansky, Lippitt and Redl, 1950) . Other "skill" factors
which seem to contribute to acceptance are "good at doing things" (Polanslq,
Lippitt and Redl, 1950) and frequency of participation in sports and extra-
curricular activities (Feinberg, 1953) .

Young and CoOper (1913+) found no relationship between acceptance and
body proportion, height and weight. However, they did report that children
with high acceptance scores tended to have nicer facial features (as rated
blind by both students and teachers). lesical attractiveness as manifested
by descriptions such as "neat", "good looking" etc. is consistently attri-
buted to children high on acceptance (Tryon, 19393 Austin and Thompson,
191483 Pepe. 1953)

W

A number of studies have attempted to define those personality
characteristics which are typical of those who are high in status in the
classroom.

1, ...] e ..g; _ 1-... 7 . ._ .9,» ,fi .. , ..
m - The major attributes of those with high acceptance seem to be
associated primarily with social skills. High acceptance children have
been reported to be more euctrovemd and self-confident (Young and Cooper,
19““) . In addition, the high acceptance child is one who "knows how to

29

have fun" and "has a good sense of humor" (Tryon, 1939 and Gold, 1958).
Pope (1953) reported that highly accepted uddfle class boys are perceived
by their classmates as being friendly, personable, studious and conform-
ing. Lower class boys who were more accepted by their peers, however,
tended to be aggressive, belligerent and domineering. Little difference
between characteristics of highly accepted middle class and lower class
girls were reported. Highly accepted girls of both classes tended to

be friendly, neat in their appearance, good-looking, outgoing and good
students. Kohlberg (1970) reported that highly accepted children are

more mature in their moral develOpment than their less popular age mates.
In contrast, those who are isolates or neglectees have been reported to

be socially uninterested, socially ineffective or withdrawn, lacking in
vitality, careless in appearance and lacking interest in peOple, activities
or the outside world (Northway, 1941+) . Pope (1953) reported that direct
expressions of aggression by girls regardless of social class and by middle
class boys were related to low acceptance.

2. WW - As yet, there is a very limited store
of knowledge concerning the correlates of classroom competence as measured
by sociometric or near sociometric techniques. As one might expect,
middle class children are perceived as being more competent in classroom
affairs than lower class children. Further, boys are perceived to be
more competent in arithmetic and games than girls (Pope, 1953) . Those
with high IQ's are generally perceived to be more able in classroom work
(Zander and VanEgmond, 1958). In addition, they reported that boys who
are viewed as being more competent in general (i.e. not only on school

work) tend to be more powerful. This was not the case for girls.

30

Interpersonal skills show low positive correlations with competence for
both boys and girls. More competent children are generally viewed as
more friendly, personable and attractive (Pope, 1953) and are more
likely to be imaginative and to have good ideas about how to have fun
(Tryon. 1939) -

3. W - Correlates of power in the classroom have
been studied in some detail by Pope (1953), Zander and VanEgmond (1958)
and Gold (1958). In each of the three studies, high power children were
viewed as more skilled in interpersonal relations--being perceived as
more attractive, more considerate and more friendly than less powerful
children. However, this relationship apparently holds only for middle
class children. Lower class children perceived as high on power are
also perceived as being belligerent, domineering, bullies and trouble-
makers. Powerml lower class girls are viewed as being tomboys, rowdy,
attention-seeking, and aggressive (Pope, 1953)

Zander and VanEgmond (1958) reported that high power girls were per-
ceived as possessing high ability on choolwork. Only high power boys who
were of high intelligence were viewed as being more competent in school-
work.

Gold (1958) found that high power boys were reported to be strong and
were able to fight. Both boys and girls high in power were perceived as
"having things you'd like to have", "doing things for you", "good at making
things", and as 'having good ideas about how to have hm". The items with
the highest relationship to power were those involving the interpersonal
skills mentioned above.

 

As mentioned earlier in this paper, acceptance, competence and power
are interrelated. Glidewell gt 3.1. (1966) present average correlations
between these three factors. The correlations represent an average based
on a summary of all available findings concerning these relationships.

The resultant correlation coefficient were: between acceptance and com-
petence 0.40: between conlpetence and power 0.303 and between acceptance
and power 0.60. Some view these correlations as indicative of a halo
effect--the highly accepted child being perceived as high on most traits
(Lindsey and Borgatta, 1954) . Others interpret these findings as suggest-
ing a "g" factor for group relations much as_Spearman's g for intelligence
(Gronlund, 1959) . It should be pointed out that if the measure of power
were based on Moreno's power-based-on-acceptance measure, the correlation
between power and acceptance reported by Glidewell is possibly inflated.
Essentially, one is using different subsets of the same set of data to
obtain both measures. In addition, as was pointed out previously, it
seems untenable to assume acceptance is the sole base or even primary
base of power in the classroom. Pope (1953) found that among lower class
children, power and acceptance were only slightly related. In general,
those factors which made for low acceptance among middle class children
(i.e. high aggressiveness, fighting, etc.) made for high power among lower
class children. 1
WWW

Implicit in many of the writings on social interaction in the class-
room is the notion that patterns of student-teacher interaction are
related to a pupil's status. (Nithall and Lewis, 1961!; Glidewell 93, 31.
1966) . In spite of the seeming importance of such a relationship, if it

32
does indeed exist, relatively little research has been devoted to this
topic. Polansky (195“) studied the relationship between the mental health
climate in the classroom and the teacher's supportiveness of group status
systems as measured by a sociometric test. Supportiveness of group status
systems was defined as the extent to which teachers had more learner-
centered contacts with high status children than with low status children.
An observation scheme developed.by“Withall was used to obtain measures
of the above. This scheme defines teacher-pupil contacts as learner-
centered, teacher-centered and neutral. It further breaks down into
categories such as praise, blame, etc. It was found that in classrooms
with "good" mental health climates teachers had more learner-centered
contacts with high status children than in classrooms with a "poor" climate.
There was no difference observed in teacher-centered and neutral contacts
between good and poor classrooms. However, no data was presented to show
the magnitude of differences in learner-centered contacts between high and
low status children. Further Polansky dealt with observations of teacher-
student interaction rather than the students' perceptions of these inter-
actions.

Flanders and Havumaki (1960) reported that teacher praise directed
toward particular students will enhance their sociometric status. Thirty-
three groups of ten high school students each were led to believe that
they had been selected as potential participants on a radio quiz show.
Each group met with a teacher-trainer or coach who was to guide each
group towards the selection of five persons who would "actually" partici-
pate in the show. The number of friendships and acquaintanceships within

the groups was low.

33

Under the experimental condition the students sitting in certain
chairs were praised continuously by the teacher and allowed to make com!
ments and suggestions freely. The students in the other chairs were not
praised and not even allowed to talk. At the end of the session, each
student was asked to choose five persons who would be good participants
on the program. Significantly more choices went to the praised group
than the other group.

Given a situation in which persons:

1. Are put into a group of strangers;

2. Have an attractive goal held up to them:

3. Observe an authority figure praising some of the members of this
group: and

h. Observe this same authority figure allow only certain members to
talk:
it seems unlikely that results other than those reported in this study
would be observed. Nature abhors a vacuum. In the absence of any other
data about the individuals in the group, it seems reasonable that those
who were singled out by the teacher, allowed to talk and were praised
would be viewed as better potential program participants or better any-
-thing--simply because the group had only this basis for making a judgment
about them.

In contrast to this experiment, the social structure of the classroom
evolves over time. It is the result of days, weeks, months and perhaps
even years of interaction among students and teachers. It is based on
multiple sources of information gleaned by the children and based on their
perceptions of classroom activities and social interaction. One source

of such information may well be teacher-pupil interaction. However, it

34
seems doubtful that due to the artificiality of the above experiment

such a relationship has been demonstrated. As was previously noted in
a study pertaining to children's perceptions of the distribution of
teacher approval and disapproval, Meyer and Thompson (1956) found that
boys received significantly more disapproval than girls. No difference
in the amount of approval between the sexes was noted.

If higher disapproval leads to lower peer status, one might eXpect
that boys would show a lower mean acceptance score. Such is not the case.
Gronlund (1959) has reported that there is no difference in the acceptance
scores obtained by each sex. This calls into question the role of teacher

disapproval as a determinant of peer group status in the classroom.

CHAPTER III - MATERIALS AND METHODS

a on

The instrument used in this stub was a 25 item scale which was a
combination of three standard sociometric items (the acceptance items)
and 22 items using a "Guess Who" format.

The selection and deve10pment of items proceeded along several
lines. Where possible, items used in previous studies were adopted.
Often, some modification was required before they were usable. For
example, most of the teacher approval-disapproval items were drawn from
a scale developed by deGroat and Thompson (1949) . Although a certain
nunber of these items were usable in their original form, some were
dated in terms of either wording or content. This necessitated the
rewarding of some items, the discarding of sale, and the writing of
new items to take their place. The items chosen for social acceptance
represented the standard triad of classroom seating, work group and
playground activities (Gronlund, 1959) . Items for power and competence
were deve10ped from scratch using the definitions cited earlier as
their basis. To be more specific, the items for competence dealt only
with the concept of classroom cometence. The power items included
one general measure of power, one item dealing with coercive power and
one dealing with reward power.

After a preliminary scale had been developed, four elementary
school teachers were asked to review the items. Specifically they were
to Judge whether an item might present reading difficulty for fourth
and fifth grade children. Each item was also rated as to whether its
content reflected a situation which might occur in their classrooms.

35‘

36

Some it‘s were revised accordirg to suggestions made by the teachers.
The scale was then given to a number of fourth and fifth grade children--
again to screen items for difficulty and ambiguity. On the basis of
this testing, additional revisions were made. The resultant scale con-
sisted of three items each for social acceptance, competence and power,
and eidxt items each for teacher approval and disapproval (see Appendix
A). The items were scrambled such that no two items falling into the
sen category were together. This was done to minimise response sets
on the part of the Ss.

In order to obtain some idea of the basis upon which the Ss made
their Judgments, an adjective check list was developed. This check
list consisted of 20 adjectives or phrases descriptive of qualities
which have been reported in the literature as being important to the
development of friendships. The items in this check list my be found
in Appendix C.

The completed scale was then further tested by administering it to
one fourth grade class. This procedure served to eliminate further
problems with the items and also served to iron out difficulties in
test administration procedures.

some

The Ss were 334 fourth, fifth and sixth grade students in a school
system which is located in a suburb of a large midwestern industrial
city. Data were collected from 12 classrooms-45 at each grade level.
The samle breakdown by grade and sea: is given in Table l.

The area in which the 3s lived could be considered a slice of middle
America. Most of the 8s were second or third generation Americans,
largely of Italian and Polish ancestry. There were no Negroes livirc

37

within the boundaries of the school district. In all but two cases
English was the spoken language at home. The area was composed almost
entirely of modest single family dwellings. Host of the Ss' fathers
were employed in semi-skilled or skilled occupations. However, at one
extreme there were a number employed in professional occupations while
at the other end there were a few on welfare or in unskilled occupa-
tions. To obtain an estimate of a child's socio-economic status the
father's occupation was obtained from school records for each child.

An occupation rating scale deveIOped by.HoGuire and White was then used
to obtain a rating (Kennedy, 1969). With this scale one can assign a
rating from one (professional) to seven (unskilled labor, unemployed,
etc.). The criteria for assigning occupational ratings may be found
in Appendix B. Table 2 contains a breakdown of the sample by parental
occupation rating and sex. Note that the mean occupational ratings for

parents of males and females were nearly the same.

TIBLE 1

Breakdown of Sample Studied by Grade and Sex

 

MW Bill

 

 

 

Grade # 5“ 57 111
Grade 5 6b #9 113
Grade 6 50 60 110

 

row 168 166 331+

 

38

TABLE 2

Breakdown of Sample Studied by Parental
Occupation Rating and Sea:

 

 

 

 

 

Sex W
L L 3 a 5 __6 7
Male 1+ 11 31+ 36 '46 27 10
Female 5 9 32 150 1+6 215 10
TOTAL 9 20 66 76 92 51 20

 

Mean Occupational Rating Hales: 1+. 37
Mean Occupational Rating Females: ll». 36
Mean Occupational Rating Total Sample: 4.36

Additional information such as IQ, achievement test scores, and more
detailed information on the Ss' backgrounds might have been desirable.
However, there is a growing reluctance on the part of both school adminis-
trators and parents to release such information for research purposes. The
school system used in this study requires that when detailed social back-
ground data is to be collected in a study, permission must be granted by
the Ss' parents. There were two reasons for rejecting this options

1. It was not felt that the additional data would be worth the

effort to obtain it.

2. In order to get an accurate picture of the social structure of

a classroom one must obtain data from as many of the students
in the classroom as possible. Traditionally a certain amount
of error has been tolerated in sociometric studies due to

absences on the day of data collection. To have had a number

39

of parents refuse to allow their children to participate in

this show would have tended to increase this error, possibly

to such an extent that the data would be rendered invalid.
W:

In addition to the writer two femle assistants were involved in
data collection. The assistants were both certified elementary school
teachers with teaching experience. They assisted in the distribution
and collection of materials, in monitoring the children as they worked
on the scale, and in answering questions raised by the children con-
cerning scale items. Before data collection began, the writer held
meetings with these assistants in order to work out a standard procedure.
A series of standard exemplars and definitions were developed for those
items which might have proved difficult for some of the 8s to understand.
For example, two of the teacher approval items had to do with who is
typically chosen to be a monitor and who is chosen to run errands for
the teacher. Monitor was defined in terms of in-classroom duties such
as taldng care of blackboards, the aquarium, classroom pets, etc. Run-
ning an errand was defined as a duty which would take the child out of
the classroom, such as taking some papers to the principal's office, etc.

2mm:

Each child was given a copy of the scale which consisted of the 25
items, the adjective check list and a page of instructions (see Appendix
C). A class list was obtained from the teacher and all names were
written on the blackboard. The Ss were then asked whether am of them
wereknownbyanickname otherthanthenamewhichwason theboard.
Although in most cases such changes were trivial-u-such as Bob for Robert
or mck for Richard-in a number of instances the nickname was quite

no

different from the 8's given name. In these cases the addition of this
practice to the procedures proved worth the effort when it came time to
score the data. Attention was then called to the blanks on the top of
the instruction page and the 8s were asked to fill in their name, age,
grade, and teacher' s name. Then the following instructions were given

orall s

y In this booklet are twenty-five items. Some questions
askyoutommethe childrenwithwhomyouwouldliketo
work, play, and sit next to in class. The others are des-
criptions which probably fit some of the children in this
classroom. Below each item are three blank spaces. We
will read each item together. Then I want you to choose
three children who are best described by the item and
write their names in the blank spaces. You must choose
children from this room. You may choose children who are
absent. Write the first name and the first letter of the
last name for each child you choose. You may only choose
three children for each item. However, you may choose the
same child for more than one item. If you cannot think of
three names for an item, write as new as you can. Your
choices will not be seen by mono else, not even your
teacher. Remember:

1. Choose three children for each item.

2. If you cannot think of three children for an

item, write down as mamr names as you can.

3. The children you choose m be in this class.

it. You may choose children who are absent.

5. You may choose the same child for more than

one item.

The children were then asked if they had am questions regarding
procedure. After questions were answered, the experimenter proceeded
to read each item allowim sufficient time between each for the 8s to
respond. Following completion of the 25 sociometric and guess who
items, the children were read the following instructions regarding
the adjective check list:

There are mam reasons why you might choose someone
to be your friend. Below are some possible reasons. Put

a check beside the ones which are most important to you when
you choose a friend.

’41

The Se were then given several minutes in which to complete the list.
It was found that most classes were able to finish the questionnaire
in 25 to 30 minutes.

CHAPTER IV - RESULTS

W
The items on the basic questionnaire were scored by counting the

number of times an individual 3 was chosen by his peers on each item.
The scoring procedures followed were those outlined by Gronlund (1959).
First, second and third choices were weighted equally. Gronlund (1955)
conducted an extensive study of the use of weighting procedures in the
scoring of sociometric data. He concluded that:

1. There is no rational basis for deciding the proper weight to
give each choice.

2. There is no evidence that differential weighting of choices
adds information over that obtained by giving all choices the same
weight.

In addition to the scores obtained from the three power items, an
additional power score was derived from the three acceptance items using
a method described by Moreno (193“). According to Moreno, power rests
upon the potential ability to influence another--ability to influence
another being based on acceptance. The basic idea underlying this scoring
method is that if an individual is chosen by another person, he is in a
position to influence that person. Further, the chooser is in a position
to influence these who choose him. Power is equal to the number of persons
choosing an individual-i.e. those persons subject to an individual's
direct influence--plus the number of persons choosing the choosers--i.e.
those persons that the individual can influence indirectly through the

ones who chose him.

#2

43

Total scores for acceptance, competence, power, disapproval and ap-
proval were obtained for each individual by sunning the scores obtained
on each of the apprOpriate items. The Moreno power score, since it was
derived from the acceptance items, was not included in the total power
score. It was felt that this score, being based on acceptance would be
more related to acceptance than would the other power items. Hence its in-
clusion would possibly inflate the correlation between acceptance and power.

 

advanced which dealt with differences between sexes and SES levels on

the status and teacher approval/ disapproval variables. These twpotheses were:

1. Girls are perceived as significantly more competent than boys.

2. Boys are perceived as significantly more powerhfl. than girls.

3. Neither sex is perceived as receiving a significantly greater
share of teacher approval than the other.

11-. Boys are perceived as receiving a significantly greater share
of teacher disapproval than girls.

5. Higher SE3 children are perceived as significantly more com-
petent than their lower BBS counterparts.

6. Lower $88 children are perceived as receiving a significantly
greater share of disapproval than higher SES children.

7. High 338 children are perceived as receiving significantly
more approval than their lower class counterparts.

Because of the few cases in the sample which fell in the extremes of
the distribution of social class ratirgs, it was necessary to combine
levels to obtain a sufficient N in the high and low 888 groups. The high
338 group was obtained by combining levels 1, 2 and 3: the low group by

2+4

combining levels 6 and 7. The two remaining levels (levels 4 and 5)
became the middle SES group.
Isai§_9£Liha_fl¥naihaaia

The general hypotheses of sex and SES differences were tested with a
two-way multivariate analysis of variance. The multivariate package
programmed by Finn was used (Finn, 1968). The independent variables were
sex and SES. The dependent variables were total acceptance, total come
petence, total power, total disapproval, and total approval--as described
in the last portion of the section on the scoring of data.

The following results were obtained:

1. The multivariate test of the equality of mean vectors between
sexes was significant (F=lz.1563, D.F.=5 and 32h, P<0.001). The means
and standard deviations for each variable for each sex may be found in

 

 

 

Table 3.
TABLE 3
Means and Standard Deviations for each Sex
on the Five Dependent Variables

‘ 32:12. ’ gas; “

323; LP. Mega . E... 9.....—
Acceptance 8.05 6.33 8.61 6.39
Competence 6.61 9.08 8.96 11.10
Power 8.25 9.01 6.76 5.87
Disapproval 33.22 38.86 8.76 16.29
Approval 15.68 16.3“ 25417 29.35

 

 

 

2. The multivariate test of significance for the SES effect was
not significant (F=l.1360: D.F.=lo and 6h8: P<0.3323).

1&5
3. The multivariate test of significance for the sex by SES inter-

action was not significant (F=0.8194: D.F.=10 and 6#8: P§0.6101).

As can be seen from the above, the data are not consonant with the
notion of social class differences. Thus, the three hypotheses concern-
ing social class differences (numbers 5, 6 and 7 above) have not been
confirmed. It is apparent, however, that the data support the notion
of sex differences.

With the Finn multivariate program, not only the multivariate test
of significance is given for each main effect and interaction, but also
univariate analyses for each combination of effect and dependent variable.
To elucidate the source of the significant multivariate test of the sex
effect, the univariate analyses of variance are presented in Table 4.

TABLE h

Univariate Analyses of Variance -
Sex Main Effect

 

 

 

 

Between
...;Eaziabls «JhuuljkLi F' JilduflLJflEBL
Acceptance 0.0365 0.0009 0.9760
Competence 195.0h80 1.7678 0.1846
Power 288.31U3 4.9564 0.0267
Disapproval M66.1135 19.3916 0.0001
Approval 1:775. 5201: 8 . 5962 0 . 0031

D.F.-1 and 328
However, as is the case always with multiple dependent tests on a single
set of data, caution must be observed in interpreting the results. This

is due to one's inability to specify the degree to which the probability

Q6

of a Type I error is compounded in such a situation. The general prac-
tice in a multiple test situation is to use a more stringent alpha level
than would normally be used (alpha divided by K where K equals the number
of contrasts). Such a practice does not allow one to specify the exact
probability of a Type I error where the contrasts are dependent, but it
does decrease this probability (Hayes, 1963, p. #88). In the present case,
it would seem to be safe to consider both the sex differences noted for
disapproval (P<0.0001) and approval (PC0.0037) the sources of the signifi-
cant multivariate test. However, although the P value obtained for power
(P<0.0267) would be adequate for most purposes, in the context of the
present situation it may or may not be safe to interpret this value as
reflecting a true difference.

In terms of the specific hypotheses preposed earlier, the following
statements can be made:

1. The data tend to support the hypothesis that girls are perceived
as being more competent than'boys, but because multiple tests were used,
this result should be interpreted with caution.

2. The hypothesis that boys are perceived as more powerful than
girls is not supported.

3. The hypothesis that there is no sex difference in the amount of
approval received is not supported by the data. Clearly, such a difference
does exist with girls perceived as receiving a higher share of approval
than boys.

llv. The data tend to confirm the hypothesis that there is a difference
in the share of teacher disapproval received by each sex. Table 3 clearly
indicates the direction of this difference--i.e. boys are perceived as

receiving a greater share of disapproval than girls.

l+7

Before moving to the next set of hypotheses, one further item should
be noted in anticipation of the discussion: no difference was found
between the mean acceptance scores for boys and girls. The mean values
of 8.05 for boys and 8.61 for girls are not significantly different.

3- WWWWfiwr
[13:11h1e3, Pearson product-moment correlation coefficients between each
variable were computed for boys, girls, and the high and low SES groups
for both boys and girls. As was the case in the previous section, the
question of the probability of a Type I error had to be considered. How-
ever, with correlations, this problem is manifested in two ways:

1. In testing whether a given correlation is significantly
greater than zero, and
2. In testing the difference between correlations.

In the case of the data obtained from.the entire sample of either
sex, the stringent alpha level used posed little problem. .A correlation
coefficient equal to or greater than 0.25 was significantly greater than
zero at the 0.002 level fer a two-tailed test.

0n the other hand, when testing the hypotheses concerning differences
in the relationships amongvariables between the high and low SES groups
of each sex, the effects were seen. For high SES boys and girls a cor-
relation coefficient of 0.45 was the minimum coefficient significantly
greater than zero (GBO.OOZ for a two-tailed test). For low SES boys
and girls, the minimum coefficient for significance was 0. 50 (d=0.002.
two-tailed). Similarly, a large difference between coefficients was
necessary in order to obtain a significant difference. In reducing the
probability of a Type I error, a price was paid-i.e. the probability of

a Type II error was increased. It is possible that significant

#8

correlations between variables, and significant differences in the cor-
relations obtained for the various high and low SES groups do exist
which are not considered as such in this study. It was decided to err on
the side of a reduction in the probability of a Type I error~~reserving
the right to point out differences in the predicted direction which might
be suggestive of further research.
a S s

§g1_fliffg;§nggg - It was hypothesized that the relationships between
the three status variables--i.e. between acceptance and competence,
acceptance and power, and competence and power~~wou1d be significantly
higher for girls than for boys. The correlations between these variables
are presented in Table 5. All three correlations for both boys and girls

are significantly greater than zero (P<0.002).
TABLE 5

Relationships among the Five Major variables
for Boys and Girls

 

’M’c 9.9m; Inna Disses

'WWWW
Comp 0.5925 0-5769
Power 0 e 5132 0 e 6960 Gelfi 0 e 5969
App 0.1:311 0.5300 0.8022 0.8810 0.2695 0.5838 -0.0726 0.1009

 

 

 

 

 

 

The difference in the correlations between competence and power for the
two sexes is significant (P<0.002). The difference between sexes on the
acceptance-power relationship--although not significant--is in the pre-
dicted direction. There is no difference between boys and girls in the

no
degree of the relationship between acceptance and competence.

W - Two hypotheses regarding differences in the rela-
tionships among status systems for high and low SES boys and high and
low SES girls were advanced:

1. The relationship between acceptance and competence is signifi-
cantly higher for high SE8 boys than low SES boys.

2. The relationship between power and competence is significantly
higher for high SE8 girls than low 888 girls.

The data for high and low 888 boys may be seen in Table 6. Note
that with the stringent alpha level being used, the correlation between
acceptance and competence is not significant for low 838 boys, while it
is for the high 838 group. The difference between these two correlations
is not significant, albeit in the predicted direction.

TABLE6

Relationships among the Five Major Variables
for High and Low SES Boys

 

t _§m 73m: Jams...—
Wmsgs Lgsgg SE LoS ass 0
Comp 0.7028 0.3833
Power 0.l¥700 0.6059 0.38514 0.l&19l

 

 

 

 

 

 

The correlations for high and low SES girls are presented in Table
7. As can be seen, the relationship between competence and power is
significantly greater than sero for both the high and low SES groups.

50
The difference between the correlations obtained for each is not signi-

ficant, but is in the predicted direction.
TABLE 7

Relationships among the Five Major Variables
for High and Low SES Girls

 

___Aaasn.L m m M—
WWWW
Comp 0.6%? 0.6131
Power 0.6“35 0.7528 0.7196 0.5958
Disapp -0.1ot:6 -0.0086 -0.1720 -0.2371 0.0522 0.0170
App 0.6222 0.5248 0.9627 0.5967 0.7251: 0.5257 41.1319 4.1345

 

 

 

 

 

 

 

 

 

A total of five hypotheses were tested concerning the relationships
between teacher approval and disapproval, and the three status variables:
1. The absolute (unsigned) values of the correlations between
teacher disapproval and acceptance, and teacher disapproval and competence

are significantly higher for girls than for boys.

2. The relationship between teacher disapproval and power is
significantly higher for boys than fer’girls.

3. There is no significant difference in the relationship between
teacher approval and acceptance for boys and girls.

8. Teacher approval is more highly related to power for girls than
fer boys.

5. The relationship between teacher approval and acceptance is
significantly'higher for high SE8 boys than for low SES boys. The data

51
obtained for boys and girls may be seen in Table 5. Considering hpothe-
sis 1—tbe relationships between teacher disapproval and competence and
teacher disapproval and acceptance—note that for both boys and girls
these correlations are insignificant. Thus, the data do not support
the notion that there is a difference in the degree to which acceptance
and competence are related to teacher disapproval for boys and girls.

Ruining the correlations between power and teacher disapproval
for boys and girls, it may be noted that while a significant positive
correlation between the two variables was obtained for boys, the cor-
relation obtained for girls was not significant. Even had the correla-
tion for girls been significant, however, the difference between the
obtained coefficients for boys and 311-1: is significant 010.002).

The correlations obtained for both boys and girls between acceptance
and approval were significant positive ones. It was Mpothesised that
there was no difference in the degree of relationship between these
variables for boys and girls. Note that the coefficient for girls is
slightly higher than that obtained for boys. However, this difference
did not prove to be statistically significant.

For both boys and girls a significant positive correlation was
obtained between approval and power. The correlation obtained for boys
could be considered low while that for girls moderate. The difference
between these two correlations proved to be significant at the 0.002 level,
supporting the notion that teacher approval and power are more highly
related for girls than for boys.

The fifth and final hypothesis concerned the relationship between
teacher approval and acceptance for high and low SES boys. It was Impo-
thesised that this relationship would be higher for high SES boys than

52
for low SES boys. An examination of the data in Table 6 indicates that
for'high SES boys a significant positive correlation was obtained between
the two variables while the correlation for low SES boys was not signifi-
cant. Again, when the difference between the two correlations is tested,
the result is not significant, but the difference is in the predicted

direction.
WW

1. The difference in the relationship between acceptance and com-
petence for boys and girls was not significant.

2. The correlations between acceptance and power for girls and
boys did not differ significantly;

3. The relationship between competence and power for girls is signi-
ficantly greater than the corresponding relationship for'bqys.

h. The relationship between acceptance and competence was not sig-
nificantly greater for high SE8 boys than for low SES boys, but was in
the predicted direction. A

5. No difference was obtained in the magnitude of the correlations
between competence and power for high and low SES girls.

6. The correlations obtained between teacher disapproval and com-
petence and teacher disapproval and acceptance were insignificant for
both boys and girls.

7. The relationship between teacher disapproval and power is signi-
ficantly greater for boys than for girls.

8. No difference was noted in the degree of relationship between
acceptance and teacher approval for boys and girls.

9. The relationship between teacher approval and power is higher
for girls than for boys.

53

10. The relationship between acceptance and teacher approval is not

significantly greater for*high SES boys than for’low SES boys.
W

The highest correlations between the five major variables were
those obtained between teacher approval and competence. Note that with
the exception of the coefficient obtained for low SES girls all cor-
relations exceeded 0.80. In terms of a comparison between boys and girls,
the difference between the obtained coefficients was not significant.
The coefficient for high SES boys was slightly higher than that obtained
for low SES boys; but again, the difference was not significant. The
difference between high and.low SES girls on this relationship did prove
to be significant (P<0.002).

C.

 

gazighlgg. Before beginning the discussion of these results, it should
be pointed out that the purpose of this section is purely descriptive.

No attempt was made to test the statistical significance of the data
which are cited. However, since this study is a descriptive-correlational
one, and since a major purpose of such a study is to provide the grist
for further research, it was felt that such an exercise might prove pro-
fiteble. This discussion will center around male-female differences.

In order to facilitate the presentation of this data, each item.has
been given a name to be used in the text. These names along with the
item number, category and a brief description of the item content may be
found in Tehle 8. The correlations between each item and each major
variable may be found in Table 9 for boys and Table 10 for girls. (Although

not discussed here, the corresponding data for’high and low SES boys and
girls may be found in.Appendices D, E, F and G.)

5“

TABLE 8

Names and Brief Content Descriptions of Each Item

on the Questionnaire

 

Item Item Item
Wm
Play 6 Accept
‘work 11 Accept
Sit 15 Accept
Make 1 Compet
well 8 Compet
Inproj 25 Compet
Fight 4 Power
Nice 17 Power
Get 20 Power
Moreno"I - Power
Attend 3 Disapp
Noise 5 Disapp
Punish 10 Disapp
Talk 13 Disapp
Susout 16 Disapp
waste 19 Disapp
Late 21 Disapp
Pickon 23 Disapp
Errand 2 App
Help 7 App
Comment 9 App
Ontime 12 App
Right 1h> App
Things 18 App
Try 22 App
Spproj 2“ App

W

 

Like to play with on playground

Like to work with

Like to sit next to

Good at making things

Does well in school

Thinks of interesting classroom projects
Gets others to do by bossing or fighting
Gets others to do because is nice

Can get others do do something

Moreno's Power Score

Scolded for not paying attention
Scolded for disturbing class

Sent to principal's office or punished
Scolded for talking in class

Suspected while teacher is out

Scolded for wasting too much time
Scolded for late assignments

Scolded for fighting or picking on others
Is asked to run errands

Asked to be monitor

Praised for comments in class

Praised for on-time assignments

Called on for right answer

Praised for bringing special things to school
Praised fer trying hard

Chosen to work on special projects

 

*Derived from the three Acceptance items

55

TABLE 9

Correlations of Each Item with Each of the
Five Major Variables for Boys

 

WW

Play 0.8884 0.4221 0.4542 0.1044' 0.3678
Wbrk 0.9231 0.5719 0.4317 0.0673 0.4236
Sit 0.9460 0.4968 0.5255 0.1362 0.3967
MIR. 0.4944 0.8493 0.2989 '0.0009 0.5185
‘Will 0.3250 0.7142 0.1641 ”01334 0.7279
Inproj 0.1905 0.8881 0.2938 -0.0688 0.7888
Fight 0.3562 0.1137 0.9316 0.5613 0.0907
N100 0.5739 0.6686 0.3406 -O.1589 0.6921
Get 0.4827 0.3064 0.9079 0.3424’ 0.2320
Hereno 0.7732 0.3781 0.4384 0.1409 0.2988
Attend 0.0452 -0.0521 0.2644 0.8858 '0.0325
Noise 0.1489 -0.0214 0.3814 0.8186 -0.0146
Punish 0.1852 '0.0611 0.5354 0.8699 '0.0417
Tllk 0.1454 0.0452 0.3742 0.8109 0.0687
Susout 0.1115 '0.0475 0.4463 0.8862 '0.0136
thto '0.0049 '0.0961 0.2053 0.8736 '0.0970
Late '0.1309 '0.2247 0.0472 0.6911 '0.2329
Pickon 0.2605 '0.0241 0.7539 0.6532 '0.0163
Errand 0.0070 0.3529 0.0232 '0.0315 0.4323
Help 0.3601. 0.11670 0.2205 0.0578 0.5847
Comment 0.3811 0.6428 0.2585 -0.0095 0.7977
Ontime 0.4206 0.5984 0.1751 -o.13#9 0.7048
Right 0.3468 0.6524 0.1723 '0.0957 0.7730
Things 0.2071 0.4495 0.1636 '0.0375 0.6393
Try 0.0730 0.1850 0.0788 0.0710 0.36116
Spproj 0.4028 0.7967 0.2631. -0.o3l+8 0.8202 ‘

 

56

TABLE 10

Correlations of’Each Item.with Each of'the
Five Major Variables for Girls

 

Play
work
Sit

0.8979

0.9391
0.9402

0.4810
0.4966
0-5523

-0.0087

0.8140

0.5681
0.8296

'0.1683

0.0641
0.0204
0.0454

-0.1206
-0.1621
-0.2331

0.1444

0.3227
0.4540
0.5250
0-5055
0.4763
0.2416
0.2195
0.5573

0.5425
0.5980
0.4660

0.8167
0.9285
0.8641

~0.0439

0.6015
0.5221
0-5179

-0.1455

-0.1537
~0.1322
-0.0943
-0.0956
-O.1951
-0.2060
-0.0427

0.5224
0.6265
0.8476
0.8936
0.8743
0.4106
0.4506
0.8673

W

0.6466
0e6683
0.6257

0.4462

0.5178
0.6239

0.4454
0.8835
0.9068
0.5755

0.0704
0.2303
0.1416
0.1824
0.0632

'0-0737
-0.1668

0.3522

0.3657
0.4466
0.5531
0.5268
0.5217
0.2968
0.3312
0.6112

88

-0.1028
-0.0645
-0.0767

-0.1583
'0.2013

-0.1230

0.4904
-0.1983
0.1045
-0.0549

0.8546

A

0.5216

0.5397
0.4195

0.5952
0.8761

0.7875

0.0128
0.6352

0-5233
0.4698

-0.1458

0.6905
0.8716
0.7545
0.7570
0.7633
0.6331
0.6071

-0.1474
-0.1122
-0.0689
-0.0908
-0.2018
-0.l901
-0.0418

-0.0981
-0.0988
-0.1940
-0.2043
-0.l680
-0.1043
-O.1075
-0.1749

0.7320
0.7906
0.8808
0.9012
0.8910
0.5453
0.5131
0.9069

 

57

W
The three acceptance items show about the same degree of relation-

ship to the total acceptance score for both boys and girls. The correla-
tions for boys and girls between acceptance and the competence items
"make" and "inproj" are essentially the same, while that with "well" is
slightly higher for girls than boys. The power item "get" shows about
the same degree of relationship for both boys and 311-1.. However, the
power item "fight" shows a moderate positive relationship to acceptance
for boys and essentially a zero order relationship for girls. 0n the
other hand, the power item "nice" is more highly related to acceptance
for girls than for boys. The correlations between disapproval and accep-
tance ten to be low-"half positive and half negative for girls, and six
positive and two negative for boys. The approval items "errand" and "try"
show essentially no relationship with acceptance for boys. All eight
approval items for girls, and the remaining six for boys show low to
moderate positive correlations with acceptance.
mm

The three acceptance items show about the same degree of relation-
ship to competence for both boys and girls. The competence item "well"
shows a higher correlation with total competence for girls than for boys.
In terms of the power items, the item "get" shows a higher correlation
with competence for girls than for boys. All eight disapproval items
show low most): negative correlations with competence for both boys and
girls.

W
The greatest differences between boys and girls may be seen in the

relationships between individual items and the total power score. The

58
three acceptance items all show moderate positive correlations with power
for both boys and girls. However, the correlations for girls tend to be
higher than those for boys. A similar result was obtained
with the three competence items, although the differences between boys
and girls tend to be larger than with acceptance items. The greatest
difference is with the competence item "well" (0.1614 for boys, 0.5178
for girls). The power items show boy-girl differences also. The second
highest correlation with total power score for girls was obtained with
the item "nice" (0.8835 versus 0.3406 for boys). For boys, the highest
correlation with power was obtained with the item "fight" (0.9316 versus
0.4454 for girls). It should be noted that on the general power item
"get" , the obtaimd correlations with the total power score were virtually
the same for both sexes (0.9079 for boys, 0.9068 for girls). 0f the
eight disapproval items, five show substantial correlations with power
for boys. The items "punish" and "pickon" show the highest correlations
of 0.5354 and 0.7539 being obtained respectively. 01‘ the correlations
between disapproval items and power for girls, only the item "pickon"
shows a substantial degree of relationship. Finally, all approval items
with the exception of "things" show positive correlations in the range
of 0.30 to 0.60 with power for girls. The highest correlation between
an approval item and power for boys was obtained with the item "spproj"
(0.2634).

D. W. Factor analyses were performed separately on
the data for boys and girls. The use of factor analysis represented both
an attempt to simplify the quantity of data obtained in this study, and
an attempt to reach a more precise understanding of the nature of the
variables being measured. Earlier it was pointed out that items were

59
classified as measuring one of the five major variehles of the study
primarily on the basis of’logical rather than empirical criteria. The
use of factor analysis can to some degree be considered a test of the
validity of these logical criteria.

In this study a principal axis analysis was performed followed by a
varimax rotation. Varimax rotation was used because a multifactor solution
with the total variance spread over a number of factors--rather than a gen-
eral factor solution-~was desired. A varimax rotation yields an orthogonal
solution. Some argue that this is unrealistic when applied to psychological
data (because psychological factors are usually correlated) others maintain
that the decision to use an orthogonal or an oblique rotation is largely a
matter of one's taste (Kerlinger, 1964, p. 670). A orthogonal solution often
lends itself to a more forward interpretation than an oblique one. Given
these considerations, it was decided to try the orthogonal rotation and
see whether the solutions obtained were satisfactory.

Initially, the number of factors to be rotated out was arbitrarily
set at five--the number of logically derived variables in the study. How-
ever, the number of factors rotated out in the final solution was depen-
dent upon the eigen values cbtained from the principal axis analysis.
Simply put, the number of factors in the final solution was equal to the
number of factors in the principal axis solution that had an eigen value
equal to or exceeding one. An item was considered to show a significant
loading on a factor if the factor loading was equal to or exceeded 0.30.

Fac r sis - Bo s
The rotated fector loadings for boys may be seen in Table 11. The

proportion of variance accounted for by each factor may be seen at the

60
bottom of the page. Note that the final solution was a five factor one.

Factor I could be considered an academic competence factor. Seven
of the eight teacher approval items, two of the three competence items
and the power item "nice" showed significant loadings on this factor.

The highest loadings were obtained for those items which logically are
associated with classroom competence: doing well in school, being called
on to give the right answer, and turning in assignments on time.

Factor II appeared to be measuring teacher disapproval. All eight
disapproval items plus the power item "fight" show significant loadings
on this factor.

Factor III is clearly the acceptance factor. All three acceptance
items have high loadings on this factor as does the Moreno power score.
Note that this is the only factor upon which the acceptance items show
significant loadings. In addition, the power items "nice" and "get" and
the competence item "make" had smaller, but still significant loadings
on this factor.

Factor IV seemed to represent a second kind of competence. Whereas
factor I was apparently tapping academic competence, factor IV seemed to
relate to special competencies. The highest loadings on this factor
were shown by the items "good at making things", "has good ideas for
special projects", "is chosen to work on special projects" and "is praised
for bringing things to school for others to see". Lesser loadings were
shown by the items “nice" and "is praised for comments in class".

The fifth factor seemed to be a power factor. The highest loadings
were obtained with the power items "fight" and "get", and the disapproval
item "scolded for picking on others". The disapproval items--getting

61
TABLE 11
Factor Analysis Boys: Factor Loadings

 

Factor Factor Factor Factor Factor

 

nag 0.2143 0.0260 -9_._§356 -o.0898 -0.1450
work 0.2274 0.0173 0862; -0.2060 -0.0753
Sit 0.1850 0.0608 m -0.1937 -0.1853
11.1.. 0.1470 -0.0029 -0, 3931 ~g,6zoo -0.0181
14.11 9% -0.1088 -0.1335 -0.1359 0.0291
Inproj 9.15155 -o.0362 -0.2832 —9, 2415 -0.0023
Fight 0.0164 9,3681 -0.2410 -0.0312 M58
Nice 915523 '0.1357 stlQEZ. 2212292. ’0-0573
Get 0.0756 0.1180 12.13538 -0.1939 1.2242
Moreno 0.0938 0.0600 -94§QQ3, -0.1258 -0.2086
Attend 0.0090 9,3556 0.0009 0.0170 -0.0 599
Noise -0.0191 2,8122 -0.0997 0.0000 -0.2312
Punish -0.0171 0.8125 —0.1075 0.0629 -g,_439_3
Talk 0.0450 91.8280 -0.0612 -0.0819 -0.2018
Susout 0.0107 1.8.61.5 -o.0325 0.0126 $3.456
Waste -0.0723 9.3356 0.0245 0.0141 0.0302
Late -0.1573 9.22262 0.0756 0.1350 0.2010
Piokon -0.0350 91% -0.1596 0.0634 W
Errand 9.392 -0.0263 0.2272 -0.2967 -0.2202
Help 9,5022 0.0480 -0.20 58 -0.2423 -0.1044
Cement 0.5991 -o.0156 -0.1625 12.3298 -0.0725
Ontime 9,8831 -0.1273 -0.2354 -0.0291 0.0167
Right 9.12958 ~o.0745 -0.1265 -0 .1580 0 . 0128
Things 0.0638 -0.03 54 -0.0261 b.8026 -0.0438
Try 9.1528 0.1894 -0.0125 —0.0623 0.2148
Spproj 9,3828 -0.0084 01%? 12.28081 0.0086

 

Proportions of Variance
1 2 3 4 5
0.1607 0.2161 0.1419 0.1166 0.0959

62
punished and being suspected of wrongdoing while the teacher is out of
the room-also showed significant loadings on this factor. Thus, five
factors appear to underlie classroom status for boys as measured by the
questionnaire: two competence factors--one tapping academic competence
and the other special competencies: a disapproval factor; an acceptance
factor: and a power factor.
W

The final solution for girls was a five factor solution. The rotated
factor loadings are given in Teble 12, with the pr0portions of variance
accounted for by each factor given at the bottom of the page.

The first factor obtained for girls seemed to be an academic compe-
tence factor. All of the three competence items including the item "make”
and seven of the eight teacher approval items show significant loadings
on this factor. In addition, two power items "nice" and "get" show small
but significant loadings on factor I. The fact that.the competence item
"make" and the approval item "things" show significant loadings on this
factor may indicate that girls make less of a distinction between academic
competency and "special" competencies than do boys.

Factor II was a relatively clearcut factor, tapping teacher disapproval.
Six of the eight disapproval items show significant loadings on this
factor. Factor II is thus similar to the corresponding factor for boys
with one notable exception--the absence of significant loadings with those
items dealing with acting out and aggression--i.e. "fight", "noise" and
"pickon".

Factor III is an acceptance-power factor for girls. The three accept-

ance items, all three competence items, three of the four power items (not

63
TABLE 12

Factor Analysis Girls: Factor*Loadings

 

Factor Factor Factor Factor Factor

 

Play 0. 2616 -0 .0982 818416 0 .0874 0 .0431
Work 0. 2618 -0.0390 -0_.§2_6.a 0 .1385 0 . 0406
31+. 0.1401 -0.o751 10452039. 0 . 0557 0 .0528
Make 0,4229 -0.0316 1.3168 0,3223 -0 .1150
won 0 646 -0 . 0449 013112 915212 -0 - 1211
Inproj g , 93 -o. 0402 91.3236 L369}: 0 0122
Fight 0. 0459 0.0860 0 .1109 o .1234 0 , 8383
N16. 0, 3%; -0.0997 11.2393 9.3256 -0 . 0248
Get 0.3009 -0.0486 -Q,9§99 9,3268. 9,4339
Moreno 0.2609 '0-1063 ’Qi§3&Z 0-0143 0-1185
Attend -o,0419 942999 0.1389 0.0220 9,3960
Noise -o.1o56 0.1898 -0.0810 -0.1318 0,2282
Punish -0. 0707 0279.43 ’0 - 0727 '0 . 0451 (b.4913
Talk 0.0637 9,3522 —0 .0656 -0 .2192 9,6929;
Susout 0 . 0202 91.6119 0 - 0857 -0 .0062 0 . 2451
wut. -0.1167 9,9933 0 . 0780 -0 . 0474 0 .1280
Late -0.1180 9113.922 0 . 1590 -0 . 0271 -o .0624
pickon -0.0581 0.2413 -0.1632 -0.0069 0.2268
Errand 9,7733 -0.0775 ~0-1140 0.0477 0.1090
Help Q‘flllj -0.1079 -O.2343 0.0693 0.1239
Comment 0,§§63 -0.0804 ‘923233. QIE§ZQ. ‘0-0564
0mm. 9 . 66; 9 -0 . 0525 821.3185 91159.4 -0 .1160
Right 0,6 539 -0 .0366 -0 .2753 01.5821 —0 .0702
Things 0,6642 -0.0453 -0.1019 -0.1056 -o.o371
Try 0.1178 -0.0133 -0.0938 0.8008 -o.o428
Spproj O I 2614 '0 . 0337 m 0 . 3862 ’0 . 0640

 

Proportions of Variance
1 2 3 4 5
0.1986 0.1320 0.1860 0.1023 0.1219

64

the power item "fight") and three approval items show significant load-
ings on this factor. Note that five of the twelve items loading on factor
III deal with academic competence, suggesting that in girls approval rests
at least to a small extent on academic competence. None of these five
items show significant loadings with the corresponding factor for boys.

Factor IV apparently represents a second competence factor, but it
does not lend itself to as clear-cut an interpretation as does the corres-
ponding factor for boys. The factor for boys is a second competence factor
involving competencies other than strictly academic ones--i.e. making things,
bringing things to school to share, etc. For girls, again the special come
petencies seem to be to some degree embedded in academic competence: hence
the significant loadings of the competence item "well" and the approval
items "ontime", "right" and "try".

Factor V seems to represent a second power factor for girls--i.e.
coercive power. Significant loadings are obtained for the power items
"fight", and "get" and five of the disapproval items including "pickon",
"noise", and "punish".

Thus, as with boys, a five factor solution is obtained for girls: a
competence factor, a teacher disapproval factor, an acceptance-power factor,
a second competence factor which seems to be more general than that ob-
tained with boys, and a second power factor based on coerciveness. One is
left with the impression (to be taken up in the discussion) that the fac-
tors are less clearrcut for girls and in some instances more difficult to
interpret.

E. W. As was noted previously, an adjective
check list was used to provide some information as to the kinds of

65
criteria used by the 83 in making their choices on the sociometric items.
A summary of the data obtained is presented in Table 13. This table pre-
sents both the percentages of boys and girls checking each item on the
list, and the rank of each item in terms of the frequency with which it
was chosen.

Boys chose an average of 15.52 items on the list, while girls chose
an average of 17.53. The item "friendly" was the most frequently chosen
item for the girls (98.2%), while it ranked as the second most frequently
chosen fer boys (86.3%). The most frequently chosen item for boys was
the item "fun to be with" (89.3%) . This item ranked second for the girls
(97%). The least chosen items were "strong" for girls (25%) and "sharp
dresser" for boys (36.3%).

Some items were chosen more frequently by one sex than the other.
These items chosen by a substantially greater proportion of girls than
boys were: cheerful, outgoing, kind, polite, considerate, honest and
patient. Only two items tended to be chosen more frequently by boys:
strong and athletic.

Looking at the overall trends, although girls tended to choose more
items than boys, their choices tended to converge on a smaller number of
items, particularly those reflecting social skills and traditional social
values. Six items--friendly, cheerful, helpful, kind, honest and fun to
be with-~were chosen by over 90% of the girls' sample. in additional seven
items were chosen by between 80% and 90%.

In contrast, the choices given by the sample of boys were Spread more
evenly over all of the items. Four--friendly, can be trusted, helpful and
fun to be with--were chosen by 80 - 90$ of the sample. No single item
was chosen by as much as 90% of the boys' sample.

66
TABLE 13

Summary of’Responses to Adjective Check List
for Boys and Girls

 

 

Smart at school

Good Sense of
Humor
Can be Trusted

Sharp Dresser

Likes Same Things

Athletic
Sure of self
Outgoing
Helpful

Considerate
Loyal

Kind
Honest

Imaginative

Good at Making

Things
Patient
Polite

Popular

IFun to be‘with

‘Hould.Like to be
Like

Strong

iBoy Rank of
86e3 2
51.2 20
64.3 13
52.4 19
74.4 7
81.5 4
66.1 12
56.5 18
“8.8 22
82.1 3
6205 15.5
67.1 11
79-2 5
75-0 6
73.2 8
63.1 14
6205 15.5
67.9 9
36.9 23
89.3 1
61.9 1?
50.0 21

sass 8528

33-398..

ages
0." oox \O\O‘F' \O O\O\U\O I-‘\O\OO VOVV

as s:

 

CHAPTER V - DISCUSSION

R a a

Among the stated purposes of'this study was the replication of
previous work in this area with an attempt to provide a sample more
representative of school children in this country, especially in terms
of SES. All too often in previous work the samples used were drawn
from laboratory schools or the school systems of major university com-
munities. These samples probably were biased in the direction of
upper middle class and professional values; thus leading one to ques-
tion the extent to which one can generalize beyond this class. The
study by Pope (1953) on characteristics of pepular children among the
middle and working classes provides sufficient cause to question whether
data drawn from children of upper middle class/professional parents is
comparable to that drawn from blue collar samples.

Unquestionably, the sample used in the present study has certain
characteristics which limit its representativeness. Virtually all of
the children in the sample were Caucasian, although a few were of
Oriental extraction. No Negroes were included because there are no
Negroes enrolled in the school system from which this sample was drawn.
In spite of the limitations, however, the occupational mix of this
sample seems to be more representative than those found in the samples
of many previous studies. The most highly represented group was the
skilled, blue collar and service occupations group. However, the
sample did include a number of individuals from both the unskilled-

unemplqyed group and the professional group.

6?

68

a a v

In the study by Meyer and Thompson (1956), cited earlier, children
perceived boys as receiving a significantly greater share of teacher dis-
approval than girls, but boys and girls were perceived as receiving the
same amount of teacher approval. Their finding was in accord with a
number of studies which used observational techniques to measure the dis-
persion of teacher approval and disapproval in the classroom (see Uithall
and Lewis, 1964) .

As in the Mbyer and Thompson study, the boys used in the present
study were perceived as receiving a significantly greater portion of
teacher disapproval than girls. However, unlike their study, girls were
perceived as receiving a greater share of teacher approval than did boys.
Such a conflict in findings could conceivably be explained on the basis of
difference in the modal SES levels of the respective samples, and differ-
ing expectations for the male role in each sample.

Before discussing this matter further, it should be pointed out that
such an explanation suggests a sex by SES interaction in regards the vari~
ables of teacher approval and disapproval. The sex by SES interaction in
the multivariate analysis of variance carried out on the data of the pre-
sent study was not significant. There are two possible explanations for
the failure to obtain a significant interaction: 1. The samples at the
extreme SES classifications used in the present study were very small. In
order to make SES comparisons, it was necessary to combine the upper three
SES levels to Obtain a high SES group and the lower two SES levels to ob-
tain a lower SES group. By combining the extremes with groups nearer to
the middle of the SES distribution it is possible that any differences

69
between the extreme groups would tend to be washed out. The upper group,
for example, would tend to be more like SES level three than SES level one;
the lower, more like SES level six than SES level seven. Such an inter-
pretation would suggest that in order to obtain significant differences
between the different SES levels one must use homogeneous samples of the
extremes rather than the more heterogeneous groups obtained by combining
several SES levels. 2. A second explanation deals with the tendency of
a child's attitudes and values to gravitate towards the modal peer attitudes
and values in a given school. There is some evidence that this tendency is
more marked in the lower than the upper SES groups (Coleman,gtmgl., 1966,
p. 305). It would seem logical, however, that the smaller the proportion
of upper SES children in a particular school the more marked the tendency
to gravitate towards the values and attitudes held by lower SES groups. In
the present study, only 29 children (less than 10$ of the total sample)
came from the two upper SES groups. It would seem likely that their atti-
tudes and values would be more like those of lower SES groups than children
from comparable SES levels in a primarily upper SES school. Thus, either
through the combining of several SES levels into more heterogeneous groups,
or through the actions of the peer group itself (via pressures to conform)
or a combination of both--it is possible that SES differences were mini-
mised and the sex by SES interaction eliminated.

Meyer and Thompson did not present the SES characteristics of their
samples, but apparently it was of upper middle class extraction. If their
sample were of upper middle class children, one would expect there to be no
difference in the amount of teacher approval perceived as going to each sex.
It is typical of white/professional groups to see a narrowing of the dif-

ferences in sex roles. Typically, the males tend to become less openly

7O
aggressive and more oriented towards school achievement. Thus, receiving a
good deal of teacher approval would be consonant with the male role in this
stratum of society. Conversely, with children of blue collar extraction, a
boy is expected to be rough-and-tumble and to be less school oriented than
girls. Thus, high teacher approval would tend to be dissonant with sex role
expectations.

In regards the difference in teacher disapproval between boys and girls
noted in both studies, even in upper middle class society a boy should not
be too much of a "goody gumshoes"--i.e. a certain amount of teacher dis-
approval is consonant with the male role for both strata of society. Role
expectations could account for differences between boys of the upper and
lower middle classes via two mechanisms:

1. It is possible that there are differences in school performance
and conduct such that lower middle class boys actually do receive less
teacher approval than their upper middle class counterparts.

2. It is possible that the stereotypes which underlie sex role expecta-
tions exert an influence directly on the perceptions of the children-~i.e.
there is no difference between the amount of approval received by lower
middle class boys and girls, but according to the stereotypes of male and
female roles there should be. Thus, such a difference is attributed even
though it does not objectively exist. This question can only be resolved by
the use of observational techniques to measure the dispersion of teacher
approval and disapproval in the classroom for children of differing SES levels.

W

The general notion that different factors are related to status for
bcys and girls is supported by the data obtained in this study. The greatest
sex differences were noted on the power variable, but acceptance and com-

petence do show differences that warrant mention.

71
29321:

There is a tendency for the individual acceptance and competence
items to show higher correlations with power for girls than for boys.
This is manifested in the significantly higher relationship between power
and competence for girls. The relationship between acceptance and power
was also higher for girls, but the difference was not significant at the
0.002 level used in this study. It was significant at the 0.01 level and
thus warrants a further look in future studies.

Examining the loadings of the individual power items on the total
power score for boys, one sees that the coercive power item.shows a high
positive correlation with power, while the social reinforcer item shows
a low positive one. The reverse is true for girls. This suggests that
power in boys is related to aggressive behavior and in girls to accept-
ance and ability as a social reinforcer. In addition, power for both
sexes is related to classroom decorum-‘but again in different ways for
each sex. Teacher approval shows a significantly higher relationship
with power for girls than for boys. Teacher disapproval for girls shows
a nonsignificant relationship with power, but a moderate positive relation-
ship for boys.

These findings suggest that the bases for power are different for
boys and girls. Thus when talking about the status variable "power", one
seems to be talking about different behaviors for boys and girls. The
generalizability of these findings, however, does require confirmation
from an upper middle class sample. Possibly with such a sample there
would be less emphasis on aggression as the base of power for boys than

with the essentially working class sample used in this study.

72
W

The major finding of this study regarding acceptance concerns the
relationship between this variable and teacher disapproval. For both
boys and girls the correlation between these two variables was not sig-
nificant. It has been anticipated that this relationship for girls
'would be statistically significant and negative. Presumably, status in
girls is related to the values espoused in the schools to a greater
degree than status in boys. Since the disapproval items used in this
study represented the antithesis of these values, it seemed the hypothesized
relationship for girls was reasonable. These results suggest that the
notion of a relationship between teacher disapproval and status is in need
of revision.

Table 14 presents the mean values on the five major variables obtained
by the 30 most accepted and 30 least accepted individuals for both boys
and girls. For girls, the high acceptance group received slightly fewer
disapproval mentions than the low acceptance group. The reverse is true
for boys--i.e. the high group received more mentions than the low group.
However, there is no consistent pattern for either group. Three of the
five highest disapproval scores were obtained by the high acceptance group
for boys. Similarly, several of the lowest disapproval scores were
obtained by boys in the same group. 'With the low acceptance group for
boys this extreme variability was also observed. A similar result was
obtained with girls: however, the high disapproval scores for girls were
not as extreme as they were for boys. These data, while only descriptive
in nature, do tend to support the earlier conclusion that there does not
appear to be a significant relationship between peer acceptance and per-

ceived teacher disapproval.

73
TABLE 14

Means on the 5 Composite Variables for the
30 Most and 30 Least Accepted Children of Each Sex

 

 

Boys Sir1§__

High Low High Low
ignighlg Ac ta ce Acce an e ance
Accept 17.32 1.30 17.47 1.57
Compet 13.11 2.38 13.11 2.66
Power 16.62 3.57 13.03 3.32
Disapp u‘Os68 25s65 9018 13s37

 

 

One further thing should be pointed out in regard to factors relat-
ing to acceptance. Those items dealing with aggressive behavior showed
low to moderate positive correlations with total acceptance for boys.

The same items had essentially zero order correlations with acceptance
for girls. This suggests that there is a tendency for highly accepted
boys to be perceived as aggressive, while there is no such tendency where
girls are concerned.

921229392222

The major finding for competence involved the relationship between
this variable and teacher approval. For both boys and girls, a high
positive correlation (r>0.80) was obtained between perceived competence
and teacher approval. Unfortunately, one of the limitations of a cor-
relational study is that it is impossible to specify cause and effect.

In the present case does high teacher approval lead to the perception of
high competence or vice versa. 1
In weighing the two alternatives--high competence causes high teacher

approval versus high teacher approval causes high competence--the author

74
'must apt for the latter. First, one must remember that this study dealt
with children's perceptions. Thus, the competence Score reflected per-
ceived competence as opposed to actual competence. Second, the teacher
is the Judge and final arbiter of performance--i.e. competence--in the
classroom. In the course of classroom discussions and other interaction
with students, the teacher makes verbal evaluations of their work and con-
duct. It would seem likely that these verbal evaluations of one's peers
would tend to influence one's perceptions of the competence of others.
Given a fellow student who receives a large share of teacher approval, it
seems likely that a child would tend to perceive this fellow student as
relatively competent.

This relationship between perceived teacher approval and perceived
competence should be further investigated by experimental manipulation of
the teacher approval variable, thus helping to establish cause and effect.
Theorists such as Sullivan and Rogers have emphasized the role of feedback
from.others in the formation of one's self-concept. This feedback is
based upon an individual's perceptions of another. A good self-concept
appears to be conducive to academic achievement; the self-concept is depen-
dent upon feedback from others: and this feedback depends upon the percep-
tions which others have of another individual. If perceived competence
depends upon the amount of teacher approval received by a child, the impli-
cations are obvious. By systematically employing approval, a teacher might
serve to raise a child's self-esteem and perhaps indirectly his academic
performance.

W
The findings regarding the relationships between teacher approval,

teacher disapproval and the three status measures have been discussed in

75

previous sections. In this section the relationship between the approval
and disapproval variables will be discussed.

For both boys and girls the correlations obtained between approval
and disapproval were not significant. This indicates that approval and
disapproval are not simply the opposites of one another. The results
of the factor analyses lend further support to this view. If approval
and disapproval were marshy the opposites of one another, one would
expect a single bipolar approval-disapproval factor with significant
loadings on this factor for all approval and disapproval items. Such
is not the case. For girls the approval items showed their highest
loadings on the same factor as did the competence items. For boys all
but one of the approval items had their highest loadings on the factor
including the competence item "well"--i.e. the item dealing with
academic competence. 0n the other hand, six of the eight disapproval
factors formed a single factor fer girls. For boys all of the dis-
approval items plus the power item "fight" formed a factor.

If one examines the content of the approval and disapproval items
used in this study, an explanation is suggested. The approval items
used in this study deal primarfky with approval of behavior reflecting
academic performance--i.e. turning in assignments on time, making com-
ments in classroom.discussions, etc. On the other hand, the content of
the disapproval items deals mostly with behaviors which, while undoubtedly
disruptive of classroom functioning, are not directly involved in
academic perfOrmance--i.e. picking on others, talking in class, etc.
The fact that the items deal with different content would help explain
the lack of a relationship between them. Further, the content of the

76
approval items would help explain the magnitude of the relationship between
approval and perceived competence.
W

The data reported in this study support the notion of a general factor
underlying status in each of the three systems studied. Further, the
data indicate that this general factor mdy be greater for girls than it
is for'bqys.

First, consider the correlations among the composite status vari-
ables for both bays and girls. There was no difference between the sexes
on the relationship between acceptance and competence (3)0.50 for both
sexes). The correlation obtained between competence and power was signi-
ficantly greater for girls than.for“bqys. However, the correlations for
both sexes were significantly greater than zero. Finally, the relation-
ship between acceptance and power tended to be of higher magnitude for
girls than for boys. Although the difference was not significant at the
alpha level used in this study (8.: 0.002), it was significant at a lower
level (<2 :- 0.01) indicating that this relationship should be investigated
more fully;

Secondly, the factor analyses tend to support both the notion of a
general status factor and the notion of a greater interrelationship
between status systems for girls. Little difference was noted between
boys and girls on the competence factors (factor 1 for each). For girls,
all three competence items, two power items ("nice" and "get") and seven
of the eight approval items (all but the item "try") showed significant
loadings on the factor. For'bqys, two competence items (not "make"), one
power item ("nice") and seven of the eight approval items (not "things")

showed significant loadings. 0f possible significance is the fact that

77
the girl who is academically competent is at least to some extent perceived
as able to "get" others to do something, whereas this is not the case for
boys. Further, the primary base of power for girls--i.e. ability as a
social reinforcer-showed a significant loading on this factor for girls.
The item "nice" likewise had a significant loading for boys, but "nice"
boys don't "get" other people to do things. The power base for boys is
aggressiveness. Thus, competence scene to some degree confounded with
power for girls.

Factor III, the acceptance factor for both boys and girls, again
shows this pattern of greater generality for girls than boys. For girls,
the three acceptance items, all three competence items, two of the three
power items ("get" and "nice"), the Moreno power item, and three approval
items ("ccmment", "ontime" and "spproj") showed significant loadings on
this factor. Foeroys, the three acceptance items, the competence item
"make", and the same three power items as for girls showed significant
loadings on the acceptance factor. Apparently, for both boys and girls,
power as a social reinforcer is related to acceptance. For girls in con-
trast to boys, however, academic competence is to some extent associated
with acceptance.

Finally; it should be pointed out that for boys apparently'a dis-
tinction is made between two kinds of competence--academic competence and
special competencies. The latter factor includes significant loadings
for the following: the competence items "make" and "inproj", the power
item "nice", and the approval items "comment", "things", and "spproj".
Such a distinction between academic competence and special competencies
is not made fer girls. Rather the competence factor is a global factor

subsuming all aspects of competence tapped in this study. This question
of’the generality of status should.be examined more closely using a

78
greater number of items dealing with more specific types of competence,

power and acceptance than have been used in the past.
h A C d - A C i

In Table 14, as was previously noted, the mean scores Obtained on
the five major variables in this study for the 30 most accepted and 30
least accepted children of each sex are presented. Note that with the
exception of disapproval for boys, the trend is what one might expect--
i.e. the children.high on acceptance tend to be perceived as higher on
competence, power and teacher approval than children low on acceptance.
It could be that there is a halo effect at work-i.e. children who are
highly accepted by their peers are perceived as being high in the other
areas also. As was pointed out in the introduction to this paper, to a
certain extent peer group behavior is guided by'the children's percep-
tions of one another vis-a-vis status. One major question pertaining to
the peer group, however, is the extent to which these perceptions and
actual behavior coincide. One urgent need is for studies dealing with
this question.

A strategy for obtaining such comparisons would involve the use of
a sociometric type instrument of the sort used in this study to determine
perceptions, and objective criteria such as observational data, achieve-
ment tests, etc. as behavioral measures. However, it would be impera-
tive that the variables being tapped by each be the same. For example,
if social reinforcement and coerciveness are used as the bases of power
on the sociometric instrument, then behavior so defined should be rated
using observational techniques. It has been difficult to compare the
results of many previous studies because of the many different definitions
of the status variables and the different methods used. .A study designed

79
as suggested above would help clarify the relationship between data
obtained with different methods.
Ed c I a 1

Most of the past studies of interaction in cthe classroom have stressed
the role of the teacher (Withall and Lewis, 1963: Schmuck and Schmuck,
1971). The teacher, by virtue of her position, was considered to be the
focal point of interpersonal behavior in the classroom. Power (i.e. influ-
ence) was viewed as largely unidirectional-~i.e. flowing from the teacher
to the students (for example the studies. of Polanslq, 1954 and Flanders
and Havumaki, 1960). However, as previously indicated, a number of inves-
tigators have found that a large portion of the variance in academic
achievement (thus in effect, the attainment of the major objective of
educational institutions) can be accounted for by the congruence of peer
values and those values associated with academic endeavor (Coleman gt 5],. ,
1966) . Further, even when there is considerable motivation to succeed
academically, peer values influence the ultimate level of academic involve-
ment, and even the specific material learned (Bushnell, 1962; Hughes,
Becker and Greer, 1962) .

A. more reasonable view of teacher influence is provided by Schmuck
and Schmuck (1971. pp- 29-31) at the beginning of the school year a teacher's
influence has three bases: 1. legitimate power-power which is hers due
to her office: 2. reward power: and 3. coercive power. Of the five
bases of power described by French and Raven (1939), these three are the
least effective in influencing others. The teachers who come to have most
influence on classroom behavior are those whose bases of power shift from
the above to power based on reference (modeling) and expertness. The most
powerful children in the classroom peer group are those children possessing

80
referent power and expert power (Gold, 1958). The above suggest that at
least when the classroom group is being formed, the peer group has a
greater potential for influence upon its members than does the teacher.

In the present study, there is substantial evidence to support the
notion that peer status is minimally related to either adherence to or
breaking of classroom rules as set down and enforced by the teacher.

This finding indicates that there is a gap between those values which

are considered important to the peer group and those considered important
to teachers. When such a situation exists, it seems likely that the
educator and classroom group will spend considerable time working at
cross-purposes to one another. Given that the peer group does have a
significant effect on academic performance, and given the estrangement of
peer and educational values in many schools, it seems imperative that the
forces of the peer group be harnessed to serve, rather than impede the
educational process.

One obvious area of conflict between peer group and teacher is in
the setting and enforcing of classroom rules. Typically, the teacher
serves the function of legislator, judge, jury and executioner in regard
to discipline. More recent works on the subject point out the need for
peer group involvement in both the making and enforcing of rules (Bronfen-
brenner, 1970: Brown, 1971).

Under such a procedure, rules are set as the need arises. The teacher
points out a maladaptive or disruptive behavior and discusses it with the
class. Using this concrete exemplar, the reasons why this behavior is
undesirable are discussed. The class as a whole decides whether a rule
is necessary. If a rule is made, then appropriate sanctions to be taken

against offenders are discussed. Such a procedure has several desirable

effects:

1. It removes the arbitrary nature of discipline as found in most
classrooms. Since responsibility for setting and enforcing rules does not
rest solely with the teacher, it is likely to lead to fewer conflicts
between teacher and peer group over discipline.

2. It should lead to the reduction of deviant behavior by bringing
to bear the influence of both the teacher and the influence of the peer
group. Often the peer group will be supportive of disruptive behavior--
even if they don't approve of it--merely to bait or get back at,the
teacher. Given the major role in setting and enforcing rules, there is
less likelyhood of peer support for such behavior.

3. A democratic disciplinary procedure as outlined above could con-
ceivabiy have a positive impact upon the children's moral development.
According to Piaget (1965), the highest levels of moral judgment cannot
be obtained in an environment in which rules are handed down and enforced
on an arbitrary basis. It is only through peer group interaction in play
settings that the child generally has a chance to make and enforce rules,
and to study the reasons underlying and the implications of rules. This
interaction leads a child from a morality of constraint to a morality of
cooperation-i.e. a morality based on mutual respect. It would seem that
the approach to classroom management would tend to foster the develOpment
of a morality of cooperation.

The use of’peer models (especially older children) as effective
agents of socialization in the educational setting has been demonstrated
in the Soviet Union (Bronfenbrenner, 1970, p. 51). Under their system an

older class "adopts" or sponsors an entering class in the same school.

82

The older children help the younger ones with homework, play with them,
teach them games, and in general provide a model of the desired behavior
in an educational setting. Bronfenbrenner (pp. 157‘58) suggests that the
adoption of such a system would be an aid to all children in the schools--
especially culturally deprived children. Although middleclass parents tend
to reinforce curiosity, and tend to serve as models for intellectual endeavor,
such is not the case with the disadvantaged child. The development of an
achievement orientation depends in part upon intense personal relationships--
i.e. the interest and involvement of the parents and others in a child's
academic and intellectual performance. Such interest and involvement is
typically lacking where the disadvantaged child is concerned. By using
older children as models and reinforcers for intellectual and academic
achievement, it seems likely that the situation as regards the disadvan-
taged child could be alleviated to some extent.

In summary, the values fostered in the educational institutions of
this country are not reinforced by the peer group. There is substantial
evidence to indicate that the attainment of educational objectives is in
large measure determined by the consonance of peer values and educational
values. When the peer group accepts and reinforces the values associated
with academic achievement, performance tends to be higher than when the
peer group rejects such values. The foregoing discussion stressed the
need for changing our approach to classroom.management by involving the
peer group in the making and enforcing of classroom rules. Such an approach
directs the power of the peer group toward educational ends. Additionally,
it was suggested that such an approach will foster a morality of cooperation
among students. Finally, it was suggested that use be made of the potential

of older children in socialising children into the academic environment.

83
It is incumbent upon educators in the United States to harness the
tremendous force of the peer group to serve the educational process.
Although the author does not advocate the coerciveness which underlies
the Soviet system, he does advocate a rethinking and rational planning

of the role of the peer group in the educational process.

BIBLIOGRAPHY

Austin, M. C. and Thompson, G. G. Children's friendships: a study of
the bases on which children select and reject their best friends.

J. mug. Igflch. 1%8. 39. 101-116e

Bonney, M. E. The relative stability of social, intellectual, and
academic status in grades II to IV, and the interrelationships
between these various forms of growth. 9, Educ 25y h. 1943, 34,
88‘102e

Bonney, M. E. Relationships between social success, family size, socio-
economic home background, and intelligence among school children
in grades III to V. §ggigmgtzy, 1944, 7, 26-39.

Bonney, M. E. A sociometric study of the relationship of some factors
to mutual friendships on the elementary, secondary, and college
I‘Vds' W0 19%0 90 21"“70

Bonney, M. E. Sociometric study of agreement between teacher judgements
and student choices. figgigmgtzy, 1947, 10, 133-146.

Bretsch, H. S. Social skills and activities of socially accepted and
unaccepted adolescents. J‘_Eggg&_£gygh. 1952, 43, 449-504.

Bronfenbrenner, U. The measurement of sociometric status, structure and

development. Sggigmggzy nonographg, No. 6, New York: Beacon House,
1945.

Bronfenbrenner, U. 259 flgzlgg g: ggilghgod; U,S, gag Q,S,S,R, New York:
Russell Sage Foundation, 1970.

Brown. 13- Web to Discipline.
Dubuque, Iowa: Wm. C. Brown Comparw Publishers, 1971.

 

Bushnell, J. H. "Student Culture at Vassar" in Ihg_Amgziggn_§gllggg, Ed.
by Sanford, N. New'York: John Wiley and Sons, 1962, 489-514.

Col-man. J- S- 21 21- W. Washington.
D. C.: 0.3. Government Printing Office, 1966.

Criswell, J. H. Social structure revealed in a sociometric test.
W. 1939. 2. 69-75.

Criswell, J. H. Sociometric methods of measuring group preferences.
W. 1943. 6. 398-408.

Criswell, J. H. Sociometric measurement and chance. §ggigmg§gg, 1944,
7, 415-421.

85

deGroat, A. F. and Thompson, G. G. A study of the distribution of
teacher approval and disapproval among sixth-grade pupils. ,1, 9f
Egg, Educ. 1949, 18, 57-75.

Durojaiye, M. O. A. Patterns of friendship. Brit, ,1, Ed, 25:192. 1969,
39. 88-89e

Evans, K. M. Sgcigmgtm 3g Education. London: Routledge and Kegan
P3311, 1962e

Feinberg, M. Relationship of background experience to social acceptance.
WM- 1953: "'8: 206-21“-

Flanders, N. A. and Havumaki, S. The effect of teacher-pupil contacts
involving praise on the sociometric choice of students. J, Educ,
Rm.- 1960. 51. 65-68-

French, J. R. P., Jr. and Raven, B. "The Bases of Social Power" in

W. Second Edition. Ed- by Cartwright.
David and Zander, A. New York: Harper and Row, 1955, 607-623.

Gardner, G. Functional leadership and popularity in small groups. 53%;;
allfisign§v 1956! 99 “91’5090

Glidewell, J. C., Kantor, M. B., Smith, L. M. and Stringer, L. A. "Social-
isation and social structure in the classroom" in W
h, Vol. 2. Ed. by Hoffman, L. w. and Hoffman,
M. L. New York: The Russell Sage Foundation, 1966, 221-256.

Gold, M. Power in the classroom. W. 1958. 25, 50-60.

Gronlund, N. E. The accuracy of teachers' judgment concerning the socio-

metric status of sixth-grade pupils. W, No. 25,
New York: Beacon House, 1951.

Gronlund, N. E. Relationships between the sociometric status of pupils
and teachers' preferences for or against having them in class.

W. 1953. 16. 142-150.

Gronlund, N. E. e r m. New York: Harper and
Brothers, 1959.

Gronlund, N. E. Generality of sociometric status over criteria in the

measurement of social acceptability. WM. 1955a,
56: 173'176-

Gronlund, N. E. The relative stability of classroom social status with
unweighted and weighted sociometric status scores. W.
1955b. “5. 345-354-

86

Grossmann, B. and Wrighter, J. The relationship between selection--
rejection and intelligence, social status, and personality amongst

sixth grade children, MM, 1948, 11, 346-355.
Harman, H. H. W. Chicago: Univ. of Chicago Press,
1960.

Hartshorne, H., May, M. A. and Maller, J. B. Studies in the Nature of
a I S S a d o . New York:

Macmillan, 1929e

Hartup, W. 91., Glaser, J. A., Charlesworth, R. Peer reinforcement and
sociometric status. gmd Dev. 1967, 38, 1017-1024.

Haber, R. F. The relation of intelligence and physical maturity to social
status of children. J, Eng, Eflc . 1956, 47, 158-162.

Honzik, M. P. Prediction of behavior from birth to maturity: a review
of W by Jerome Kagan and Howard Moss. 219121.11-

Wo. 1965. 2. 77-88.

Hughes, E. C., Becker, H. S. and Greer, B. "Student Culture and Academic
Effort" in W, Ed. by Sanford, N. New York: John
Wiley and Sons, 1962, 515-530.

Jennings, H. H. W. New York: Longmans, Green
and Co., 1943.

Kennedy, W. A. A follow-up normative study of Negro intelligence and
achievement. W, No. 126, 1969, 34, Appendix 2.

Kohlberg, L., LaCrosse, J. and Ricks, D. "The predictability of adult

mental health from childhood behavior" in the W
W. Ed. by Wolman, B. New York: McGraw-Hill, 1970.

Lindsey, t}. and Bergatta, E. F. "Sociometric Measurement" in W
WWW- 301- by Lindzey. G-
Cambridge, Mass.: Addison-Wesley Publishing Co. Inc. 1954, 405—448.

Lippitt, R. and Gold, M. Classroom social structure as a mental health
problem. ,1, Socigl IS§gg§, 1959, 15, 40-60.

Meyer, W. J. and Thompson, G. G. Sex-differences in the distribution
of teacher approval and disapproval among sixth-grade children.

LEW- 1955. 47, 385-396-

Moreno, J. L. W Washington, D. C.: Nerzggs and Mggm
Monograph “Ge 58’ 193ne

Moreno, J. L. W Fomdaflgns g; Sgcigmetm, Group
W. New York: Beacon House. 1953-

87

Mouton, J. 8., Blake, R. R. and Fruchter, B. The reliability of socio-
metric measures. §ggigmgtzz, 1955, 18, 7-48.

Neugarten, B. L. Social class and friendship among school children.
Mgr gs 5091010 1952. 51’ 305-313e

Northway, M. L. Outsiders, figgdddetzz, 1944, 3, 10-25.

Pepinsky, P. N. The meaning of "validity" and "reliability" as applied
to sociometric tests. Edug, and nggh. Meas. 1949, 9, 39-49.

Piaget. J - WWW- TranS- by Mo Gabain. New
York: The Free Press, Free Press Paperback Edition, 1965.

Polansky, L. Group social climate and the teacher's supportiveness of
group status systems. J, Educ, Sgciol. 1954. 28, 115-123.

Polansky, N., Lippitt, R. and Redl, F. The use of near sociometric data
in research on group treatment processes. Sociomgtry, 1950, 13, 39-62.

Pepe, B. Prestige values in contrasting socioeconomic groups of children.

Egychiatzx, 1953, 16, 381-385.

Potashin, Reva. A sociometric study of children's friendships. §Qgigd§try.
1%. 9' “8-70 0

Proctor, C. H. and Loomis, C. P. "Analysis of Sociometric Data" in

W- Ed- by Johoda. Mu Deutsch. M-
and Cook, S. W. New York: Dryden Press, 1951, Chap. 1?.

Remmers, H. H. "Rating Methods in Research on Teaching" in fldgdbggk_d£
3W. Ed. by Gage, N. L. Chicago: Rand McNally &
Co. 1963, 329-378.

Roff, M. and Sells, S. B. Relations between intelligence and sociometric
status in groups differing in sex and socioeconomic background.

151211.321” 1965. 15. 511-5-16-

Schmck. R. A. and Schmck. P- A- W.
Dobuque, Iowa: ‘wn. C. Brown Company Publishers, 1971.

Shoobs, N. E. Sociometry in the classroom. §ggdgmetgy, 194?, 10, 154-164.

Singer, A. ’Certain aspects of personality and their relation to certain
group models and constancy of friendship choices. W
1951. “5. 33-42-

Smith, G. H. Sociometric study of least-liked and best-liked children.
Wu 1950. 51. 77-85-

Taylor, E. A. Some factors relating to social acceptance in eighth-grade
classrooms. J, Educ, ngch. 1952, 43, 257-272.

88

Tiktin, S. and Hartup, W. W. Sociometric status and the reinforcing

effectiveness of children's peer. W. 1965, 2,
206-315.

Tryon, C. McC. Evaluations of adolescent personality by adolescents.
- _ d A':W--~.. , Vol. 4,

 

No. 4,1939.

Wertheimer, R. R. Consistency of sociometric status position in male
and female high school students. ,1, Ego, Psych. 1957, 48, 385'90.

Winch, R. F. W. Rev. Ed. New York: Holt, Rinehart and
Winston, 1963.

Withall, J. and Lewis, W. W. "Social Interaction in the Classroom" in
W. Ed- by Gage. N- L- Chicssm
Rand McNally and Co., 1963, 683-714.

Witryol, S. L. and Thompson, G. G. An experimental comparison of the
stability of social acceptability scores obtained with the partial-
rank-order and paired-comparison scales. Wu 1953,
M, 20’30e

Young, L. L. and Cooper, D. H. Some factors associated with popularity.
AM- 1944. 35. 513-535-

Zander, A. and VanEgmond, E. Relationship of intelligence and social
power to the interpersonal behavior of children. M.
1958. 49. 257-268.

 

L W

A. What children do you like to play with on the playground?

B. What children would you enjoy working with in class?

C. What children would you choose to sit next to you in class?

1'1- W

A. Hero is someone who is good at making things.

B. Here is someone who always does well in school.

C. Here is someone who thinks of may interesting classroom
projects to do.

III- Em:

A. Here is someone who gets other children to do something by
bossing them or fighting with them.

B. Here is someone who can get others to do something because
he or she is a nice person.

C. Here is someone who can get other children to do what he or
she wants them to do.

IV- 1211mm].

A. Here is someone who is often scolded by the teacher for not
paying attention in class.

B. Here is someone whom the teacher often scolds for disturbing
the class in some way (shooting paper wads, passing notes,
making noises, etc.)

C. Here is someone who sometimes has to go to the principal's

office or has to be punished by the teacher.

89

V.

90

D. Here is someone whom the teacher often scolds for talking too
much in class.

E. Here is someone who is often suspected by the teacher when some-
thing happens while she is out of the class.

F. Here is someone who is often ecclded by the teacher for wasting
too much time in class.

G. Here is someone who is often scolded by the teacher for not
turning in assignments on time.

H. Here is someone whom the teacher often scolds for starting
fights or picking on someone else.

Apprgval

A. Here is someone whom the teacher often asks to do errands for
her.

B. 'Here is someone whom.the teacher often chooses as a monitor.

C. Here is someone who is praised by the teacher for his or her
comments during classroom.discussions.

D. Here is someone who is praised by the teacher for always having
their assignments done on time.

E. Here is someone who is called on by the teacher when she wants
the right answer.

F. Here is someone who is often praised by the teacher for bringing
special things to school for the class to see.

G. Here is someone whom the teacher praises for trying hard, even
if the work isn't easy for him or her.

H. Here is someone who is chosen to work on special projects in class.

APPENDIX B

McG -Wh Sc e

LuelJ.

Professionals.-9Lawyer, judge, physician, engineer, professor,
school superintendent, 91; gl.

Proprietors.-cLarge businesses valued at $100,000 or more,
depending on community.

Businessmen.--Top executive president,'g;‘gl., of corporations,
banks, public utilities.

White collar.--CPA: editor of newspaper, magazine: executive secre-
tary of status organization.

Blue collar.--None.

Service.--None.

Farm people.--Gentlemen farmers or landowners who do not supervise
directly their preperty.
2191.2

Professionals.--High school teacher, librarian, and others with four
year degrees.

Preprietors.-€Businesses valued at $50,000 to $100,000.

Businessmen.-2Assistant office and department managers or super-
visors; some manufacturing agents.

White collar.-9Accountant: insurance, real estate, stock salesmen;
editorial writers.

Blue collar.--None.

Service.-None.

Farm people.-€Land operators who supervise preperties and have an
active urban life.

91

92

W

Professionals.--Grade school teacher, registered nurse, minister
without four year degree.

Proprietors.--Business or equity valued from $10,000 to $50,000.

Businessmen.-Managers of small branches or buyers and salesmen of
known merchandise.

White collar.~~Bank clerks, auto salesmen, postal clerks, railroad
or telephone agent or supervisor.

Blue collar.~-Small contractor who works at or supervises his jobs.

Service.-None.

Fam people.-~Farm owners with "hired help," Operators of leased
property who supervise.
Luau!

Professional s . --N one .

Proprietors.--Business or equity valued from $5,000 to $10,000.

Businessmen and white collar. "Stenographer, bookkeeper, ticket
agent, sales people in department stores, 31, 51.

Blue collar.--Foreman; master carpenter, electrician, gt 51.3
railroad engineer.

Service.--Police captain, tailor, railroad conductor, watchmaker.

Farm people. “Small landowners, operators of rented property hiring
"hands."
LmLS

Professionals.-None.

Proprietors.--Business or equity valued from $2,000 to $5,000.

Businessmen and white collar.-Dime store clerks, grocery clerks,

telephone and beauty operators, :3, 31.

93

Blue collar.-2Apprentice to skilled trades, repairmen, medium
skilled worker.

Service.--Policeman, barber, practical nurse, brakeman,[gt‘gl,

Farm people.-‘Tenants on good farms; feremeng owners of farms who
"hire out."
1:31:11

Professionals.--None.

Preprietors.--Business or equity valued at less than $2,000.

Businessmen, white collar, and blue collar.--Semiskilled factory
and production workers, assistants to skilled trade, warehousemen, watch-
men.

Service.--Taxi and truck drivers, waiter, waitress, gas station
attendant.

Farm people. SharecrOppers, established farm laborers.
LmLZ

Professionals.-None.

Proprietors.--None.

Businessmen.--None.

'Hhite collar.--None.

Blue collar.--Heavy labor, odd-Job men, mine or mill hands, unskilled
workers.

Service.--Domestic help, busboy, scrubwoman, janitor, Janitor's
hclper.

Farm people.-‘Migrant'workers, "squatters" and "nesters."

Plus.-4The reputed lawbreakers and the unemployed.

APPENDIX C

chg

NAME AGE GRAIE

 

TEACHER ' S NAME

In this booklet are twenty-five items. Some questions ask you to
name the children with whom you would like to work, play, and sit next
to in class. The others are descriptions which probably fit some of
the children in this classroom. Below each item are three blank
spaces. We will read each item together. Then I want you to choose
three children who are best described by the item and write their names
in the blank spaces. You must choose children from this room. You
may choose children who are absent. Write the first name and the first
letter of the last name for each child you choose. You may only choose
three children for each item. However, you may choose the same child
for more than one item. If you cannot think of three names for an item,
write as many as you can. Your choices will not be seen by anyone else,
not even your teacher.

REMEMBER:

1. Choose three children for each item.

2. If you cannot think of three children for an item, write
down as many names as you can.

3. The children you choose M be in this class.

1+. You may choose children who are absent.

5. You may choose the same child for more than one item.

91.;

1.

2.

3.

5.

7.

8.

95

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Here is someone who is good at making things.

(a) (b)

(c)

Here is someone whom the teacher often asks to do errands for her.
(a) (b)

(0)

Here is someone who is often scolded by the teacher for not paying
attention in class.

(a) (b)

(0)

Here is someone who gets other children to do something by bossing
them.or fighting with them.

(a) (b)

(c)

Here is someone whom the teacher often scolds fer disturbing the
gigs; in some way (shooting paper wads, passing notes, making noises,
(a) (b)

(c)

‘What children do you like to play with on the playground?

(a) (b)

(0)

Here is someone whom the teacher often chooses as a monitor.

(a) (b)

(6)

Here is someone who always does well in school.

(a) (b)

 

 

(c)

 

9.

10.

ll.

12.

13.

115.

15.

16.

96

Here is someone who is praised by the teacher for his or her com-
ments during classroom discussions.

(0) (b)
(0)

Here is someone who sometimes has to go to the principal's office
or has to be punished by the teacher.

 

 

 

 

 

 

 

(a) (b)

(c)

What children would you enjoy working with in class?
(a) (b)

(c)

 

Here is someone who is praised by the teacher for always having
their assignments done on time.

(a) (b)
(c)

Here is someone whom the teacher often scolds for talking too much
in class.

(a) (b)
(c)

Here is someone who is called on by the teacher when she wants the
right answer.

 

 

 

 

 

 

 

 

 

(a) (b)

(c)

What children would you choose to sit next to you in class?
(a) (b)

(c)

 

Here is someone who is often suspected by the teacher when something
happens while she is out of the class.

(a) (b)
(c)

 

 

17.

18.

19.

22.

23.

2h.

97

Here is someone who can get others to do somethim because he or
she is a nice person.

(a) (b)
(0)

Here is someone who is often praised by the teacher for bringing
special things to school for the class to see.

(a) (b)
(c)

Here is someone who is often scolded by the teacher for wasting
too much time in class.

(a) i_ (b)

(0)

Here is someone who can get other children to do what he or she
wants them to do.

(a) (b)
(0)

Here is someone who is often scolded by the teacher for not turn-
ing in assignments on time.

(a) (b)
(0)

Here is someone whom the teacher praises for trying hard, even if
the work isn't easy for him or her.

(a) (b)
(0)

Here is someone whom the teacher often scolds for starting fights
or picking on someone else.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

(a) (b)

(0)

Here is someone who is chosen to work on special progects in class.
(a) (b)

 

 

(c)

 

98

25. Here is someone who thinks of man interesting classroom projects
to do.

(a) (b)
(c)

 

 

99

There are mam reasons win you might choose someone to be your

friend. Below are some possible reasons. Put a check beside each

one which is important to you when you choose a friend.

Friendly
Good-looking
Cheerful, good natured

Strong

Smart at school

Good sense of humor

Canbe trusted

Sharp dresser

___Likes the same things as you
___Athletic
Sure of himself (herself)

“Outgoing

...—301151“!-
Considerateuthinks of others
__.L°!tl
...—Kind
___Bonest

___Imsginative (can think up
things that are fun to do)

____Good .1: nking things

____Patient-does not get mad
easily

Polite
__P°pu1tr
__Fun to be with

Someone you would like to
be like

 

 

APPENDIX D

Correlations of’Each.Item'with Each of the
Five Major Variables for High 833 Boys

 

 

Errand 0.0551 0.3343 0.1305 0.0171 0.4144
Help 0.0192 0.5720 0.3073 -0.0359 0.6301.
Comment 0.5160 0.6212 0.3511 “0.0280 0.8158
Ontime 0.6983 0.7040 0.3003 ‘0.1600 0.7335
Right 0.66% 0.7889 0.3493 -0.1u63 0.8313
Try 0.3604 0. 01:05 0.2209 0.1050 0.1.716
Spproj 0.5187 0. 7871 0.3210 -0.1623 0.8951

 

100

APPENDDCE

Correlations of.Each.Item with Each of the
Five'thor Variables ferlLow'SBS Boys

 

WW

Play 0.8750 0.1810 0. 5981 0.2370 0.1686
‘Work 0.9068 0.0650 0.5035 0.0967 0.0387
Sit 0.9532 0.3655 0.5725 0.1598 0.3823
Make 0.0757 0.8509 0. 4941 0.2018 0.7121
we11 0.0819 0.6900 0.1550 0.0181 0.7272
Inproj 0.3658 0.9078 0.3806 0.0708 0.8387
Fight 0.5319 0.2695 0.9583 0.5679 0.3370
Nice 0.0898 0.6991 0.6039 0.1570 0.6521
Get 0. 5073 0.3860 0.8922 0.3005 0.3001
Attend 0.1213 0.2770 0.2060 0.8859 0.2922
Noise 0.2121 0.0211 0.3157 0.8307 0.0698
Punish 0.2031 -0.0031 0.0855 0.9006 0.0297
Talk 0.2116 0.0896 0.0000. 0.7802 0.0873
sheeut 0.1500 0.1232 0.3985 0.930? 0.1260
Late .1600 -0.1100 0.1251 0.7336 -0.1037
Pickon 0.3203 0.0270 . 0.6596 0.7065 0.0053
Cement 0.2639 0.7916 0.2739 0.1531 0.8653
Ontime 0.2155 0.5127 0.2232 0.0193 0.6707
Right 0. 0501 0. 5975 0 .0979 0 . 0008 0. 7310
Things 0.0000 0.7803 0.5725 0.1932 0.7571
Try 0.2760 0.3205 0.3970 0.0017 0.5005
Spproj 0.0197 0.8778 0.0003 0.1050 0.8175

 

101

v—i

APPENDIX F

Correlations of Each Item with Each of the
Five Major Viriables for’High SES Girls

 

 

71.37 0.8881 0.5726 0.6190 -0.1100 0.5777
Work 0 . 9391 0 . 6897 0 . 6530 0 . 0039 0 . 6723
Sit 0.9553 0.5245 0.5281 “0.1809 0.4890
Well 0.5900 0.9356 0.7187 -o.1790 0.9328
InprOJ 0 e $33 0 e 9116 0 e 7159 .0 e 0787 0 e ”00
Fight -0.1160 -0.0093 0.2913 0.0290 -0.0008
Nice 0 e 7801 0 e 7591 0 e 9066 '0 e 0751 0 a 71077
Get 0.3808 0.5740 0.8742 0.0732 0.5919
Moreno 0 . 8371 0 . 5373 O . 5750 0 . 0255 0 . 5310
Attend “0.1011 “0.1506 “0.1200 0.6381 “0.1591
Noise “0.1202 “0.1935 0.0280 0.7053 “0.1762
Punish “0.0110 “0.0888 0.0592 0.7664 “0.0510
Talk 0.0512 “0.0215 0.2660 0.7385 0.0472
Susout “0.0818 “0.1129 0.0617 0.6446 “0.0378
WC,“ ‘Os1223 001300 '0e1203 006777 "00133"
Late “0.0689 “0.0807 0.0278 0. 5406 “0.0929
Pickon 0.1036 0.0548 0.2831 0. 5137 0.1437
Errand 0.4625 0.6538 0.4854 0.0259 0.7450
Comment 0.6175 0.9303 0.7255 “0.2208 0.9357
Ontime 0.6385 0.9516 0.7286 -0.l5l0 0.9517
Things 0.1109 0. 3859 0.2012 -0.0290 0.0821

 

102

APPENDIX G

Correlations of’Each Item with Each of the
Five Major Variables for Low SES Girls

 

 

Play 0.9506 0-5793 0.7525 0-0577 0.5375
Work 0.8887 0.7078 0.6007 -0.1029 0.5160
Sit 0.8642 0.3824 0.6482 “0.0214 0.3100
Make 0.4324 0.8464 0.3701 “0.2181 0.3616
we11 0.4830 0.8824 0.4292 “0.2955 0.6869
Inproj 0.5935 0.6827 0.6869 -0.0318 0.3719
Nice 0.8251 0.6423 0.8330 “0.1906 0.6913
Get 0.5915 0.6378 0.9121 0.0115 0.5154
Moreno 0.8100 0.0838 0.5290 0.0701 0.0018
Noise 0.2807 -0.0832 0.1333 0.6858 0.0205
Punish 0.2513 “0.0778 0.0491 0.5867 0.1100
Talk 0.1104 “0.1488 “0.0088 0.7255 “0.0168
.Susout “0.2146 “0.1377 0.0053 0.6306 “0.1859
W‘ato -002767 ”002328 .001167 006826 .00 2585
LQtQ .0 e 3711 .0 0 3&3 .0 e 1959 O 0 60% -0 e 20 32
P1¢k°n 002857 000539 003299 006132 -00074‘5
Errand 0.3796 0.3343 0.3362 0.0149 0.7210
Help 0.5940 0.4717 0.5328 0.1009 0.7910
Comment 0.5742 0.7281 0.5315 0.1227 0.6917
Right 0.5633 0.8579 0.5399 -0.1703 0.7087
Things 0.4814 0.5269 0.4820 “0.2064 0.7774
Try’ 0.0305 0.6311 0.0032 -0.0608 0.2220
Spproj 0.5928 0.7596 0.6970 -0.2209 0.8353

 

103

0mm

222

mfifi

1&1

MWfl

see
232

w8®%&%

00000000
33232221

%&fi®%fi&%

ik1ii$zm

nfifimfiuMm

3321”...)2210)

 

APPE‘DIX H
Standard

for Both Boys and Girls

Boys;

 

5 5 36
8W1 M03

0 . O .
6 56 56666

$M$fi$mww

M101Mmmm

meflfiwsa

22323hm23

zfifim7wsw

1121.]..222

 

Mean Scores and Standard Deviations on Each Item

 

 

Play
Work
Sit

Well
Inpro j

Make

Fight
Nice
Get
Moreno

Attend
N oi as
Punish
Talk
Susout
Waste
Late
Pickon

....
.m ne 8 .3

t8 0
an an r
re nimmw
EHCOR S

100

 

 

"ITllfllfliflfljll'ififl'ifl'flmflllfiflifl’1'“