-; ——v

~rv.v

H

 

 

 

 

 

 

 

 

u

 

 

m

 

RUM!“

   

N

   

55521:

 
 

a};

Haul]!

 

”951095;:

 

 

 

 

 

 

 

. ....,vq~. ‘-.-fiq . . 34,4.

» ‘ .

:CONCEPT LEARNING
FUNCTEO‘N or .5!

 

OF YOUNG cm

[gm

 

 

Vp-RELA’TII

\‘Vu \‘V‘V‘V'V’

.N‘SstP "TO? I

 

" ,CHILD-J'EACHERsLfi .

 

 

axis for the Degree of ?h.D.r ' r .

h .
MCHEGANSTATE UNIVERSITY ” .. fl ‘ 3  

VICTORG. ma g _ ‘ '
1971 ‘ r.~ ' ; .

   
   
   
    

  
 
 
 
 

  
      

  
    

‘ ' ' ' - J' .. ' D-MQ‘ r r-,. P.
- ' V . u . v v ", ,"M v'
‘, , ' ‘ ‘ . ,. ...' . Fr, ' r "»‘r' r»~';7,.
. . . ' ,7. ,-.r
. . a: "i
.
' ' ,.. V . . , I: '53., may:
‘ , ‘ ' ' ' ' ' ‘ . " " ' - - ,w, —|-Ifi"
. ». . .. . , , , , ..

Wants-am ha 1.81 ~-' 3

 

. n ”‘0 LIBRARY E
"E" EEEEEEEEEEE' h :5” FE

 

This is to certify that the
thesis entitled
Concept Learning of Young Children
As A
Function of Sibling Relationship to Child-Teachers

presented by

Victor G. Cicirelli E

has been accepted towards fulfillment
of the requirements for

Ph .D. degree in Psychology

W,

Major professor

Date—lamamw

0-7639 1
l

 

 

éoaKNL’C‘LT

g,.:;:,- .

 

 

ABSTRACT

CONCEPT LEARNING OF YOUNG CHILDREN AS A
FUNCTION OF SIBLING RELATIONSHIP
TO CHILD-TEACHERS

BY

Victor G. Cicirelli

This study asked the question: Is concept learning
of a younger child taught by an older child affected by
sibling relationship? Each of the 120 first-grade children
had an older sibling in third-grade; equal samples of 30
sibling pairs were drawn from the population of boys with
older sisters, boys with older brothers, girls with older
sisters, and girls with older brothers. For half of these
children, the older sib served as teacher of the younger
sib; the remainder were re-paired so that the older child
taught an unrelated first-grade child. Once taught the
concept, the older child taught it to the younger child
in a 10-minute session.

The children scored higher on the concept attain-
ment test when taught by a female sib than when they were
taught by a male sib or by a female nonsib--boys scoring

higher than girls. Girls teaching their sibs used a

Victor G. Cicirelli

deductive teaching method more than did other teachers.
Children with higher concept attainment scores also
showed greater awareness of the relevant attributes of
the concept. Results can be interpreted in terms of
role theory and sibling rivalry and have implications
for the selection of older children for use in school

tutoring programs.

CONCEPT LEARNING OF YOUNG CHILDREN AS A
FUNCTION OF SIBLING RELATIONSHIP

TO CHILD-TEACHERS

BY

Victor G. Cicirelli

A THESIS

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

DOCTOR OF PHILOSOPHY

Department of Psychology

1971

is.-.

ACKNOWLEDGMENTS

My sincere appreciation to my chairman, Dr.
Charles Hanley, and to other members of my doctoral
committee, Dr. Lauren Harris, Dr. John McKinney, and
Dr. Ellen Strommen, for their constructive suggestions
and encouragement.

I also wish to thank the following persons and
organizations: Dr. Herbert Klausmeier and other staff
members of the Research and Development Center for Cog-
nitive Learning, University of Wisconsin, for making
available facilities necessary for carrying out this
study, and Superintendent Len Brittelli, Assistant to
the Superintendent Lyman Olson, and the principals and
teachers of the Beloit School System, for graciously
allowing this study to be conducted in the Beloit Ele-

mentary Schools.

ii

Chapter

TABLE OF CONTENTS

I 0 INTRODUCTION 0 O O O O O O O O 0

II. DES
III. RES
IV. DIS

Sibling Structure, Intelligence, and

Achievement . . . . . . . . .
Problem. . . . . . . . .
Significance of the Study. . . . .

IGN AND METHOD . . . . . . . .
Population and Sample . . . . . .
Dependent Variable . . . . . . .
Procedure . . . . . . . . . .

Pilot StUdy O O O O O O O O O
ULTS C O O O O O O O O O O 0

Comparisons Between Cell Means . . .
Concept Attainment Score and Reported
Reasons for Identification of Examples
of the Concept . . . . .
Analysis of Variables in the Teaching
Situation . . .

Teaching Session Variables and Concept

Attainment. . . . . . . . .
Sib Status and Sex and Teaching Session
Variables . . . . . . .
Teaching Method and Other Teaching
Situation Variables. . . . . . .

CUSS ION O O O O O O O O O O O

Sibling vs. Parent-Child Interaction .
Directions for Future Research . . .
Educative Implications. . . . . .

V. SUMMARY 9 O I O O O O O O O O 0

REFERENCES

iii

Page

ll
14

15
18
21
22
27
28

36

36
42
47
48
56
62
69
72
74
76

82

Page

APPENDICES
Appendix
A. Trapezoid Concept Attainment Test . . . . 85
B. Frequency Distribution of Concept Attain-
ment Scores . . . . . . . . . . . 96
C. Data Sheet. . . . . . . . . . . . 97

iv

LIST OF TABLES

Table Page

1. Summary of Analysis of Variance for Trape-
zoid Concept Attainment . . . . . . . 29

2. Means and Standard Deviations of Trapezoid
Concept Attainment for Various Sub-
groups. . . . . . . . . . . . . 31

3. Simple Main Effects in the Analysis of
Variance Following a Significant
Interaction Between Sex and Sibling
Status of the Teacher. . . . . . . . 34

4. Newman-Keuls Tests for the Significance of
Differences Between All Possible Pairs
of Subgroup Means . . . . . . . . . 38

5. Learners Concept Attainment and Reasons for
Identification of Concept Examples . . . 4O

6. Reasons for Identification of Concept Examples
and Other Variables . . . . . . . . 41

7. Variables in the Teaching Situation and Con-
cept Attainment Score. . . . . . . . 49

8. Sex of the Learner and Variables in the
Teaching Situation. . . . . . . . . 51

9. Sex and Sibship of the Teacher and Variables
in the Teaching Situation . . . . . . 53

10. Variables in the Teaching Situation and Sex
of Teacher and Learner for Two Extreme
Subgroups. . . . . . . . . . . . 57

11. Teaching Method and Other Variables in the
Teaching Situation. . . . . . . . . 59

E

Table Page
12. Frequency of Use of Various Teaching Methods
and Mean Concept Attainment Scores of
Learners O I O O O O O O O O O O 63

B-l. Frequency Distribution of Concept Attainment
Scores. . . . . . . . . . . . . 96

vi

LIST OF FIGURES

Figure Page

1. Interaction Between Sex and Sibling Relation-
ship of Teacher . . . . . . . . . . 35

2. Learners Concept Score as a Function of Sibling

Relationship, Sex of Teacher, and Sex of
Learner. O O I O I O I O O O O O 37

vii

INTRODUCTION

The academic achievement of children varies from
child to child within the same family as well as from one
family to another. Many investigators explain this phe-
nomenon, at least partially, in terms of different
parent-child interactions. Presumably, parents treat
their children differently depending on the child's
position in the family, and this in turn leads to dif—
ferences in achievement. However, other investigators,
while recognizing the influence of the parent, hold that
the achievement of each child in the family may be
related to the direct influence of sibling-sibling inter-
actions on the child. Presumably, siblings also treat
each other differently depending on their positions in
the family, and this contributes to differences in
achievement.

The achievement of each child in the family might
be understood better if more attention were given to
sibling structure and interaction as variables influencing
the child directly or in conjunction with parent-child
interactions. (Two children have a sibling relationship

1

when they share the same parents. The sibling structure
of a family is the network of positions for children in
the family defined by the number of children, birth
order, sex of the children, their ages, plus the age
spacing between the children. The sibling status of two
or more children is their position within the network,
defined by the age, sex, and birth order of each, plus
the age spacing between them. "Sibling interaction"
means the reciprocal interchange of nonverbal, emotional,
and intellectual communication between siblings.) Intel-
lectual communication (for example, the instruction of

. a younger child by an older) as one aspect of sibling
interaction is of particular interest in relation to
children's achievement, as also is the way such inter-
action may be patterned depending on the sibling status
of the children. This study investigates the effect of
the sibling relationship (for children with different
sibling statuses) on the learning of a concept by a
young child taught by an older child, and demonstrates
the importance of sibling interaction as an educative
process influencing the child's achievement.

There is evidence in the literature for the dif—
ferent treatment of children in different positions in
the family by the parent. For example, Bossard and Boll
(1960) describe the distinct role expectations for chil—

dren depending on their position in the sibling structure.

In an interview study, Dean (1947) reported that mothers
perceived their older child differently from their
younger child of the same sex, and, in an observational
study, Lasko (1954) noted that mothers were less warm
emotionally and more restrictive and coercive toward
their first child than toward their second. Similarly,
Hilton (1967), in a laboratory study, found that mothers
of first-born children were more likely to interfere
with and direct the behavior of their first-born than
their second-born child.

There is, however, little evidence relating inter-
action with siblings to the child's characteristics,
behavior, and development. Irish (1964) has stated that
in general the interactions between and among children in
the home have been given relatively little heed, and,
according to Bossard and Boll,

The role of siblings has been considered chiefly
in the light of 'displacement' and 'rivalry.’ It is
rarely that one finds any but the negative aspects
of sibling relationships, and warnings how to deal
with them (1954, p. 532).

Yet Irish has surveyed the sociological literature to
point out the strength and positive values of the sibling
bonds throughout life, rating them second only to mother—
child ties. He states,

Sibling relationship can perform a number of
functions. Brothers and/or sisters spend many hours
together and share a wide range of activities. . . .
Interactions with siblings function as one avenue for

the socialization of children . . . on occasion, sib-
lings may act as substitutes for parents. They turn

to eaCh other when sufficient attention is not shown
by an indifferent, harried, or uncomprehending parent.
Sometimes siblings are more effective teachers than
adults, particularly if youthful skills are involved.
Siblings may often understand childhood problems and
new situations better, in some ways, than do the
parents they share. . . . Siblings may serve as role
models for one another; particularly may the younger
observe the older siblings of the same sex. They can
serve as challengers and stimulators (1964, p. 282).

From interview-questionnaires administered to
kindergarten children from two-child families, Koch (1960)
found that 70% of the children would rather play with
their sibs than play alone (even though, paradoxically,
second-born children said that they would be happier if
they had no sib).

Adams (1963) found that siblings maintain fairly
frequent contact throughout life, and that sibling
rivalry or comparison between brothers remains important
in adulthood.

From a comparative viewpoint, Harlow's (1969)
studies with monkeys demonstrate the importance of peer
interaction for adequate socialization (where peers are
age-mates reared with the infant monkey). Cooperative
behavior, control of aggression, and appropriate sex
behavior develop optimally when both maternal care and
peer play are available in contrast to maternal care

alone. In fact, peer interactions under optimum con-

ditions may fully compensate for lack of mothering.

Sutton-Smith and Rosenberg state,

Until now most of the work on siblings has
attempted to show how parents make siblings dif-
ferent. The major point to be made in this book,
however, is that siblings also make each other
different (1970, p. 2).

Although the foregoing studies and comments make
the reader appreciate the importance of a child's siblings
for his development, they do little to clarify the nature
of the sibling interaction.

In this regard, one study by Sutton-Smith (1966)
asked fifth-grade children what games they played with
their sibs and with their nonsib playmates, and who was
usually "boss" in the play. He found that first-born
children took high power roles with their younger siblings,
and equal or lower power roles with their friends, while
later-born children took low power roles with their older
siblings but high power roles with their friends.

Sutton-Smith and Rosenberg (1970) reported the
results of an interview questionnaire administered to
upper-grade elementary school children in regard to tactics
to get their sibling to do what they wanted him to do.
Boys used attack and offense more often, while girls used
reasoning, defense, and making the sibling feel obligated;
certain tactics were more typical of first-born than of

later-born children. Also same-sex siblings used more

powerful tactics than did opposite-sex siblings.

Unfortunately, not much is known about the effect
of sibling interaction on the child's achievement. There
are no studies demonstrating that different sibling
interactions are related to differences in achievement;
studies relating sibling structure to achievement have
not dealt with intervening sibling interactions. There
are many studies relating birth order to children's
achievement, but the literature is relatively small when
one considers other sibling variables (age, sex, age
spacing between siblings) and families with more than
two children. Nevertheless, it seems important to sum-
marize the results of certain studies in this area, as
they demonstrate that a child's achievement is rather
strongly influenced by his position in the sibling

structure.

Sibling Structure, Intelligence, and Achievement
In the investigation of the relationship of sib-

ling structure to intelligence and achievement, the
earliest research efforts centered around family size
and IQ. In general, a negative correlation has been
found, which may diminish or disappear with increased
socioeconomic status (Anastasi, 1956). Apparently
family size as such need not limit the intellectual
functioning of family members, provided the socioeconomic
level of the family is high enough to provide adequate

care for each child.

A second major question has been the effect of
birth order on ability and achievement, spurred both by
the psychoanalytic conception of the unusual role of the
first-born and by observation of the over-representation
of the first-born among the eminent (Altus, 1966;
Schachter, 1963). Studies of birth order among older
children and college students have generally demonstrated
the superiority of the first-born over the later-born
child on measures of IQ and school achievement (Altus,
1966; Lees & Stewart, 1957; Maxwell & Pilliner, 1960;
Rosenberg & Sutton-Smith, 1964; Schachter, 1963; Walker
8 Tahmisian, 1967), although the situation appears to be
reversed with preschool and kindergarten children (Abe,
Tsuji, & Suzuki, 1964; Koch, 1954). It might be noted
parenthetically that Bayley (1965) found no relation
between birth order and sex of the child and mental and
motor test scores from 1-15 months of age. Harris (1964),
in a detailed analysis of the work of eminent and highly
creative men, concluded that those among them who were
first-born children tended to produce work of an abstract
verbal nature, while the work of those who were later-born
children was characterized by practical inventiveness
and precise, discriminating attention to detail. The
superiority of the first-born child, particularly in
verbal skills, is usually accounted for by his unique

relationship with his mother.

Studies of more complex family patterns engendered
when the child's birth order, sex, sex of sibling(s), and
sibling spacing are taken into consideration have found
their effect on the child's intelligence and achievement
to be less simple than that claimed for birth order alone.
Such findings are characterized by a great deal of inter-
action between variables, most frequently involving the
sex of the child in relation to the sex of the sibling.
Both Koch (1954) and Schoonover (1959) reported that, in
the two-child family, boys and girls with a male sibling
were superior in IQ and achievement to children with a
female sibling. They concluded that, perhaps because of
his greater freedom, the male is somehow more stimulating
to his sib. Rosenberg and Sutton-Smith (1964), in a study
of college students, confirmed this finding for quanti-
tative scores on the ACE, but found that language scores
were enhanced by the presence of a female sibling. When
the three—child family was considered (Rosenberg & Sutton-
Smith, 1966), the quantitative ACE scores were enhanced
for girls who had two brothers, and for boys who had
heterogeneous siblings (rather than two brothers or two
sisters). In a third study (Rosenberg & Sutton-Smith,
1969), college males from two-child families scored
higher on the ACE when there was a large (4-6 year) age
spacing between them and their sibs, while females scored

higher when they had a female sib with a close age spacing

(1-3 years). Cicirelli (1967), in a study of sixth-grade
school children, found that in the two-child family
first-born girls and second-born boys scored higher on IQ
than second-born girls or first-born boys, while reading
and arithmetic achievement was enhanced for children with
a sibling of like sex close in age. In a portion of the
study concerned with larger families, birth order was not
significantly related to abilities or achievement in
three- or four-child families; however, in the three-child
family, IQ and reading achievement were significantly
depressed for children who had two brothers (compared to
children who had two sisters, or a sister and a brother).

Only a small amount of evidence is available
regarding other kinds of cognitive abilities and traits.
Stewart (1967) discovered the first-born male to be more
field-dependent than the last-born male, while Eisenman
(1967) found that first-born males and later—born females
prefer greater stimulus complexity than do later-born
males or first-born females. Cicirelli (1967) reported
that, in the two-child family, verbal creative abilities
were enhanced for children who had a sibling of like sex
close in age.

Even though any attempt at summarizing the complex
interactions between sibling structure variables found in

the Koch, Rosenberg and Sutton-Smith, and Cicirelli

10

studies will surely result in over-simplification, certain
general trends exist in their results:

1. Ordinal position in the family bears less
relation to intellectual ability and achieve—
ment in the three- and four-child family than
in the two-child family.

2. There is some evidence (Cicirelli, 1967) for
a developmental trend in the effects of birth
order and sibling sex; that is, a sibling
status associated with enhanced development of
a child's abilities at a certain age may be
less facilitating at a later stage in the
child's development.

3. Sibling spacing may affect intellectual ability
in interaction with sex and position variables,
but the effect is not clear or consistent.
Effects of sex and position variables seem
most pronounced for spacings of 2-4 years.

4. Intellectual ability seems to be affected by
the sex of siblings (female sibs associated
with verbal ability, male sibs with quanti-
tative).

The previous research has made it evident that a

child's learning and achievement are rather strongly
influenced by the siblings who surround him during his

early years. Some writers (Bossard & Boll, 1960; Harris,

11

1964) attribute these effects largely to the different
parental treatment accorded a child by virtue of his age,
sex, and position in the family, and to the resulting
role specializations within the family. Others (Irish,
1964; Koch, 1960; Sutton-Smith & Rosenberg, 1970) recognize
the influence of the parent, but hold that the quality of
the interaction between siblings is an important factor;
sibling interactions, like parent-child interactions,
vary according to the participants' positions in the sib-
ling structure. A child's achievement has been related
to his sibling status, and his sibling status has been
related to the nature of his sibling interactions; there
is as yet no evidence demonstrating sibling interaction
to be an intervening mechanism between sibling status and

sibling achievement.

Problem

Inasmuch as the sibling relationship is one of
extensive intimate daily contact, it seems reasonable to
assume that a sibling pair has established customary
patterns of communication and responsiveness to each
other, i.e., an enduring, characteristic manner of inter-
acting. One aspect of this characteristic interaction
pattern might be an educative function where information
is transmitted from one sibling to another and styles of
learning and abilities are gradually shaped and resulting

levels of achievement modified.

12

If a sibling relationship leads to characteristic
and enduring ways of interacting, and one aspect of this
interaction is educative, then a young child might learn
more when taught by an older sibling (with whom he has
well established communication patterns) than when taught
by a similar child who is not his sibling. In other words,
children who have a sibling relationship may learn more
in a teaching-learning interaction situation than children
who do not have a sibling relationship. Of course, the
sexes of the children (and other sibling status character-
istics) may influence both the interaction and the learn-
ing achievement.

On the other hand, the effects of sibling rivalry
and competition are not clear, although Rosenberg and
Sutton-Smith (1969) suggest that sibling competition may
have a negative effect on males when their sib is close
in age.

Studies of social reinforcement (Stevenson, 1965)
while not involving a teaching-learning situation have
shown that social reinforcement by peers was effective
in increasing task performance. However, effects were
greater when reinforcement was delivered by a disliked
rather than by a liked peer. Also social reinforcement
was more effective when delivered by strangers than by
parents. If one considers the older child in the teaching-

learning situation as having some of the properties of

13

the social reinforcer, then one might predict that the

more familiar sibling would be less effective than the

nonsibling.

pair of
given a
younger
come of
concept

teacher

In the present experiment, the older child of a

children (both sibling and nonsibling pairs) was

set period of time to teach a concept to the

child. Although the primary focus was on the out-

the teaching session (as reflected in the learner's

attainment relative to the sexes of learner and

and their sibling relationship), certain behaviors

in the teaching session were also measured. The study

asked the following questions:

1.

Is the concept learning of a young child when
taught by an older child affected by the fact
of their being or not being siblings?

Is there a difference in the concept attain-
ment of children when taught by older children
who are male or female?

Is there a difference in the concept attainment
of young boys and girls when they are taught

by older children?

Does the effect of the sib or nonsib as teacher
depend on his sex and also on the sex of the

younger child?

l4

Significance of the Study

 

The study should contribute to a greater under-
standing of the family, and, more specifically, to the
understanding of the relation of sibling status to sibling
interaction and concept attainment. If young children do
learn more (or less) from their older siblings than from
nonsiblings this would suggest the important effects of
sibling interaction.

From a practical viewpoint, the study may also
extend understanding of an area in which there is as yet
little knowledge: the effectiveness of older children as
teachers of younger children. Bronfenbrenner (1970) has
described the wide use of older children as teachers of
younger children in the Soviet Union, and feels the United
States has failed to exploit this potential source of
teaching talent. The findings may have application to
the more effective use of older primary grade children
as tutors of younger children in the primary grades in
this country. Since approximately 60% of elementary
school children have sibs in the same school, it would
be important to know whether sibs or nonsibs function

more effectively as teachers.

II

DESIGN AND METHOD

A 2-x-2-x-2 factorial study was designed. The
three factors in the design were sex of the learner, sex
of the teacher, and whether the teacher was a sibling.
Below is a diagram of the experimental groups indicating

the number of subjects in each.

 

 

 

 

Male teacher Female teacher
Sib Nonsib Sib Nonsib
Male learner n = 15 n = 15 n = 15 n = 15
Female learner n = 15 n = 15 n = 15 n = 15

 

Subjects could not be assigned at random to the
factors sex of learner and sex of teacher, since they are
properties of the subjects themselves. Consequently,
other factors associated with sex may not be balanced
out, and may cause differences in concept attainment.
However, such a design does allow one to determine if any
relationship exists between sex of the learner and concept

attainment.

15

16

Age of teacher and learner and the spacing between
them were held approximately constant by using only first-
grade learners with an older sib in the third grade, so
that on the average the learner was six years old with a
sib two years older. Because of the small population of
sibling pairs available for the study, it was not possible
to control for the number of children in the family or
for the ordinal positions of learner and teacher therein.
However, within the basic four groups of the sample (boys
with older brothers, boys with older sisters, girls with
older brothers, and girls with older sisters), there was
no reason to believe that size of family and ordinal
position would vary systematically.l Assignment to the
sib and nonsib teaching conditions was random, so that
uncontrolled sib structure variables and other individual
differences would not be expected to vary systematically
between these two groups.

Analysis of covariance was originally proposed to
control for the effect of individual differences in the
learning abilities of teacher and learner. Both the num-

ber of trials needed for the older child to attain

 

1Of the 120 sibling pairs in the study, 107 pairs
were found to come from families where the older and
younger child were "next" to each other in ordinal position;
12 pairs came from families where there was one other sib
between the first and third grades, and the remaining
pair had twins between them in the family. The average
family size in the study was 4.

l7

criterion on the trapezoid concept and the younger child's
score on the Metropolitan Readiness Test were considered
as possible covariates, but since correlations with the
dependent variable were extremely low (see page 30), the
analysis of covariance was not used.

In carrying out the study, the experimenter first
taught the trapezoid concept to the third-grade child
who in turn taught the first-grade child the same material.
The same experimenter was used throughout the study,
thereby minimizing any personality differences that might
have influenced the study. A standardized procedure
(developed in a pilot study) was used in teaching the
trapezoid concept to the third-grade child; the sequence
of events in this teaching session was always the same.
After the lO-minute session in which the older child
attempted to teach the younger child the trapezoid concept,
a standardized concept attainment test was administered
to the younger child.

A list of subjects to be tested at each school was
prepared by an associate and placed in a sealed enveIOpe
to be given to the school secretary who summoned the
children from their classrooms for the experimental
session. Thus, when the older child came for the teaching
session, the experimenter did not know whether the child
with whom he was paired was a younger boy or a younger

girl, or whether the younger child was a sibling or a

l8

nonsibling. Since the list of subjects was randomly
ordered within each school, no attempt was made to control

time of testing.

Population and Sample

 

The population of the study consisted of all the
sibling pairs in 10 of 14 elementary schools in Beloit,
Wisconsin, such that the younger sib was in the first
grade and the older sib was in the third grade. There
were 134 such sibling pairs. The subjects were primarily
lower middle class. Beloit is a small city of 35,000
people on the Illinois border, about 85 miles northwest
of Chicago. It is the site of a small liberal arts col-
lege. Industries in the city include woodworking, and
the manufacture of papermaking and grinding machines,
diesel engines, power pumps, machine tools, shoes,
hosiery, and building and carton paper. Even though the
city is surrounded by farmland, it reflects an industrial-
urban rather than a rural culture.

Of the 120 families actually represented in the
study, 6 fathers held executive or higher professional
positions, 10 fathers were business managers, proprietors
of medium size businesses, or lesser professionals, 11
fathers were administrative personnel or owners of small
independent businesses, 9 fathers were clerical or sales

workers or technicians, 35 were skilled manual workers,

19

33 fathers were machine operators or semiskilled workers,
and 16 were unskilled workers.

The population was divided into the following four
groups: first-grade boys with third-grade sisters (34
pairs), first-grade boys with third-grade brothers (31
pairs), first-grade girls with third-grade sisters (31
pairs), and first-grade girls with third-grade brothers
(38 pairs). A sample of 30 sib pairs was randomly drawn
from each of these four subgroups of the population; 15
pairs were assigned to the sib teaching condition and 15
pairs were assigned to the nonsib teaching condition.
(This random sampling and assignment was accomplished by
assigning each sib pair in a given category of the popu-
lation a number from a table of random numbers; these
numbers were then arranged in ordinal sequence and the
first 15 pairs assigned to the sib condition and the next
15 pairs assigned to the nonsib condition.)

The 15 pairs in a given nonsib group were then
re-paired so that a first-grade learner would be taught
by someone other than his sib. However, since children
could not be transported between schools for the experi-
ment, the re-pairing had to be done within those sib
pairs in the sample from each school. If there were only
two sib pairs at a particular school, these were simply
re-paired; if there were three or more pairs at a partic-

ular school, each pair was assigned a number from a table

20

of random numbers, then the numbers were arranged in order
and the older sib of the lowest number pair in the series
was assigned to the younger sib of the next highest pair,
and so on.

In two cases, there was only one pair out of the
subgroup of 15 that was to be re-paired at a particular
school; hence there was no other pair available at that
school for re-pairing. This was handled in the following
manner. Assume that four schools were represented in the
sample of boys with older brothers, and that after random
assignment to sib and nonsib conditions the distribution

of the pairs was:

Sib Nonsib

School A 4 5
School B 4 4
School C 4 5
School D 3 1

One of the sib pairs from school D was randomly selected
and transferred to the nonsib group; then one of the
nonsib pairs from schools A, B, and C was randomly
selected and transferred to the sib group. Once this was
done, the re-pairing was done as before. Such a procedure
disrupts the original random assignment of pairs to sib
and nonsib categories; however, since this happened only
twice it was felt that the results would not be distorted

to any great degree.

21

Once the re-pairing of the nonsib groups was done,
an associate prepared lists for each of the ten schools
containing all of the sib and nonsib pairs selected from
that school arranged in random order. This list was then
given to the school secretary who summoned children for
testing; the experimenter did not know the subgroup iden-
tification of the children he was working with prior to

the teaching session.

Dependent Variable

 

The dependent variable was the learner's score
on a concept attainment test involving the trapezoid con-
cept. (A trapezoid is a four-sided figure with two
parallel sides.) The trapezoid concept was selected
as being of sufficient difficulty and unfamiliarity so
that first-grade learners would not be able to identify
it immediately. In other words, the trapezoid concept was
not only a concept to be learned, but one of such diffi-
culty that it would allow the third-grade teacher to
interact with the first-grade learner over a considerable
period of time. (This would allow one to gather data on
the teaching—learning process.)

Various types of discriminably different trapezoids
were constructed, varying in the length of the sides, size
of the angles, and orientation of the figure. Of these,
15 different examples were drawn on cards. Fifteen other

geometric figures (circle, rectangle, square, hexagon,

22

etc.) to be used as negative examples were also drawn on
cards. The positive and negative examples were placed in
a random sequence for presentation to the learner. (Appen-
dix A is a copy of the test.)

Instructions for administering the test were as
follows:

Look at each figure. If it is a trapezoid, say

"yes."

If it is not a trapezoid, say "no."
In administering the test, each card was shown in sequence.
The experimenter asked for each card, "Is this a trape-
zoid?" The experimenter also asked each subject why he
answered the way he did and recorded his reasons.

The learner's score was the number of positive
and negative examples correctly identified. Content
validity was assumed; the internal consistency relia-
bility (Kuder-Richardson formula 20) was computed to be

0.63 for the sample of the study.

Procedure

 

The third-grade child was brought to the experi-
mental room by the school secretary or her assistant, and
introduced to the experimenter. During the first part of

the session, the third-grade child was taught to master the

23

trapezoid concept. While a standardized teaching procedure
was used, it included a variety of teaching techniques.2

The instructional materials were 40 geometric
figures on cards: 30 were trapezoids (constructed as
previously described) and 10 were other geometric figures.
A set of 12 thin black sticks about 6 inches in length
was available for extending the lines of the figures.

The subjects were first shown three examples of
trapezoids and three negative examples (circle, triangle,
hexagon) to establish a contrast. They were asked if
they noticed anything similar about the three trapezoids.
Some subjects immediately noticed the four sides (or four
lines). Others did not, and attempted to find some simi-
larity on the basis of size or shape. Finally the exper-
imenter asked them to count the number of sides. After
they did so, the experimenter asked them again what was
common to the trapezoids, to which most would then say
that they have four sides. Another three examples of

trapezoids were then shown, along with three more negative

 

2In a pilot study, third-grade pupils tended to
imitate the way they were taught when teaching the younger
learner. A variety of teaching techniques was then used
by the experimenter so that the third-grade pupil would not
have time to imitate all of these in the short session in
which he taught the younger child. In other words, since
there was not sufficient time for him to duplicate all of
his own experience, he was forced to select certain of the
experimenter's teaching methods. This selective imitation
revealed his preferences and allowed for individuality and
variability in teaching behavior.

24

examples. This time they were shown other geometric
figures which had four sides: a rectangle, a square,

and a parallelogram. The sticks were then used to extend
the lines of the figures to demonstrate that when it was
a trapezoid one pair of sides was parallel (maintained
the same distance between them and did not meet when
extended). It was pointed out that these negative
examples (rectangle, etc.) had two pairs of lines in
which this occurred. Subjects were asked to judge the
distance between pairs of sides of the figures.

The children were then asked to define a trape-
zoid (assuming that they were, by this time, familiar
with the attributes). If the child could not give a
definition he was given hints about the relevant attri-
butes. Finally, if necessary, he was told the attributes
and then asked to tell what a trapezoid was in his own
words. Coaching was continued until he was able to give
a definition.

Then the deck of cards containing the 40 examples
was placed on the table, and the subject was asked to
sort them into two piles: trapezoids and nontrapezoids.
After the subject finished, the experimenter told him
which of his sorts had been correct, which wrong.

The child then was asked to draw a trapezoid and

a figure that was not a trapezoid. If the drawings were

25

not acceptable (four sides and one pair of parallel lines),
the experimenter identified his errors and asked him to
try again.

Finally, the child took the concept attainment
test. After each trial the eXperimenter informed the
child of his errors, and the test was repeated until he
made no more than two errors.

This teaching session with the older child lasted
from 30 to 45 minutes, depending on his speed of learning.

When the older child had reached the learning
criterion, he was asked to be the teacher of the younger
child, and was given the following instructions:

Now you know how to tell a trapezoid when you

see one. We have someone from first grade coming in,
and I would like you to teach him (or her) what a
trapezoid is. I want you to be the teacher now. You
can teach any way you like. You do not have to teach
in the same way that you were taught. Feel free to
try new ways of teaching. It may be that the first
grader will be your brother or sister. If so, teach

him like you would teach anyone else. You may use my
cards, sticks, pencil, paper, or anything else you

 

3The experimenter recorded on his data form (Appen-
dix C) the older child's performance at various stages in
the teaching session: whether or not he had difficulty in
grasping the attribute of four sidedness, and the attribute
of parallel lines, whether he could verbalize the attri-
butes in giving a definition, time needed to sort the deck
of example cards into trapezoids and nontrapezoids, number
of negative examples falsely identified as trapezoids,
number of positive examples falsely identified as nontrape—
zoids, number of trials needed to make an adequate drawing,
as well as item scores for each of the presentations of
the concept attainment test.

26

would like. Now you take my chair and think for a
minute on how you want to teach while I get the
first-grade student.

Then the younger child was brought into the room and intro-
duced to the older child as follows:

I want you to meet . He (or she)
is going to be your teacher. He has been working very
hard and he has learned what a trapezoid is. Now he
would like to teach you the meaning of a trapezoid.
Okay, , now you are the teacher.

 

 

The teaching session between the older and the
younger child then took place; it was timed to last 10
minutes. The experimenter sat in a chair off to the side
and manually recorded the ongoing session. The session
was also tape recorded, but the experimenter's recording
sheet was used as the primary source of data. This was
found to be more fruitful, since the experimenter could
record both verbal and nonverbal behavior (pointing,
sorting, etc.) of the children. The tape alone could not
reveal nonverbal behavior, and many children relied
heavily on it. The tapes were used to cross-check the
verbal responses recorded manually. (A copy of the exper-
imenter's recording sheet is in Appendix C).

After ten minutes, the experimenter interrupted
the teaching session, thanked the older child for his
help, and dismissed him to return to his classroom. Then
the experimenter administered the concept attainment test
to the younger child. The younger child then was thanked

and dismissed to return to his classroom.

27

Pilot Study

 

Before the study was undertaken, a small pilot
study with 20 sibling pairs was carried out in an elemen-
tary school in Mount Horeb, Wisconsin. The pilot study
provided a basis for selecting one of two tasks proposed
for the study; for determining the level of difficulty of
the task, and also whether the subjects responded appro-
priately to the instructional materials; and for estimat-
ing the grade level of children suited to the task.
Neither kindergarten children nor first-grade learners
identified by the principal as "disadvantaged" (children
from extremely poor large families, many from broken
homes) were able to learn the trapezoid concept with the
short teaching session. Teaching procedure, recording
procedure, materials, and instructions were modified and
then standardized. The suitability of the concept attain-

ment test was determined (range and variability of scores).

III

RESULTS

The learner's scores on the trapezoid concept
attainment test were tabulated. Frequency distributions
of scores for the total group and four main subgroups are
presented in Appendix B. Scores ranged from 11 to 30,
the latter being the highest possible score. (The proba-
bility of an individual score of 20 or greater on the
basis of chance is less than .05. Over half of the
children achieved scores of 20 or above, indicating that
some learning of the trapezoid concept had taken place.)

A fixed-effects analysis of variance in the fac-
torial design was used to analyze the learners' concept
attainment scores, and provide answers to the above

4,5

questions. Table 1 is a summary of the analysis of

 

4Cochran's C statistic was calculated to test for
homogeneity of variance. It was .206 for 8 groups and 14
degrees of freedom for each cell. This failed to reach
the .05 level of significance. Since the individuals in
the study were randomly selected from a defined population,
the assumptions concerning random samples from populations
and the independence of numerator and denominator of the
F ratio should be met. Distributions for the group as a
whole and the four main subgroups are given in Appendix B.
Visual inspection of these distributions does not reveal

28

29

TABLE 1

SUMMARY OF ANALYSIS OF VARIANCE FOR TRAPEZOID
CONCEPT ATTAINMENT

 

 

Source df Mean Square F p
Sib-nonsib status of teacher
(S-NS) l 1.88 .13 .72
Sex of teacher (Sex T) l 99.01 6.92 .01
Sex of learner (Sex L) 1 110.21 7.70 .006
S-NS x Sex T 1 126.08 8.81 .004
S-NS x Sex L 1 3.68 .26 .61
Sex T x Sex L l 6.08 .42 .52
S-NS x Sex T x Sex L 1 20.01 1.40 .24
Within Cells 112 14.31 — -

 

30

variance; the means and standard deviations of the groups
involved in the analysis of variance are given in Table 2.
There was no significant difference between
teachers who were siblings or nonsiblings but there was a
significant main effect of the sex of the teacher. Both

of these results must be qualified in the light of the

 

a gross distortion; therefore the assumption of normal
distribution appears to be satisfied. In any event, vio—
lations of these assumptions are not serious when the cell
frequencies are equal as in the present study.

5 . . .

The or1ginal plan of analys1s was to use as
covariates the learner's MRT scores and the number of
trials needed for the teacher to reach a predetermined
level of mastery of the trapezoid concept. (Teachers took
from one to five trials to attain the learning criterion;
the mean number of trials to criterion was 2.87.) However,
when correlations of these variables with the concept
attainment score of the learner were computed, the teacher's
number of trials correlated -.122 with the learner's con-
cept attainment, and the learner's Metropolitan Readiness
Test score correlated .021 with his concept attainment
score. Neither of these correlations was large enough to
be statistically significant, and so the analysis of
covariance was not used. (It might be added that the
teacher's IQ correlated .05 with the number of trials
needed for him to attain the trapezoid concept, and .04
with the learner's concept attainment score.) The exper-
imenter recorded on his data form (Appendix C) the older
child's performance at various stages in the teaching
session: whether or not he had difficulty in grasping
the attribute of four-sidedness and the attribute of
parallel lines, whether he could verbalize the attributes
in giving a definition, time needed to sort the deck of
example cards into trapezoids and nontrapezoids, number
of examples falsely identified as trapezoids, number of
examples falsely identified as nontrapezoids, number of
trials needed to present an adequate drawing, as well as
item scores for each of the presentations of the concept
attainment test. Since none of these measures was sig-
nificantly associated with the learner's concept attain-
ment, no further analysis of this material was made.

31

 

HH.v hv.mH om nflmcoz I mHmEmm .va
mo.m mo.om om nfimcoz I mam: .MH
ah.m hm.ma om nflm I mHmEmm .NH
mm.v mm.om om new I cams .HH
mv.m ma.mH om nflmcoz mamfimm I .OH
om.m hm.ma om nflmcoz OHMZ I .m
mN.m mv.HN om QHm mHmEmh I .m
ha.v hm.ha om QHm OHMZ I .h
mm.m mv.mH om I I mHmEmm .m
mn.m mm.om om I I OHMZ .m
vm.m mN.om ow I mHmEmm I .v
mo.v 5v.ma om I OHME I .m
oo.m mm.mH om Qflmcoz I I .N
ma.v om.ma om new I I .H
am 2 z newcoz Hmcomma Hmcummq
no new no xmm mo xwm

 

mmDQmDmDm mom BZHEZHdBBd EmmUZOU

QHONmm<MB m0 mZOHB¢H>mQ QM<QZ¢Bm 02¢ mZGMZ

N mgmflfi

32

 

no.v

mv.m
mm.v
ov.v
vm.m
om.m
mm.~
mv.v
mm.H

mb.m
mm.m
mv.m
mH.v

mm.mH

no.5a
no.mH
ov.ma
mm.mH
hm.ma
mn.mH
mm.om
om.mm

oa.mH
mn.hH
5v.HN
om.mH

oma

ma
ma
ma
ma
ma
ma
ma
ma

new
new
newcoz
nflmnoz
new
newcoz
nflmaoz
new

can:
can:
mam:
onEmm
onEmm
onEwm
mam:
mamaoh

mHmEmm
can:
mamfiom
can:

Hmpoe

mamfimm
was:
mamfiom
onEmm
mamfimm
mam:
mam:
mam:

onEwm
mamfimm
man:
can:

.hm

.mm
.mm
.vm

.mH
.hH
.ma

33

significant interaction between the sex of the teacher and
the teacher's sib status. In other words, the differences
in effectiveness of sibs or nonsibs as teachers depends
upon their sex, and the differences in effectiveness of

male or female teachers depends upon their sibling rela-

 

tionship to the learner. El
In order to interpret this interaction, tests of %
simple main effects in the analysis of variance were i
carried out to answer the question: Is there a difference E
in effectiveness of male and female teachers (a) when those EV

teachers are siblings, (b) when those teachers are non-

siblings? Conversely, is there a difference in effective-

ness of sibling and nonsibling teachers (a) when those

teachers are males, (b) when those teachers are females?
The tests of simple main effects are found in

Table 3.6

The results are: (a) There is a significant
difference in the effectiveness of sibling and nonsibling
teachers when they are females, but not when they are
males; (b) There is a significant difference in the

effectiveness of male and female teachers when they are

siblings, but not when they are nonsiblings.

 

6Ad hoc comparisons of male sib teachers vs.
female nonsib teachers, and of male nonsib teachers vs.
female sib teachers were also made, using Tukey "A" method.
Neither of these differences was significant.

34

TABLE 3

SIMPLE MAIN EFFECTS IN THE ANALYSIS OF VARIANCE FOLLOWING
A SIGNIFICANT INTERACTION BETWEEN SEX AND
SIBLING STATUS OF THE TEACHER

 

Source df Mean Square F p

 

Sib - nonsib (male

 

teachers) 1 48.59 3.40 .06
Sib - nonsib (female

teachers) 1 79.34 5.54 .02
Sex of teacher

(sibs) 1 224.26 15.67 .001
Sex of teacher

(nonsibs) l 0.81 0.07 .99
Within cells (from

Table 1) 112 14.31 - -

 

Table 2 indicates the magnitude and direction of
the simple main effects, and Figure 1 provides a visual
representation of the form of the interaction. The
results indicate that:

1. Sisters are more effective than brothers
when teaching younger siblings irrespective
of the sex of the younger child.

2. Sisters are more effective in teaching
younger siblings than girls are in teaching
unrelated younger children.

3. Boys tend to be more effective in teaching
unrelated younger children than in teaching
younger sibs.

4. Boys and girls do not differ in effectiveness

as teachers of unrelated younger children.

Mean
Concept
Attain-
ment
Score

status of teacher.

24

23

22

21

20

l9

18

17

 

3S

 

Ar—-—-—1§ Male teacher
O——-O Female te ache r
O
A.
0
IA
_1 l
' v]
Sib Nonsib
Teacher

Figure l.--Interaction between sex and sibling

11?

'! ”(Immi- finmnfl'h‘ (.D‘l'fili‘“

Him-3'.

36

Comparisons Between Cell Means

Even though there was no significant higher order
interaction between sex of teacher, sex of learner, and
sibling relationship of the pair, it is of interest to
compare the eight cells or subgroups of the study.
Table 2 (lines 19 to 26) gives means and standard devi-
ations for these groups, and Figure 2 depicts the
relationship graphically. The Newman-Keuls test was
used to assess the significance of differences in all
pairwise comparisons between cell means. The results of
these tests (Table 4) indicate that older sisters teaching
younger brothers are significantly more effective (in
terms of the learner's concept attainment score) than:

1. older brothers teaching younger brothers

2. older girls teaching younger girls

3. older boys teaching younger girls

4. older brothers teaching younger sisters
Other differences between subgroups were not significant.

Concept Attainment Score and Reported Reasons for

Identification of Examples of the Concept

During the concept attainment test, the learner
was asked his reason for identifying a geometric figure
as either a trapezoid or a nontrapezoid. The learners'
answers were coded into the following categories:

1. Does not know.

2. Irrelevant reason. Learner used irrelevant

or nonessential attributes to decide that

@1517.“ P - .-‘l.- Ea,
-:
- u M

‘1‘. "m

‘3.“

Is '-

37

 

 

 

 

24 E. A. A.Male teacher, male learner
Au—-wA Male teacher, female learner
0 0—0 Female teacher, male learner
23 “' (D—-—~{) Female teacher, female learner
22 «I-
Mean
Concept 21.--
Attainment
Score
20 +-
19 ‘-
18 1F-
17'"
1 l
r l
Sib Nonsib
Teacher

Figure 2.--Learners concept score as a function of
sibling relationship, sex of teacher, and sex of learner.

38

TABLE 4

NEWMAN-KEULS TESTS FOR THE SIGNIFICANCE OF DIFFERENCES
BETWEEN ALL POSSIBLE PAIRS OF SUBGROUP MEANS

 

Difference Between Means

 

 

Means M2 M3 M4 M5 M6 M7 M8
M1 = 23.20 2.93 3.47 3.57 4.67* 4.80* 5.18* 6.13*
M2 = 20.33 - 0.60 0.66 1.80 1.93 2.26 3.26
M3 = 19.73 - - 0.06 1.20 1.33 1.66 2.66
M4 = 19.67 - - - 1.14 1.27 1.60 2.60
M5 = 18.53 - - - - 0.13 0.46 1.46
M6 = 18.40 - - - - - 0.33 1.33
M7 = 18.07 - - - - — - 1.00
M8 = 17.07 - - - - - - -

 

M4 is the mean for sisters teaching younger
brothers; M is the mean for boys teaching younger boys;
M3 is the m an for girls teaching younger boys; M4 is the
mean for sisters teaching younger sisters; M5 is the mean
for girls teaching younger girls; M6 is the mean for boys
teaching younger girls; M is the mean for brothers teach-
ing younger brothers; M8 18 the mean for brothers teaching
younger sisters.

Significant at .05 level.

39

the figure was a trapezoid, e.g., a long
side, points at the end of lines.

3. Same shape. Learner said example was a

 

trapezoid because it had the "right shape."

4. Four sides. Learner cited one of the essential

 

attributes for identifying a trapezoid.

5. Parallel lines. Learner cited the other and

 

more difficult essential attribute for identi-
fying a trapezoid.

6. Both essential attributes.

 

In Table 5, the coded reasons for concept identif-
ication have been tabulated in a bivariate distribution
with the learner's concept attainment score. On inspect-
ing the frequencies, it is apparent that many learners
can identify the concept on an intuitive basis without
being able to say why (as indicated by the "does not know"
category), while others are able to identify the concept
on the basis of only partial learning (as indicated by
the use of only one of the essential attributes).

The first three categories of reasons for concept
identification were combined into a category called "inad-
equate reasons," and the last three categories combined
into a category called "adequate reasons" so that a chi-
square test for association could be computed between
reasons for identifying examples and the learner's con-

cept identification score. The results, presented in

40

TABLE 5

LEARNERS' CONCEPT ATTAINMENT AND REASONS FOR
IDENTIFICATION OF CONCEPT EXAMPLES

 

Reason Concept Attainment Score

 

Low (1-19) High (20-30)

 

1. Does not know 21 13
2. Irrelevant reason 25 5
3. Same shape 7 5
4. Four sides 4 36
5. Parallel lines 1 1
6. Both 4 and 5 1 l

 

Table 6, indicate a significant association between ade-
quate reasons for identification of concept examples and
concept attainment score.

Chi-square tests were also computed between
factors which were significant in the analysis of variance
and the adequacy of reasons for identification of
examples. These are also presented in Table 6. While
there is no significant difference in the reasons pre-
sented by male and female learners, there is a signifi-
cant tendency for learners taught by their older sisters
to present more adequate reasons for identification of
examples. Also, when two extreme subgroups are compared,
brothers taught by their older sisters presented signifi-
cantly more adequate reasons than sisters taught by

older brothers.

41

.Hm>oa Ho. can um DQMOAMHcmHm
as

.Hm>mH mo. men um ucmoflmnamflm
an

 

 

 

..mm.m H a ma mumumfim mcfleommu muwnuoum .n
OH m muonuoun mcwnomop muoumflm .m
mmsoquSm oEmHuxm .v
«om.m m Ha ma newcoc onEom .c
OH om nflmcoc mam: .0
ma ma new mamEmm .n
m em new mam: .m
"Hozommu mo
mflzmcoflpmHmu mafianwm can xmm .m
mm.m a ma me meEmm .Q
mm mm mam: .m
umcumma can no xmm .m
«smo.mm H mm mm AcmoE m>onmv omIom .Q
m mm Adams Bonmv aHIH .m
muoom ucmECwmuum ummocou .H
mumsvm c0mmmm commom m Hum
Iago mp mumsvmc¢ mumsvoomcH and. >
mocmsvmum

 

mmqdem¢> mmmBO 024 mmqm2¢xm
BmmUZOU ho ZOHBflUHmHBZMDH mom mzomflmm

m mam¢9

42

Analysis of Variables in the Teaching Situation

The lO-minute teaching session was tape recorded.
The observer supplemented the recording with notes on the
number of examples of the trapezoid concept presented,
manner of presentation, and so on. While the original
intent had been to count the occurrences of various kinds
of teaching behavior, in actual practice the teachers in
many of the teaching sessions spent so much time simply
presenting examples that there was little variation in the
frequency of other kinds of behavior. Therefore, many
kinds of teaching behaviors were recorded by simply noting
whether or not they occurred in the session.

The behaviors recorded were as follows:

1. Teaching method. A judgment was made by the
observer as to whether the teaching method was deductive,
inductive, or mixed.7 In the deductive method, the
teacher began by defining, describing, explaining, or
demonstrating, and then presented examples. In the
inductive method, the teacher began by presenting examples,
asking the learner if the figure was a trapezoid. Mixed
methods were those in which the teacher began with either
a deductive or inductive approach and then clearly shifted

to the other approach. Of the 120 teachers, 27 teachers

 

7A graduate student also judged the teaching method
on the basis of the observer's record of the teaching
session for each of the 120 teachers; there was 86% agree-
ment between the two sets of judgments.

43

used the deductive approach, 45 the inductive approach,
and 48 the mixed approach.

2. Explained, described, or defined the concept.
Scored "yes" if the teacher made any statement which
served to explain, describe, or define what a trapezoid
was, otherwise "no." Some 53 teachers scored "yes," 67
"no."

3. Demonstrated or illustrated attributes.
Scored "yes" if the teacher made statements or gestures
pointing out the four sides (or angles) of the trapezoid,
or the pair of parallel lines. Scored "no" if such state-
ments were absent. There were 67 "yes" and 53 "no."

4. Number of examples presented. A count of the
number of examples, both positive and negative, which the
teacher presented to the learner in the course of the

teaching session; its frequency distribution is as follows:

 

 

Total No. of examples Frequency
1-10 25
11-20 45
21-30 41
31-40 5
41-50 1
51-60 3

5. Number of positive examples. A count of the
number of positive examples which the teacher presented to
the learner during the teaching session; its frequency

distribution is as follows:

44

 

 

 

Number of Fre uenc
Positive Examples q y
1-10 51
11-20 60
21-30 3
31-40 6

6. Choice of examples. Scored as "random" if the
teacher presented examples to the learner in the order in
which they happened to appear in the deck of example cards
left by the experimenter; scored as "selective" if the
teacher searched through the deck to find particular
examples for presentation. Eighty-five teachers used
examples randomly, and 35 were selective.

7. Presentation of examples. Scored as
"sequential" if the teacher showed examples to the
learner one at a time; scored as "simultaneous" if the
teacher placed two or more examples before the learner at
one time; scored as "other" if the teacher switched mode
of presentation during the lesson. Some 48 teachers
presented examples sequentially, 51 simultaneously, while
21 were in the "other" category.

8. First example. Scored as "positive" if the
first example was positive; scored as "negative" if the
first example was negative; scored as "not applicable" if
the teacher presented two or more examples simultaneously.
Some 32 teachers scored "positive," 26 "negative," and

62 "not applicable."

45

9. Teacher identified examples. Scored "yes"
if the teacher himself identified the examples which he
presented and "no" if he did not. There were 40 teachers
scoring "yes," and 80 scoring "no."
10. Teacher asked learner to identify examples.
Scored "yes" if the teacher asked the learner to state :3
whether examples were trapezoids and "no" if the learner
was not asked to identify examples. There were 99 "yes"
and 21 "no."
11. Amount of feedback. When the learner identi- «J
fied an example presented to him, the teacher may have
told him whether or not he was right. The amount of
feedback was scored either as none, limited (whether
correct or incorrect), extensive (correct or incorrect,
plus reason for error), or "not applicable" (teacher did
not ask learner to identify examples, or shift in method
did not allow for feedback). Number of "none" scores was
17, number of "limited" scores was 63, number of "exten-
sive" scores was 17, and number of "not applicable" scores
was 23. The "none" and "not applicable" categories were
combined for further analysis, as were the "limited" and
"extensive" categories.
12. Kind of feedback. The feedback was scored
"correct" when all of it was accurate, and "incorrect"
if any of it was erroneous. Cases of no feedback were
scored as "not applicable." There were 35 "correct" scores,

47 "incorrect" scores, and 38 "not applicable" scores.

46

13. Teacher asked questions. Scored "yes" if
the teacher asked questions of the learner (other than to
ask him to identify an example), otherwise "no." There
were 33 "yes" scores and 87 "no" scores.

14. Teacher told learner what to do. Scored "yes"
if the teacher directed the learner to do certain things
in the teaching situation (such as count the number of
sides of geometric figures, compare two figures, use
sticks to measure lengths of sides, etc.); otherwise "no."
There were 21 "yes" scores and 99 "no" scores.

15. Teacher asked learner to draw a trapezoid.
Scored "yes" if the learner was asked to draw, otherwise
"no." Since there were only 6 "yes" scores and 114 "no"
scores, this item was not used in any further analysis.

16. Teacher gave hints. Scored "yes" if the
teacher gave hints (e.g., it has four sides) to the
learner regarding the identification of examples, other-
wise "no." Since there were only 5 "yes" scores and 115
"no" scores, this item was not used in any further
analysis.

(Without exception, the third-grade teachers
were cooperative and businesslike in undertaking their
tasks. However, some were more assured than others,
some had doubts about their ability to learn and teach
the concept, and some expressed curiosity about why they
were there. Many became enthused when they were informed

that they were to teach a younger child what they had

47

learned. When the first-grade learners entered the
experimental room, they were a little quiet and reserved,
but not necessarily ill at ease. Sibs, of course, knew
each other, and in most cases nonsibs were acquainted,
knowing the other as a friend of their own sister or
brother.

Like the older children, the first graders were
cooperative on approaching their learning task. In
general, the atmosphere was businesslike and task
oriented; there was little or no interaction of a per-
sonal nature, such as silliness, fooling around, bossi-
ness, showing off, and so on.)

Teaching Session Variables and
Concept Attainment

 

 

Correlational methods were used to investigate
whether variables measured in the teaching situation
were related to the learner's concept attainment score.
The product moment correlation between concept attain-
ment and the total number of examples was 0.06, while the
correlation between concept attainment and the number of
positive examples presented was 0.01, neither correlation
being statistically significant. For nonquantitative
teaching session variables, chi-squares tested the
relations between the variables and the concept attain-
ment score. (Subjects were dichotomized on the basis

of their concept attainment scores: those scoring above

48

the mean and those scoring below the mean.) The results
appear in Table 7. There were only two significant chi-
squares. One reflected a tendency for the initial pres-
entation of a positive example to lead to more effective
learning. The other reflected the greater number of

questions teachers asked of poor learners than of better

learners.

Sib Status and Sex and TeachipgiSession Variables

 

Chi-squares for the relation between sex of
learner and each of the teaching situation variables
appear in Table 8. Only one was significant, a tendency
for the teacher to ask more questions of female learners
than of male learners.

Chi-squares between each of the teaching situation
variables and the compound "sex of teacher and sib status
of teacher," which was categorized into male-sib teachers,

female-sib teachers, male-nonsib teachers, and female-
nonsib teachers, are presented in Table 9. The following

six relationships were statistically significant:

1. Girls teaching their sibs tended to use the
deductive method more often than did other
teachers, while boys teaching their sibs
tended to use the inductive method more often

than did other teachers.

49

 

 

NH m Hmzwo
em.m m om ma Hmeucmsgmm
ma mm msomcmuasfiflm "moamfimxm mo coHUMpcomoum .n
m~.o a ma ma m>HuooHom
me me Eoccmm “moamfimxm wo moaono .m
am.o H mm am an M ”mmadsmxm
mm om OHIH m>flpflmom mo umnfisz .m
mm.m a om om Hm M
Hm mm omIH "monmem mo Honfisz .v
"mmusnfluuum
oa.o H mm em 02 coumuumsHHH Ho
mm mm mow poumuumcosoo umnomoe .m
“umoocoo
va.o H mm em oz omcflmop no .ponwuommc
mm mm mm» .cochmeo Honomoa .m
om.m m mm mm pmxez
om mm m>euosccH
ma m m>Huoscoa "conuoE mcflzomoe .H
mmMMWm mo nmon mzoq manmenm> mceeomme
. mocmswmnm

 

mmOUm BZMEZH498¢ BmMUZOU
924 ZOHBfiDBHm UZHmofimB mmfi ZH mmqmfiHm¢>

h mamfle

50

 

.Hm>ma Ho. mzu #6 DGMOHMHcmLm

es

.Ho>oa mo. may no Damoflwflcmflm
.1

.umou ucmEchupm ummocoo so some one w>onm mcfluoom muocnmmq u mem

n

.ummu ucoEchuum umoocoo so some onu soaon mcwuoom mumcummq H.304o

 

mm.o

«mm.m

Ho.o

mm.o

No.0

mo.o

swo.m

mm
m

om
Ha

ma
vm
ma

ma
mv

Ha
om

ov
Hm

mm
ma
am

he
NH

mm
mm

ma
mm
ha

NN
hm

OH
me

ow
ma

hm
HH
HH

02
mo»

02
mo»

manmoflad
Inc #02
uoouuoocH
uoouuoo

wEom
mcoz

OZ
wow

02
mm»

mfinmoaad
Ian #02
m>Hummmz
m>flufimom

“0c ou umnz
Hocnmoa paoe Honomme

“Hmcumma mo
mcoeummsv coxmo Honomms

"xomnemmm mo cans

“xomnpmmm mo u:soE«

unmanamxm hmwucwcw ou
Hmcumma ooxmm Resumes

"mwamfimxm
pmfiwwucopw Honomoa

"mamfiwxm umufim

.va

.ma

.NH

.HH

.OH

; A
r. I. .... .1." M. .H..u.wui.t "n

51

 

 

mo.m m NH m Monuo
mH om HMHucmsgmm “mmHmexm
om Hm msomcmuHsEHm mo coHumucmmmum .n
nm.o H mH oH m>Huomem
Hv we Eoccmm “moHQmeo mo oOHOEU .m
om.H H mm Hm HH M "mmHmmem
mm mm OHIH o>HHHmom wo Hmnfidz .m
me.m H om om Hm M
ow om omIH "mmHmmem mo umnEsz .v
mwuannuum
vm.o H mm em oz poumuumsHHH no
Hm mm mow coumuumcoEop Hmsomoe .m
"ummocoo cochoo
mm.o H mm Hm oz no .ponHuomoo
em mm mm» .UmcHMmem uwcowme .m
m~.m m mm om cotz
mm mm o>HposocH
0H 5H o>Huozch "cozuofi mcHnommB .H
oumnvm umcumoq umaummq 6 Hum mcH 0mm
IHQU mp mHmEmm mHmz Ham. > .2 B
mocmswwum

 

ZOHBéDBHm DZHmudma mmB ZH mmndem<> 02¢ mmzmfimn HEB m0 xmm

m mHm¢B

1.
PE .Mr. Hrwmgpigud HE?

.Hm>mH H0. we» on HcMOHchmHm
I.

.Hm>oH m0. on» no pcmonHcmHm
an

 

52

N0.N H mm we oz "Co on 06:3
n VH mow HocummH UHOH Honomme
«*m0.n H em on oz ”HochmH mo
mm 0H mow mcoHumosw coxmm Hosomos
mHnmoHHm
mm.H m Hm RH Ida uoz
0m 5m uomuuoocH
0H 0H uoonuou "Homecoom mo chM
00.0 H mm me oEom
mm 0H ocoz ”Homaomom mo unsofim
00.0 H 0H HH oz "mmHmmew mMHucooH on
0m av mow HmchmH coxmo umsomos
mH.0 H am He oz "mmHmmem
Hm 0H mow coHMHucmpH Hmcomoa
mHanHHd
HH.e N mm mm Ida uoz
HH mH m>Hummmz

NH 0H o>HuHmom umemew umHHm

53

 

 

NN.e 0 a e 0 N 06000
0H MH 0 NH HMHucmsvmm "memmeo
HH NH 0H HH msomcmHHsEHm mo coHumucmmmHm .N
.om.a m 0 s 4H HH m>HHomHmm
em 0m 0H 0H Eoccmm "mmHQmem wo moHonO .0
me.m m 4H NH HH NH HH M “mmHHENxm
0H NH NH HH OHIH m>HHHmom mo smassz .m
me.m N mH OH OH mH HN M
mH ON oN mH ONIH "mdesmxm mo amnesz .e
“mmuflflfluuuw
«.VN.NH m NN m m mH oz emumuumsHHH so
0 NN NN mH mos emumuumcosme umeomme .N
"ummocoo
.HN.OH m NH ON a NH oz emanme .0mnHuomme
HH OH HN HH mm» .0mcHMmeo umeomwe .N
.ImH.mN 0 mH NH NH 0 emtz
NH NH N NH m>HuoseaH
N 0 4H m m>Huosemo "cornea mcHzomme .H
mumsvm mp pmzm omzz nmm mm:
IHgo «HQMHum> mcHnomme
mocmswmnm

 

20HB¢DBHm 02H20¢WB 029 2H

m000¢Hm¢> 02¢ mmmo¢mB 029 00 0HmmmHm 02¢ 20m

m 000¢B

54

1
A —
r.W‘I\IIcu§.I“Ii ii u.

.Ho>oH Ho. men so HoooHMHzmHm
k.

or

.Ho>oH mo. ozu no HoooHMHcmHm
.1

.Hocomou nHmcoc onEom u mzmc «Hosomop

 

nHmcoc onE u mz20 “Honomou QHm onEom n mmn “Hocomou nHm mHmE u 020
0H.N m 0N NN NN NN oz "00 on pone
N m m m mm» HOCHMOH UHOu. HOEDMOB .fiH
”HmcumoH
0e.N m HN HN 0N mN oz 00 mcoHumooo
m m 0H 0 mo» ooxmo Hozoooe .mH
oHoooHHo
I.¢N.NN 0 N 0 NH 0 now ooz
0H NH 0 4H uoouuoocH
0 mH 0 0H uoouuoo “xomnpomm mo UcHz .NH
«Hov.NH m HN 0N HH mN ozom
m m ma N. @002 "xowflfiwmm MO HEDGE¢ .HH
"monmeo
00.N m N N m m oz zMHuzooH on
mN 0N HN NN mow monsooH ooxmo Horoooe .OH
om.0 m 0N HN 0H 4N oz "moHdamxo
0H m 0H 0 mo» 0oHOHHcooH nozomoe .m
oHoooHHo
No.0 0 HH 0H 0H NH Ido uoz
m 0 0 0 o>Humooz
HH N m 0 o>HuHmom "oHoeoxo umHHz .m

55

2. Girls teaching their sibs tended to do more
explaining, describing, and defining of the
concept than did other teachers.

3. Girls teaching their sibs and boys teaching
nonsibs tended to do more demonstrating and
illustrating attributes of the trapezoid than
did other teachers; girls teaching nonsibs
did least of all.

4. Teachers of sibs tended to be more selective
of the examples presented than teachers of
nonsibs.

5. Girls teaching their sibs tended to give less
feedback than other teachers, perhaps because
of their greater use of deductive method made
feedback less relevant.

6. Girls teaching nonsibs tended to give incorrect
feedback more than other teachers.

The two extreme subgroups (in terms of their con-
cept attainment scores) were sisters teaching brothers
and brothers teaching sisters. Chi-squares were computed
to determine if there was any association between member-
ship in an extreme subgroup and each of the variables in
the teaching situation. Unfortunately, for 7 of the 14
teaching situation variables, the expected cell frequen-

cies were too small to permit computation of the

56

chi-square.8 The results for the remaining variables are
presented in Table 10. Only one of the relationships was
statistically significant; older sisters teaching younger
brothers explained, described and defined the concept more
often than did older brothers teaching younger sisters.
(Although the chi-square could not be computed, older
sisters teaching younger brothers tended to use a more
deductive approach and brothers teaching younger sisters
tended to use a more inductive approach.)9

Teaching Method and Other Teaching
Situation Variables

 

 

The results of the chi-square tests of the relation
between the teaching method used throughout the teaching
session and the more specific teaching behaviors appear
in Table 11.

Teaching method was significantly associated with
9 of the 13 teaching situation variables for which chi-
squares were computed. The use of the deductive method
was associated with a tendency to:

1. Explain, describe, or define the concept.

 

8However, for 3 of these variables, it was pos-
sible to calculate Fisher's Exact Test. For variable 3,
the probability was .080; for variable 13, it was 1.00,
and for variable 14, it was .168.

9The chi-square computed when the inductive and
mixed teaching method categories were combined was 6.64,
significant at the .01 level.

57

 

 

momou N m m Monuo
300 009 m m HMHucosgom
N N msomcouHSEHm "monmeo 00 coHuoucomon .N
mH.0 H m m o>HuooHom
N 0 Eowcmm "monmeo 00 oOHQEU .0
NHS H a. N HH m
0 0 0HIH "monmew o>HuHmom mo Honesz .0
HS H N N HN m
m m 0mIH "mononm 00 Hogans Hmuoe .v
mommu H 0 H oz "mouannuum omumuumsHHH
Bow 009 m 0H mow no poumnumcoEop nosomoe .m
summocoo
««Hm.0 H 0H m oz pwchop .conHuomoc
m mH mow .pocHMmeo Honomoe .m
mommu N v 0 cotz
Bow 008 0 0 o>HuosocH
m 0H m>Huosch "conuoE mcHsomoB .H
ooumsvm 00 Q00 mmm
IHBU mocmswouh oHQMHHm> mcHsomoB

 

20HB¢DBHm UZH20¢0B 009 2H m000¢H0¢> 2003800 20HB¢HUOmm¢ 000 mBm0B 00¢DOmIH00

mmbommem 0200820 ORB 000 0020¢00 02¢ 00mo¢09 00 20m 02¢

OH 000¢B

58

mcHnomou Honuoun HopHo

.Ho>oH Ho. map so oomoHoHomHm
k.

k.

.Ho>oH mo. one om ooooHoHcon
«.

.muHschcoo How oouoouuoo oHoB monmswmIHsuo «HoumHm Homcsoz

"m0

“Honuoun

nomadoz mcHzooou HoumHm HopHo n mmm

 

mommo
Bow ooe

mommu
sow ooe

mommu

sow ooe

0H.N

mm.N

vH.0

momou
sow 005

500 con 00 NLDG)

NVON

02
wow

02
wow

oHnooHHmmm uoz
uoouuoocH
uoounou

oEom
ocoz

02
wow

02
wow

oHnmoHHmmm uoz
o>Hummoz
o>HuHmom

"op ou umcz
HocumoH UHou Honomos

"HocnmoH mo
mCOHumosg ooxmm Honoooa

"zooooooo mo oon

"Roonpoom mo ucsoޢ

"monono 00HucopH
ou HocuooH coxmo uonoooa

"moHQmeo
ponHucopH Honoooa

"onono umHHm

.VH

.MH

.NH

.HH

.0H

59

 

 

NN.N a NN NH NH oHoooHHooo ooz
m 0H m o>Huomoz
mH HH 0 o>HuHmom "oHQono umuHm .0
MH.N 0 0H 0 8 uoauo
NH «N N HoHucoooom
mm NH HH msoocmquEHm "moHQono mo coHHMHcomon .N
*NN.N N NH N NH o>HoooHom
mm mm «H Eoocmm ”monono 00 ooHoao .0
.NN.N N HN HN NH HH M “moHNENxo
NN NH NH OHIH o>Honoo 0o Honeoz .N
.NN.N N NH NN N HN M
NN NH NH ONIH umoHNzoxo No Hoosoz .e
"mouanuuum
«evm.0m N HH Hm m 02 UoHMHHmSHHH HO
mm «H mm mm% UmflmuflmCOEmmv Hmflumwh. .m
"umoocoo
NNHN.NH N NN 0N NH oz oooHooo .oonHuomoo
NN N NH mow .oooHoHoxo uozoooe .N
onmswm 00 oz QH on
lHflU mHQMHHm> 00800608
. mucosvonm

 

20H8¢D8Hm 02H00¢08 008 2H m000¢H0¢> 00080 02¢ 000802 02H00¢08

HH 000¢8

60

u z

 

.Ho>oH Ho. one no HomoHNHoNHm
k.

k.

.Ho>oH No. or» no ozoonHoNHm
k.

uposuoe mcHnomou o>HuosocH u H “conuoE mCHnomou o>HuosooU u D

.mponHoE mcHnooou coxHE

 

o n o
Nm.m N mm 00 NH oz now on 00:3
0 0 0 mo» HocuooH pHou uocomoe .vH
N«00.0H N 0m NV NH oz ”HocnmoH 00
0H m NH mow mQOHumoso ooxmm uonoooe .mH
«.No.NH N NH N NH oHoooHHoNo ooz
NH 0N v uoouuoocH
0H NH N uoouuou "Homnpoom mo ccHz .NH
ssHo.MH N mm mm HH oEOm
0H 0 0H ocoz "Homnooom 00 0:5084 .HH
HN.m N 0 m 0 oz "monmeo NMHHQoUH 0»
0e 00 NH mow HocumoH ooxmo Honomoa .0H
«NHN.0H N NN 00 NH oz “moHdeoxo
HN 0 0H mow coHMHucocH Honoooe .N

61

2. Demonstrate or illustrate attributes of the
concept.

3. Present fewer examples of the concept.

4. Present fewer positive examples.

5. Be more selective of examples presented.

6. Identify examples for the learner.

7. Give no feedback to the learner.

8. Give more correct feedback, when given.

9. Ask questions of the learner.

Inductive method was associated with the opposite
tendencies. Thus, the relation of specific teaching
behaviors to teaching method supports the observer's glo-

bal judgment of teaching method.

 

IV

DISCUSSION

The discussion of the results takes up the central

finding of the study: the interaction between the sex of

the teacher and the sibling relationship to the learner.

 

Why are sisters more effective than brothers in teaching
younger siblings? Why are sisters more effective in
teaching their younger siblings than girls are in teaching
unrelated younger children? Why are brothers not only
less effective than sisters in teaching their younger
siblings, but also less effective than boys teaching
younger unrelated children? And finally, why is there
little difference in the effectiveness of boys and girls
as teachers of younger unrelated children, when girls
(sisters) are so much more effective than boys (brothers)
as teachers of their sibs? How can all these findings
be explained?

The data on teaching method may provide an
initial clue to an explanation. Table 12 presents data
taken from Tables 2, 9, 10; it can be seen that there is
a tendency for the groups making the greatest use of
inductive teaching method to have lower scores on the

62

63

 

 

 

N0.NH 0 v m muoumHm
mCHnomop muonuoum

0N.mN 0 0 0H muonuoun
0cHnomou muoumHm

Nm.NH NH 0 0 mnHm
mcHnomou muonpoum

mH.NH NH 0H N mnHmcoc
mcHsomou mHHHw

Nm.NH NH NH 0 mnHmcoc
mchomoH ozom

mv.HN 0 MH 0H mQHm
mcHsooou muoumHm

ouoom o>Huosch potz o>HpooooQ
ucoEchund moouw

umoocou coo:

ponuoz mcHzomoB mo mucosvoum

 

m0020¢00 00 0000m 82022H¢88¢
8000200 2¢0E 02¢ 0000802 02H00¢08 mDOH0¢> 00 0m0 00 002000000

NH 000¢8

64

concept attainment test. This tendency is exaggerated
when the two "extreme" subgroups (sisters teaching
brothers and brothers teaching sisters) are considered.
These data suggest the hypothesis that a deductive
teaching method is more effective than an inductive
method when older children teach young children a con-
cept attainment task.

In other words, sisters teaching siblings were
most effective, and made greater use of a deductive
teaching method than did other teachers. On the other
hand, brothers teaching sibs were least effective and
made the most use of inductive teaching method. (Girls
teaching younger unrelated children used a deductive
approach less often than brothers teaching younger sibs,
but they used a mixed approach almost twice as often.
The conclusion here is that the amount of deductive
teaching used in the mixed method is sufficient to
account for the results.)

Table 5 indicates that those learners who had
the highest concept attainment scores also gave more
adequate reasons for identification of the concept
examples. Deductive teaching may be more effective in
concept attainment with young children in that it enables
the young child to become aware of the relevant attri-
butes of the concept at an age when he may be unable to

abstract these attributes for himself from presentation

 

65

of examples. Also, a deductive teaching method provides
the learner with appropriate verbal labels for concept
attributes when these might not be in the young child's
repertoire. (A deductive approach is not to be equated
with a verbal approach to teaching. Both deductive and
inductive methods can involve concrete objects, manipu-
lation, examples, applications, and so on. However, a
deductive approach tells and gives the rules followed
by examples, while an inductive approach gives the
examples and the learner must abstract the rules with
varying degrees of guidance by the teacher.)

While the results of the study appear to give
support for the position that a deductive approach is
more effective than an inductive approach in teaching
concepts to primary grade children, it must be cautioned
that the relationship between teaching method and con-
cept attainment is probably a limited one. The chi-square
for teaching method and concept attainment score (dicho-
tomized into high and low scores) was not significant
(Table 7). However, in this case, the chi-square may
not have been sensitive enough to detect a difference
as other evidence does suggest the existence of at
least a weak relationship.

Another finding was that on the average boys
scored higher than girls on the concept attainment test.

While this may be regarded as an instance of general

 

66

male superiority on tests involving mathematical and
spatial ability (Tyler, 1965), it too fails to account
for the findings since the male learner's achievement
is not consistently superior but varies with the sex of
the teacher and sibling relationship to the teacher in
a similar pattern for both sexes (Figure 2). Another

hypothesis is that the learner's abilities are different

.V. U

.N-r-z
00--

depending on his sex and the sex of the sibling (pp. 6—10),
but this fails to account for the different results

obtained when learners are taught by sibs or nonsibs;

 

the general ability of the learner was found to be non-
correlated with concept attainment (p. 30), and also
there were no significant interactions involving the sex
of the learner. The prediction from social reinforce-
ment literature that nonsibs would be more effective
than sibs was not supported.

However, the finding that female sibs were more
effective as teachers of the younger child than were
female nonsibs or male sibs can be interpreted in terms
of role theory and sibling rivalry.

Girls tend to be more identified with their
mothers and with their female school teachers than are
boys (Koch, 1960; Sutton-Smith & Rosenberg, 1970), and
thus it would seem that they are more ready to assume a
teaching role. Also an older sister tends to have more
responsibility for care of younger sibs than does an

older brother, often being delegated caretaking or

67

teaching roles (Bossard & Boll, 1960; Mead & Heyman,
1965). Sisters consequently develop the role of teacher
of younger siblings to a greater extent than do brothers;
in fact, the older sisters in this study may have used

a deductive teaching method as frequently as they did
simply because they were more accustomed to telling
younger siblings what to do. Also, according to Sutton-
Smith and Rosenberg (1970), reasoning was a tactic used
by older girls interacting with younger sibs.

Conversely, learners at the elementary school
level may be more ready to accept a girl in the role
of teacher, especially if she is their sister. (One
of the first-grade learners in this study remarked that
her third-grade teacher was no good because "boys can't
be teachers.") It is hypothesized that the younger sib
of an older sister is more motivated to learn from her
because he perceives her in the role of a teacher, and
his learning is more effective because he has developed
habits of responding to the manner in which his older
sister communicates.

This hypothesis is consistent with the research
literature and can account for both the superiority of
the sisters when they are teaching younger brothers and
the greater effectiveness of the deductive teaching

method.

 

I: .d ' .» 3-17,.

68

However, it does not seem to account for the dif—
ference between brothers teaching younger siblings com-
pared with boys teaching younger unrelated children.

Why are brothers the least effective teachers? Perhaps
the effects of sibling rivalry are greatest when the
teacher is a male sib. There is a tendency for boys to
react more intensely to sibling displacement, and sibling
displacement by a newborn that occurs within 2-4 years
is most intense (Koch, 1960; Rosenberg & Sutton-Smith,
1969). Thus older brothers, close in age to younger
siblings, would demonstrate more hostility, competitive-
ness, and jealousy than would older sisters. Such sib-
ling rivalry could well lead the younger sibs in turn

to be defensive and resist learning from an older brother
Both Sutton-Smith (1969) and Adams (1968) speak of the
effects of sibling rivalry between males of college age
or older; Cicirelli (1967) found IQ and reading achieve—
ment to be depressed in children from three-child
families who had two brothers. On the other hand, both
Koch (1954) and Cicirelli (1967) found that in the two—
child family, achievement was enhanced for boys with
older male sibs. Thus, the situation is not entirely
clear, but it may be that when sibling rivalry exists,
the male sib is less motivated to teach his younger sib,
or the younger is more resistant to learning. Sutton—

Smith and Rosenberg (1970) reported that older boys

69

typically used "attack and offense" as power tactics in
interactions with their younger sibs.

In summary, the younger sib's motivation to
learn is increased when taught by an older sister
because she is perceived as fitting the teacher's role,
and younger siblings are less motivated to learn when :0
taught by older brothers because of the greater sibling .
rivalry. (Boys and girls who taught younger unrelated .

children should be relatively equivalent in effective-

ness; boys should be more effective than brothers, since 1}

 

less sibling rivalry would be involved, and girls should
be somewhat less effective than sisters because the
deductive method was not so frequently used and the
learner may not be so motivated to accept a female non-

sib as teacher.)

Sibling vs. Parent-Child Interaction

 

A further issue can be raised: Why should sis-
ters use the deductive teaching method so markedly when
they are teaching sibs? If this were a general approach
to teaching developed from a parent-child relationship,
then sisters should consistently apply this approach to
the teaching of related and unrelated children. (All
the girls in this study were actually older sisters
with younger siblings; half of the "sisters" were ran—
domly re-paired to teach unrelated children.) As we

see from Table 12, 14 of 30 older sisters used the

7O

deductive method while teaching sibs (and only 3 used the
inductive method), while only 2 of 30 older girls used
the deductive method while teaching nonsibs (and 13 used
the inductive method). This difference suggests that the
sister's use of a deductive teaching method is not a
general characteristic of her, but is relative to the
situation. Extending this idea, if parents treat older
sisters and brothers differently, resulting in the
development of different characteristics, then one
would expect sisters to teach in a consistently different
manner from brothers, regardless of who was the learner.
But this is not the case. Not only do girls shift in
their teaching approach depending on their sibling
relationship to the learner, but this is also true of
the older boys as teachers. Concept attainment by
learners varied similarly depending on the sibling
relationship to the teacher.

This shifting in teaching approach by the
teacher and achievement by the learner depending upon
the sib relationship strongly suggests that sibling
interaction itself is an important determiner (along
with parent-child interactions) of the role behavior
of siblings. The sister's or brother's teaching role
emerges out of the ongoing interaction between sibling
pairs as they establish their customary patterns of

communication and responsiveness.

 

 

71

As was pointed out in Chapter 1, sibling inter-
action itself has never been adequately studied. Using
an interactional framework, one might think of the
family as a system involving the following three sub-
systems: parent-parent interactions, parent-child
interactions, and sibling-sibling interactions. Per-
haps the sibling-sibling interactions can be studied in
and of themselves, and their relations to the other sub-
systems. One might think of the peer group area of
study as being subdivided into family peer groups and
non-family peer groups. Within this framework, one can
View sibling interactions as one example of peer group
interactions, and hence bring to bear the existing
research in that area for formulating hypotheses. Cer-
tainly this is extremely important, as sibling inter-
actions may lead to the development of roles that
transfer to nonfamily peer group interactions. Sutton-
Smith (1966) has noted the reversal of power roles which
occurred depending on whether a child was playing with
his sibs or his nonfamily peers.

In today's world, interactions between peers
are important to understand since the peer group is
becoming an ever stronger agent of socialization. (The
Russians and Israelis use the peer group to inculcate
adult values; and in America peer group socialization is

in defiance of adult values.) In any event, the sibling

 

ifi- Jr's-wu- a
. .

‘13

.- -29

72

relationship is one of extensive intimate daily contact,
and it seems reasonable to assume that a sibling pair
has established customary patterns of communication
which would also serve for the educative function of
their interaction. Thus, by studying interaction, one
should be able to get valuable insight into the manner

in which they influence each other's learning.

Directions for Future Research

 

Further studies are needed to determine whether
the results can be extended to different children under
different conditions. For example, the children in
this study were two years apart. This means a great
amount of interaction, and hence optimum conditions for
sibling rivalry (Rosenberg and Sutton-Smith, 1969). Sib-
ling rivalry should be less when the sibling spacing is
greater, and therefore one might predict that fifth-grade
boys would be relatively more effective teachers of
first-grade brothers than were the third-grade boys of
this study. Also, the effectiveness of sibs as teachers
may be related to family size. One might speculate that
as family size increases, the amount of mother-child
interactions with a given child would decrease and the
influence of older siblings would increase.

A series of studies, recently reported by Hess
and Shipman (1965), may help point the direction in

which future research on siblings should move. Hess

‘T‘Efi‘

 

'.

( “kn-11'»

73

and Shipman attempted to observe and isolate aspects of
maternal behavior (maternal teaching styles) related to
the child's cognitive development. In these studies, a
preschool child and his mother were observed in an inter-
action session, during which the mother was asked to

teach three simple tasks to the child. Mothers from

r 2.“..- ‘H
I '1

four different socioeconomic status groups were found
to differ on language usage (particularly on use of
abstract terms and complex syntactic structures), pre-

ferred modes of classificatory behavior, and various

[TR-m“: .-

aspects of teaching style (e.g. praise, criticism), even
though there was relatively little difference in the
affective element of their interaction with their chil-
dren. Both the children's performance on the tasks and
their verbalization of the underlying principles were
found to be correlated with the responses of the mothers.
Working class mothers and children, for example, tended
to act without taking time for reflection and planning,
while the middle class mothers were more likely to
encourage the child through verbal and nonverbal cues

to reflect, to anticipate the consequences of his action,
and in this way to avoid error.

This kind of approach may well lead to a better
understanding of intrasibling influence. Particularly
where the preschool child is concerned, it would seem
that the older sibling (or siblings) would have the

greatest influence on the younger sibling (rather than

74

vice versa), at least in intellectual tasks. Assuming
this to be true, then the cognitive and teaching styles
of the older sibling may well be important variables
influencing the learning and cognitive development of

the younger sibling. And, analogous to the Hess and
Shipman work, such characteristic styles may be uncovered
through observation of sibling interaction in an intel-
lectual task. The task in the current study was perhaps
too highly structured to allow this kind of individual

variation in teaching.

Educative Implications

 

In summary, this study provides evidence to
support the idea that sibling interaction can be a con-
tributing factor to the concept attainment of younger
children. Such results may have practical implications
for schools using older children as tutors for younger
children (provided the generalizations are cautiously
made). For example, Cloward (1967) reports a successful
program using high school students as tutors of fourth-
and fifth-grade students. The Wisconsin Research and
Development Center for Cognitive Learning conducted a
study in which sixth-grade pupils helped six- and seven-
year-olds with arithmetic with encouraging results
(Report from Project MODELS, 1968). Also, the Center
for Research on Utilization of Scientific Knowledge at

the University of Michigan's Institute for Social Research

75

is actively promoting the establishment of "cross-age
helping programs" in the schools (Lippitt, 1969; Lippitt
& Lippitt, 1968). While not all outcomes of such pro-
grams are positive, children are being used more than
ever as teachers of other children (Bronfenbrenner,
1970). Since it is common for children to have at least ”1
one sibling in the same school, the current findings
would suggest that it would be advantageous to use girls

as tutors of their younger siblings in such programs

 

where possible, but to use boys or girls as tutors of NJ
nonsiblings. Of course such a recommendation should

serve merely as a guideline for action type research,

in which the results are evaluated and actions revised

accordingly.

SUMMARY

Previous research has demonstrated that the
achievement of a child is related to his position in

the sibling structure of the family. This is frequently

 

explained by differences in parental treatment, but
interaction with siblings may also be an important
factor.

The purpose of this study was to investigate the
effect of older sibs and nonsibs of both sexes as
teachers of younger children on a concept learning
task. The specific questions of the study were:

1. Is the concept learning of a young child
when taught by an older child affected by
the fact of their being or not being
siblings?

2. Is there a difference in the concept attain-
ment of children when taught by older
children who are male or female?

3. Is there a difference in the concept attain-
ment of young boys and girls when they are
taught by older children?

76

77

4. Does the effect of the sib or nonsib as
teacher depend on his sex and also on the
sex of the younger child?
Each of the 120 first-grade children in the study
had an older sibling in third-grade; equal samples of 30
sibling pairs were drawn from the population of boys with {1
older sisters, boys with older brothers, girls with older
sisters, and girls with older brothers. For half of the

children in each group, the older sib served as the

 

teacher of his or her younger sib; the remaining half K]
were re-paired so that the older child taught an unre-
lated first-grade child.

The experimenter trained the older child in the
trapezoid concept to a given learning criterion in a
30-45 minute session. A standardized teaching procedure
involving a variety of teaching techniques was used.
Then the older child was asked to teach the trapezoid
concept to a younger child. (The ensuing lO-minute
teaching session was recorded on an observation schedule
by the experimenter as well as on magnetic tape.) The
younger child was then given a concept attainment test
to determine his mastery of the trapezoid concept.

An analysis of variance in the 2 x 2 x 2 fac—
torial design was carried out on the learner's concept
attainment scores. Also chi-square tests of association

were made to determine whether any relationship existed

78

between variables measured in the teaching situation and

the learner's concept attainment.

The main results of the study can be summarized

as follows:

1.

Concerning the attainment of the trapezoid

concept,

a. Regardless of their teachers, young boys
achieved significantly greater attainment
of a trapezoid concept than young girls.
This is not unexpected in view of boys'
superior spatial abilities.

b. Older sisters were significantly more
effective teachers of younger siblings
(boys or girls) than older girls were of
unrelated younger children (boys or girls).
On the other hand, older boys tended to
be more effective teachers of unrelated
younger children (boys or girls) than
older brothers were of young siblings.
However, the results were not statisti-
cally significant.

c. Older sisters were significantly more
effective teachers of younger siblings
(boys or girls) than older brothers.
Older boys and girls showed no significant

difference in their effectiveness as

 

79

teachers of unrelated younger children
(boys or girls).
When the 8 subgroups of the study were com-
pared, the subgroup in which the learners
showed the highest concept attainment score
was that in which older sisters were IN
teachers of younger brothers. This sub-

group differed significantly from the lowest

x- ‘un- A

scoring subgroup, that in which older

brothers were teachers of younger sisters,

 

Hill?
0’ ' .

as well as from subgroups in which older
brothers were teachers of younger brothers,
older boys were teachers of younger girls,
and older girls were teachers of younger
girls. No significant differences were

found in other subgroup comparisons.

In subsequent analysis of the data, it was
found that learners who had higher concept
attainment scores also showed greater aware-
ness of the relevant attributes of the con-
cept, in contrast to those who either did
not know the attributes, invented erroneous
attributes, or who used irrelevant attributes.
An analysis of variables in the teaching sit-
uation indicated that these variables were
not in general related to the learner's con-

cept attainment score. However, it was found

80

that girls teaching their sibs used the
deductive teaching method and its associated
teaching behaviors (explaining and describing,
demonstrating and illustrating attributes,
selection of examples) more than did other
groups. When the two extreme subgroups were E1
compared, older sisters teaching younger
brothers tended to use the deductive approach

while older brothers teaching younger sisters

 

tended to use the inductive approach. L}

By way of interpretation of the results, it was
suggested that high concept attainment scores were
related to the use of a deductive method of teaching by
older sisters (which was part of a well-established com-
munication system serving the educative function); that
the learners were more ready to accept older girls than
older boys in the role of teacher, and that girls were
more ready to assume the teaching role than boys. Sib-
ling rivalry between the older brother and his sibs was
advanced as a possible reason for the relative ineffec-
tiveness of male sibs as teachers.

Finally, it was suggested that sibling inter-
action itself is an important factor along with parent-
child interaction in determining the role behavior of
siblings. The sister or brother's teaching role relative

to another sibling emerges out of the ongoing interaction

81

between the pair as they establish their customary pat-
tern of communication and responsiveness.

Although the results of the study need to be
extended in further research, they may find application
to those school programs using older children as tutors
of younger learners. These findings suggest that it
would be best to avoid selecting boys and instead use
girls as tutors of their younger siblings where possible,

but to use either boys or girls as tutors of nonsiblings.

 

REFERENCES

 

REFERENCES

Abe, K., Tsuji, J., & Suzuki, H. The significance of WN
birth order and age difference between siblings E}
as observed in drawings of prekindergarten
children. Folia Psychiatrica & Neurologica
Japanica, 1964,717, 315-325.

 

Adams, B. N. Kinship in an urban setting. Chicago: :
Markham, 1968. j

 

 

Altus, W. D. Birth order and its sequelae. Science,
1966, 151 (3706), 44-49.

 

Anastasi, A. Intelligence and family size. Psyghological
Bulletin, 1956, 53, 187-209.

 

Bayley, N. Comparison of mental and motor test scores
for ages 1-15 months by sex, birth order, race,
geographical location, and education of parent.
Child Development, 1965, 36, 379-411.

 

Bossard, J. H. S., & Boll, E. S. Security in the large
family. Mental Hygiene, 1954, 38, 529-544.

 

Bossard, J. H. 8., & Boll, E. H. The sociologygof child
development, 3rd Ed., New York: Harper, 1960.

 

 

Bronfenbrenner, U. Two worlds of childhood. New York:
Russell Sage Foundation, 1970.

 

Cicirelli, V. G. Sibling constellation, creativity, I.Q.
and academic achievement. Child Development,
1967, 38, 481-490.

 

Cloward, R. D. Studies in tutoring. Journal of Experi-
mental Education, 1967, 36, 14-25.

 

 

Dean, D. A. The relation of ordinal position to perso—
nality in young children. Master's Thesis.
State University of Iowa, 1947.

Eisenman, R. Complexity-simplicity: II. Birth order and
sex differences. Psychonomic Science, 1967,
8, 171—172.

 

82

 

83

Harlow, H. F. Age-mate or peer affectional system. In
Lehrman, D. S., Hinde, R. A., & Shaw, E. (Eds.),
Advances in the study of behavior, Vol. II.
New York: Academic Press, 1969. Pp. 333-383.

 

Harris, I. D. The_promised seed. New York: Free Press
of Glencoe, 1964.

 

Hilton, I. Differences in the behavior of mothers toward
first- and later-born children. Journal of
Personality and Social Psychology, 1967, 7,
282-290.

 

 

' $11111
b’

——-- r.

Irish, D. P. Sibling interaction: A neglected aspect in
family life research. Social Forces, 1964, 42,
279-288. I

 

 

Koch, H. L. The relation of primary mental abilities in
five- and six-year-olds to sex of child and
characteristics of his sibling. Child Develop—
ment, 1954, 25, 209-223.

 

Koch, H. L. The relation of certain formal attributes of
siblings to attitudes held toward each other
and toward their parents. Monographs of the
Society for Research in ChiId Development,
1960, 25, 1-124.

 

 

Lasko, J. K. Parent behavior toward first and second
children. Genetic Psychology Monographs,
1954, 49, 97-137.

 

Lippitt, P. Children can teach other children. Instructor,
May, 1969.

 

Lippitt, R., & Lippitt, P. Cross-age helpers. NEA
Journal, March, 1968.

Lees, J. P., & Stewart, A. H. Family or sibship position
and scholastic ability. Sociological Review
(British), July, 1957, 85-106.

 

Maxwell, J., & Pilliner, A. E. G. The intellectual
resemblance between sibs. Annals of Human
Genetics, 1960, 24, 23-32.

 

Report from Project MODELS. Technical Report No. 46,
Research and development activities in R & I
Units of two elementary schools of Milwaukee,
Wisconsin, 1966-67. Madison, Wisconsin:
Research & Development Center for Cognitive
Learning, 1968.

84

Rosenberg, B. G., & Sutton-Smith, B. The relationship of
ordinal position and sibling sex status to
cognitive abilities. Psychonomic Science, 1964,
1, 81-82.

Rosenberg, B. G., & Sutton-Smith, B. Sibling association,
family size, and cognitive abilities. Journal
of Genetic Psychology, 1966, 109, 271—279.

Rosenberg, B. G., & Sutton-Smith, B. Sibling age spacing
effects upon cognition. Deve10pmental Psycho- t1
logy, 1969, 1 (6), 661-668. '

 

Schachter, S. Birth order, eminence, and higher education.
American Sociological Review, 1963, 28, 757-768.

 

Schoonover, S. M. The relationship of intelligence and .
achievement to birth order, sex of sibling, and t
age interval. Journal of Educational Psychology, rm
1959, 50, 143-146.

 

Stevenson, H. W. Social reinforcement of children's
behavior. In Lipsitt, L. P., & Spiker, C. C.
(Eds.), Advances in child development and
behavior. Vol. II. New York: Academic Press,
1965. Pp. 97-126.

Stewart, R. H. Birth order and dependence. Journal of
Personality and Social Psychology, 1967, 6,
192-194.

 

 

Sutton-Smith, B. Role replication and reversal in play.
.Merrill Palmer Quarterly, 1966, 12, 285-298.

 

Sutton-Smith, B., & Rosenberg, B. G. The sibling. New
York: Holt, Rinehart and Winston, 1970.

 

Walker, C. E., & Tahmisian, J. Birth order and student
characteristics: a replication. Journal of
Consultipg Psychology, 1967, 31, 219}

 

 

 

APPENDICES

 

APPENDIX A

TRAPEZOID CONCEPT ATTAINMENT TEST

APPENDIX A
TRAPEZOID CONCEPT ATTAINMENT TEST

Wisconsin Research & Development Center E;
for Cognitive Learning '

This is an achievement test designed to assess the E
child's attainment of the trapezoid concept. It consists ‘
of 30 geometric figures for the subject to identify, 15
trapezoids and 15 non-trapezoids arranged in a random
sequence. Figures selected as examples of trapezoids
varied in length of lines, size of angles, and the spatial
orientation of the figure. Nonexamples were selected from
such figures as hexagons, pentagons, quadrilaterals, paral-
lelograms, rectangles, squares, triangles, and circles.

 

II 4:01;
“II—*-

The subject is instructed to tell whether each
figure is or is not a trapezoid after viewing it. His
score is the number of items correct.

Content validity is assumed; internal consistency
reliability is .63.

Instruction to subject: Look at each figure. If it is a
trapezoid, sayflYes." If it is not a trapezoid, say "No."

 

85

1. Is this a trapezoid?

Yes

No

 

2. Is this a trapezoid?

Yes

w

No

 

' 3. (Is this a trapezoid?

Yes

 

No

 

86

 

 

 

 

 

I?"

4. is this a trapezoid?

Yes

 

No

*—

5. Is this a trapezoid?

Yes

 

No

 

2M" ‘ I

6. Is this a

trapezoid?
Yes

 

No

 

 

 

 

 

 

I-Z "-

 

.7
I o

8.

9.

Is this a trapezoid?
Yes

No

Is this a trapezoid?

Yes

 

No

Is this a trapezoid?
Yes

No

 

 

 

 

C

10. Is this a trapezoid?

Yes

 

No

 

11. Is this a trapezoid?

Yes.

 

No

 

12. Is this a trapezoid?

'Yes

 

N0

89

 

 

 

- r .1...

 

14“. 0.45.212 [Ts-3‘ Nulffi .-

‘-

13. Is this a trapezoid?
Yes

No

 

14. Is this a trapezoid?
Yes

No

15. Is this a trapezoid?
Yes)

No

 

 

16. Is this a trapezoid?

Yes

No

 

 

17. Is this a trapezoid?

Yes

 

No

 

 

18. Is this a trapezoid?

.Yes

No

 

 

 

‘fi _--_~ “‘4'
i-, g

19.

20.

21.

Is this a trapezoid?
Yes

No

 

Is this a trapezoid?
Yes

w

No

Is this a trapezoid?
Yes

No

 

 

93

22. is this a trapezoid?
Yes

No

 

23. Is this atrapezoid?
Yes

No

24. Is this a trapezoid?

Yes

 

No

 

 

 

25. Is this a trapezoid?

Yes

 

No

26. Is this a trapezoid?
Yes

No

27. Is this a trapezoid?
Yes

No

94

 

 

 

'F

I]
.l'zii
U

r- -z 7“ .5.”

 

'13?"

 

28. is this a trapezoid?

Yes

No

 

29. Is this a trapezoid?

Yes

No

30. Is this a trapezoid?

Yes

 

No

 

95

 

 

 

I

 

APPENDIX B

FREQUENCY DISTRIBUTION OF CONCEPT ATTAINMENT SCORES

96

APPENDIX B

TABLE B-l

FREQUENCY DISTRIBUTION OF CONCEPT ATTAINMENT SCORES

 

 

 

Concept Total Females Males Males Females
Attainment Group Teaching Teaching Teaching Teaching
Score Males Males Females Females
30 1 l
29
28
27
26 3 2 l
25 6 3 l l 1
24 12 6 2 l 3
23 9 5 2 l 1
22 14 4 2 2 6
21 6 3 2 l
20 10 3 2 3 2
19 10 2 3 5
18 8 3 4 1
17 ll 4 l 3
16 7 3 l
15 8 2 5 1
l4 4 l 1 2
l3 4 2 l 1
12 4 l l 2
ll 3 l 2
Total N 120 30 30 30 30

 

APPENDIX C

DATA SHEET

 

iJ on
b

II.

APPENDIX C

DATA SHEET-Older Child

School

 

Background data

a.

Name b. Otis IQ

 

 

Observation Schedule--Experimenter teaching older child.

a.

Presentation of examples; identification of 4 sides.
Identifies 4 sides on first trial
Does not identify 4 sides on first trial
Uses irrelevant attributes instead of 4 sides
(e.g. size, shape)
Mentions no attributes

Presentation of examples: Identification of paral-
lel and nonparallel lines of trapezoid. (Subject
judges distance between pairs of parallel and non-
parallel lines as being equal or unequal.)
Subject makes all judgments correctly.
Subject makes one or two judgments incor-
rectly.
Subject makes more than two judgments incor-
rectly.

Definition of trapezoid.
Defines trapezoid on first trial
Defines after given hints on relevant
attributes
Defines after being told relevant attributes
Defines after extensive coaching

Sorting
Time to complete sorting
Number of errors; nonexamples falsely iden-
tified as trapezoids
Number of errors; examples falsely identi-
fied as nontrapezoids.

Drawing

Number of trials needed to present adequate
drawing (4 sides & parallel lines)

97

_J'1’.l- .f‘.

 

'9r
2..

“Ha-nae- ,

98

III. Concept Attainment Test

5.4
i

 

Trial 1 Trial 2 Trial 3
1 16 1 16___ 1 l6___
2— 17: 2: 17 2: 17___
3: 18 3_ 18: 3_ 18____
4 19: 4__ 19_ 4__ 19___
5: 20___ 5___ 20_ 5__ 20__ k
6___ 21_ 6_ 21_ 6__ 21___ ,.
7___ 22 7__ 22__ 7__ 22_____
8___ 23: 8 23__ 8 23___
9 24 9— 24 9‘— 24
10"— 25‘— 10— 25— 10— 25—
11“— 26— 11““ 26— 11— 26—
12—— 27"“ 12““ 27“- 12‘“— 27—
13: 28— 13: 28: 13: 28:
14__ 29"- 14 29__ 14 29__ -
15___ 3o: 15: 3o___ 15: 3o__
Trial 4 Trial 5
1 l6 1 16
2— 17— 2"" 17—
3— 18_ 3"— 18‘—
4— 19— 4— 19—
5"- 20: 5: 20““
6— 21 6 21_—
7"— 22"— 7_ 22—
8— 23— 8— 23—
9— 24_ 9— 24—
10‘— 25— 10— 25—
11: 26— 11_ 26‘—
12“" 27— 12"“ 27—
13_ 28— 13_ 28—
14— 29— 14: 29:
15: 3o: 15__ 30_

Number of trial to criterion

 

II.

III.

wmqmmvbWNH
o o o o o o o o

HHHl—‘H
obUNI-‘o
o

15.

DATA SHEET - Younger Child

Background Data

a. Name

School

 

d. Father's Occupation

 

b. Sex

 

 

c. Birthdate

e. MRT

 

 

f. Siblings, Birthdate and Sex

 

 

 

 

 

.l
i‘ I
V
”Quail .

 

 

 

 

 

Recording of lO-minute teaching session.

Concept Attainment Test.

Score Reason

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

99

Score Reason

16.

 

 

l7.

 

 

18.

 

 

l9.

 

20.

 

 

21.

 

 

22.

 

 

23.

 

 

24.

 

 

25.

 

 

26.

 

 

27.

 

 

28.

 

 

29.

 

30.

 

 

“111111111111