Racial identification and preference in young children as a function of race and sex of the experimenter and child

MISSING IMAGE

Material Information

Title:
Racial identification and preference in young children as a function of race and sex of the experimenter and child
Physical Description:
ix, 69 leaves. : ill. ; 28 cm.
Language:
English
Creator:
Kirn, Katrine Geha, 1948-
Publication Date:

Subjects

Subjects / Keywords:
African Americans   ( lcsh )
Race awareness   ( lcsh )
Genre:
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )

Notes

Thesis:
Thesis--University of Florida.
Bibliography:
Bibliography: leaves 60-63.
Statement of Responsibility:
by Katrine G Kirn.
General Note:
Typescript.
General Note:
Vita.

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 000577527
notis - ADA5222
oclc - 13987366
System ID:
AA00003954:00001

Full Text


















RACIAL IDENTIFICATION AND PREFERENCE
IN YOUNG CHILDREN AS A FUNCTION OF
RACE AND SEX OF THE EXPERIMENTER AND CHILD







By


KATRINE GEHA KIRN


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


OF


UNIVERSITY OF FLORIDA
1973















DEDICATION


This study is dedicated to the day when no one asks which doll

is a nice color.















ACKNOWLEDGMENTS


The author extends sincere thanks to the kindergarten teachers

and the principals of Glen Springs, Rawlings, and Prairie View

Elementary Schools in Gainesville, Florida. Also, special thanks go

to Doug Freeman, Vicki Johnson, Karen Maitland, and Keith Williams

for their hard work as experimenters. Every committee member is

thanked for his help and interest, especially Marvin Shaw who overcame

considerable obstacles to offer his assistance.

And finally, to Steve, who lived with this study as well as his own,

and yet was able to serve as a fine critic and a calm source of support.















TABLE OF CONTENTS



Page

ACKNOWLEDGMENTS-------------------------------------- iii

LIST OF TABLES--------------------------------------------.......v

LIST OF FIGURES-------------------------------------------vi

ABSTRACT--------------------------------------------------vii

I. INTRODUCTION AND HISTORY- -------------------------------1

Introduction---------------------------------------------------

Review of the Literature----------------------------------------2

Statement of the Problem and Hypotheses-----------------------20

II. METHOD------------------------------------------------------22

Subjects-------------------------------------------------------22

Experiments--------------------------------------------------22

Experimental Design------------------------------------------22

Procedure-----------------------------------------------------23

III. RESULTS------------------------------------------------------27

IV. DISCUSSION N---------------------------------------------------52

REFERENCES--------------------------------------------------60

BIOGRAPHICAL SKETCH---------------------------------------64













LIST OF TABLES


Page


1. Experimental Design -----------------------------------------24

2. Clark and Clark (1958) Data Compared with

Current Data ---------------------------------------------28

3. Results of Analysis of Variance on Initial

Preference Scores-----------------------------------------30

4. Results of Analysis of Variance on Preference

Scores, Condition I-----------------------------------------35

5. Results of Analysis of Variance on Preference

Scores, Condition II ---------------------------------------37

6. Choice of Pretty versus Not Pretty t-test Results ----------------39

7. Results of Analysis of Variance on Accuracy .-..----.-

Scores, Condition I --------------------------------------- 40

8. Results of Analysis of Variance on Accuracy

Scores, Condition II ---------------------------------------44

9. Results of Analysis of Variance on Self-

Identification Scores, Condition I ---------------------------47

10. Results of Analysis of Variance on Self-

Identification Scores, Condition II -------------------------- 49













LIST OF FIGURES


Page


1. Subject Race by Subject Sex by Experimenter

Sex, Initial Preference-----------------------------------31

2. Subject Race by Experimenter Race by

Experimenter Sex, Initial Preference----------------------32

3. Subject Sex by Experimenter Race by

Experimenter Sex, Initial Preference----------------------34

4. Subject Race by Subject Sex by Experimenter

Sex, Preference, Condition I------------------------------36

5. Subject Race by Experimenter Race, Accuracy,

Condition I ---------------------------------------------42

6. Subject Race by Subject Sex by Experimenter

Race by Experimenter Sex, Accuracy, Condition I-----------43

7. Subject Sex by Experimenter Race, Accuracy,

Condition II ---------------------------------------------45

8. Subject Race by Subject Sex, Self-Identification,

Condition I ---------------------------------------------48

9. Subject Race by Subject Sex, Self-Identification,

Condition II ---------------------------------------------51













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


RACIAL IDENTIFICATION AND PREFERENCE
IN YOUNG CHILDREN AS A FUNCTION OF
RACE AND SEX OF THE EXPERIMENTER AND CHILD


By

Katrine Geha Kirn

December, 1973



Chairman: Marvin E. Shaw, Ph. D.
Cochairman: Paul Satz, Ph. D.
Major Department: Psychology



A considerable amount of research concerning racial awareness,

identification, and preference in young children has been conducted

since 1939. The results of this research included the findings that

children were able to make racial discrimination as early as age three

and that a high degree of racial awareness was developed by age five.

It was also found that black children rejected black dolls and puppets and

photographs of blacks and attributed negative characteristics to blacks.

Black children identified themselves as white significantly more often

than they identified themselves as black, and they often became upset

when asked to make a racial self-identification. These results were








interpreted to mean that black children have a negative self-image.

The major purposes of the present study were to determine how recent

social changes have affected this self-image of black children and to de-

termine whether or not changes in the experimental situation (i. e. race

and sex of the experimenter) could affect experimental results.

Subjects were fifty-six black and fifty-six white children including

an equal number of males and females who had been enrolled in an inte-

grated kindergarten for almost a full academic year. A black female,

a white female, a black male, and a white male experimenter were used.

Each subject was presented with four identical dolls, two black with

black hair and two white with yellow hair. Subjects were first invited to

pick up any doll to play with. Then each subject was asked to give the

experimenter the doll he (she) would "like to play with, and the doll "that

is a nice doll, "that looks bad, "that is a nice color, "that looks like a

white child, "that looks like a colored child, "that looks like a Negro

child, "that looks like you. Subjects were then presented with a second

set of dolls identical to the first except that one black doll and one white

doll were dressed in clean, new smocks and one black and one white doll

were dressed in torn, dirty smocks to indicate a socioeconomic dif-

ference.

It was hypothesized that black subjects would prefer the black dolls

and identify themselves with the black dolls significantly more often

than in previous studies. It was also hypothesized that black subjects

would show a greater preference for the doll of their own race when the


viii








experiment er was of the same race. It was also expected that all sub-

jects would prefer the high status doll. No hypothesis could be made

regarding the effects of the interaction of race and sex of the experi-

menter and child.

The results supported the first two hypotheses. Several conclu-

sions were made on the basis of the results. A significant increase in

preference for black by black subjects was found in relation to a previous

study. Black subjects preferred black significantly more than white

subjects. White subjects identified themselves with the appropriately

colored doll significantly more often than black subjects. Racial pref-

erence, accuracy, and identification were found to vary in a complex way

as a function of the race and sex of the experimenter and child. It was

concluded that the results at least in part indicate the subject's response

to the experimental situation, since the results varied with changes in the

situation. Therefore, it would seem that conclusions regarding the self-

image of black children on the basis of such data must be made cautiously

and it should be kept in mind that generalizations on the basis of such

data may be tenuous.















Chapter I

INTRODUCTION AND HISTORY


Introduction



"In any society in which ethnic group membership carries some

connotation of inferiority or superiority, the social hopelessness of

those in a discredited minority group is transmitted to their children"

(Landreth, 1967, p. 291). An assumed inferiority of the Negro

certainly pre-dates the American society. In 1758, Linnaeus described

the African variety of man as being phlegmatic, indulgent, crafty, lazy,

and negligent (Borgatta and Lambert, 1968). Although knowledge of

racial characteristics has increased since the eighteenth century, some

beliefs about the inherited inferiority of the Negro still persist. Sev-

eral studies (Clark and Clark, 1939; Radke and Trager, 1950) have shown

that black children as young as four years of age seem to be aware of

the undesirable quality of their skin color. Black children preferred

white to black companions (Landreth and Johnson, 1953); they said a

white doll was prettier than the black one and they would prefer to be like

the white one (Radke and Trager, 1950).

However, in the years since the Supreme Court decisions making

integration compulsory in schools and other public places, much has









happened to the American Negro. Early civil rights sit-ins have pre-

sumably resulted in the development of the concepts of pride and

dignity in blackness. The notion of Black Power, "black is beautiful",

and the Afro-American movement have been aimed at improving both

the actual position and the self-image of the black. The most obvious

result of these efforts has been the acceptance of the preference to be

called black rather than Negro or colored. However, there is some ques-

tion as to whether or not the apparent changes in the self-attitudes of

blacks have yet been internalized. Have the self-images of black chil-

dren changed through the efforts to develop black pride? Do they no

longer reject the black doll, a behavior which has been interpreted as a

rejection of their own blackness?



Review of the Literature

The reality of color is an inescapable fact for the American Negro

in that it becomes an important part of the concept of the self. As

Seward pointed out, "color is inherent in the concept of 'self', as aware-

ness emerges in a race-conscious social context which assigns values to

the perception of color" (1956, p. 129). Allport (1954) and Proshansky and

Newton (1966) all considered the preschool and early elementary school

years (ages three to seven) as the crucial period for the development of

the child's feelings about himself and about those who are ethnically

different. During these years the child becomes aware of racial differences,







3

learns the labels, and learns the evaluative connotations of these labels.

Thus, the research on racial identification and awareness has largely

concentrated on these early years.

The awareness of racial differences has been found as early as age

three (Clark and Clark, 1958; Stevenson and Stewart, 1958; Morland,

1958; Stevenson and Stevenson, 1960). The degree of awareness increased

steadily until age six or seven; by this age, all children were able to make

accurate racial identifications (Clark and Clark, 1958; Morland, 1958).

While the ability to make racial discrimination has occurred as early as

age three, it was during the fourth year that a high awareness of race was

found (Goodman, 1952). Goodman (1952) found in a study of nursery

school children that a high degree of awareness of racial differences did

not occur before the age of four years, three months, while low aware-

ness did not occur in subjects older than four years, eleven months.

Clark and Clark (1958) also proposed that the time between four and five

years of age may be the crucial period in the development of racial

attitudes. Thus, it has been established that children are aware of racial

differences and can make racial discrimination by age five.

As Clark (1955) pointed out, the concept of racial awareness must

be considered conjointly with the concept of racial preference: "The

child... cannot learn what racial group he belongs to without being

involved in a larger pattern of emotions, conflicts, and desires which are

part of his growing knowledge of what society thinks about his race"

(p. 23). Thus, as the child learns racial labels, he also learns that







4
positive and negative values are attached to those labels. Research has

shown that black children apparently have learned the negative value of

their own racial identity and have assimilated this into their self-concepts.

Kenneth and Mamie Clark studied racial awareness and identifi-

cation in young children and the value attached to races through what

they termed the Dolls Test (Clark and Clark, 1958). Four dolls were

used, two black and two white with black and yellow hair respectively. The

dolls were clad only in diapers, so that there could be no sex identifica-

tion. Children were asked to give the experimenter the doll "that you

would like to play with, that is a nice doll, that looks bad, .

that is a nice color, that looks like a white child, that looks

like a colored child, that looks like a Negro child, .that looks

like you" in that order. Questions one through four were to indicate

racial preferences, questions five through seven racial awareness, and q

question eight self-identification. The preference questions were given

first, because it was found that once the child had identified himself with

a doll, he tended to show preference for that doll; the experimenters

postulated that in these cases they were not measuring racial preference,

but rather ego-involvement (Clark and Clark, 1958).

Black children of ages three through seven were tested with this

method; southern children were taken from segregated schools, northern

children from mixes schools. The results showed that ninety-four per-

cent of the children correctly identified the white doll, ninety-three per-

cent identified the colored doll, and seventy-two percent correctly chose







5

the Negro doll. Thus, it was concluded that the concept of race had been

formed. On the self-identification question, thirty-three percent of the

children identified themselves with the white doll, while sixty-six percent

of the children chose the black doll. The difference was significant and

the experimenters concluded that this indicated that some black children

wished to be white. The criticism has been made that some of the lighter

children were actually closer to the white doll in color; however, even

among the darkest children, only seventy-five percent identified them-

selves with the black doll. Thus, it is clear that among some children

there was a problem in racial self-identification. Furthermore, the chil-

dren were found to consistently reject the black doll and prefer the white

doll. There were more northern children who saw the black doll as bad

(seventy-one percent), while forty-nine percent of the southern children

saw the black doll as bad. Rejection of the child's own race was still the

norm. No testing was done with white children.

Although criticisms may be made of this study, the spontaneous com-

ments of the children in the testing situation left little doubt that the

research revealed some important facts about the self-image of black chil-

dren. In some cases, children ran out of the room crying when asked to

make the self-identification. There were numerous comments about black

being dirty and white being clean. Children tried to explain away their

blackness; one said he had burned himself that morning, another said he

had gotten suntanned on vacation.

In another study (Clark and Clark, 1950), a coloring test was used







6

to give children a wider range of responses. Children were given a sheet

of paper with drawings of a leaf, an apple, an orange, a mouse, a boy,

and a girl, plus a box of twenty-four colored crayons, including brown,

black, white, yellow, pink, and tan. Children were asked to color the

objects to det ermine whether or not they knew what color things really

were; children were then instructed to color the appropriate sex drawing

the color that they were and the opposite sex drawing the color that they

would like little girls or boys to be. Many of the children spent a long time

choosing the crayon for skin color, sometimes choosing one and then ex-

changing it for a lighter color. Some of the children who had been coloring

carefully, scribbled over the drawing representing themselves. A slight

majority of the children did not use brown or black, but rather used a

lighter color or some bizarre color such as red or green. This study also

indicated that black children were rejecting their skin color or feeling

some conflict over skin color.

Clark (1955) hypothesized a conflict between the need for self-esteem-

and the awareness that dark skin color is a basis for rejection by society.

Children assimilate prevailing social attitudes, see that they are identi-

fied with something which is rejected, and consequently reject themselves.

The result is a conflict about oneself quite early in development, as

early as three years. Older children were found to prefer light brown

over either black or white. Clark saw this as a compromise to resolve the

conflict of one's inevitable blackness. Clark continued to build his case

by stating that black psychiatric cases may have delusions of being white









or may deny their racial ancestry. Thus, Clark concluded that a self-

hatred develops which may result in aggression, delinquency, submissive-

ness, over-compensation in achievement, but always a general defeatist

attitude and lowering of aspiration.

Regardless of the theoretical explanations, some very specific con-

clusions can be made from the preceding studies. It was demonstrated

that children were able to correctly identify racial differences as early as

age three. In addition, black children rejected black and preferred white

as a skin color for dolls and for drawings. Finally black children some-

times became upset when asked to make racial self-identifications.

One may feel it is necessary to accept Clark's work with some caution

since he is black himself. However, other investigators also concluded

that there is a poor self-image among black children. Horowitz (1939)

studied racial preferences in a rural area using black children from

grades one through five. Children were given a group of photographs of

black, white, and Filipino boys and girls; they were asked to show the ex-

perimenter all those photographs of persons they would like to play with,

have for a cousin, sit next to at school, not be allowed to play with and so

on. Photographs of black children were consistently rejected and race

was found to be even more important than-sex in making choices; a white

boy would prefer a white girl as a playmate rather than play with a black

boy.

Johnson (1941) studied blacks between the ages of twelve and twenty,

living in the rural South. Subjects were given six racial categories and









were asked to use these categories to describe people in thirty value-

judgment situation. Johnson found that black was used disproportionately

for negative judgments. For instance, forty percent of the boys chose

black for the "ugliest girl you know, while eleven percent chose yellow,

and seven percent chose light brown. Forty-three percent of the boys and

twenty-three percent of the girls indicated black as the color of the "mean-

est boy (girl) you know. The experimenters reported spontaneous com-

ments by subjects such as "black is ugly, "black people are mean, "

"black people are evil. The conclusion was that the judgments made in

the experimental situations were based on stereotypes rather than actual

experience.

Radke and Trager (1950) and .Radke, Sutherland, and Rosenberg (1950)

found that black children showed a preference for white dolls and gave un-

desirable descriptions to photographs of blacks more frequently than did

white children, so that black children seemed to be attributing more nega-

tive value to themselves than did the white children. Goodman (1952)

found that nine percent of the black children in her sample expressed hos-

tility toward whites, while twenty-four percent expressed hostility toward

their own race; thirty-three percent of the white children expressed hostility

toward blacks, but none of the white children studied was hostile toward his

own race. Stevenson and Stewart (1958) found that black children

perceived other black children as aggressive, bad, and those "whom

other children fear" significantly more often than white children attributed

these characteristics to other white children. Both black and white










subjects in this study most often chose white children as "winners in a

game. Morland (1962) found that sixty percent of the black children

studied preferred to play with white children, while only ten percent of

the white children preferred to play with blacks.

Thus, it can be seen that other researchers in addition to the Clarks

have found that black children preferred white dolls and photographs and

rejected black dolls and photographs. Black children exhibited hostility

toward blacks and attributed negative characteristics to blacks, saying

they preferred to play with white children.

It was also found that children associate a socioeconomic class with

race. Radke and Trager (1950) asked five-to eight-year-old children to

choose a house for a white and a black doll, given a "good" house and a

"poor" house. Eighty-two percent of the white children and sixty-seven

percent of the black children gave the "poor" house to the black doll;, -

seventy-seven percent of the white children and sixty percent of the black

children gave the "good" house to the white doll. Thus, it would seem

that children of both races had assimilated and accepted the inferior-

social status of the black.

Landreth and Johnson (1953) found socioeconomic class to be a

major factor in children's racial attitudes. Five-year-old children were

invited to complete a picture of a person engaged in some activity by

adding one of three insets; the insets were identical except that the skin

color of the person was either white, brown, or black. Children of

professional parents showed no significant differences in choice of color.









Children of unskilled workers, both black and white, chose the white-

skinned inset significantly more often than the dark-skinned ones. The

experimenters concluded that value associated with color may depend

largely on experience; they determined that working class parents are

more likely to instill ethnic prejudice than are parents with more educa-

tion.

In summary, studies in the 1940's and 1950's have consistently

shown that black children rejected black images in the forms of dolls and

photographs. In addition, black children assigned themselves negative

personality attributes and associated black skin color with poor socio-

economic conditions. White children were also rejecting of black skin

color; however, some studies found white children less negative towards

blacks than the blacks themselves. One study found that children of

professional parents did not discriminate among skin colors, while both

black and white children of unskilled workers preferred white.

In the numerous attempts in the literature to explain the phenome-

non of negative self-images in black children, part of the problem has

been attributed to the experience of being poor and lower class. Sewell

(1961) found that healthy personality traits on the California Test of

Personality were more frequent in children of higher socioeconomic class.

However, Watson (1966) reported that children were unable to ascribe

class or class-related activities before the sixth grade; the inferior

status of the black could be detected by age four. A study of Amish chil-

dren found that they felt persecuted, inferior, and not as strong as other







11

children (Clark, 1955). Children of Italian-born parents demonstrated

a sense of inferiority and rejection, poor social adjustment, and emo-

tional instability (Clark, 1955). Lewin (1948) has reviewed the phenome-

non of self-hatred among Jews. It seemed that all minority children

suffered some damage -- a sense of inferiority, personal humiliation,

and confusion about self-worth (Clark, 1955). Of course, the situation

is especially acute for the black child, since he can never escape his badge

of inferiority.

In Lewin's (1948) theoretical system the black was the most extreme

case of self-hate resulting from belonging to a group. Belonging to a

socially underprivileged group may seem to be an impediment to reaching

some future goal; however, the individual is forced to remain a part of

that group by external forces. The individual may try to set himself apart

from the group; when he is unable to, self-hatred develops. The hatred

is a product of internalized values of the higher strata, including the low

esteem of the majority for the lower class group. Lewin interpreted the

problem to be pathological, but in persons of normal mental health he

accepted the problem as social psychological. Self-hate can be overcome

when the feelings of group inferiority and inequality are no longer a

reality. Lewin judged neurotic trends to stem from the lack of adjustment

to group inferiority.

Much of the self-hate and negative self-image among blacks appears

to have been learned from the family as well as from the outside white

world. A child builds his sense of self from responses made to him by







12

other people; he learns how others perceive him through their acts and

attitudes, and comes to perceive himself in the same way (White, 1964).

Many of the black families are disrupted; the mother or maternal grand-

mother is the dominant family figure. Children from broken homes

frequently feel unwanted, since they may well be a burden on the mother

and are often shuffled from place to place; these children develop an

insatiable need for reassurance (Kluckhohn and Murray, 1967). The black

home may often be fatherless because the father was unable to meet the

image of being a good provider and left out of desperation (Grossack, 1963).

Even if the father is physically present, he is likely to be unimportant,

since the mother or mother-substitute serves as the dominant family figure.

For male children, there is no strong male figure to serve for sex role

identification. Furthermore, mothers consistently prefer their daughters

in these homes, leaving their sons to feel inferior. To overcome this,

boys may identify with the maternal authority and subsequently feel some

sense of inadequacy in terms of masculinity. The girls, in an attempt

to live up to their mothers' high demands, may show extreme conformity

to the maternal expectations (Kluckhohn and Murray, 1967). In addition,

parents serve as primary models for in-group identification. Black chil-

dren cannot see their parents as capable and powerful protectors; in the

face of white society the parents appear intimidated and impotent

(Shane, 1960).

Being black iii a black family and subculture, the child finds few

black leaders with whom to identify; the white culture has obvious







13

advantages. The development of a secure self-system in blacks is ob-

structed by the acceptance of the white value system, and thus the indi-

vidual and the group measured worth by closeness to white features. The

lighter the skin, the straighter the hair, the narrower the nose and mouth,

the more self-esteem the black once developed (Kluckhohn and Murray,

1967). A child who was black and very "African" in appearance may have

been looked upon with some scorn by family, friends, society at large,

and even by the child himself. This may have changed in the light of

black pride. However, in addition to a poor self-image and a disrupted

family, the black child is likely to live under marginal social conditions,

with a lack of privacy, limited opportunities to explore the outside world,

the lack of an aesthetically pleasing environment, and so on (Goldstein,

1967).

Thus, it can be seen that the poor self-image in black children de-

veloped for a variety of reasons. One suggested cause was the experience

of being poor and of the lower socioeconomic class; this experience was

found to have generated feelings of inferiority among Amish, Italian-

Americans, and Jews as well as blacks. Lewin (1948) hypothesized that

the black perceived himself as a member of a socially underpriviled

group and tried to set himself apart from that group. When this was im-

possible, the black internalized the values of the privileged majority and

self-hate developed. Black families have often been disrupted, with no

strong male figure for sex role identification. Black children saw their

families, especially fathers, as impotent in the face of white society, so







14

that in-group identification and pride was hindered. Finally, the black

child saw whites and those closer to being whites as having more social

and economic advantages.

The black child may react to these factors in ways other than self-

hate or a poor self-image. For some, the result has been extreme

submissiveness, playing a passive and servile role. However, passive

acceptance of the role dictated by the white society is a poor adjustment.

A report by the Group for the Advancement of Psychiatry (1957) sug-

gested that the compliant behavior merely masks anger, fear, and re-

sentment. In studying Thematic Apperception Test responses, Mussen

(1953) found that white boys saw the world as a friendly place, while

black boys saw it as hostile and threatening. Palermo (1959) found black

children in the fourth to the sixth grades expressed more anxiety than

white children. Kardiner and Ovesey (1962) reported denial and dis-

tortion of reality evidenced in the Rorschach protocols of black adults.

Another reaction to the frustration of the situation has been an attempt

to escape through drugs or alcohol (Drake, 1965). The light-skinned

black sometimes completely escaped his group by "passing" as white

(Drake, 1965). Another means of dealing with rejection has been aggres-

sion against the out-group; Clark (1955) asserted that aggression through

crime has been the black's way of rejecting white society. This seems

to ignore the fact that crime is generally associated with poverty.

Nevertheless, delinquency and crime are not uncommon in black neigh-

borhoods.










The problems of the black child have been seen initially in the

schools. The child comes to school with poor verbal skills, the one

ability which is extremely important in our education system.

Goldstein (1967) reported that low-income blacks came to the first grade

well-dressed and eager; by the fourth grade they were exhibiting problems

in behavior and attitude. The child goes to school having experienced

only punishment as discipline and is unaccustomed to the middle class

system of social rewards. The child quickly learns that bad behavior brings

punishment and is his only way of getting the teacher's attention. Other chil-

dren get special privileges from the teacher, while the black child may

be rejected and denied privileges because of his poor behavior, his

clothes, his dialect, or perhaps his skin color. Thus, the child's self-

image worsens, he becomes aggressive and eventually wants to quit school

(Davis and Dollard, 1940). The teacher may expect the child to be a

poor student and this expentancy has been demonstrated to increase the

probability of failure (Janis et al., 1969). And so, a developing child

who already has a poor sense of self-worth is put into a school situation

which continually shows him that he is not as good as middle class white

children; the black child becomes caught in a whirlpool of failure, re-

jection, and feelings of inferiority.

And so, the reactions to being black and a member of an under-

privileged minority may include passive acceptance of an inferior role,

anxiety, and distortion of reality. There may be an attempt to escape

through alcohol, drugs, or by "passing. Aggression in the form of crime







16

may result. Such reactions may lead to a further worsening of the self-

image.

One of the assumptions underlying school desegregation was that

it would improve the position of the black child by providing better educa-

tion; it was also expected that interracial contact would improve race

relations. One might also assume that good relations with white students

would decrease inferiority feelings among black children. Williams and

Byars (1970) found trends toward increased self-esteem among black

eleventh-graders after integration. However, there has been no con-

sistent evidence to show that integration has improved racial attitudes;

simple interracial contact has apparently not brought about greater toler-

ance or acceptance in the schools (Carithers, 1970). Thus, while one

study may find increased self-esteem among black children through white

acceptance and positive experience with desegregation, another study

might just as easily find decreased self-esteem or no change as a result

of integration. Therefore, it is difficult to determine as yet just what

effect desegregation has had in relation to self-image among blacks.

Thus, the question remains of what effect the social changes of

recent years have had on the self-image of black children. Findings of

recent studies have been somewhat inconsistent and contradictory.

Butts (1963) reported that the misperception of skin color correlated

with low self-esteem; children high in self-esteem did not incorrectly

identify their skin color. In a replication of the Dolls Test, Gregor

and McPherson (1966) found black children rejecting the black doll.







17

Greenwald and Oppenheim (1968) added a Mulatto doll and found that more

black children were able to correctly identify themselves than in the

Clark studies, which used only black and white dolls. However, the ma-

jority of the children still rejected the black (and brown) dolls; moreover,

the experimenters were white and the Clark questions were changed, mak-

ing this an inexact replication. Using puppets and slightly different ques-

tions, Asher and Allen (1969) also found that black children preferred

white and rejected black. Crooks (1970) also replicated the Clark studies

and found essentially identical results; however, when children who had

been in an interracial pre-school were tested, black children showed a

greater preference for black than did children who had not been in any pre-

school. Blakely and Somerville (1970) questioned children in grades one

through twelve concerning their preference of racial names; most blacks

preferred Negro or Afro-American to black. Sciara (1971) asked black

males in the fourth grade to assign photographs of light, medium, and

dark-skinned individuals to low and high status jobs; results were highly

significant with light-skinned persons being assigned high status jobs and

dark-skinned persons low status ones.

Only two recent studies have shown a positive change in racial pref-

erence. Using a perfect replication of the Clark (1955) method, Hraba and

Grant (1970) found significant differences between their results and those

of the Clark study. At all ages tested, black children preferred the black

dolls and identified themselves with the black doll. The researchers

stated that the race of the experimenter was found to be unimportant, but







18

they did not specify how this was determined. Hraba and Grant viewed

their findings with some limitations, since the only testing was done in

Lincoln, Nebraska; they pointed out that the city may be somewhat unique.

Also, the city had just completed a long black pride campaign, suggesting

that such programs might have a positive effect for some time period.

The researchers did find that light children were just as likely as darker

children to choose the black doll. Similar findings were reported by

Harris and Braun (1971). Using puppets and slightly different questions

and order of questioning, they found that black children chose the black

puppets regardless of sex or socioeconomic class. The experimenter was

a black female and research was conducted in the Philadelphia, Penn-

sylvania area. Thus, on the basis of these two studies, it might be con-

cluded that a complete reversal of racial preference has occurred in black

children. However, variations in methods and geographical areas may

have had some effect.

To summarize the research of the 1960's and early 1970's, several

studies have found that black children prefer white dolls or puppets, while

two studies found that black children prefer black. In one study, more

children were able to correctly identify themselves by skin color when a

brown doll was added to the black and white dolls; however, on preference

questions the black children still preferred the white doll. Children who

had been in an integrated pre-school indicated a slightly higher preference

for black than children who had not been in any pre-school. Black chil-

dren associated photographs of dark-skinned people with low status jobs.







19

Variations in method, including stimulus materials, questions asked, sex

of the child, and race and sex of the experimenter, combined with incon-

sistent results across experiments make conclusions on the basis of this

more recent data somewhat difficult.

Recent research on the effects of race and sex of the experimenter

on experimental results has led to the conclusion that this can be an im-

portant factor. In a study involving imitation of television models,

Nicholas et al. (1971, a) found that white girls modeled adult males of both

sexes, while black girls imitated black females. Nicholas et al. (1971, b)

also found that black subjects imitated female models less than white sub-

jects; no race-sex interactions were found. Doll, Fagot, and Himbert

(1971) in a study of sex-role preference and black and white lower class

males, found a significant interaction between sex of the experimenter and

race of the child. Strickland (1972) reported that black children were

significantly more willing to delay receiving a reward from a black ex-

perimenter than they were with a white experimenter; white children pre-

ferred the delayed reward regardless of experimenter race. These -

studies clearly indicate that the race and sex of the experimenter can have

a significant effect on experimental situations. Although these studies do

not deal with racial identification and preference, the indication is that

some experimenter effect can be expected.

In summary, the research of the last thirty years has established

some facts in the area of racial identification and preference in young chil-

dren. Children as young as three years old were able to make racial










discrimination; by five years a high degree of racial awareness had

developed. Children had also learned the values attached to races, so

that black was associated with mean, ugly, poor houses, and low status

jobs. Black children rejected dolls with black skin, photographs of

black children, and black puppets and preferred white; only two recent

studies found black children preferring black. Black children often become

upset when asked to identify themselves in relation to race; a significant

number of black children identified themselves as white. Recent studies

in areas other than racial preference have found that race and sex of

the experimenter had a significant effect on experimental results; but this

variable was not examined systematically in relation to racial identifi-

cation and preference. In addition, recent research in the area has

varied the methodology used to study racial identification and preference,

so that it is difficult to make comparisons with earlier research.


Statement of the Problem and Hypotheses

The present study had two major purposes. The first purpose was

to further examine the self-image of black children in order to determine

whether social changes have affected this self-image. To this end, an

exact replication of the Clark and Clark (1958) study was conducted so

that a measure of change could be made. The second major purpose was

to determine how the experimental situation could affect the results, since

the results of previous studies have been taken to be an indication of self-

image. To examine experiment er effects, male and female, black and

white experimenters were used. To further study how changes in the







21

experimental situation could affect the results, the test questions were

asked a second time with the dolls dressed to indicate high and low socio-

economic class. An equal number of white children were tested so that

a comparison of black and white responses to the experimental situation

could be made.

The following hypotheses were offered as alternatives to the null

hypothesis:

1. Black subjects will prefer the black doll and

identify themselves with the black doll more

often than subjects in the Clark and Clark (1958)

study.

2. Black subjects will show a higher preference for

the black doll than will white subjects.

3. Subjects will show a higher preference for dolls of

their own race when the experiment er is of the

same race.

4. All subjects will prefer the high status doll which is

of the same race preferred during the first set of

questions.

No predictions could be made regarding the effects of sex of the experi-

menter or the interactions of race and sex of the experimenter and child.















CHAPTER II

METHOD


Subjects


Subjects were fifty-six white and fifty-six black children enrolled

in three Gainesville, Florida, elementary schools which have attempted

to maintain a racial balance. All subjects were in the kindergarten,

five years old, and had been in an integrated school environment for ap-

proximately one academic year.


Experimenters


Four experimenters were used -- a black female, a white female,

a black male, and a white male. All experimenters were first-year

graduate students, except for the black female who was a senior in the

undergraduate college. Each experimenter tested twenty-eight children

(seven black females, seven white females, seven black males, and

seven white males). Each experimenter was given written instructions

regarding the procedure; no information concerning experimental hy-

potheses was given.


Experimental design

A 2X2X2X2 factorial design was used, with seven subjects in each







23

cell (see Table 1). Scores measuring four dependent variables were as

follows:

(1) Initial Preference -- number of children picking

up the black doll during the Initial Preference

condition.

(2) Preference Score -- number of times the black doll

was chosen in response to questions one, two and

four plus the number of times the white doll was

chosen in response to question three.

(3) Accuracy Score -- number of correct responses to

questions five, six, and seven.

(4) Self-Identification -- number of subjects choosing

the correct doll for their race.

Data were always examined from the viewpoint of preference for

black except when accuracy was involved.


Procedure


Experiment ers were introduced to the children by the classroom

teacher, who explained that the children would be going with one of the ex-

perimenters one at a time to play a short game. The experimenter took

each child to a corner of the school library, where they sat at a table and

four dolls were laid out in front of the child. The dolls were identical

except for skin color; two were brown-skinned with black hair and two

were white-skinned with blond hair. The dolls were clad in a flannel















Table 1

Experimental Design


Black Subject White Subject
Male Female Male Female

Black Male 7 7 7 7
Experimenter Female 7 7 7 7

White Male 7 7 7 7
Experimenter Female 7 7 7 7










blanket during Condition I; no sex identification could be made. Dolls

were placed in front of the child in a randomly varied order, either

black-white-black-white or white-black-white-black. The child was told

that he could play with one of the dolls while the experimenter prepared

recording sheets and got ready to "play the game. As an attempt to

obtain a purely behavioral measure of preference, the experimenter re-

corded whether the child picked up a white doll, a black doll, or both;

this was termed initial preference.

The child was then asked the following questions, which are identi-

cal to those used by Clark and Clark (1958):

1. Give me the doll that you would like to play with.

2. Give me the doll that is a nice doll.

3. Give me the doll that looks bad.

4. Give me the doll that is a nice color.

5. Give me the doll that looks like a white child.

6. Give me the doll that looks like a colored child.

7. Give me the doll that looks like a Negro child.

8. Give me the doll that looks like you.

The dolls for Condition I were then removed and the dolls for Condition

II were laid out. These four dolls were identical to the first four except for

dress. One black and one white were clad in a clean new smock which were

trimmed with lace, while one black and one white dollwere clad in smocks

which were torn and smeared with dirt. Thus, a socioeconomic status of

the dolls was implied. Dolls in the new smocks were termed Pretty (P) and
I







26

dolls in the torn smocks were termed Not Pretty (NP). The procedure

using this second set of dolls was identical to the procedure of Condition I;

the same set of eight questions were asked.

Throughout the testing, spontaneous comments made by the children

were recorded in writing by the experimenters. Children were thanked

and returned to their classroom when testing was completed.















CHAPTER III

RESULTS


Two of the four hypotheses were partially supported by the

results. In addition, some new information was obtained.

Hypothesis 1: Black subjects will prefer the black doll and iden-

tify themselves with the black doll more often than in the Clark and

Clark (1958) study.

In order to compare the current results with the Clark and Clark

(1958) data, percentages were computed for the response to each question;

percentages for questions one through four were based on the number of

times the black doll was chosen and percentages for questions five through

eight were based on the number of accurate responses. The Test For

Significance of Difference Between Two Proportions (Bruning and Kintz,

1968) was computed for each question (see Table 2). The only significant

differences in the results of the two studies were in response to question

one, "Give me the doll you would like to play with, and question two,

"Give me the doll that is a nice doll.." The present study showed a sig-

nificant increase in the preference for black on these two questions. On

the remaining preference questions (three and four) there was an increase

in the preference for black, but the difference was not significant. There

was not a significant increase in self-identification with the black doll.















Table 2


Clark and Clark (1958) Data Compared with Current Data

Clark Current Current
Preference: Study Black White

1. like to play with 32% 54%* 29%**

2. a nice doll 38% 48%* 35%**

3. ---a bad doll 59% 57% 53%

4. --- a nice color 38% 43% 34%


Accuracy:

5. --- a white child

6. --- a colored child

7. --- a Negro child

Self-Identification:

8. --- you


94%

93%

72%


89%

91%

66%


89%

89%

72%


65% 69% 91%

*Significantly Different from Clark
**Significantly Different from Current Black









Clark and Clark used only black subjects; percentages for the current

white subjects are given in Table 2 for the purpose of comparison.

Whites identified with the appropriate color doll significantly more often

than black subjects.

Hypothesis 2: Black subjects will show a higher preference for the

black doll than will white subjects.

Results of Analysis of Variance on the Initial Preference scores are

shown in Table 3. Subject race was the only main effect found to be sig-

nificant, with black subjects choosing the black doll more often than white

subjects (Black mean = 2. 07, White mean = 1. 41). Three second order

interactions were significant. The interaction of Subject Race by Sub-

ject Sex by Experimenter Sex (Figure 1) showed that Black subjects did

not differ significantly with a Male or a Female Experimenter. White sub-

jects, however, showed a greater preference for the black doll when the

experimenter was of the opposite sex.

The significance of the Subject Race by Experimenter Race by Ex-

perimenter Sex interaction revealed that the response evoked by a White

Female experimenter and a Black Male Experimenter was essentially the

same; this was also true for the response evoked by the Black Female

experimenter and the White Male experimenter (see Figure 2). All sub-

jects showed the same preference for the black doll with a Black Female

or White Male experimenter; Black subjects, however showed consider-

ably greater preference than White subjects with the White Female and

Black Male experimenter. Subject Sex by Experimenter Race by












Table 3


Results of Analysis of Variance

S Race

S Sex

E Race

E Sex

S Race by S Sex

S Race by ERace

S Sex by E Race

S Race by E Sex

S Sex by E Sex

E Race by E Sex

S Sex by S Race by

ERace

S Race by S Sex by

E Sex

S Race by E Race by

E Sex

S Sex by E Race by

E Sex

S Sex by S Race by

E Race by E Sex


PC


.05,
.01,


on Initial Preference Scores
F
8.15**

1.72

0.01

0.15

0.05

0.15

0.48

0.15

3.72

0.01




1. 34




5.01*




5.01*




5. 72*




1.34


F> 3. 94
F 6.90






























2. 0


Mean Number
of Times
Black Chosen


- a ~


Black Female S


White Male S
Black Male S




% White Female S


Figure 1 Subject Race by Subject
Sex by Experimenter Sex,
Initial Preference





























2.0.


Mean Number
of Times
Black Chosen


1.0.


S. Black Female E
-- White Male E


White
Black


Black S


Female E
Male E


White S


Figure 2 Subject Race by Experimenter
Race by Experimenter Sex,
Initial Preference










Experiment er Sex interaction also shows that subjects responded dif-

ferently to different experimenters (see Figure 3). When the experi-

menter was Black, responses of Male and Female subjects were

essentially the same regardless of experimenter sex. However, when the

experimenter was White, subjects showed a higher preference for black

when the experimenter was of the opposite sex.

Results of the analysis on Preference scores for Condition I are

reported in Table 4. The main effect of subject race was again signifi-

cant with Black subjects preferring the black doll more often than White

subjects (Black mean = 1. 89, White mean = 1. 32). The interaction of

Subject Race by Subject Sex by Experimenter Sex was significant at the

.05 level (see Figure 4). Black Female subjects were most likely to pre-

fer the black doll, while White Male subjects showed a greater preference

for black when the experimenter was Male. Black Male subjects were

especially affected by experimenter sex, showing almost as high a

preference for the black doll as Black Female subjects when the experi-

menter was Male, but decreasing in preference with a Female experi-

menter so that only White Male subjects had lower preference scores.

No significant differences were found among preference scores for Con-

dition II; results are reported in Table 5.

Hypothesis 3: Subjects will show a higher preference for dolls of

their own race when the experimenter is of the same race.

This hypothesis was not supported by the data, which are reported

above. The interactions of race and sex were found to have more effect




























2.0


Mean Number
of Times
Black Chosen


1.0


Male E
Female E
Male E


Female E


Male S Female S


Figure 3 Subject Sex by Experimenter
Race by Experimenter Sex,
Initial Preference







35






Table 4

Results of Analysis of Variance on Preference Scores,

Condition I

F
S Race 4.94*

S Sex 1.24

E Race 0.02

E Sex 2.78

S Race by S Sex 0.69

S Race by E Sex 0.48

S Race by E Race 1.56

S Race by E Sex 0.31

S Sex by E Sex 0.31

E Race by E Sex 3.26

S Race by S Sex by E Race 0.02

S Race by S Sex by E Sex 4.94*

S Race by E Race by E Sex 0.94

S Sex by E Race by E Sex 0.48

S Race by S Sex by E Race

by E Sex 0. 17


*p < .05, F > 3.94
#p < .01, F > 6.90





























2.0.


Mean Number
of times
Black Chosen


1. 0


Male E


< Black Female S


White Female S

Black Male S
White Male S


Female E
Female E


Figure 4 Subject Race by Subject Sex
by Experimenter Sex,
Preference, Condition I















Table 5


Results of Analysis of Variance on Preference Scores,

Condition II

F
S Race 0.28

S Sex 2.49

E Race 0. 28

E Sex 1.73

S Race by S Sex 1. 11

S Race by E Race 1. 11

S Sex by E Race 0.28

S Sex by E Sex 0.07

S Race by E Sex 3.39

E Race by E Sex 1.73

S Race by S Sex by E Race 1.11

S Race by S Sex by ESex 0.62

S Race by E Race by E Sex 0.69

S Sex by E Race by E Sex 0.62

S Race by S Sex by E Race by E Sex 1.73


p < .05, F > 3.94
p-< .01, F> 6.90









than race alone. In fact, the highest performance scores were found

when the experimenter and subject were of different races. For Black

subjects, the highest preference for black was found with a White Female

experimenter; for White subjects the highest preference for black was

found with the Black Female experimenter. While the highest preference

was found with a Female experimenter of the opposite race as the subject,

the lowest preference was found with a Male experimenter of the opposite

race.

Hypothesis 4: All subjects will prefer the high status doll which is

of the same race preferred during the first set of questions.

This hypothesis was not supported. In examining the results of Con-

dition II, t-tests were performed on the data for each subject group to

determine whether subjects chose Pretty (P) or Not Pretty (NP) dolls more

often. None of these t-tests was significant except for Black Female sub-

jects (see Table 6). Black Female subjects preferred the Pretty doll at

the 05 level of significance. Therefore, it was decided that P and NP

dolls did not constitute an important difference and P and NP dolls of each

color were considered together.

No hypotheses were formulated regarding the effect of sex and the

interactions of race and sex of the experimenter and child. It was found

that both race and sex interacted to form an important factor. Effects on

preference are cited above. The effects were seen especially in the

analysis of accuracy. Results of the analysis using accuracy scores for

Condition I are reported in Table 7. No main effect was significant















Table 6

Choice of Pretty versus Not Pretty t-test Results

White Male subjects t = 0. 489

White Female subjects t = 0. 740

Black Male subjects t = 0. 023

Black Female subjects t = 2. 504*



*p. 05, df=54














Table 7


Results of Analysis of Variance on Accuracy Scores,

Condition I
F
S Race 0.89

S Sex 0.89

E Race 0.89

E Sex 3.20

S Race by S Sex 0.36

S Race by E Race 5.69*

S Sex by E Race 3.20

S Sex by E Sex 0.09

S Race by E Sex 0.09

E Race by E Sex 0.09

S Race by S Sex by E Race 0.80

S Race by S Sex by E Sex .1.42

S Race by E Race by E Sex 0.36

S Sex by E Race by E Sex 0.*36

S Race by S Sex by E Race by E Sex 7.20**


*p < .05, F > 3.94
**p< .01, F 6.90









The interaction of Subject Race by Experimenter Race was significant

at the 05 level (see Figure 5). Subjects were more accurate when the

experimenter was of the same race. The interaction of Subject Race

by Subject Sex by Experimenter Race by Experimenter Sex was signifi-

cant at the 01 level (see Figure 6). With White experimenters, the

subject groups which were most and least accurate did not vary with the

experimenter sex. With White experiments, White Female subjects were

most accurate, followed by Black Females, White Males, and finally

Black Males. With a Black Male experimenter, White Male subjects were

least accurate. With a Black Female experimenter, Black Males were

most accurate and White Female subjects were least accurate. Overall,

the least accurate group was the Black Male subjects with the White

Female experimenter. The most accurate groups were White Male sub-

jects with a Black Male experimenter, White Female subjects with a

White Male experimenter, and Black Male subjects with a Black Female

experimenter.

The analysis of the accuracy scores for Condition II are reported

in Table 8. Two main effects were significant. The effect of subject

race was significant at the 01 level with White subjects more accurate

than Black subjects (White mean = 0. 91, Black mean = 0. 696. The effect

of subject sex was significant at the 05 level with Female subjects more

accurate than Male subjects (Female mean = 0. 88, Male mean = 0. 73).

In addition, the interaction of Subject Sex by Experimenter Sex was signi-

ficant at the 05 level (See Figure 7). Female subjects were equally

























2.0

Mean number of
Correct
Responses

1.0


White S

>< Black S


Black E White E


Figure 5 Subject Race by Experimenter
Race,
Accuracy, Condition I





















2. 0


Mean Number of
Correct
Responses


1.0


- -


Black Male S

Black Female S
White Female S
White Male S


Black
Male E


2. .


Mean Number of
Correct
Responses


1.0


Black
Female E


White Female
Black Female
White Male

Black Male


White
Male E


White
Female E


Subject Race by Subject Sex
by Experimenter Race by
Experimenter Sex,
Accuracy, Condition I


Figure 6


I


"q" -














Table 8


Results of Analysis of Variance on Accuracy Scores,

Condition II
F
S Race 9. 00**

S Sex 4.00*

E Race 2.25

E Sex 1.00

S Race by S Sex 2.25

S Race by E Race 0.00

S Sex by E Race 4. 00*

S Sex by E Sex 2.25

S Race by ESex 0.25

E Race by E_ Sex 0.00

S Race by_S Sex by E Race 0.25

S Race by S Sex by E Sex 1.00

S Race by E Race by E Sex 0.25

S Sex by E Race by E Sex 0.25

S Race by S Sex by E Race by E Sex 1.00


*p < .05, F> 3.94
**p< .01, F 7 6.90





























2.0.

Mean Number of
Correct
Responses
1.0.


Female S
Male S


Black E White E


Figure 7 Subject Sex by Experimenter
Race,
Accuracy, Condition II










accurate regardless of the race of the experimenter. However, Male

subjects were less accurate with Black experimenters than with White

experimenters.

The effects of race and sex were also seen in the analysis of the

self-identification question. The results of the analysis of Self-

Identification in Condition I are reported in Table 9. Two main effects

were significant at the 01 level. White Males were more accurate on

Self-Identification than Black Males (White mean = 0. 89, Black mean =

0. 50). In addition, Female subjects were more accurate in Self-Identifi-

cation than Male subjects (Female mean = 0. 82, Male mean = 0. 57). The

interaction of these two variables, Subject Race by Subject Sex, was also

significant at the .01 level (see Figure 8). White subjects were equally

accurate whether Male or Female; Black Female subjects were more ac-

curate than Black Male subjects.

The analysis of Self-identification for Condition II is reported in

Table 10. Three main effects were significant at the 01 level. Subject

race had a significant effect, with White subjects more accurate in Self-

Identification than Black subjects (White mean = 1.91, Black mean = 1. 50).

Subject sex also had a significant effect, with Female subjects more ac-

curate than Male subjects (Female mean = 1. 82, Male mean = 1. 59).

Finally, the effect of sex of the experimenter also had a significant effect,

with more accurate Self-Identification being made with a Male experimen-

ter than with a Female experimenter (Male mean = 1. 80, Female mean =

1. 61). The interaction of Subject Race by Subject Sex was also significant














Table 9

Results of Analysis of Variance on Self-Identification Scores,

Condition I


Race

Sex

Race

Sex

Race by S Sex

Race by E Race

Sex by E Race

Race by E Sex

Sex by E Sex

Race by E Sex

Race by S Sex by E Race

Race by S Sex by E Sex

Race by E Race by E Sex

Sex by E Race by E Sex

Race by S Sex by E Race by E Sex


*p .05,
**p< .01,


F
27.92**

11.31*

0. 23

0.92

11.31*

0. 23

0.23

0.92

0.00

0.92

0. 23

0.92

0.92

0.92

0.00


F> 3.94
F> 6.90





























2.0-

Mean Number of
Correct
Responses
1.0 Black S

White S





Male S Female S


Figure 8 Subject Race by Subject Sex,
Self-Identification, Condition I







49










Table 10

Results of Analysis of Variance on Self-Identification Scores.

Condition II

F
S Race 35.36**

S Sex 11.27**

E Race 1.66

E Sex 8.07**

S Race by S Sex 15.00**

S Race by E Race 1.67

S Sex by E Race 0.07

S Sex by E Sex 1.67

S Race by E Sex 0.60

E Race by E Sex 0.60

S Race byS Sex by E Race 0.60

S Race by S Sex by E Sex 0.07

S Race by S Sex by E Sex 0.60

S Sex by E Race by E Sex 0.60

S Race by S Sex by E Race by E Sex 0.07


*p < .05, F 7 3.94
**p< .01, F7 6.90









(pee Figure 9). White subjects were not affected by sex of the experi-

menter, while Black subjects made more accurate Self-Identification

responses with the Female experimenter than with the Male experimen-

ter. These results are consistent with those of Condition I.

A chi-square was performed on the responses to "Give me the

doll that is a nice doll" and "Give me the doll that looks like you" to see

if children preferred and identified with the doll of the same color. It

had been assumed that a child who prefers one color and identifies with

a different color doll is revealing a poor self-image based on skin color.

No relationship was found between the responses to these questions (chi-

square = 5. 984, df = 9). Therefore, it can be concluded that the re-

sponses to these two questions should not be used as a measure of a

child's self-image.

Finally, none of the spontaneous comments made by the children

had to do with their color or the color of the dolls. In fact, the children

said very little, but they were extremely cooperative in "playing the

game". No children cried, left the situation, or refused to participate.




























2.0.
Mean Number of
Correct
Responses

1.0 White S
Black S





Male S Female S


Figure 9 Subject Race by Subject Sex,
Self-Identification, Condition II







52







CHAPTER IV

DISCUSSION


There are some very definite conclusions which can be made

from the present study.

1. There has been a significant increase in preference for

black by black children over the Clark and Clark (1958)

data.

2. Black subjects preferred black significantly more than

White subjects.

3. There was no significant difference in response to the

self-identification question in comparison with the Clark

and Clark data, while White subjects were significantly

more likely to identify themselves with the doll of their

color.

4. Racial preference, accuracy, and identification were found

to vary in an unpredicted and complicated way as a func-

tion of the interaction of race and sex of the experimenters

and subjects.

5. The current data (and previous data) indicate a response

to the experimental situation, in that changes in the situation

resulted in different responses.






53

In a direct comparison with the Clark and Clark (1958) data it is

interesting to note that there has been a significant increase in the pre-

ference for black on two of the four preference questions. This could

lead to the conclusion that the social changes of the past few years have

in fact improved the way blacks see themselves. There was no signifi-

cant difference in self-identification, however, so that it would seem

that the black child is still rejecting his own color. When comparing

the data of black and white subjects of the current study in the same

fashion, it can be seen that whites and blacks differ significantly on the

same two preference questions as those which differ from the Clark data.

It is clear that Black subjects did show a greater preference for black,

further supporting the conclusion that black self-image may have im-

proved. The White subjects were significantly more likely to identify

themselves with the doll of their own color, which may also indicate

that response of Black subjects in the self-identification question does in-

dicate self-rejection. All of the three groups were equally accurate in

racial identification. Although this data supports some hypotheses, it

should be considered with some caution, given the major results of this

study. It can be accepted that Black subjects showed an increased pre-

ference for black over the Clark data with no increase in self-identification

with black; however, it must be kept in mind that variables in the situation

did significantly affect the responses of the subjects.

Given the choice of picking up any doll (Initial Preference), Black

subjects did choose the black doll significantly more often than did White







54

subjects. The three interactions which were significant are complex and

difficult to interpret. It does appear that Black subjects were not affected

by experimenter sex, while White Male subjects showed a greater pre-

ference for black with Female experimenters, and the same is true for

White Female subjects with Male experimenters. This result may be re-

lated to the fact that blacks often only have a maternal figure in the home

and are not socialized in the same way to sex roles as are whites. Un-

fortunately, in the Subject Race by Experimenter Race by Experimenter

Sex interaction, Black and White subjects respond in the same way to Black

Female and White Female experimenters; Black subjects, however, show

a high preference for Black with the White Female experimenter and

Black Male experimenter, while White subjects show a low preference

with the same experimenters. Regardless of whatever dynamic explanation

may be attempted, it is clear that something as simple as picking up a doll

can be affected by the interaction of race and sex of the experimenter

and child.

On the preference questions in Condition I, Blacks again showed a

higher preference for black than did Whites, as was hypothesized. Also,

it is no surprise that White Male subjects least preferred the black doll

and Black Female subjects showed the greatest preference for the black

doll. Black girls are simply the most likely to play with black dolls. All

subjects preferred black more with a Male experimenter in this condition.

Black Male subjects in particular showed a radical decrease in preference

for black with a Female experimenter. It could be that the black boy's









environment has such a dearth of male role models, that the result was

an extremely positive response to Male experimenters. By simply

adding another variable to the situation in Condition II (i. e., dress of the

dolls), the significant effects of Condition I preferences were wiped out.

Thus, it can be concluded that the response of racial preference in this

situation is sufficiently delicate as to be affected by external variables.

Therefore, although it cannot be denied that there are differences in the

preference scores, one must be very cautious in making conclusions from

this study and similar racial preference data. It appears that the child's

response to the immediate situation is being measured in addition to

racial preference.

This is true to such an extent that even accuracy was affected by the

situation. No main effects had a significant effect on accuracy in Con-

dition I. This is to say that White subjects were not more accurate than

Black subjects, nor Males than Females. However, the interaction of

subject race with experimenter race did affect accuracy. Subjects of each

race were more accurate with an experimenter of their own race. This

may be attributed to the effect of modeling or social desirability. It is

important to remember also that children were less accurate with an

experimenter of the opposite race. There can be little question that re-

sponse to the experimenter was as important as racial awareness and

knowledge of racial labels.

Furthermore, the interaction of subject race and sex with experi-

menter race and sex was also significant. When the experimenter was






56

White, subjects were more accurate with Male experimenters and less

accurate with Female experimenters; when the experimenter was Black,

the results became more complicated. White Male subjects were more

accurate with the Black Male experimenter than with the Black Female

experimenter; the exact opposite was true for Black Male subjects. Black

Female subjects decreased in accuracy with a Black Female experimenter,

while White Female subjects increased in accuracy with the Black Female

experimenter (in relation to performance with the Black Male experimen-

ter). These results are not consistent with the preference results. How-

ever, the complexities of the interaction reinforce the hypothesis that much

of what is being measured is the response to an external situation rather

than a degree of knowledge.

In Condition II, where another external variable was added, the ques-

tion of accuracy becomes much more clear cut. White subjects were

significantly more accurate than Black subjects; Female subjects were

more accurate than Male subjects. Female subjects were equally accu-

rate regardless of experimenter race, while Male subjects were con-

siderably less accurate with a Black experimenter than with a White

experimenter. These results could be explained by saying that females

are traditionally more school oriented and therefore might be more

likely to do well on a test of accuracy. However, more important than

such an explanation is the added evidence that the experimental situation

can affect the subjects' accuracy responses.

On the self-identification question, White subjects identified


_ 1___1_1___(






57

themselves with the correct color significantly more often than did

Black subjects; this was also true for Female subjects in relation to

Males. Furthermore, White subjects were equally accurate whether

Male or Female. Black Male subjects, however, were far less

accurate in self-identification than all other subjects. Black Female

subjects were almost as accurate as White subjects on this question.

The identical interaction was found in Condition II. Thus, it would seem

that only Black Male subjects are now rejecting their race. Perhaps

this can be explained by the opportunities of the Female subjects to

positively identify with maternal figures and perhaps black female teach-

ers. An additional significant finding in Condition II was that more ac-

curate Self-Identification responses were made with a Male experimenter

than with a. Female experimenter. Once again, the results are very

complex and it would seem that the concept of self-rejection is tenuous

to some degree.

Since no relationship was found between responses to the request for

a "nice doll" and a "doll that looks like you, it may be concluded that

the whole concept of self-rejection on the basis of the Dolls Test has

been overinterpreted. In fact, it is questionable whether or not the

internal states of racial preference and identification can be measured by

the choice of a doll. The Dolls Test can be used with confidence as a

behavior showing preference for dolls of different colors; as behavior,

these preferences can be modified by changes in the experimental situa-

tion. Generalizations of racial preferences and identification may be made,









but this should be done with qualifications.

Before further generalizations are made from the present study,

some further research would be useful. The effect of varying experimen-

ters was obviously powerful. There could be some question as to whether

the personalities of the experimenters had any effect in addition to the

effects of race and sex. Given the complexities of the race and sex inter-

actions, this would seem unlikely. Nevertheless, ideally the research

should now be replicated with several experimenters of each race and sex.

A further investigation of the effects of varying the dolls could be done.

One possibility is a replication with half of the subjects experiencing Con-

dition II before Condition I to check for any order effect. Also, other

modifications of the dolls' dress and color and also sex might be made.

Finally, a study correlating the various methods used to measure racial

preference among children should be used with the same children; thus,

some firmer conclusions as to what is being measured might be possible.

All of the above does not say that previous studies have provided

worthless information. Rejection by blacks of their identity and position

has been a reality; it would seem that the situation is now somewhat al-

leviated. However, it seems that excessive interpretations have been made

on the basis of such data in the past and insufficient consideration has been

given to the effects of the experimental situation. It has become very clear

that the children were reacting very differently in the situation depending

on who the experimenter was. This leads to the consideration of some

implications for schools. Further information is needed on how the









factors of race and sex affect the children's responses to the teacher

and the school situation, where the teacher is often a white female, How-

ever, considerably more evidence would be needed before scwh a ques-

tion can be seriously considered














REFERENCES

Allport, Gordon. The Nature of Prejudice. Doubleday and Co.,
Garden City, New York, 1954.

Asher, S. R. and V. L. Allen. "Racial preference and social com-
pa prison processes. J. Social Issues, 1969, 25, 157-166.

Blakely, Karen and Addison Somerville. "An investigation of the
preference for racial identification terms among Negro and
Caucasian children. Journal of Negro Education, 1970 (Fall)
39 (4), 314-319.

Borgatta, Edgar and William Lambert. Handbook of Personality:
Theory and Research.. Rand McNally: Chicago, 1968.

Bruning, James and B. L. Kintz. Computational Handbook of
Statistics, Scott, Foresman, and Company: Glenview, Illinois,
1968.

Butts, H. P. "Skin color perception and self-esteem. Journal
of Negro Education, 1963, 32, 122-128.

Carithers, Martha. "School desegregation and racial cleavage:
A review of the literature. J. Social Issues 1970, 26 (4),
25-47.

Clark, Kenneth. Prejudice and Your Child. Beacon Press:
Boston, 1955.

Clark, Kenneth and Mamie Clark. "Emotional factors in racial
identification and preference in Negro preschool children. "
Journal of Negro Education, 1939, 19, 341-350.

Clark, Kenneth and Mamie Clark, "Emotional factors in racial
identification and preference in Negro children. Journal
of Negro Education, 1950, 19, 341-350.

Clark, Kenneth and Mamie Clark. "Racial identification and
preference in Negro children. (Published in E. E. Macoby,
T. M. Newcomb, and E. L. Harley, eds. Readings in Social
Psychology. Henry Holt and Co.: New York, 1958. )








Crooks, Roland. "The effects of an interracial preschool program
upon racial preference, knowledge of racial differences, and
racial identification. J. Social Issues, 1970, 26, 4, 137-144.

Davis, Allison and John Dollard. Children of Bondage: The
Personality Development of Negro Youth in the Urban South.
American Council on Education: Washington, D. C., 1940.

Doll, Paddy A., Hacker Fagot, and Joanne Himbert. "Experimenter
effect on sex-role preference among black and white lower-
class male children. Psychological Reports, 1971 (Dec.),
29 (3, Pt. 3), 1295-1301.

Drake, St. C. "The small and economic status of the Negro in the
United States. Daedalus, 1965, 94, 771-814.

Goldstein, Bernard. Low Income Youth in Urban Areas: A
Critical Review of the Literature. Holt, Rinehart, and
Winston, Inc.: New York, 1967.

Goodman, M. E. Race Awareness in Young Children. Collier
Books: New York, 1952.

Greenwald, H. J. and D. B. Oppenheim. "Reported magnitude of
self-misidentification among Negro children -- artifact? "
Journal of Personality and Social Psychology. 1968, 8, 49-52.

Gregor, A. J. and D. Angus McPherson. "Racial attitudes among
white and Negro children in a deep-south standard metro-
politan area. Journal of Social Psychology, 1966, 68, 95-100.

Grossack, Martin (ed). Mental Health and Segregation. Springer
Publishing Co., Inc.: New York, 1963.

Group for the Advancement of Psychiatry. Psychiatric Aspects of
School Desegregation. New York: Group for the Advancement
of Psychiatry, 1957.

Harris, Susan and John Braun. "Self-esteem.and racial preference
in black children. Proceedings American Psychological Associa-
tion Convention, 1971, 6 (Part 1, 259-360.

Harowitz, R. "Racial aspects of self-identification in nursery school
children." Journal of Psychology, 1939, 7, 91-99.

Hraba, J. and G. Grant. "Black is beautiful: a reexamination of
racial preference and identification. Journal of Personality
and Social Psychology, 1970, 16, 398-401.








Janis, I., G. Mahl, J. Kagan, and R. Holt. Personality:
Dynamics, Development, and Assessment. Harcourt,
Brace, and World, Inc.: New York, 1969.

Johnson, C. S. Growing Up in the Black Belt. Washington, D. C.:
American Council on Education, 1941.

Kardiner, A. and L. Ovesey. The Mark of Oppression.
Cleveland, Ohio: World, 1962.

Kluckhohn, C. and H. A. Murray. Personality in Nature, Society,
and Culture. Alfred A. Knopf: New York, 1967.

Landreth, Catherine. Early Childhood: Behavior and Learning.
Alfred A. Knopf: New York, 1967.

Landreth, Catherine, and B.C. Johnson. "Young children's
responses to a picture inset test designed to reveal reactions
to persons of different skin color. Child Development,
1953, 24, 63-79.

Lewin, Kurt. Resolving Social Conflicts. Harper and Row:
New York, 1948.

Morland, J.K. "Racial recognition by nursery school children in
Lynchburg, Virginia." Social Forces, 1958, 37, 132-137.

Morland, J.K. "Racial acceptance and preference of nursery
school children in a southern city. Merrill Palmer
Quarterly, 1962, 8, 271-380. _

Mussen, Paul. "Differences between the TAT responses of Negro
white boys. Journal of Consulting Psychology, 1953, 17,
373-376.

Nicholas, Karen B., R. E. McCarter, and R. V. Heckel. "The
effects of race and sex on the imitation of television models."
Journal of Social Psychology, 1971 a, (Dec. ), 85 (2), 315-316.

Nicholas, Karen B., R. E. McCarter, and R. V. Heckel. "Imi-
tation of adult and peer television models by white and Negro
children. Journal of Social Psychology, 1971 b, (Dec.), 85
(2), 317-318.

Palermo, D. S. "Racial comparisons and additive normative data on
the children's Manifest Anxiety Scale. Child Development,
1959, 30, 53-57.








Proshansky, Harold and Peggy Newton. "The nature and meaning of
Negro self-identity. (Published in Martin Deutsch, Irwin Katz,
and Arthur Jensen eds. Social Class, Race, and Psycholo -
gical Development. Holt, Rinehart, and Winston, Inc.:
New York, 1968.)

Radke, M. and H. G. Trager. "Children's perceptions of the social
roles of Negros and whites. Journal of Psychology, 1950,
29, 3-23.

Radke, J. Sutherland, and P. Rosenberg. "Racial attitudes of
children. Sociometry, 1950, 13, 254-171.

Sciara, Frank J. "Perceptions of Negro boys regarding color and
occupational status. Child Study Journal. 1971 (Sum), 1 (4),
203-211.

Seward, G. Psychotherapy and Culture Conflict. New York: Roland
Press, 1956.

Sewell, W. H. "Social class and childhood personality. Sociometry,
1961, 24, 350-356.

Shane, Norton. "Some subcultural considerations in the psychotherapy
of a Negro patient. The Psychiatric Quarterly, (Jan. ), 1960,
34, 1-19.

Stevenson, H. W. and N. G. Stevenson. "Social interaction in an
interracial nursery school. Genetic Psychology Monographs,
1960, 61, 37-75.

Stevenson, H. W. and E. C. Stewart. "A developmental study of
race awareness in young children. Child Development, 1958
29, 399-410.

Strickland, Bonnie P. "Delay of gratification as a function of race
of the experimenter. Journal of Personality and Social
Psychology, 1972 (Apr), 22 (1), 108-112.

Watson, G. Social Psychology: Issues and Insights. J. B. Lippincott
Co.: New York, 1966.

White, R. W. The Abnormal Personality. The Ronald Press Co.:
New York, 1964.

Williams, Robert and Harry Byars. "The effect of academic
integration in the self-esteem of southern Negro students."
Journal of Social Psychology, 1970, 80, 183-188.













BIOGRAPHICAL SKETCH


Katrine Geha Kirn was born on August 21, 1948, in Detroit,

Michigan. She later lived in Johnstown, Pennsylvania, where she

graducated from Johnstown Central High School in 1966. She received

the degree of Bachelor of Arts from Chatham College in Pittsburgh,

Pennsylvania, in 1970. At Chatham, she received the Psychology

Department Award for Outstanding Involvement in Psychology. Ms. Kirn

received the degree of Master of Arts in psychology from the Univer-

sity of Florida in June, 1971. During that year she served as a

graduate teaching assistant for the psychology department. In Septem-

ber of 1971, she was awarded a fellowship by the United States Public

Health Service. In August, 1973, she completed a year-long intern-

ship in clinical and community psychology at the Malcolm Bliss Mledital

Health Center in St. Louis, Missouri. She received the degree of

Doctor of Philosophy from the University of Florida in Decemnber, 1973.

She is married to Steven P. Kirn.

















I certify, that I have read this study- and that in my opinion it
C(uoir'lins to ac piabli standard-s of scholarly prDiesentation and is
It [ll adequate, in scope and quality, as a dissertation for the degree
of Doctor of Philosophy.






Marvin E. Shaw, Chairman
Professor of Psychology






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



)

----------------1--------
Paul Satz, Cochairman
Professor of Psychology





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












1 c 'rt ily that I[ ave read thLs study and tL.'al in m( opi) ion tl St
(conform' s tu acc tpablt s(,n!L i(i.- oI sch lauio pr '. antatio an i1
filly adequate, inii scope alnd alit y, as a disrtationi for tie degree
of Doctor of Philosophy.






Norman N. Markel
Associate Professor of Speech





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






Robert C. Ziller
Professor of Psychology





This dissertation was submitted to the Department of Psychology
in the College of Arts and Sciences and to the Graduate Council, and
was accepted as partial fulfillment of the requirements for the degree
of Doctor of Philosophy.

December, 1973


Dean, Graduate School







































UNIVERSITY OF FLORIDA
II3 1262 I I08553 6976
3 1262 08553 6976