Journal of American Indian Education-Arizona State University

Journal of American Indian Education

Volume 12 Number 3
May 1973

SELF-CONCEPT COMPARISONS OF ANGLO AND INDIAN CHILDREN

Roger Martig and Richard DeBlassie

Roger Martig is a graduate student and Richard DeBlassie is Professor and Head, Department of Counseling and Educational Psychology, New Mexico State University, Las Cruces.

STUDIES REPORTED by several researchers concerning Indian and Anglo children show conflicting results in relationship to self concept and Indian ethnic group membership (see Note 1). Consequently, it appears that there is a definite need to (a) increase our general knowledge about Indian ethnic group membership with respect to self-concept and (b) examine self-concept among primary Indian children. Other variables which may be considered as worthy of study include grade level and sex differences. Rationale for these two variables comes from studies by Hathhorn, who contended that the Indian’s low self-esteem is further reinforced as a result of the educational process (number of years in school), and Havighurst who reported evidence of sex differences in self-concept among Indian children.

From this background the following questions emerged as primary concerns of this study: (1) To what significant degree does Indian ethnic group membership influence self-concept when compared to self-concept and Anglo group membership? (2) To what significant degree does grade level influence self-concept among both Indian and Anglo elementary school children? (3) To what significant degree does sex determine self-concept among Indian and Anglo elementary school children?

Methodology

THE SAMPLE. The sample consisted of 129 elementary children from the Tularosa, New Mexico, Public School District. The study included both Indian and Anglo children in the first and fourth grades. Within the sample there were 22 first-grade Indian boys, 15 first-grade Anglo boys, 13 first-grade Indian girls, 14 first-grade Anglo girls, 11 fourth-grade Indian boys, 13 fourth-grade Anglo boys, 23 fourth-grade Indian girls, and 16 fourth-grade Anglo girls. The sample of Indian children included all Indian children in the first and fourth grades at the Bent-Mescalero Elementary School. The sample of Anglo children was composed of first- and fourth-grade children taken at random from the Tularosa Elementary School.

THE INSTRUMENT. The 23-item Primary Self-Concept Scale (Muller and Leonetti) was used as the self-concept instrument for this study. The scale consists of 23 sets of drawings of boys or girls. The subject responds by circling the child most like himself or herself. The scale yields a total of nine scores including:

1. Total self-concept. The PSCS is composed of 23 items that depict at least one child in a positive role and one child in a negative role. The student’s total self-concept score is determined by summing the number of positive choices which he or she has selected for the entire PSCS.

2. Peer aggressiveness vs. cooperation. This variable determines the child’s view of himself in relationship to peer cooperation.

3. Peer ostracism vs. acceptance. This factor measures the child’s view of his acceptance by his schoolmates.

4. Intellectual self-image. The student’s intellectual self-image is measured by this variable.

5. Helpfulness. The helpfulness factor assesses the child’s view of himself as one who helps others/one who is helped by others.

6. Physiological self. A child’s view of his body and bodily functions.

7. Adult acceptance vs. rejection. The student’s view of himself as being accepted or rejected by adults is measured through this factor.

8. Emotional self. The emotional self variable assesses how the child views his emotional self, i.e., happy or sad.

9. Success vs. nonsuccess. This factor measures the child’s view of himself to success at task-oriented pursuits.

COLLECTION OF DATA. The PSCS was administered separately to groups of first-grade Indian children, first-grade Anglo children, fourth-grade Indian children, and fourth-grade Anglo children. The senior researcher administered all the first-grade instruments with the help of one teacher and two teacher’s aides. Administration of the PSCS to fourth-grade children was completed by the researchers with the help of one teacher and one teacher’s aide. Before the administration of the PSCS, all teachers and teacher’s aides took part in a short orientation to the test conducted by the senior researcher. All students were given PSCS booklets in which they directly recorded their responses.

The PSCS was handscored. The results were then recorded on IBM keypunch cards. Each IBM card was checked twice to verify the accuracy of the recorded data.

TREATMENT OF THE DATA. Means and standard deviations were first calculated for all dependent variables in order that homogeneity of variance, analysis of variance, and after F comparisons could be computed. After all data were tested for homogeneity, a three-factor analysis of variance was performed on the nine dependent variables (PSCS variables). A t after F test was computed to compare means when significance was found through analysis of variance.

The three-factor analysis of variance for each of the PSCS variables examined significance in terms of (a) main effects (grade level, sex, and ethnic group) and (b) interaction effects.

Results and Discussion

Table I indicates the results of the statistical analyses of each of the nine PSCS variables. The following inferences can be gleaned from Table 1: (1) Most of the main effects analyses were found to show nonsignificant F-ratios. Only five F-ratios for main effects were found significant. (2) Only two interaction effects were found significant.

After F-tests (t-ratios) were computed for those main and interaction effects that were found significant. Tables II-VI show the results of these analyses.

The study revealed that grade level did not appear to influence self-concept of any of the factors measured by the PSCS. None of the nine measures of self-concept assessed by the PSCS was affected by grade level changes.

Analysis of variance did indicate significant differences with respect to sex for dependent variables 1 (total self-concept), 3 (peer ostracism vs. acceptance), 5 (helpfulness), and 8 (emotional self). In general, girls were found to have higher self-concept mean scores than boys on the above dependent variables. Havighurst concluded that Indian girls appear to be more self-critical than Indian boys. In dependent variables 1 and 3, t tests computed on the mean scores revealed the opposite; statistically the first-grade Indian boys had significantly lower self-concepts than the fourth-grade Indian girls.

Analysis of variance on dependent variable 7 (adult acceptance vs. rejection) indicated significant differences in terms of ethnic group membership. The t after F test indicated that Indian boys and girls at the first-grade level saw themselves rejected by adults to a significantly greater degree than first- and fourth-grade Anglo boys.

In general, the results of this study are in contradiction to those of studies done by Coleman and Hathhorn, both of whom found Indian children to have lower self-concepts than white children. The results of this research also contradict Hathhorn’s finding that Indian children enter school with lower self-concepts than white children. According to Hathhorn’s study, the educational process reinforces the Indian’s low self-concept. The researchers found no statistical evidence to substantiate either Coleman’s or Hathhorn’s conclusions. The researchers, however, did find results similar to those of Lammers and Havighurst who indicated no significant differences in self-concept between Indian and white students.

Table I
Statistical Analyses of PSCS Variables

	Variables¹ -- Results of Analysis of Variance (F Tests)
Source of Variation	1	2	3	4	5	6	7	8	9
Grade Level (A)	NS	NS	NS	NS	NS	NS	NS	NS	N
Sex (B)	*	NS	*	NS	*	NS	NS	*	N
Ethnic Group ©	NS	NS	NS	NS	NS	NS	*	NS	N
AxB	NS	NS	NS	NS	*	NS	NS	NS	N
AxC	NS	NS	NS	NS	*	NS	NS	NS	N
BxC	NS	NS	NS	NS	NS	NS	NS	NS	N
AxBxC	NS	NS	NS	NS	NS	NS	NS	NS	N

NS=Not significant at .05 level of probability.
*=Significant at .05 level of probability.

The nine PSCS variables
1. Total self-concept
2. Peer aggressiveness vs. cooperation
3. Peer ostracism vs. acceptance
4. Intellectual self-image
5. Helpfulness
6. Physiological self
7. Adult acceptance vs. rejection
8. Emotional self
9. Success vs. nonsuccess

Table II
Significant t Ratios for Dependent Variable 1

Group A as Compared to Group B
Group A	Group A mean	Group B	Group B mean	dt	t
First grade Anglo boys	18.73	Fourth grade Indian girls	20.26	36	2.17*
First grade Indian boys	18.86	Fourth grade Indian girls	20.26	43	2.04*