BIA EDUCATION RESEARCH BULLETIN, YEAR 1973

ACADEMIC ACHIEVEMENT OF HIGH SCHOOL STUDENTS IN FEDERAL AND PUBLIC SCHOOLS

Willard P. Bass

The impetus for this study came from the United States Senate Appropriations Subcommittee during hearings held in February, 1965. At that time the Chairman of the Subcommittee asked the Commissioner of Indian Affairs to have a study made of the gain in academic achievement of Indian students in public schools compared to that of Indian students in the same span of time in Federal schools. The Commissioner agreed to initiate such a study beginning with Ninth Grade students and continuing through their graduation.

Numerous studies have been made of academic achievement of American Indian students. Only a few, however, have examined levels of achievement in various types of schools. Notice among these is the extensive study by Coombs (1958) in which it was found that the achievement of Indian students enrolled in public schools was higher on the average than that of Indian students enrolled in Bureau of Indians Affairs (BIA schools. However, since this was not a longitudinal study and initial individual differences were not controlled, differences in group achievement could not be attributed to the educational experiences provided to students by the schools. Although Coombs was careful to point out that differences in socioeconomic backgrounds of the students in the groups being compared may have accounted for the disparity in achievement levels, it became accepted as almost axiomatic, as a result of the findings of the study, that an Indian pupil would make greater academic progress in a public school than in a BIA school.

THE PROBLEM

The major concern of this longitudinal study was to determine whether academic achievement differs significantly for American Indian high school students enrolled in four types of schools: 1) Federal on-reservation, 2) Federal off-reservation, 3) public on- reservation, and 4) public off-reservation. Other important interests were whether achievement differs by geographic area, grade and sex. In addition, it was the purpose of the study to gather a variety of data on other psychological and sociological variables and to investigate the relationship of some of them to achievement and to each other.

The major independent variables were: school type, grade level, geographic area and sex. The principal dependent variable was academic achievement. Other dependent variables were: self-concept, achievement motivation, value orientation and school interest.

METHOD

The Sample. In the fall of 1966 a sample of Indian high school students was drawn from 21 schools located in the seven states of Alaska, Arizona, Nebraska, New Mexico, Oklahoma, South Dakota and Utah. Approximately equal numbers were drawn from each of the four school types. The sample also was stratified on the basis of sex, grade and geographic area, with approximately equal numbers of male and female; with 34% from 5 Grade Nine, 28% from Grade Ten, 20% from Grade Eleven, and 18% from Grade Twelve. The sample included all Indian students enrolled in certain schools and a random selection of Indian students from other schools. It was drawn so as to provide representation from five selected BIA administrative areas proportionate to the numbers of students enrolled in Federal schools in those areas. This sample, numbering 3,346 students, was pre-tested in the Fall of 1966. In the spring of 1967, testing sessions were held again in all of the same schools, at which time usable test results were obtained for 2,584 of those who had been pre- tested in the Fall. This group of 2,584 subjects, who were administered both pre-tests and post-tests, comprised the sample for the first year of the study.

In the Fall of 1967, a total of 3,375 students was tested. Of these, a substantial number were Ninth Grade students, new to the sample. The others were students who had been tested the previous year. In the Spring of 1968, a total of 2,556 students was post-tested.

The next testing session was held in the Spring of 1969. Data were sought only on Tenth, Eleventh and Twelfth Grade students who had been tested at some prior time in the study. Data were obtained for 1,337 students.

The final testing was accomplished in the Spring of 1970 and was confined to Eleventh and Twelfth Grade students who had been tested previously. Of the 1,337 students tested in the Spring of 1969, it was possible to test 837 again in the Spring of 1970.

Measuring Instruments. The following tests were administered during the course of the study.

Fall, 1966 California Achievement Tests (CAT), Advanced, Complete Battery, 1957 Edition, 1963 Norms, Form W.

California Short Form Test of Mental Maturity (CTMM), 1963, Level 4. Mooney Problem Check List (Abbreviated version), Form J-SH Questionnaire.

Spring, 1967 CAT, Form X.

Fall, 1967 CAT, Form Y.

CTMM. Administered to all Ninth Grade students and to others in Grades 10 through 12 new to the sample. Questionnaire. Administered to all Ninth Grade students and to others in Grades 10-12 new to the sample. Semantic Differential.

Spring, 1968 CAT, Form W. School Interest Inventory, by William Cottle, published by Houghton Mifflin Company, 1966.

Spring, 1969 CAT, Form X. California Psychological Inventory - Five Measures: CS (Capacity for Status) , SP (Social Presence) , AC (Achievement via Conformance), SA (Self-Acceptance) and AI (Achievement via Independence). Value Orientation Scale.

Spring, 1970 CAT, Form Y. Vocational Aspiration Scale.

Testing Procedures. Each of the six testing sessions was completed in one day at each school. All testing each Fall was accomplished within a period of about two weeks during late September and early October. Spring testing was done during the latter half of April.

In each geographic area, testing was under the supervision of a trained and experienced psychometrician, who either administered the tests or trained and supervised others, all of whom had previous experience in testing.

ACADEMIC ACHIEVEMENT

In comparing academic achievement of various groups, initial individual differences were taken into consideration. Post-test California Achievement Test (CAT) raw scores were used as a criterion and differences in means were tested for significance by analysis of co- variance. Since individual differences in scholastic aptitude and academic ability could conceivably influence criterion scores, pre-test intelligence and achievement scores were used as control variables. The CTMM intelligence quotient scores were used as a scholastic aptitude control, and the pre-test CAT scores were used as a prior achievement control.

Achievement by School Types. Using pre-test and post-test scores gathered over a span of four years, it was possible to assess achievement for four periods of one school year each, three periods of two school years each, two periods of three school years each, and one period of four school years. When this was done by grades for each period, using CAT scores for reading, mathematics, language and total battery, there were 52 assessment categories for one-year periods, 32 for two-year periods, 16 for three-year periods, and four for the four-year period, a total of 104 in all.

Controlling for individual differences in scholastic aptitude and academic ability, treatment of the data by analysis of covariance revealed that differences in achievement between the four types of schools were significant at the .05 level of confidence for 18 of the 52 assessment categories for one-year periods, 17 of the 32 assessment categories for two- year periods, 10 of the 16 assessment categories for three-year periods, and none of the four categories for the four-year period. On the basis of adjusted criterion means, which were calculated for those categories were significant differences occurred, Federal schools ranked higher than public schools on one-year analyses; public on-reservation schools ranked highest, and Federal off-reservation schools lowest on two-year analyses. Public schools ranked slightly higher than Federal schools on three-year analyses. However, when subjected to the Friedman Test, the differences in ranks of the four types of schools were found not significant at the .05 level of confidence for one-year, two-year or three-year analyses.

Altogether, the four types of schools were compared on 104 measures of academic achievement. Of this total of 104 categories of measure, differences in achievement between school types were found to be significant at the .05 level of confidence for 45 categories. The ranking of the four school types on the 45 significant categories based on adjusted criterion means are shown in Table 1, page 7.

Table 1
TOTAL RANKINGS OF SCHOOL TYPES
BASED UPON ADJUSTED CRITERION
ACHIEVEMENT MEANS
All Time Spans

Applying the Friedman Test to the above data yields an X² of 3. With 44 degrees of freedom, this falls far short of the X² of 60 + necessary for significance at the .05 level.

In summary, significant differences in achievement between types of schools were found for less than one-half of the categories measured and no significant hierarchical pattern of achievement emerged for those categories where significant differences in achievement did exist. The evidence indicates that, when initial individual differences in scholastic aptitude and academic ability were controlled, academic achievement of Indian students did not differ significantly in the four types of schools.

Academic Achievement by Areas. Designated geographic areas, which correspond to BIA administrative areas, were Aberdeen, Muskogee, Navajo, Phoenix and Juneau.

Analyses of academic achievement by areas were accomplished in the same manner as were the analyses by school types. Differences in achievement between the areas were found to be significant at the .05 level of confidence for 71 of the 104 categories of assessment. Rankings of areas, based upon adjusted criterion means, revealed that the Juneau area ranked first by a wide margin and the Aberdeen area ranked second for every time span. No clear pattern of rank was evident for the other three areas. Rankings based upon the adjusted criterion means for the 71 categories for which significant differences were found are presented in Table 2.

Table 2
TOTAL RANKINGS OF AREAS
BASED UPON ADJUSTED CRITERION
ACHIEVEMENT MEANS
All Time Spans

Applying the Friedman Test to the data in Table 2 yields an X² of 107.14. With 70 degrees of freedom this value is significant at the .01 level of confidence.

In summary, when individual differences in scholastic aptitude and academic ability were controlled, differences in achievement between areas were significant, with the Juneau area ranking highest, followed by Aberdeen. No clear pattern of superiority emerged for the other three areas.

Academic Achievement by Grades. Based upon CAT total battery scores, Ninth Grade students ranked at the 27th percentile and Twelfth Grade students at the 14th percentile when compared to national norms. In comparing actual grade placement with achievement grade placement, Indian students on the average were one year below grade level academically when entering Ninth Grade, but were more than two and one-half years below when i1 about to graduate from high school. Scores were consistently highest in language and lowest in mathematics. However, the greatest regression occurred in reading. Ninth Grade students ranked at the 34th percentile in reading and Twelfth Grade students at the 12th percentile.

Academic Achievement by Sex. During the course of the study, the CAT was administered to all grades (9-12) four times in the first two years, once to grades 10-12 in the third year, and once to grades 11 and 12 in the fourth year. This provided 21 comparative mean scores for boys and girls on each sub-test - reading, mathematics and language. With remarkable consistency, boys scored higher than girls in all 21 cases in mathematics and in 20 of 21 reading cases. Girls, on the other hand, scored higher than boys in every case in language. The evidence clearly indicates a slight but reliable superiority of boys over girls in reading, and a considerable superiority in mathematics. However, girls demonstrated a substantial superiority over boys in mastery of English language skills.

RESPONSES TO OTHER MEASUREMENT INSTRUMENTS

Responses to a number of self-report instruments yielded a variety of data which are the bases for certain conclusions.

A native language, rather than English, was the principal medium of oral communications used in the homes of two-thirds of the students. About 50% of the homes had television sets, but only 15% had telephones. Only 50% of the students' fathers and about 20% of the mothers were regularly employed. Less than 30% of parents were high school graduates and about 60% had completed eighth grade.

A higher percentage of public high school than Federal high school students had telephones, TV sets, and daily newspapers in the home. Also, more public school students had started to school at six years of age or younger, more of their parents were high school graduates, and more were regularly employed. Public off-reservation schools had the highest percentage in whose homes English was the primary language, as well as the highest percent- age who spoke English when they started school.

In general, Indian students appeared to value education highly, like school, be greatly concerned about grades, have confidence in their scholastic ability, and respect their teachers. But they also indicated that school is skipped frequently and many gave evidence of being high dropout risks. Indian students expressed optimism about the future and indicated that they have a healthy pride in racial and cultural heritage by rating Indian high among ten concepts, and much higher than White Man, on a Semantic Differential. The latter concept was given the lowest rating and Myself As A Person the next lowest rating. However, since mean ratings of these two concepts were in the positive range, highly unfavorable attitudes toward self and the white man do not seem to be implied.

Type and location of schools seemed to be factors related to certain student attitudes. Students attending off-reservation schools registered greater optimism concerning their future than did on-reservation students. Those in the most integrated situation (public off- reservation schools) scored highest on measures of self-esteem and those in the most segregated situations (Federal on-reservation schools) scored lowest. Those in the segregated Federal schools rated Indian higher than did those in the integrated, public schools.

Muskogee area students scored highest on measure of self-esteem, but also registered the lowest opinion of teachers and the least interest in school. Juneau area students evidenced the greatest interest in school and the highest regard for teachers and for education. Phoenix area students rated themselves lowest of the five areas on measures of self-esteem and on attitudes toward their present life, their future, their school success, education and college. The data indicate that students in the Aberdeen, Juneau and Muskogee areas were more oriented to, values of the dominant culture and possessed greater social presence than those in the Navajo and Phoenix areas. Students in public off-reservation schools rated highest and those in Federal on-reservation schools lowest on social presence and on orientation to values of the dominant culture (value orientation). Significant gains on social presence and value orientation for each successive grade (10-12) suggest that the school has a socializing and acculturating effect upon Indian students.

Responses to a Vocational Aspiration Scale indicated a lack of information among Indian high school students about occupations, particularly those occupations in the higher prestige range. Also, vocational expectations were lower than vocational aspirations. Boys, more than girls, thought that they would be unable to obtain jobs as high ranking as they desired.

RELATIONSHIPS BETWEEN VARIABLES

In order to explore relationships between achievement and certain other variables measured by the various instruments administered during the study, Pearson product moment coefficients of correlation were computed. A number of these correlations proved to be significant at the .01 level of confidence. As expected, there was a strong correlation between mental ability and achievement (+ .668) . There was a positive significant correlation between orientation to the values of the dominant culture and achievement (+.359). A moderately high significant correlation between self-concept and achievement (+ .264) suggests that strong positive feelings about self are reflected to some degree in higher levels of achievement for Indian students.

Differentials between the students own rankings of their desired and expected occupations were obtained and were found to have a significant positive correlation with achievement (+ .294) , as well as with mental ability (+ .290) and acculturation as measured by responses to certain items on a questionnaire (+ .259) . It is hypothesized that Indian students scoring high on the latter three actors tend to have relatively high vocational aspirations but also tend not to raise their actual expectations accordingly, possibly for reasons having to do with minority group status, and cultural influences.

ACADEMIC ACHIEVEMENT AND LANGUAGE OF THE HOME

The relation between principal language spoken in the home and academic achievement was explored. This was done by comparing home language of high achievers with home language of the entire sample. In the sample as a whole, 33 % of the students were from homes in which English was the principal language spoken and 67% were from native speaking homes. But of the high achievers (those ranking at or above the 50th percentile in achievement), 54 % from native speaking homes. The differences in expected and actual frequencies proved to be significant far beyond the .001 level of confidence when tested by chi-square. The evidence clearly indicates that there was a definite relationship between the language of the home and academic achievement.

SPECIAL NOTE. This is a special summary report of the original study made by Dr. Willard P. Bass entitled, “ Analysis of Academic Achievement Of Indian School Students In Federal And Public Schools.” It was completed in 1971 and published by the Southwestern Cooperative Educational Laboratory in Albuquerque, New Mexico.