AN ANALYSIS OF ACADEMIC ACHIEVEMENT OF INDIAN HIGH SCHOOL STUDENTS IN FEDERAL AND PUBLIC SCHOOLS

I. INTRODUCTION

The major concern of this longitudinal study was to determine whether academic achievement differs significantly for American Indian students enrolled in four types of schools: (1) federal on-reservation, (2) federal off-reservation, (3) public on-reservation, (4) public off-reservation. Other important interests were to examine differences in academic achievement 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.

Of the numerous studies that have been made of academic achievement of American Indian students, only a few have examined levels of achievement in various types of schools, notably the extensive study by Coombs who found that Indian students who were enrolled in public schools achieved at a higher level on the average than did those enrolled in Bureau of Indian Affairs (BIA) schools. However, since initial individual differences were not controlled statistically, 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 almost axiomatic, as a result of the findings of the study, that an Indian student would make greater academic progress in a public school than in a BIA school. ¹

In this present study, individual differences were taken into consideration in comparing academic achievement of various groups. To provide a measure of control of individual differences influencing achievement the statistical technique of analysis of covariance was employed so that the differences in achievement could be attributed to the treatments being tested.

The Sample

In the fall of 1966 a sample of American 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: federal on-reservation, federal off-reservation, public on-reservation, and public off-reservation. The sample was also stratified on the basis of sex, grade, and geographic area, with approximately equal numbers of male and female, and with 34% coming from 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 students from other schools and was drawn so as to provide representation from certain Bureau of Indian Affairs administrative areas proportionate to the numbers of students enrolled in federal schools in those areas. This sample, numbering 3,346 students, was pretested in the fall of 1966. In the spring of 1967 testing sessions were held again in all of the same schools, at which time it proved possible to obtain usable post-test results for 2,584 of those who had been pretested in the fall. This group of 2,584 subjects, who were administered both pretests and post-tests, then comprised the sample for the first year of the study.

Insufficient time to make necessary arrangements, considering certain difficulties encountered, made it impossible to include public school native Alaskan students in the sample for the first year of the study. However, this situation was corrected and students enrolled in two public schools were added to the sample for the succeeding years of the study.

In the fall of 1967 a total of 3,375 Indian students was tested. Of these, a substantial number were ninth grade students brought into the sample for the first time, while the others were principally students who had been tested the previous school year. In the spring of 1968 a total of 2,556 Indian students was tested. Of this number, complete and usable data for both the fall pretest and the spring post-test were obtained for 1,928 Indian students.

The next testing session was held in the spring of 1969. No new ninth grade students were added to the sample at this time. Data were sought only on students who had been tested at some prior time in the study. This, of course, limited the spring of 1969 sample to tenth, eleventh, and twelfth grade students. Data were obtained for 1,377 students in the 1969 spring testing.

The final testing was accomplished in the spring of 1970, and again was confined to students who had been tested previously, thus limiting the sample to eleventh grade and twelfth grade students. Of the 1,377 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 Alaska public school students, grades 10-12, new to the sample.

Questionnaire. Administered to all ninth grade students and to Alaska public school students, 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.

Analysis of Data

In comparing groups within the sample on the basis of academic achievement, post-test California Achievement Test (CAT) raw scores were used as a criterion and differences in means were tested for significance by analysis of covariance. Since individual differences in scholastic aptitude and in academic ability could conceivably influence criterion scores, pretest intelligence and achievement scores were used as control variables. The California Test of Mental Maturity (CTMM) intelligence quotient scores were used as a scholastic aptitude control, and the pretest California Achievement Test (CAT) raw scores were used as a prior achievement control.

III. ACADEMIC ACHIEVEMENT BY SCHOOL TYPES

One-Year Analyses

Since achievement tests were administered at six different points within a span of four school years it was possible to analyze achievement for the following one-year periods:

Fall 1966 (pretest) - Spring 1967 (post-test)

Fall 1967 (pretest) - Spring 1968 (post-test)

Spring 1968 (pretest) - Spring 1969 (post-test)

Spring 1969 (pretest) - Spring 1970 (post-test)

Table 1 presents the mean raw scores of the criterion and control variables for reading, mathematics, language, and total battery, by school types, for ninth grade students who were pretested in the fall of 1966 and post-tested in the spring of 1967. Also presented in Table 1 are analysis of covariance figures and adjusted criterion means.

The F scores of 12.82, 18.73, 4.67, and 19.84 with 3 and 868 degrees of freedom are all significant beyond the 1% level. Therefore, there is little doubt that the ninth grade students enrolled in the four types of schools differed significantly in achievement during the 1966-67 school year. Since significant F values were found, it is appropriate to compute adjusted criterion means for each school type. In similar succeeding tables, whenever differences in criterion means are not found to be significant, adjusted means are not presented.

In order to avoid burdening the body of the report with tables, the remaining mean raw scores, analysis of covariance, and, where appropriate, adjusted criterion means, by grade and school type, for each of the one-year measurement periods are presented in Tables A1- A12 in Appendix A.

A compilation of the adjusted criterion means found in Table 1 and in Tables A1 - A12 is presented in Table 2. Rankings, by school types, are indicated in parentheses. Those categories for which achievement differences between school types were found not to be a significant (11th grade reading in 1966-67, 12th grade reading in 1966-67, 10th grade reading in 1967-68, etc.) are marked with the letter x.

An inspection of Table 2 reveals that significant achievement differences between school types occurred for only 18 of the 52 categories. A summary of the rankings appearing in Table 2 is presented in Table 3. The sums of the ranks (∑R) reveal that the general ranking of school types from highest to lowest was: federal on-reservation (31),federal off-reservation (39), public on-reservation (53), public off-reservation (57).

To test the differences in ranks for significance the Friedman Test, a form of rank order analysis of variance was employed. The formula is

where k is the number of rankings made (18) and n is the number of objects being ranked (4).

Then

Reference to an X² table reveals that a value of 14.66 with k-1=17 degrees of freedom is not significant at the .05 level.

To summarize, when individual differences in scholastic aptitude and academic ability were controlled, significant differences in one-year academic achievement between school types were found for only 18 of 52 categories of measurement and differences in the rankings of school types on those 18 categories were not significant. Apparently, academic achievement did not differ significantly between the four types of schools for the one-year time periods.

Two-Year Analyses

Achievement, by school types, was analyzed for the following two-year spans:

Fall 1966 (pretest) - Spring 1968 (post-test)

Fall 1967 (pretest) - Spring 1969 (post-test)

Spring 1968 (pretest) - Spring 1970 (post-test)

Tables A13 - A20 in the Appendix present the means of criterion and control variables, analysis of covariance and, where appropriate, adjusted criterion means, by school types, for each of the two-year measurement periods.

A summary of adjusted criterion achievement means by school types for two-year periods, taken from Tables A13 - A20 in the Appendix, is presented in Table 4. Significant differences in achievement between types of schools were found for 17 of the 32 categories of measurement, while differences were found not to be significant for 15 categories.

Table 5 presents a summary of the rankings from Table 4. Based upon the sums of the ranks (∑R) in Table 5, the general ranking of school types from highest to lowest was: public on-reservation (38) public off-reservation (42), federal on-reservation (43), federal off-reservation (47) However, differences in ranks which obviously are very slight, proved to be non-significant. Use of the Friedman Test yields an X² of only 1.44. With 16 degrees of freedom, this value falls far short of the X² of 26+ necessary for significance at the .05 level.

The evidence indicates that when individual differences in scholastic aptitude and academic ability were controlled there were not significant differences in two-year academic achievement between school types for 15 of the 32 categories measured, while for the 17 categories for which significant differences were found the rankings of school types were so mixed that no significant pattern of superiority emerged. Obviously, the two-year analyses do not indicate that academic achievement differed significantly between the four types of school.

Three-Year and Four-Year Analyses

Achievement by school types was also analyzed for the following three-year and four-year spans of time:

Fall 1966 (pretest) -Spring 1969 (post-test)

Fall 1967 (pretest) - Spring 1970 (post-test)

Fall 1966 (pretest) - Spring 1970 (post-test)

Tables A21 - A25 in Appendix A present the means of criterion and control variables, analysis of covariance and, where appropriate, adjusted criterion means, by school types, for each of the three-year measurement periods and for the four-year period.

A summary of adjusted criterion achievement means of school types for three-year and four-year time spans is presented in Table 6. There were no significant differences in achievement between the four types of schools for the four-year period from the fall of 1966 to the spring of 1970. The three-year analyses yielded significant F scores for 10 of the 16 categories. However, the orders of rank on the 10 significant categories are very mixed and favor the two public school types only slightly, as can be seen by reference to the sums of the ranks in Table 7.

Testing the differences in ranks for significance with the Friedman Test yields an X² of only 0.96, which falls far short of the figure of 16.9 necessary for significance at the .05 level. Since differences in achievement between the four school types were found to be non-significant for 6 categories and the ranks of the school types did not differ significantly on the 10 categories for which significant achievement differences were registered, it appears that academic achievement did not differ significantly between types of schools during the three-year periods.

Summary of Analyses of Academic Achievement by School Types

On the basis of adjusted criterion means, which were calculated for those categories having significant differences, federal schools ranked higher than public schools on one-year analyses, public on-reservation schools ranked highest and federal off-reservation schools lowest by small margins on two-year analyses, and public schools ranked slightly higher than federal schools on three-year analyses. However, rankings of school types were so mixed on those categories for which significant differences were found that differences in ranks were not significant for one-year, two-year, or three-year analyses. No significant differences in achievement between the four types of schools were found for the four-year period.

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 only 45 categories. The rankings of the four school types on the 45 significant categories are shown in Table 8, which is a composite of Tables 3, 5, and 7.

Applying the Friedman Test to the data in Table 8 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 hierarchal pattern of achievement emerged for those categories where significant differences in achievement did exist. The evidence, therefore, does not indicate that the academic achievement of American Indian students was superior or inferior in any particular type of high school when individual differences in scholastic aptitude and academic ability were controlled.

IV. ACADEMIC ACHIEVEMENT BY AREAS

Analyses of academic achievement by geographic areas also were made, similar to the analyses of achievement completed for school types. One-year, two-year, three-year, and four-year analyses were made.

Designated areas correspond to Bureau of Indian Affairs administrative areas and include the following: Aberdeen, Muskogee, Navajo, Phoenix, and Juneau. The numbers of students drawn from each area were based upon the numbers of students from the area enrolled in BIA schools. Therefore, as might be expected, numbers of subjects varied greatly for the different areas.

In testing differences in achievement between areas by analysis of covariance, post-test achievement scores were used as the criterion and pretest achievement and intelligence scores were used as control variables, just as they were in analyzing achievement by school types.

One-Year Analyses

Achievement by areas was analyzed for the following one-year periods:

Fall 1966 (pretest) - Spring 1967 (post-test)

Fall 1967 (pretest) - Spring 1968 (post-test)

Spring 1968 (pretest) - Spring 1969 (post-test)

Spring 1969 (pretest) - Spring 1970 (post-test)

Table 9 presents the means of criterion and control variables, analysis of covariance, and adjusted criterion means, by areas, for ninth grade students who were pretested in the fall of 1966 and post-tested in the spring of 1967. The remaining data for one-year analyses of academic achievement by areas are presented in Tables B1 - B12 in Appendix B.

A compilation of the adjusted criterion means gathered from Table 9 and Tables B1 - B12 is presented in Table 10. Rankings by areas are indicated in parentheses, and categories for which significant differences in achievement were not found are marked with the letter x.

A study of Table 10 reveals that achievement differences between the five areas were found to be significant for 31 of the 52 categories. Based upon the adjusted criterion mean scores for the 31 categories for which significant differences were found, the Juneau area ranked first in 24 of 31 categories. A summary of the rankings from Table 10 is presented in Table 11. On the basis of sums of ranks (?R), the Juneau area ranked highest by a wide margin (42), Aberdeen ranked second (85), followed in order by Phoenix (102), Navajo (117), and Muskogee (119).

Using the Friedman Formula to test differences in ranks for significance yields an X2 value of 50.42. With 30 degrees of freedom, 50.42 is significant at the .05 level, indicating that achievement differed significantly between areas on one-year analyses.

Two-Year Analyses

Achievement by areas was analyzed for the following two-year periods:

Fall 1966 (pretest) - Spring 1968 (post-test)

Fall 1967 (pretest) - Spring 1969 (post-test)

Spring 1968 (pretest) - Spring 1970 (post-test)

Tables B13 - B20 in Appendix B present the means of criterion and control variables, analysis of covariance, and, where appropriate, adjusted criterion means, by areas, for each of the two-year measurement periods.

A summary of adjusted criterion achievement means by areas for two-year intervals is presented in Table 12. An inspection of the table reveals that differences in achievement between areas were found to be significant for 25 of the 32 categories, while for only 7 categories were differences found non-significant. Again, as was true for the one-year analyses, the Juneau area ranked first by a large margin, followed by Aberdeen. Next in order were Navajo, Phoenix, and Muskogee.

A summary of the rankings from Table 12 appears in Table 13. On the basis of sums of ranks (∑R), the Juneau area ranked highest (40), Aberdeen and Navajo were second (75 and 76), followed by Phoenix (86) and Muskogee (98).

Application of the Friedman Test for differences in ranks yields an X² value of 30.01, which is not significant at the .05 level with 24 degrees of freedom. Thus, analyses indicate that achievement did not differ significantly between areas for the two-year measurement periods.

Three-Year and Four-Year Analyses

Analysis of achievement by areas was also analyzed for the following three-year and four-year spans of time:

Fall 1966 (pretest) - Spring 1969 (post-test)

Fall 1967 (pretest) - Spring 1970 (post-test)

Fall 1966 (pretest) - Spring 1970 (post-test)

Tables B21 - B25 in Appendix B present the means of criterion and control variables, analysis of covariance and, where appropriate, adjusted criterion means, by areas, for each of the three-year time spans and for the four-year period.

A summary of adjusted criterion means by areas for three-year and four-year time spans are presented in Table 14. Differences in achievement for the areas were significant for 11 of the 16 three-year categories and for all 4 of the four-year categories. Juneau ranked first in every three-year category, and was followed in order in overall rankings by Aberdeen, Navajo, Muskogee, and Phoenix.

A summary of rankings from Table 14 for three-year analyses is presented in Table 15. On the basis of sums of ranks the order of rank from highest achievement to lowest is: Juneau, Aberdeen, Navajo, Muskogee, and Phoenix.

The Friedman Test yields an X² of 23.27, which is significant at the .O1 level with 10 degrees of freedom, indicating that differences in achievement between areas were significant for the three-year measurement periods.

A summary of rankings from Table 14 for the four-year analyses is shown in Table 16. On the basis of sums of ranks it can be seen that achievement was highest in the Juneau area, followed in order by Aberdeen, Muskogee, Phoenix, and Navajo.

Computation of X² by the Friedman Formula gives a value of 10, which is significant at the .05 level. Achievement of Indian students in the five areas seemed to differ significantly over the four-year measurement period.

Summary of Analyses of Academic Achievement by Areas

Based upon adjusted criterion means, the Juneau area ranked first and the Aberdeen area second for every time span. Rankings for the three other areas varied for the different measurement periods.

Of the total of 104 categories on which achievement was measured, differences in achievement between areas were found to be significant at or beyond the 5% level of confidence for 71, or more than two-thirds, of the categories. Rankings based upon all 71 categories are presented in Table 17, which is a composite of Tables 11, 13, 15, and 16.

Applying the Friedman Test to the data in Table 17 yields an X² of 107.14. With 70 degrees of freedom this value is significant at the .O1 level of confidence.

In summary, when individual differences in scholastic aptitude and academic ability were controlled, differences in achievement between areas appeared to be 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

Academic achievement data for each grade are presented in Table 18. While it is evident that academic achievement of Indian students, as measured by the California Achievement Test, is progressive from grade 9 through grade 12, it is also evident that achievement is regressive when compared to national norms. For example, the difference in grade placement in reading for students at the actual 9.1 grade level and those at the 12.8 grade level was 2 grades rather than the 3.7 grades considered normal. For mathematics the grade placement differences were even smaller, registering 1.3 grades for the 1966-67 school year and 1.9 grades for the 1967-68 school year. Language showed somewhat greater differences, with 2.5 grades for 1966-67 and 2 grades for 1967-68. In comparing actual grade placement with achievement grade placement as measured by the California Achievement Total Battery mean scores, it is seen that Indian students were about one year retarded academically when entering ninth grade but were more than two and one-half years retarded when about to graduate from high school.

Percentile rankings demonstrate this progressive retardation very strikingly. Based upon total battery mean scores, ninth grade students ranked at percentile 27, while twelfth grade students ranked at percentile 14. Similar regression occurred for each of the separate subject areas. Scores were consistently highest in language and lowest in mathematics.

Academic Achievement by Sex

When achievement scores are compared by sex, it can be seen from an inspection of Table 19 that boys consistently scored slightly higher than girls in reading and considerably higher than girls in mathematics, while girls consistently scored substantially higher than boys in language. Attesting to the consistency of the above achievement pattern is the fact that the only exception to the pattern in the 63 comparisons presented in Table 19 was for 12th grade reading in the spring of 1969 testing.

The evidence clearly indicates superiority of Indian boys over girls in the mastery of reading and mathematics skills and the superiority of girls over boys in the mastery of English language skills.

VI. RESPONSES TO OTHER MEASUREMENT INSTRUMENTS

Responses to Questionnaire

Each student was asked to respond to a questionnaire as a means of obtaining personal and familial data. Total affirmative responses to each question are presented in percentage form for each school type in Table 20, and for each area in Table 21.

Student responses to the questionnaire, as presented in Table 20, indicate that higher percentages of public high school than federal high school students have telephones, TV sets, and daily newspapers in the home. Also, more started school at six years of age or younger, more of their parents are high school graduates, more of their parents are regularly employed, and fewer students know how to take part in tribal ceremonies.

Public off-reservation schools had the highest percentage in whose homes English is the primary language, as well as the highest percentage who spoke English when they started school. Public on-reservation schools had the highest percentage of students residing on a reservation.

When questionnaire responses are tabulated by areas, as presented in Table 21, certain differences and similarities between areas become evident. Some of these that seem most apparent are:

1. A majority of Indian high school students in the Aberdeen (75%), Navajo (80%), and Phoenix (90%) areas live on reservations, while very few in the Muskogee (5%) and Juneau (5%) areas do.

2. Only 1°/ of the students tested in the Juneau area claimed to have a TV set at home, while substantial numbers in other areas, ranging from 48 percent in the Navajo to 87 percent in the Muskogee areas, claimed them.

3. The Juneau and Navajo areas ranked lower than other areas in telephones, daily newspapers, parents who graduated from high school, parents divorced or separated, and families receiving welfare assistance.

4. The Navajo area has much the lowest percentage of homes in which English was the primary language (16%), and in percentage of students who spoke English when they started school (56%), while Aberdeen was highest in both categories (66%, 92%).

5. The Navajo area had the highest percentage who had dropped out of school for 1/2 year or more (15%), and the Muskogee area the lowest (4%).

6. The Navajo area ranked highest in percent of full-blood Indian students with 91 percent, followed closely by the Phoenix area at 86 percent.

7. The Muskogee area had the lowest percentage of fathers regularly employed (37%) and Phoenix the highest (54%), while the Juneau area had the lowest percentage of employed mothers (8%) and the Aberdeen area the highest (27%).

8. Knowledge of how to take part in tribal ceremonies seems to differ little from one area to another, ranging only from a low of 31 percent in the Juneau area to a high of 38 percent in the Aberdeen area.

Questionnaire responses by grade were also tabulated but are not presented because of the similarity of responses from grade to grade. For the same reason, it did not seem profitable to present responses by sex.

Responses to Mooney Problem Check List

Tables 22-25 present responses of students to the abbreviated version of the Mooney Problem Check List administered in this study. Students were asked to check each item that they felt described a problem for them. Figures in the tables are percentages of all students in each category who checked the particular item.

Table 22 reveals that more Indian high school students-of both sexes were concerned about grades than about any other problem listed. This item, number 16, was checked most frequently, both by boys (58%) and by girls (73%). It is interesting to note that the sexes agreed on the ten problems that troubled them most, although not always in the same order of rank. The other items ranking in the toy ten, in order of total frequency checked, are: (30) oral reports, (4) having to ask parents for money, (36) not interested in certain subjects, (25) deciding what to take in high school, (18) needing to find a part time job now, (38) not knowing what I really want, (1) not getting enough sleep, (31) needing to know my vocational abilities, and (41) lying without meaning to.

Other items which ranked high in frequency of responses for both sexes were: (6) bashful, (9) trouble with writing, (11) not knowing how to buy things wisely, and (27) awkward in meeting people.

Girls also checked frequently: (8) poor complexion or skin trouble, and (7) getting too excited. Apparently girls felt that they had more problems than did boys, since 32 of the 41 items were checked by a higher percentage of girls than boys. However, the following problems seemed to concern more boys than girls: (23) school too strict, (29) smoking, (12) girls don't seem to like me, and (15) underweight.

Those problems checked least frequently were: (3) being an only child, (35) trouble with my feet, (22) afraid I may need an operation, (37) Mother, (19) wanting to know more about boys, and (2) being treated like an outsider.

The following are some observations based upon inspection of Table 23:

1. In the Juneau area, parents not liking students' friends (24), and not being allowed to run with certain friends (5) are not as frequent problems as for students of other areas.

2. A higher percentage of Juneau area students found smoking (29) and needing to know their vocational abilities (31) to be causes for concern than did students of other areas.

3. Higher percentages of students in the Navajo area, with Phoenix running a close second, checked the following items: (2) being a grade behind in school, (7) getting too excited, (25) deciding what to take in high school, and (34) afraid God is going to punish me.

In making comparisons of Mooney data between school types it can be seen in Table 24 that higher percentages of students in federal schools than in public schools considered the following items to be problems: (1) don't get enough sleep, (2) being a grade behind in school, (9) trouble with writing, (25) deciding what to take in high school, (29) smoking, (32) needing to know my vocational abilities, and (38) not knowing what I really want. It is interesting that students in federal on-reservation schools checked about twice as frequently as other students item 34, "afraid that God is going to punish me." They also checked item 22 more frequently, "afraid I may need an operation."

Public school off-reservation students checked less frequently than others item 23, "school is too strict."

In examining Mooney data by grades there are a number of problems that seem to become increasingly critical as students progress from grade 9 through grade 12. As can be seen from inspection of Table 25 these problems are: (1) don't get enough sleep, (11) not knowing how to buy things wisely, (18) needing to find a part-time job now, (26) ill at ease at social affairs, (27) awkward in meeting people, (31) needing to know my vocational abilities, and (38) not knowing what I really want. Other problems seem to remain at a relatively consistent level, with one rather interesting and striking exception being the relatively high percentage of twelfth grade males who considered item 23, "school is too strict," a problem.

It is possible to analyze the Mooney Problem Check List responses by problem areas, as well as by individual responses. Items may be grouped together into seven problem areas. When total responses made to all items in a problem area are figured as percentages of total possible responses for all items in that problem area and this is done for each of the seven areas, it is possible to see the areas of greatest concern. The general problem areas, with percentages of items that were checked for each problem area, are as follows:

Problems of greatest concern to Indian high school youth appear to be in areas IV and II, having to do with money, work, the future, and school. Next appear to be those concerning self (VII) and relations to people (VI). Of least concern seem to be problems of home and family.

The Semantic Differential

A Semantic Differential was administered in the fall of 1967. In this instrument students were asked to react to ten concepts: SCHOOL, TEACHERS, MY SUCCESS IN SCHOOL, MYSELF AS A PERSON, INDIAN, WHITE MAN, MY PRESENT LIFE, MY FUTURE, EDUCATION, COLLEGE. Under each concept, twelve bipolar seven-point scales, using adjective pairs, were presented, three for each of four major factors. The four major factors and their opposite adjective pairs were as follows: Evaluation (cognitive)?good-bad, valuable-worthless, important-unimportant; Evaluation (affective)?pleasant-unpleasant, ugly-beautiful, nice-awful; Potency?weak-strong, shallow-deep, influential-powerless; Activity?fast-slow, busy-idle, active-passive.

The following is the general format used:

SCHOOL

1. Good ==== ==== ==== ==== ==== ==== ==== Bad

2. Weak ==== ==== ==== ==== ==== ==== ==== Strong

etc.

Each scale was scored as follows:

A score of 1 on the above scale indicates a rating of very unpleasant, 2 - quite unpleasant, 3 - slightly unpleasant, 4 - neutral, 5 - slightly pleasant, 6 - quite pleasant, 7 - very pleasant.

Table 26 presents mean scores of factors under each of the concepts for school types and also for each grade. The score for each factor was derived by averaging the mean scores of the factor's three scales.

Comparing Semantic Differential Scores by Concepts

A comparison of total sample mean scores for the various concepts in Table 26 reveals that Indian high school students have a high regard for education. EDUCATION was given the highest overall rating of the ten concepts with highest mean scores on both of the Evaluation factors and second highest on the Potency and Activity factors. COLLEGE was rated second highest overall, with the third highest score on Cognitive Evaluation, second highest on Affective Evaluation, and top scores on Potency and Activity. SCHOOL was rated second highest on the Cognitive Evaluation factor, but only sixth on Affective Evaluation. Apparently, school was liked less than it was valued.

Overall rankings of the ten concepts, from highest to lowest, with ratings on each factor shown in parentheses, were as follows:

EDUCATION (1,1,2,2); COLLEGE (3,2,1,1); MY FUTURE (4,3,3,3);
INDIAN (5,4,5,4); SCHOOL (2,6,4,7); MY PRESENT LIFE (8,5,7,5);
TEACHERS (7,8,6,6); MY SUCCESS IN SCHOOL (6,7,8,8);
MYSELF (9,9,9,9); WHITE MAN (10,10,10,10).

As can be seen, there was great consistency in factor ratings. When subjected to Friedman's rank order of analysis test it was found that differences between concept ratings were significant at the .O1 level of confidence.

Apparently, Indian students were quite optimistic about their future, since they rated the concept MY FUTURE third highest. However, a comparatively low self-concept is indicated by the next to last rating of MYSELF AS A PERSON on all four factors. The concept WHITE MAN scored lowest on every factor.

Comparing Semantic Differential Scores By School Types

Differences between ratings assigned to the concepts by the four school types proved to be significant beyond the .05 level of confidence for only three of the ten concepts -- INDIAN, WHITE MAN, and MY FUTURE.

Federal off-reservation school students gave INDIAN a higher rating on every factor than did students of the other types of schools. Federal on-reservation school students rated INDIAN next highest, public-off reservation students next, and public on-reservation students rated INDIAN lowest on every factor. With only one exception, federal schools in the study are segregated, while all public schools are integrated. Apparently, Indian students attending all-Indian schools hold a higher opinion of Indians than do those in integrated school situations.

Public on-reservation school students rated WHITE MAN higher than did other students on every factor, federal on-reservation students rated WHITE MAN next highest overall, federal off-reservation next to lowest overall, and public off-reservation students lowest on every factor. Those students attending school on reservations in an Indian dominated society tended to rate WHITE MAN higher than did those attending school off-reservation in the white man's world.

On the concept MY FUTURE, federal off-reservation school students scored highest overall, followed in order by public off-reservation, public on-reservation and federal on-reservation. Apparently, off-reservation Indian students are more optimistic about the future than are on-reservation students.

Comparing Semantic Differential Scores By Grades

Based upon scores on all factors, differences between the ratings of ninth, tenth, eleventh, and twelfth grade students were significant for three of the ten concepts--WHITE MAN, MY PRESENT LIFE, and EDUCATION. For each of the three concepts, ratings tended to be higher for each successively higher grade.

When the Cognitive Evaluation factor scores are examined by themselves, the pattern of progressively higher scores for each successive grade is noticeable for all concepts except MYSELF AS A PERSON, INDIAN, and COLLEGE. However, scores on Affective Evaluation do not show the same increase. As Indian students progress through high school it appears that they place an increasing value on school, teachers, education, their success in school, their present life, their future, and white people, but experience no increased positive feeling toward them.

Scores on Potency and Activity factors vary only little by grades for most concepts. Exceptions are increases in Activity ratings for WHITE MAN, and in Potency ratings for WHITE MAN, EDUCATION, and COLLEGE.

Comparing Semantic Differential Scores By Areas

Semantic Differential scores are presented by areas in Table 27. Ratings by the Indian students in the five geographic areas differed significantly on every concept except SCHOOL. The following are some observations on the ratings:

1. The Juneau area gave TEACHERS and WHITE MAN the highest ratings, followed closely in each case by the Navajo area. The lowest ratings were given these two concepts by the Muskogee area. The Aberdeen area also rated WHITE MAN very low, with next to lowest scores on all four factors.

2. The Muskogee area rated INDIAN highest and the Juneau area rated INDIAN lowest.

3. Muskogee area students seemed to evidence greater self-esteem and confidence than students in other areas by registering the highest scores on all factors for MY SUCCESS IN SCHOOL, MYSELF AS A PERSON, INDIAN, and MY FUTURE. The Muskogee area also scored highest on three factors and second on the fourth factor for the concept MY PRESENT LIFE.

4. The Aberdeen area scored high on MYSELF AS A PERSON, MY PRESENT LIFE, and MY FUTURE, ranking second to Muskogee on each concept.

5. The Phoenix area rated the following concepts lower than did the other areas: MY SUCCESS IN SCHOOL, MY PRESENT LIFE, MY FUTURE, EDUCATION, and COLLEGE.

6. The Navajo and Juneau areas rated EDUCATION higher than did the other three areas.

7. COLLEGE was rated highest by the Muskogee area.

Comparing Semantic Differential Scores By Sexes

It is evident in Table 27 that females generally rated the concepts higher than did males. Girls rated TEACHERS, WHITE MAN, EDUCATION, and COLLEGE higher on every factor, and SCHOOL and MY FUTURE higher on three of the four factors. Exceptions to the general trend appear for the concepts MY SUCCESS IN SCHOOL and MYSELF AS A PERSON, which were rated higher by boys than by girls on all four factors, and MY PRESENT LIFE, which was rated higher by boys on three factors. Indian boys seem to have a better self-concept and greater confidence in themselves than do girls, but may have less optimism about the future.

Responses to School Interest Inventory

In the spring of 1968 the School Interest Inventory was administered to 2164 Indian high school students. On this instrument each student was asked to respond to 150 statements by marking them true or false. Table 28 presents percentages of true and false responses to certain items that have been selected for presentation because of their information value. Some of the items presented in Table 28, like numbers 31 and 73, are not used in scoring the tests but do provide valuable personal and familial data. Other items, like numbers 8 and 93, are meaningful for individuals but not for group analysis, and are omitted. Item numbers in Table 28 correspond to item numbers in the instrument. It will be noted that the percentages do not always add to 100 percent, because some items received no response from a small percentage of students.

Comparing School Interest Inventory Scores by School

Types and Areas

The School Interest Inventory can be scored to obtain either weighted or unweighted totals. The weighted method, which assigns values of 1 to 9 for each item, is used in this study. Boys and girls are scored on different scales and, therefore, their scores are not comparable. The scale for boys contains 90 items and has a potential total score of 375, while the scale for girls has 86 items and a potential score of 337. There are 72 items common to both scales. Some items in the Inventory are not used for scoring on either scale.

As in golf and cross-country, the lower score is the better score. High scores on the School Interest Inventory indicate lack of interest in school and high probability of dropout. In this study, mean weighted scores are used to compare the interest in school of Indian students in different types of schools and in different geographic areas. These scores are presented in Table 29. Since scores registered by boys and girls are not comparable they are presented separately.

Inspection of Table 29 reveals that the mean scores for males differ considerably for the four school types and also for the five areas, while female scores for areas differ somewhat less than do male scores, and differ even less for school types. To test the differences for significance, analysis of variance was used. The results are presented in Table 30.

Interest in school, as measured by the School Interest Inventory, differed significantly for boys in the four types of school, with those in public on-reservation schools registering the greatest interest and those in federal on-reservation schools the least. Differences were also significant for boys in the five geographic areas, with those in the Aberdeen area registering the greatest interest in school and those in the Navajo area the least.

Differences for girls by school types were not significant. However, differences for girls by areas were significant, with those in the Juneau area registering the greatest interest and those in the Phoenix area the least.

When male and female scores are considered together and a combined ranking is determined for school types, the order from greatest interest to least interest is as follows: public on-reservation, public off-reservation, federal off-reservation, federal on-reservation. Similarly, the order for areas is as follows: Juneau, Aberdeen, Phoenix, Navajo, Muskogee. It is interesting that the order of rank of areas on the School Interest Inventory is identical to the academic achievement ranking appearing in Table 17.

There are no tables of normative data for the School Interest Inventory. However, some comparison can be made of mean scores for Indian students in this study with mean scores for non-Indian students in other studies. A study in one high school found that the mean weighted score for male students who stayed in school was 51.98, while the mean weighted score for male students who later dropped out was 116.52. For females the scores were 56.91 for stay-ins and 103.77 for dropouts. A study of students in four other schools found mean scores of 72.69 for male stay-ins, 137.20 for male dropouts, 60.49 for female stay-ins, and 110.02 for female dropouts.1

It is evident from the above figures that mean scores for Indian students tend to run high, almost approaching dropout levels. This, of course, is consistent with the high dropout rates for Indian students, which have been found to be 39 percent in the Southwest2 and 48 percent in the Northwest3 from enrollment in grade eight to graduation from high school.

The School Interest Inventory has proved to be a useful instrument for identifying potential school dropouts. However, its value with American Indian students was not known. In the interests of investigating the predictive value of the SII for Indians some further analyses were attempted on a small scale.

Since the SII was administered in the spring of 1968, those students who were enrolled in grade nine at that time normally would have graduated in the spring of 1971. Computer printouts of names of ninth grade students who had taken the SII in 1968 were mailed to selected schools with the request that the students listed be identified as graduates, dropouts, or transfers. Unfortunately, some school officials were unable to classify a majority of those who had withdrawn as either definite dropouts or transfers, and a fourth classification of "unknown" was added. Furthermore, as no attempt was made to follow-up those who were identified as transfers, it was not possible to determine whether they were eventually dropouts or graduates.

Responses were received from four BIA schools and six public schools as widely scattered in location as Alaska, Arizona, Utah, New Mexico, Oklahoma, Nebraska, and South Dakota. Table 31 shows the number of students in each classification and the average weighted score registered on the SII by the students so classified.

It is evident from the average scores in Table 31 that the SII does discriminate to some degree between Indian graduates and dropouts. However, an examination of single scores leads one to the conclusion that identification of dropouts on an individual basis would be difficult.

An inspection of test items reveals some that seem inappropriate for Indians. An item analysis was made of the responses of the graduates and dropouts to determine which items seem to discriminate and which do not. Some items that discriminate between graduates and dropouts for the general school population but do not for Indian students are the following:

4. I have more than one older brother or sister.

16. To get a job like my fathers I will have to finish high school.

30. My parents are active in community affairs.

33. My mother does a lot of church work.

47. I like love scenes on television.

50. My mother did not complete eighth grade.

63. My father did not complete high school.

76. My mother completed high school.

84. My father works with his hands.

94. Counting my parents and me, there are more than five people in my family.

141. I have more friends of the opposite sex than of my own sex.

145. My father did not complete eighth grade.

Other discrepancies appeared in responses by boys to questions about clothes and by girls to questions about age. Items 25 and 132, "I would be happier in school if I could buy better clothes;" and "Most of the people in my homeroom have better clothes than I do;" did not discriminate for Indian boys, although they did for girls. Also, items 88, 93 and 138, "I am one of the oldest in my homeroom;" "Most of my friends are older than I;" and "I am older than most of the people in my class;" did not discriminate for Indian girls, although they did for boys.

Among the items that seem to discriminate for Indian students but do not for others are the following:

6. It would be more fun to go to an art gallery than to a showing of new cars (Indian graduates tended to answer this "true", and dropouts to answer it "false."

31. My father earned more than $3000 last year.

52. I am not at ease with others.

140. My mother encourages me to do well in school.

It appears that the School Interest inventory could be a useful instrument for identifying high dropout risks among Indian students if some modifications were made in scoring based on the item analysis of responses of dropouts and graduates.

The California Psychological Inventory and the Value Orientation Scale

Five measures from the California Psychological Inventory (CPI) were used: CS (Capacity for Status), SP (Social Presence), AC (Achievement via Conformance), SA (Self-Acceptance), and AI (Achievement via Independence). According to the test manual these measures were designed to assess characteristics of personality as described below.

CS To serve as an index of an individual's capacity for status (not his actual or achieved status). The scale attempts to measure the personal qualities that underlie and lead to status.

SP To assess factors such as poise, spontaneity, and self-confidence in personal and social interaction.

AC To identify those factors of interest and motivation that facilitate achievement in any setting where conformance is a positive behavior.

SA To assess factors such as personal worth, self-acceptance, and capacity for independent thinking.

AI To identify those factors of interest and motivation that facilitates achievement in any setting where autonomy and independence are positive behaviors.4

Also, a value scale,5 developed by Strodtbeck, was used as a measure of acculturation to middle class values. This Value Orientation Scale consisted of eight true and false questions as follows:

1. Planning only makes a person unhappy since your plans hardly ever work out anyway.
2. When a man is born, the success he's going to have is already in the cards, so he might as well accept it and not fight against it.
3. Nowadays, with world conditions the way they are, the wise person lives for today and lets tomorrow take care of itself.
4. Even when teenagers get married, their main loyalty still belongs to their fathers and mothers.
5. When the time comes for a boy to take a job, he should stay near his parents, even if it means giving up a good job opportunity.
6. Nothing in life is worth the sacrifice of moving away from your parents.
7. The best kind of a job to have is one where you are part of an organization, all working together, even if you don't get individual credit.
8. It's silly for a teenager to put money into a car when money could be used to get started in business or for an education.

The first three questions have to do with time orientation and mastery over one's destiny. Questions 4-6 measure familism versus individualism, and loyalty to extended family versus loyalty to nuclear family. Question 7 tests for group versus individual orientation. Question 8 deals with immediate versus postponed gratification.

It was hypothesized that those holding values of the dominant culture would tend to answer the first seven questions, "false," and the last question, "true," while those more oriented toward traditional Indian values would tend to answer the questions in the opposite way. In using the Value Orientation Scale as a measure of orientation to values of the dominant culture, scores were computed by totaling the number of "middle class" answers. Thus, a higher score indicated a greater degree of orientation to middle class values.

Table 32 presents mean scores, by types of school and by sex, for each of the five scales of the CPI that were used, and for the Value Orientation Scale. Differences in means were tested for significance by analysis of variance.

Differences between scores for school types were not significant, either for males or females, on CAPACITY FOR STATUS, or on ACHIEVEMENT VIA CONFORMANCE. Significant differences were found, for both males and females, on SOCIAL PRESENCE and on ACHIEVEMENT VIA INDEPENDENCE, for females on SELF-ACCEPTANCE, and for females on VALUE ORIENTATION.

Public off-reservation students scored highest on SOCIAL PRESENCE, while federal on-reservation students scored lowest.

Public off-reservation and federal on-reservation students scored highest on ACHIEVEMENT VIA INDEPENDENCE, public on-reservation students ranked next, and federal off-reservation students ranked lowest.

On SELF-ACCEPTANCE, female students in public off-reservation schools scored highest, public on-reservation next, then federal off-reservation, and federal on-reservation lowest. While differences in scores for males were not significant, it is interesting to note that they follow the same pattern.

Females in public off-reservation schools indicated a higher degree of acculturation than those in other types of schools by scoring highest on VALUE ORIENTATION, while those in federal on-reservation schools scored lowest. Differences were not significant for males.

Differences between scores for areas were significant in most instances, as can be seen in Table 33.

When area scores in Table 33 are examined and compared, the following facts become apparent:

1. Juneau males and Muskogee females scored higher than their counterparts from other areas on CAPACITY FOR STATUS.
2. Muskogee students, both male and female, scored substantially higher than students from other areas on SOCIAL PRESENCE. Next highest were Aberdeen and Juneau, followed by Phoenix and Navajo.
3. Muskogee and Juneau students ranked highest on ACHIEVEMENT VIA CONFORMANCE.
4. Muskogee area female students ranked highest on SELF-ACCEPTANCE, followed by Juneau, Aberdeen, Phoenix, and Navajo. Muskogee and Juneau male students also ranked highest, although differences were not significant.
5. Differences in scores for ACHIEVEMENT VIA INDEPENDENCE were not significant, either for males or females. However, it is interesting that both male and female students in the Navajo area scored highest on this factor.
6. The orders of rank on VALUE ORIENTATION are very similar for males and females. Considering the scores of boys and girls together on this measure of acculturation, Juneau students rank highest, followed by Muskogee, Aberdeen, Phoenix, and Navajo.

An examination of Table 34 reveals that scores tend to increase for each successive grade, with twelfth grade students scoring higher than tenth in every instance, and higher than eleventh with only one exception. Differences in scores between grades were significant in six of the twelve cases. Attention is directed to gains on SOCIAL PRESENCE and VALUE ORIENTATION from grade ten to grade twelve. Apparently the school has a strong socializing and acculturating effect upon students.

Males scored higher than females on all scales for which differences in scores were significant, except on VALUE ORIENTATION. The scores indicate that boys in the sample are more ambitious and self-seeking, feel more self-confident in social interaction, and have a greater sense of personal worth than girls, but that girls are more oriented to the values of the dominant culture.

The Vocational Aspiration Scale

A vocational aspiration scale was devised to measure the differential between level of occupational desire and level of occupational expectation of Indian high school students. The instrument was administered in the spring of 1970 to 1,286 students in grades eleven and twelve.

In constructing the instrument, 110 occupations were selected for each sex and listed in order of general standing as determined by reference to rankings appearing in various studies. Eleven groups of occupations were then formed from the list, each group containing one occupation from the ten highest occupations listed, one from the next ten highest on the list, and so on down to one from the lowest ten on the list. In each of the eleven groups, then, ten occupations were listed, each one representing a different level of occupational standing from high to low.

Each group of ten occupations was presented to the examinee at three different points in the instrument. At one point the examinee was asked to rate the occupations from one to ten on the basis of general standing, at another point to indicate which job in the group he would choose to have in the future if free to have any one he wished, and at still another point to check the job which he feels is the best one he is really sure that he can get in the future.

Table 36 presents the eleven occupation groupings appearing in the Vocational Aspiration Scale for Males and the order of rank in each group as determined by the mean ratings of the 635 male respondents. Table 37 presents the same information for the 651 female respondents.

Inspection of Tables 36 and 37 reveals a number of ratings that would seem to suggest either a lack of information about occupations or possibly some cultural bias. For example, boys rated computer programmer higher than sociologist or psychologist, electric or telephone lineman higher than veterinarian, factory machine operator and carpenter higher than newspaper editor, auto or airplane mechanic higher than certified public accountant, heavy machine operator higher than actor, oilfield roughneck and mail carrier higher than undertaker, electrician higher than clergyman, and welder higher than TV announcer. Girls rated medical technician higher than physician or sociologist, physical therapist higher than lawyer, practical nurse higher than biologist, file clerk higher than radio announcer, computer programmer higher than architect, stenographer higher than dietician or TV announcer, bank teller higher than newspaper editor, office clerk higher than veterinarian or draftsman, airline stewardess higher than speech therapist, and teacher aide higher than postmistress.

In general, the Indian high school students rated the occupations of lower standing quite accurately but did not display the same judgment in ranking the occupations considered to be of higher standing. This may reflect the low socioeconomic backgrounds of most of the students and their consequent lack of first hand acquaintance with many of the higher prestige occupations.

In determining the differential between occupational desire and occupational expectation for each examinee, the difference was computed between the ranking of the preferred job in each group and the best expected job in that group, using the examinee's own job ratings. A total score for each subject was then computed by subtracting the sum of the desired occupation scores from the sum of the expected occupation scores. Scores ranged from +84 to -32. If expectations were lower in rank than preferences, a positive differential resulted. There were many cases of preference for jobs having lower standings than the jobs the examinee thought he could get, resulting in negative scores for about one-third of the examinees. However the mean differential score for girls was +7.6 and for boys was +9.4, indicating that, in general, vocational expectations were somewhat lower than were aspirations, and that boys, more than girls, felt that they would not be able to procure jobs of as high ranking as they would like.

REFERENCES

1. Cottle, William C. Examiner's Manual for the School Interest Inventory. Houghton Mifflin Company, Boston, 1966.

2. Owens, Charles S. and Willard P. Bass. The American Indian High School Dropout in the Southwest. Southwestern Cooperative Educational Laboratory, Albuquerque, 1969.

3. Selinger, Alphonse D. The American Indian High School Dropout. Northwest Regional Educational Laboratory, Portland, Oregon, 1968.

4. Gough, Harrison G. California Psychological Inventory Manual. Consulting Psychologists Press, Inc., Palo Alto, California, 1957.

5. Strodtbeck, Fred L. "Family Integration, Values, and Achievement," in A. H. Halsey, et. al., Education, Economy, and Society. The Free Press of Glencoe, Inc., New York, 1961.

VII. RELATIONSHIPS BETWEEN VARIABLES

Variables Measured

In order to explore the possible relationships between certain variables measured by the various instruments administered during the study, Pearson product moment coefficients of correlation were computed. The following factors were used.

1. Academic Achievement, as measured by the total battery score on the California Achievement Test (CAT).
2. Mental Ability, as measured by the California Test of Mental Maturity (CTMM).
3. Value Orientation, as measured by the Value Orientation Scale (VOS).
4. Self-Concept, as measured by the combined scores of three California Psychological Inventory (CPI) scales--Capacity for Status, Social Presence, and Self-Acceptance.
5. Self-Concept, as measured by responses on the Semantic Differential (SD) to the concepts MY SUCCESS IN SCHOOL, MYSELF AS A PERSON, INDIAN, MY PRESENT LIFE, and MY FUTURE.
6. Acculturation, as measured by 16 questions on the Questionnaire (Q).
7. Acculturation, as measured by 23 selected items from the School Interest Inventory (SII).
8. Achievement Motivation, as measured by the Achievement Via Conformance (AC) scale of the California Psychological Inventory (CPI).
9. Achievement Motivation, as measured by the Achievement Via Independence (AI) scale of the California Psychological Inventory (CPI).
10. Vocational Aspiration-Expectation Differential, as measured by the Vocational Aspiration Scale (VAS).

Pearson product moment coefficients of correlation were computed in one operation between all of the above ten variables in order to provide a matrix as presented in Table 38. This procedure limited the sample to 391 students for whom there were data on every variable.

Correlations Between Achievement and Other Variables

As can be seen in Table 38, academic achievement proved to be correlated significantly in the positive direction with six of the nine other variables. The coefficient of .667 indicates, as expected, that there is a strong positive correlation between mental ability and academic achievement. Probably there are factors other than innate mental ability that are being measured by the CTMM, such as reading skill and ability to work quickly and accurately, but, in any event, it is a strong predictor of academic success.

Of the remaining variables, value orientation has the highest correlation with achievement, and self-concept, as measured by the CPI, also appears to have definite relationship to achievement. These two variables will be discussed at greater length later.

The moderate correlation of .294 between achievement and the vocational aspiration-expectation differential indicates that there was some tendency for better students to have greater differences between occupational desires and actual expectations than did poorer students. It will be noted that the coefficient of correlation between mental ability and vocational aspiration-expectation differential is almost identical to that between achievement and vocational aspiration-expectation. Possibly the Indian high school juniors and seniors of lower mental ability and academic standing are cognizant of their limitations and, therefore, do not aspire as high vocationally as do the more intelligent students of higher academic rank. Many capable Indian students, aware of their potential, may desire high ranking occupations, but, because of problems endemic to their minority group status and, possibly, because of cultural influences that bear upon them, they may be doubtful about ever making their aspirations an actuality.

Acculturation (Q) shows some positive correlation with achievement, indicating a definite but moderate relationship between the two. Apparently, degree of acculturation, as measured by the sum total of such characteristics as living off of a reservation, speaking English in the home, having a TV set, parents having a high school education, etc., has a positive relation to achievement, but is not a highly potent factor. The factor of home language was isolated from other factors included in acculturation (Q) and its relationship to achievement was investigated. Findings from this analysis are presented later in this chapter.

Some positive correlation is indicated between mental ability and value orientation. It would appear that to some degree students of higher intelligence have internalized more of the values usually associated with the dominant culture than have students of lower intelligence.

A moderate and rather substantial positive correlation, as indicated by the coefficient of .373, exists between mental ability and acculturation (Q). The explanation may be that students from backgrounds indicative of greater acculturation simply do better on standardized tests, like the CTMM, than do students from less acculturated backgrounds which are generally conceded to produce educational disadvantage.

The coefficient of correlation of .259 between acculturation (Q) and vocation aspiration-expectation differential is indicative of a small but significant and definite relationship. It appears that the more acculturated students had greater differentials between occupational desires and occupational expectations than did less acculturated students. Perhaps students with a relatively high degree of acculturation have had more of the experiences that tend to raise occupational aspirations but are dubious of their chances of realizing them, while those of a lesser level of acculturation tend to expect and be satisfied with lower prestige occupations.

Other correlations showing significant relationships are interesting but not very enlightening, except perhaps as suggestions for further steps in analysis. For example, the correlation between acculturation (Q) and acculturation (SII) is moderately high and might be combined into one measure for further analyses. Substantial correlations between achievement motivation (AC) and achievement motivation (AI), and between self-concept (CPI) and achievement motivation (AC) were not unexpected. However, the almost zero correlation between self-concept (CPI) and self-concept (SD) was surprising, as were the low correlations between value orientation and the two acculturation variables.

Relationships of Value Orientation and Self-Concept to

Academic Achievement

The relationships between value orientation and academic achievement, and between self-concept (CPI) and academic achievement were analyzed further, using a sample of all students for whom data had been gathered on all three variables. This raised the number of subjects to 1,664. Pearson product moment coefficients of correlation were computed, as presented in Table 39.

As can be seen in Table 39 there was a coefficient of correlation of .359 between academic achievement and value orientation. This is somewhat higher than the coefficient of .314 appearing in the matrix constituting Table 78 that was computed on a smaller sample. The coefficient of .359 indicates a moderately high positive relationship between degree of orientation to values of the dominant culture and level of academic achievement.

A significant relationship, but of lesser magnitude, also was evident between self-concept (CPI) and achievement. The coefficient of correlation again was higher for the larger sample than that shown in Table 78. A coefficient of correlation of .264 suggests that strong positive feelings about self on the part of Indian students are reflected to a moderate degree in higher levels of achievement.

Academic Achievement and the Language of the Home

The relation between principal language of the home and academic achievement was explored. This was done by comparing home language of high achievers with home language of the entire sample.

An examination of individual achievement test scores registered by students in the fall of 1966, and in the fall of 1967 by all new students added to the sample at that time, revealed that only 349, or less than ten percent, scored at or above the 50th percentile. It was found that 189, or 54 percent, of these 349 high achievers came from homes in which English was spoken most of the time, while 160, or 46 percent, came from homes in which a native Indian language was predominant. For the sample as a whole, 33 percent came from English speaking homes and 67 percent from native speaking homes. Thus, if the home language pattern were the same for the high achievers as for the whole sample, it would be expected that only 33 percent of the 349 high achievers, or 115 instead of 189, would be from English speaking homes and 67 percent, or 234 instead of 160, would be from native speaking homes.

To test for significance in difference of home language between high achievers and the entire sample chi-square was employed. Actual and expected frequencies for high achievers are shown in Table 40.

Computation of chi-square yields an X² value of 63.307. This value is much greater than the 6.635 necessary for significance at the .O1 level of confidence, and, in fact, is far beyond the value of 10.8 necessary for significance at the .001 level. Evidence clearly indicates that a significantly greater number of students from English speaking homes and fewer from native speaking homes were high achievers than the numbers of each in the whole sample would warrant. Apparently, there is a definite relationship between the language of the home and academic achievement.

VIII. SUMMARY AND CONCLUSIONS

Academic Achievement by Types of Schools

The primary concern of the study was to determine whether there were significant differences in academic achievement between Indian students in four types of high schools--federal on-reservation, federal off-reservation, public on-reservation, and public off-reservation. Using pretest and post-test scores gathered over a span of four years it was possible to assess achievement for four one-year periods, three two-year periods, two three-year periods, and one four-year period. When this was done by grades for each time period, using the California Achievement Test scores for reading, mathematics, language, and total battery, 104 categories of assessment resulted.

Controlling for initial individual differences in scholastic aptitude and academic ability, treatment of the data by analysis of covariance revealed that differences in achievement between the four school types, significant at or beyond the .05 level of confidence, had occurred for only 45 of 104, or less than one-half, of the categories. Significantly, no reliable differences in achievement between the four types of schools were found for the four-year period from the fall of 1966 to the spring of 1970. For the 45 categories for which significant achievement differences were registered, rankings were so variable that no hierarchal pattern, or evidence of particular superiority or inferiority, emerged.

In the light of the above findings it can confidently be concluded that when initial individual differences in scholastic aptitude and academic ability were controlled there was no evidence that academic achievement of American Indian students was greater in one type of school than in another.

Academic Achievement by Areas

Differences in academic achievement of Indian students in the Aberdeen, Muskogee, Navajo, Phoenix, and Juneau areas were found to be significant at or beyond the .05 level of confidence for 71, or more than two-thirds, of the 104 categories of assessment.

Students in the Juneau area demonstrated a marked superiority in achievement, ranking first by wide margins for one-year, two-year, three-year, and four-year analyses. Aberdeen ranked second for every time span. The other three areas varied in achievement rank for the various categories to the extent that no pattern of overall rank could be determined.

When initial individual differences in scholastic aptitude and academic ability were controlled, the evidence clearly established the academic superiority of the Juneau area, with Aberdeen a distant second, and no clear pattern of rank evident for the other three areas.

Academic Achievement by Grades

The data show that the academic achievement of Indian students is progressive from grades 9-12 but regressive when compared to national norms. Based upon California Achievement Total Battery mean scores, Indian students were one year retarded academically when entering ninth grade and more than two and one-half years retarded when about to graduate. In terms of percentiles, they ranked at the 27th percentile at the ninth grade level and at the 14th percentile at the twelfth grade level.

Scores were consistently highest in language and lowest in mathematics. However, the greatest regression in comparison to national norms occurred in reading. Ninth grade students ranked at the 34th percentile in reading and twelfth grade students at the 12th percentile.

Academic Achievement by Sexes

There were 21 comparative scores for boys and girls on reading, 21 on mathematics, and 21 on language. Boys scored slightly higher in 20 of the 21 reading cases. Boys scored higher than girls in all 21 cases in mathematics. Girls, on the other hand, scored higher than boys in every case in language.

The evidence clearly indicates a slight but reliable superiority of Indian boys over girls in reading, and a considerable superiority in mathematics. However, girls demonstrate a substantial superiority over boys in the mastery of English language skills.

Responses to Other Measurement Instruments

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

A native language, rather than English, was the principal medium of oral communication 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 the parents were high school graduates and about 60% had completed eighth grade. Families were comparatively large; three-fourths of the students' families numbered five or more.

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 indicated.

Type and location of schools seemed to be factors related to certain student attitudes. Indian students attending schools in off-reservation settings registered greater optimism concerning their future than did on-reservation students. Those in the most integrated situations (public off-reservation schools) scored higher on measures of self-esteem than did those in the most segregated situations (federal on-reservation schools). Those in the segregated, federal schools indicated a greater regard for the Indian than did those in the integrated, public schools. Students attending schools located in the Indian-dominated society of the reservation registered a higher opinion of the white man than did those attending schools located in the off-reservation, Anglo-dominated society.

Students in the Muskogee area had the highest opinion of the Indian of any of the five areas and the lowest opinion of the white man, while those in the Juneau area registered the highest opinion of the white man and the lowest opinion of the Indian. Muskogee area students scored highest on measures of self-esteem, but also registered the lowest opinion of teachers and the least interest in school. Juneau area students, on the other hand, evidenced the greatest interest in school and the highest regard for teachers and education. The students in the Phoenix area rated themselves lowest of the five areas on measures of self-esteem and of 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. Also, students in public off-reservation schools rated highest and those in federal on-reservation schools lowest on value orientation and social presence. 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.

There was an evident lack of information among Indian high school students about occupations, particularly about those 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 of as high ranking as they desired.

Relationships Between Variables

Pearson product moment coefficients of correlation showed that mental ability, value orientation, self-concept as measured by three scales of the California Psychological Inventory, acculturation as measured by a questionnaire, and achievement motivation via conformance were significantly related to achievement. Mental ability was highly related to achievement, as was expected. There was a substantial positive relationship between orientation to the values of the dominant culture and achievement. The moderately high positive correlation between self-concept and achievement suggests that strong positive feelings about self are reflected to some degree in higher levels of achievement for Indian students. Acculturation and achievement motivation via conformance were reliably, but not highly correlated with achievement.

The proportion of Indian students from English speaking homes who were high achievers was significantly greater than from native speaking homes. Apparently, there is a definite relationship between the language of the home and academic achievement.

A differential between desired and expected occupations was obtained for Indian students and was found to have a significant-positive correlation with achievement, as well as with mental ability and acculturation. It is hypothesized that Indian students scoring high on the latter three factors tend to have relatively high vocational aspirations but also tend not to raise their actual expectations correspondingly, possibly for reasons having to do with minority group status and cultural influences.

1 Coombs, L. Madison, et al. The Indian Child Goes to School. U. S. Department of the Interior, Bureau of Indian Affairs, Washington, D.C., 1958.