Volume 32 Number 3
May 1993

John S. Backes

Research indicates that students at all levels find greater motivation and perform, at higher levels academically when instructional methods complement student personal learning style (Gentry & Ellingson, 1981; Gregorc & Butler, 1983). Yet, traditional practice in the public school classroom assumes little difference in learning style among students. Consequently, teachers may employ learning activities without regard to learning style. This makes failure a probable experience in the classroom.

Some American Indian students are said to be deductive or holistic in their approach to learning (Barwell, 1981; Kleinfeld, McDiarmid, Grubis, & Parrett, 1983). Nearly 20 years ago, John (1972) stated, "When working with Indian children, educators choose methods of instruction that zero in on what they wish to accomplish instead of methods that reflect the developmental stages of children" (p. 332). Swisher and Deyhle (1989) also suggested that the teaching method chosen has a significant effect on whether students learn or fail. They further indicated that "American Indian students come to learn about the world in ways that are different from mainstream students" (p. 5). Consequently, a mismatch between learning style and teaching methods may influence the dropout rate among American Indians.

Numerous studies have been conducted and many theories proposed to explain the factors that influence dropouts. These theories include such factors as socioeconomic status, substance abuse, legal problems, peer pressure, health, personal crisis, and the perceived value of education by the student and parent (Rumberger, 1987; Orshansky, 1967; Wehlage, Rutter, Smith, Lesko, & Fernandez, 1989). Analyses of the learning styles of dropout students in selected high schools have revealed that students at risk of dropping out learn in a processing style and with instructional strategies that differ significantly from those students who perform well (Dunn & Griggs, 1988). The national dropout rate for American Indian students is estimated to be approximately 30% (Swisher, Hoisch, & Pavel, 199 1) to approximately 50% (Chavers, 1991). Notwithstanding the problems associated with the specific and most "accurate" definition of dropout, the significant problem of American Indian students not completing high school is acute.

The purpose of this study was to determine the effect that learning style has on the learning success or failure of American Indian Chippewa (Metis) high school students as compared to non-American Indian high school students. The personal learning styles of American Indian Chippewa (Metis) students in Belcourt, North Dakota were identified and compared to those styles of non-American Indian students from Crookston, Minnesota. Crookston was chosen as the parallel study group to the American Indian Chippewa (Metis) group because of its relative size match (population and geographic area), its educational program match (K12 structure and community college), its contrasting homogeneous population and its dropout data representative of other upper midwest communities. Specific research questions to be answered were:

The term dropout is used to refer to a student who discontinued attendance in school and did not graduate or transfer to another public, private, or state approved education program (Instruction Manual for Completing MISOI and MIS02 Fall Reporting Forms for North Dakota Public Schools, 1991-1992). A related term often used by officials in schools with high populations of American Indian students is the non-completion rate. The term non-completion is used to refer to those students who entered the ninth grade but did not graduate with their class 4 years later and had not officially enrolled in another public, private, or state approved education program. The non-completion rate included those students who 1) became parents and discontinued schooling, 2) told host school officials of their intention to transfer schools but failed to ever enroll in the designated school, 3) developed attendance and discipline problems that resulted in discontinuation of schooling, or 4) would be considered traditional dropouts as defined in the MISORMIS02 Manual for 1991-92. Within this study, the term dropout is broadly defined to include non-completion students as well as traditional dropouts.

The term Metis means mixed blood among the American Indian Chippewa and persons of European descent, with French being the dominant European blood.

The term learning style means the basic channels through which the mind receives and expresses information. In addition to receiving and expressing information, learning style includes the gestalt of cognitive, affective, and environmental elements.

The town of Belcourt is the largest town located on the reservation. Belcourt has a population of 2,565 with a 6-year average senior class size of 95 students. The ethnic breakdown of the student population at Turtle Mountain Community High School during the period of this study was 99.94% American Indian Chippewa (Metis), 0.06% Caucasian, and 0.00% Hispanic, Black, and Asian. The American Indian subject student population in this study consisted of 98 Chippewa (Metis) students, representing 100% of the 1991-92 graduates from Turtle Mountain Community High School in Belcourt, North Dakota and 135 Chippewa (Metis) students, representing 40.3% of the total dropouts over the past 4 years.

Central High School is a public high school located in northeastern Minnesota. The Crookston, Minnesota community has a population of 8,628 with a 6-year average senior class size of 141 students. During the period of this study, the ethnic breakdown of the Crookston student population was 0.05% American Indian Chippewa (Metis), 91.7% Caucasian, 6.26% Hispanic, 0.07% Black, and 0.07% Asian. The non-American Indian subject population consisted of 94 non-American Indian student graduates from Central High School in Crookston, Minnesota, representing 92.1 % of the 1991-92 senior class and 31 student dropouts, representing 64.6% of the total dropout population over the past 4 years.

The instrument that was used to individually determine personal learning style was the Gregorc Style Delineator - Research Edition (Gregorc, 1982). The reliability of the delineator was assessed in terms of internal consistency using standardized Alphas as the statistics and a test-retest correlation coefficient. The standardized Alphas range is 0.89 to 0.93. The correlation coefficients between the first and second tests were from 0.85 to 0.88 for the four learning style scales.

The validity of the Gregorc Style Delineator was assessed in terms of construct validity by interview and of predictive validity by correlation between Gregorc Style Delineator scores and attribute scores, and responses to the descriptions resulting from the delineator. Correlations between Gregorc Style Delineator scores and ratings of attributes were from .55 to .76 for the four learning style scales.

The readability of the directions for the delineator is at a grade level of 6.5 on the Fry Readability Formula (Fry, 1968). The 40-word vocabulary based Gregorc Style Delineator matrix has a grade equivalent of 8.2 with a word range from grade 4 to grade 13 (Dale & O'Rourke, 1979).

The instrument identifies four channels, labeled concrete sequential, abstract sequential, abstract random, and concrete random. The channels reveal themselves through characteristics which Gregorc calls "style." They are described by the following attributes:

The Gregorc Style Delineator contains 40 words separated into ten 4-word sets. The subject assesses the relative value of the four words in each set, using self as a reference point. Subjects then rank in order the ten sets of four words. A rank of 4 is given to the word in each set which best and most powerfully describes the subject. Subjects then give a 3 to the word which is next most like them, a 2 to the next, and a 1 to the word which is the least descriptive of the subject. Each word in a set must have a ranking of 4, 3, 2, or 1. No two words in a set can have the same rank. There are no right or wrong answers as the person ranks the four words. Once the subjects have completed the 40-word delineator, a total score for each style is determined. Of the four final combined scores, the largest numerical score is representative of the dominant learning style for that individual subject. Subjects were informed to go with first impressions when ranking words.

Administration of the delineator to graduates at both Turtle Mountain Community High School and Crookston High School was conducted by the researcher onsite at the two high schools as part of regularly scheduled social studies class. Administration of the delineator to the American Indian Chippewa (Metis) dropout subject group was conducted by the researcher at the Turtle Mountain Community College General Educational Development (GED) building where dropouts agreed to meet after subjects were contacted by phone or as members of the regular GED program at the community college. Administration of the delineator to the non-American Indian dropout subject group was conducted at the Crookston Learning (Alternative) Center after subjects were contacted by phone or in person at the Learning Center. Other non-American Indian dropouts were administered the delineator at the Polk County Jail Education Program for inmates in Crookston, Minnesota.

After an explanation of the procedure printed on the back side of the Gregorc Style Delineator, the subject subgroups were asked to identify themselves as graduates or non-graduates at the top of their form. No other demographic graphic information was requested. The administration of the Gregorc Style Delineator took approximately 10 to 15 minutes, including instructions, self-assessment, and scoring.

The data were analyzed using mean style scores for the total subject populations as well as by comparison of graduate and dropout populations (subgroups). Dominant learning style comparisons by total subject populations were tallied. Percentages of dominant learning styles were calculated for each style and for each subgroup. Annual enrollments, numbers of dropouts, and dropout rate percentages were calculated for each of the subgroups from the 1985-86 school year through the 1990-91 school year. Additionally, the 1986-1991 annual high school completion and non-completion rates for students enrolled in the Turtle Mountain Community High School were calculated and charted.

As stated earlier, the national high school dropout rate for American Indian students is approximately 30% (Swisher et al., 1991) to approximately 50% (Chavers, 1991). The mean annual dropout rate in grades 9-12 at Turtle Mountain Community High School, Belcourt, North Dakota over the past 6 years was 16.20% (Table 1). This represented a mean of 79 students (grades 9-12) dropping out annually.

Table 1
Annual High School (grades 9-12) Enrollments, Dropouts, and Dropout Rate
Percentage for Turtle Mountain Community High School, Belcourt, ND from
1985-86 to 1990-91.

These figures contrast with the North Dakota state mean dropout rate of 1.75% (R. Torgeson, personal communication, March, 1992). The 5 year mean non-completion rate at Turtle Mountain Community High School was 59.56% (Table 2).

Table 2
Annual High School Completion and Non-Completion Rates of Those Students
Who Enroll in Ninth Grade and Graduate With Their Class 4 Years Later
for Turtle Mountain Community High School, Belcourt, ND
for Graduating Classes 1987-91.

The mean annual dropout rate in grades 9-12 at Central High School in Crookston, Minnesota over the past 6 years was 2.90%. This represented an annual mean of 16.6 students in grades 9-12 (Table 3).

Table 3
Annual High School (Grades 9-12) Enrollments, Dropouts, and Dropout Rate
Percentage for Central High School, Crookston, MN from 1985-86 to 1990-91.

Table 4 shows the results of a four-way composite comparison of the four subgroups related to dominant learning style. The dominant learning style which had the highest percentage among the total subject American Indian Chippewa (Metis) population was abstract random (40.3%). The dominant learning style which had the highest percentage among both of the American Indian Chippewa (Metis) subgroup populations was also abstract random with 35.7% of the graduates and 43.7% of the dropouts exhibiting a dominant abstract random style. The learning style with the lowest percent among the American Indian Chippewa (Metis) population was abstract sequential (15.4%).

The dominant learning style that had the highest percentage among the non-American Indian student population was concrete sequential (38.4%). Both the graduate and dropout subgroups of non-American Indian students showed concrete sequential to be their dominant learning style with 37.2% of the graduates and 41.7% of the dropouts having concrete sequential as their dominant style. The personal learning style of abstract sequential was the least dominant for the total non-American Indian population with 8.0% of the population.

Table 4
Composite Dominant Learning Style Comparison of Subject Groups
by Population and Percentage

The composite mean scores for learning styles of American Indian Chippewa (Metis) and non-American Indian graduates and dropouts are presented in Table 5. No significant differences were found among the mean learning styles of the American Indian Chippewa (Metis) subgroups. The non-American Indian subgroups showed a significant difference in the mean learning style of abstract sequential for graduates and dropouts (t-value = +2.54). A significant difference was also found between the mean learning style of concrete random for non-American Indian graduates and dropouts (t-value = -1.93). These results are significant at the .05 level using a one-tailed test. These results indicated that the non-American Indian dropout subgroup showed a significantly stronger dominant abstract sequential learning style than did the graduate subgroup. On the other hand, the graduate subgroup showed a significantly stronger dominant concrete random learning style than the dropout subgroup. The mean learning styles of concrete sequential and abstract random did not differ significantly among the non-American Indian subgroups.

The composite mean scores of learning styles for the total American Indian Chippewa (Metis) and non-American Indian populations are presented in Table 6. Significant differences were found in composite learning styles of the American Indian Chippewa (Metis) and non-American Indian populations in three of the four learning styles. The difference between mean scores of the two populations for the concrete sequential learning style yielded a t-value of -2.22, indicating that the non-American Indian population had a significantly higher concrete sequential mean score than that of the American Indian Chippewa (Metis) population. The difference between the mean scores of the two populations for the abstract random learning style yielded a t-value of +2.69, indicating a significantly stronger abstract random learning style among the American Indian Chippewa (Metis) population. The difference between the mean scores of the two populations for the concrete random learning style yielded a t-value of -2.56, indicating a significantly stronger concrete random learning style among the non-American Indian population. These results are significant at the .05 level using a one-tailed test. No significant difference was found between the mean scores of the two populations for the learning style of abstract sequential.

Table 5
Composite Learning Style Mean Scores of Graduates and Dropouts of American
Indian Chippewa (Metis) and Non-American Indian Subject Populations

*Significant at the. 05 level, one-tailed test.

Table 6
Composite Learning Style Mean Scores of Total Surveyed Subject Population of
American Indian Chippewa (Metis) Students and Non-American Indian Students

*Significant to the .05 level, one-tailed test.

The dominant personal learning style of both of the American Indian Chippewa (Metis) subgroups (dropouts and graduates) was found to be abstract random. The dominant personal learning style of both of the non-American Indian subgroups (dropouts and graduates) was found to be concrete sequential. Thus, an overall difference in dominant personal learning styles was found between the American Indian Chippewa (Metis) subjects and the non-American Indian subjects.

No significant differences were found in the learning style mean scores between American Indian Chippewa (Metis) dropouts and graduates for any of the four learning styles. No significant differences were found in the learning style mean scores between non-American Indian dropouts and graduates for the concrete sequential and abstract random learning styles. The non-American Indian dropout and graduate subgroups did show significant differences in the mean scores of the abstract sequential and concrete random learning styles. This finding should be interpreted with caution because of the relative low number (8) of non-American Indian dropouts who were either abstract sequential or concrete random. Composite mean scores were significant for the two dominant personal learning styles found for the total subject populations [abstract random for the American Indian Chippewa (Metis) subjects and concrete sequential for the non-American Indian subjects]. In addition, statistical significance was found between the composite mean scores of the two populations for concrete random. Although statistical significance was found, the overall importance of this finding is negligible because concrete random was the third most frequently reported dominant style for both populations.

The importance of these findings lies in the fact that a dominant learning style of the total population of American Indian Chippewa (Metis) students was shown, and that learning style (abstract random) coincides with a deductive, holistic instructional methodology. On the other hand, the dominant learning style of non-American Indian students (concrete sequential) coincides with inductive, linear methodology that dominates traditional classrooms. Consequently, the dominant personal learning style of the American Indian Chippewa (Metis) students does not match the teaching style of many classrooms, and this mis-match is a determining factor in their learning success or failure (Swisher & Deyhle, 1989).

That no significant difference existed between the dominant personal learning styles of the dropout and graduate subgroups of either population may indicate a limitation of this study. The single factor with a possible effect on dropout rate that was addressed in this study was that of learning style. Other variables that can affect dropout rates [such as socioeconomic status, substance abuse, legal problems, peer pressure, health, personal crisis, and the perceived value of education by the student and parent (Rumberger, 1987; Orshansky, 1967; Wehlage et al., 1989)] were not controlled in this study.

The most effective application of learning style theory lies in the greater understanding of it by teachers. A teacher's ability to adapt to individual differences, recognizing, and building on the strengths of students, American Indian or otherwise, will likely increase learning, motivation, and school completion (graduation). Learning depends on an appropriate discrepancy between the learner's present capacities and the difficulty of the task.

American Indian Chippewa (Metis) students in this study exhibited a dominant abstract random learning style, both for graduates and dropouts. The characteristics of the abstract random learning style include: sensitivity, emotion, personalization, imagination, interpretation, holistic view, aesthetic appreciation, part of a social group, discussion, reflection upon feelings, flexibility, and adaptability. With the dominant learning style and learning preferences of the subject American Indian Chippewa (Metis) population identified, the question of how to best address the individual and group learning needs, and thus reduce the disproportionately high dropout rate, is paramount. Through teacher awareness (staff development) of learning style theory, teacher observation of student traits need ultimately to produce an instructional strategy that is uniquely fitted to the individual student traits/styles.

The teacher's role in providing a style fit for the instruction of American Indian Chippewa (Metis) students requires the creation of a classroom environment that is open to sharing rather than a more traditional teacher-controlled setting. Teacher behaviors that acknowledge students as equals and demonstrate a sense of interest and purpose for each individual student will increase motivation and create a sense of community. A tolerance of differences and diversity can be realized through the integration of many ideas, possibilities, and theories, allowing for style differences and cultural adaptation. Acceptance, respect, and a positive focus on the American Indian Chippewa (Metis) culture and heritage results in increased self-worth as well as the creation of a holistic environment.

Learning through doing, rather than through verbal description, engages the student in the social context of learning. Assessment of student-generated portfolios would document student achievement and reflect student initiative and responsibility for learning. Necessary skills needed for mastery are attained more readily through visuals and demonstrations. Flexible and variable educational instruction would allow for informal processes to occur and for greater locus of control.

Teaching effectiveness needs to be measured in terms of the teacher's response to American Indian Chippewa (Metis) student learning style characteristics and ultimate success of the student in the classroom. The goal of implementing culturally/individually appropriate instructional strategies is to assure that the strengths of the students and various cultures be the factors that give shape to content, context, and instructional mode employed. The pedagogy of poverty (Haberman, 1991) approach (inclusive of such teacher acts as giving information, asking questions, giving directions, making assignments, giving tests, assigning homework, and giving grades, etc.) that often exists in reservation schools must change. This behavioristic (concrete sequential) instructional approach must be integrated with the constructivist/cognitive approach to teaching. Acknowledging the prior knowledge and inherent worth that the learner brings to the classroom, constructivist methods of teaching tend to be indirect, involving the creation of a learning environment. This approach to teaching, consistent with the abstract random style characteristics, is founded in the importance of what the students are doing rather than in the observable actions of the teacher.

Whereas the pedagogy of poverty includes teacher functions that, taken separately, may be helpful for learning; when performed together, they systematically exclude other student learning activities. As a result, the classroom atmosphere created by constant teacher direction and student compliance results in a passive resentment or overt resistance to learning/dropping out (Haberman, 1991; Hart, 1983). American Indian Chippewa (Metis) students, like all students, desire an enriched educational environment that includes examination of human differences. This examination includes differences of race, culture, socioeconomic status, religion, ethnicity, and gender. These issues are examined constantly by youth as they strive to make sense of the world and their place in it. The exploration of major concepts and multiple ideas, in contrast to pursuit of isolated facts, enables graduates to use various and competing ways to understand themselves, their culture, and the world. The involvement in the planning of educational activities enables students to learn to make informed choices.

Where students are involved in reflecting on their own lives and how they have come to believe and feel as they do, the process of building a probing, experimenting, reflecting environment is taking place. This environment, absent from the schools studied by Haberman (1991) and reflected sparingly in the subject reservation school, is congruent with the style characteristics of the abstract random learning style which was found to be dominant among the subject American Indian Chippewa (Metis) subject populations.

These findings and recommendations are consistent with previous research by Barwell (1981) and Kleinfeld et al. (1983) who characterized American Indian students to be more responsive to an overview or "big picture" approach to learning new material. This "orientation" approach to learning proceeds from the general to the particular. Kleinfeld et al. (1983) indicated that American Indian children are more apt to participate actively and verbally in group projects and in situations where they have control in volunteering. Additionally, the authors found evidence of a predisposition of American Indian children to participate in group or collaborative situations. These results also support the assertion made by Swisher and Deyhle (1989) that learning style differences were found in the literature they reviewed.

Studies have revealed that students at risk of dropping out of school learn in a style and with instructional strategies that differ significantly from those students who perform well (Gentry & Ellingson, 1981; Gregorc & Butler, 1983). This study found no significant difference between the dominant learning style of American Indian Chippewa (Metis) graduates and American Indian Chippewa (Metis) dropouts. However, a significant difference between the dominant learning style of American Indian Chippewa (Metis) subjects and the non-American Indian subjects was found. The dominant learning style of the American Indian Chippewa (Metis) subject population was abstract random, whereas the dominant learning style of the non-American Indian subject population was concrete sequential.

Because the dropout rate of American Indian students has been shown to be higher than that of non-American Indian students, and since the dominant learning style of American Indian Chippewa (Metis) students has been shown to be significantly different from that of non-American Indian students overall, it may be concluded that learning style is one factor that affects the learning success or failure of high school students. Furthermore, when the teaching style of instructors matches the learning styles of students, the probability of learning success increases.

Instructional methods that complement American Indian Chippewa (Metis) students' personal learning style are recommended as a means to increase academic success and ultimately lower the American Indian Chippewa (Metis) high school dropout rate. Recommendations are made for moving away from instructional strategies that focus primarily on teacher behaviors that control and dictate the learning environment to instructional strategies that are more indirect and cognizant of learner behaviors. The American Indian Chippewa (Metis) students' stake in maintaining the pedagogy of poverty is of the strongest possible kind. It systematically absolves them of responsibility for learning and puts the burden on the teachers, who must be accountable for making them learn (Haberman, 1991). Therefore, the recommendations in this article should be implemented knowing that students will initially resist changes to a system in which they have traditionally been able to predict the outcomes.

Although learning style differences did exist between and among the multicultural total populations in this study, it is imperative that educators not teach solely to groups (ethnic), but teach to individuals' learning strengths. This will ensure increased academic achievement and improved attitudes among students toward learning. To assert that certain aspects of academic performance are beyond the American Indian Chippewa (Metis) capacities is setting the stage for cultural disillusionment and educational deprivation.