ACF OPRE: American Indian-Alaska Native Head Start Research and Outcomes Assessment (AIAN)

CONDUCTING RESEARCH

Currently, there is literature on research among American Indian and Alaska Native children, fostered largely by Native researchers, that argues cogently for four changes in the lenses through which questions are formed, services are devised, and research is conducted (Deyhle and Swisher, 1997; Moreno, 1991; Running Wolf et al., 2002; Strang and von Glatz, n.d.):

Research should emphasize understanding the strengths these children bring to their educational experiences and life paths, rather than examining information and knowledge to determine “what’s wrong” with AI-AN children and youths. In short, this suggests changing the paradigm from a deficit model to an assets model.
Research should focus on examining the institutions and practices that serve these children, rather than examining data and drawing conclusions that something in the AI-AN experience leads to academic underperformance and school failure. Supporters of this perspective suggest that if institutions and practices more closely built on the learning styles and life experiences of AI-AN children, those children would achieve far greater success.
Research should carefully consider whether measures and expectations are accurate for the population under study, especially since they are often compared against measures and expectations derived from other groups (e.g., middle-class majority culture).
Research should empower American Indian and Alaska Native communities through efforts such as community advisory groups to guide the work and capacity-building among local populations to conduct their own studies.

Below, we present information about conducting research among AI-AN populations and then conclude with comments on the content and quality of the existing knowledge base.

The Practice of Conducting Research Among American Indian and Alaska Native Populations

Contemporary research among American Indian and Alaska Native populations is affected by the experience of those communities (“Our Voices, Our Vision,” cited in Swisher, 1996):

Just as the exploitation of American Indian land and resources is of value to corporate America, research and publishing is valuable to non-Indian scholars. As a result of racism, greed, and distorted perceptions of Native realities, Indian culture as an economic commodity has been exploited by the dominant society with considerable damage to Indian people. Tribal people need to safeguard the borders of their cultural domains against research and publishing incursions.

Existing publications are replete with discussions about ways to conduct research in “Indian Country.” A number of recent materials carry very clear messages that to be considered valid and reliable, research must be conducted by those who fully comprehend the historical experiences of AI-AN populations, recognize that tribes are sovereign nations, respect the culture, and actively demonstrate respect for tribal members (e.g., Swisher, 1996). Other materials, particularly those discussing health-related research point out additional factors to consider: the need for researchers to obtain permission from the tribe to conduct the work, the role of local Institutional Review Boards or similar entities, the importance of engaging tribal and community input, compensation for tribes and participants, confidentiality of tribes and individuals, and responsiveness to questions of interest to the tribal community (Norton and Manson, 1996).

As part of the report from the Indian Nations at Risk Task Force, strategies for appropriately evaluating early childhood education were listed as follows (Paul, 1991, p. 12):

Study participants should use their language of choice.
People affected by early childhood education should have their perceptions incorporated into studies.
High priority should be attached to cultural values, and evaluation procedures should respect them.
Research should incorporate cultural relevance.
Evaluation procedures should be developed by, or in consultation with, people involved in the education process.
The evaluation design should be created within the processes being studied.
Results should not be published as negative reflections on the groups being studied.
Shared, culturally driven values should be used for community education.
Evaluations should establish guidelines for the next wave of research.

A recent report from a Federal Interagency Task Force responsible for Executive Order 13096, American Indian and Alaska Native Education, contains a research agenda, along with a set of guidelines for conducting research among AI-AN populations. The assumptions for conducting studies include the following: research should include a focus on success, researchers must respect tribal sovereignty, and research needs to be sensitive to tribal differences (Strang and von Glatz, n.d.).

One study on an evaluation of a Head Start program, although conducted more than 20 years ago, contains the steps that the researchers followed to determine whether Head Start met the needs of children, whether the Head Start program was successful, the problems teachers had, and if the tribe should administer Head Start itself (Tigges and Zastrow, 1981). To conduct the project, the evaluators discussed it with tribal officials, reviewed literature, and developed the research design. The research design included (1) identifying issues to be studied that were the priority of tribal officials; (2) determining the methods that would be used (e.g., time frame for interviews, use of verbal rather than written questionnaires, confidentiality, and avoiding comparisons of Head Start classrooms); (3) clarifying the effect that the allowed methods would have on the evaluation; and (4) establishing procedures for gathering information. The evaluators concluded that the project responded to the information needs of the tribe and was conducted within the rules established by those requesting the study. They note that their approach (Tigges and Zastrow, 1981, p.12):

. . . was successful because evaluators frequently do find themselves in [a] social context different than the one in which they were trained to be evaluators, and occasionally, very different from the social context in which the concepts of social science research were developed. However, this difference is rarely accounted for in the evaluation they are asked to do. That is, evaluators are asked to do acceptable social science evaluations of programs whose participants are from a culture with basic assumptions different than those of Western science.

Some Native researchers and commentators have expressed very strong opinions about the way research should be conducted. They note, for example, that much of the qualitative work conducted is “presented from an outsider’s perspective” (Swisher, 1996):

If non-Indian educators have been involved in Indian education because they believe in Indian people and want them to be empowered, they must now demonstrate that belief by stepping aside. They must begin to question their motives beyond wanting to do something to improve education for Indian people. In writing about Indian education, they must now defer to Indian authors, or at least co-author in a secondary position . . . Much of what has been written is historically accurate and not harmful or offensive; it is sensitively, and in some cases beautifully, done. What is missing is the passion from within and the authority to ask new and different questions based on histories and experiences as indigenous people.

Even members of the American Indian-Alaska Native community have experienced challenges in conducting research among indigenous populations. For example, (Brayboy and Deyhle, 2000):

Bryan McKinley Brayboy (Lumbee Cheraw): I have worked hard to develop a balance between being a good researcher and a “good Indian” simultaneously. Many traditional methods of conducting research directly conflicted with my sense of being an Indian. [In one research location] where I knew many people and had relationships with the participants, the transition from being a friend and colleague to researcher was difficult. Conversations became more structured; I asked more questions, often had a notebook in my hand, and I watched the students through a different lens than before.

Donna Deyhle: Although [my] ethnic heritage is part Choctaw, [I] was raised in a European-American cultural environment and therefore speak from a hybrid identity . . . Although I could never become a complete “insider”—no matter how much I have been taught, I will never become Navajo—I have used my position . . . to become a “broker of sorts.” This has brought the “insider” voices and experiences of Navajo youth and their families to the “outside” as evidence in legal battles for social justice.

Content and Quality of the Existing Knowledge Base

Despite the many references cited in this synthesis, the field of early childhood education for American Indians and Alaska Natives is woefully understudied. Calls for more systematic and substantial research have been issued regularly.

In 1998, President Clinton signed Executive Order 13096, American Indian and Alaska Native Education, which set forth six goals: improving reading and math skills, increasing high school graduation and postsecondary attendance, reducing the influence of factors such as poverty and substance abuse that impede educational performance, creating safe and drug-free schools, improving science education, and expanding the use of educational technology. The Executive Order established an Interagency Task Force on American Indian and Alaska Native Education to oversee the plans and implementation of the order. The Task Force directed the U.S. Department of Education to develop a research agenda, and a Research Agenda Working Group fulfilled the assignment. General conclusions from the Working Group include the following: (1) detailed national data are needed, (2) definitions should be resolved by tribes, (3) research on AI-AN students should involve those with demonstrated knowledge of their cultures, (4) the quality of research should be of high priority, (5) generalizable findings are needed, and (6) a clearinghouse in Native education should be established. For early childhood education, the Working Group identified two priority research topics: (1) the status of infants and preschool-aged AI-AN children on school readiness domains and (2) the programs and services available for infants and preschool-aged AI-AN children.

Still, these calls for additional research have not yet resulted in a substantial research base. Obtaining materials for this document required digging deeply into sources outside the usual social science knowledge base, particularly into materials other than refereed journal articles, published books, and other types of commonly accepted social science resources. The debate about what constitutes scientifically valid research in the social and behavioral sciences is a long one and will not be addressed here, except to note that there may be competing expectations for what exactly constitutes valid research regarding AI-AN children. For some, the fact that findings make good common sense and resonate with experience may be sufficient; for others, systematic data collection and careful hypothesis testing may be necessary before they consider the findings valid.

In its current state, the literature on services for young AI-AN children relies heavily on qualitative methods, including personal histories, case studies, descriptive analyses, and ethnographic approaches. There are, however, a number of studies that use systematic measures and tools, suggesting that some researchers and AI-AN communities have found it suitable to use other forms of social science or education research practices, such as evaluations and standardized assessments. Among the research studies identified for this synthesis, a variety of data collection methods were used (see Appendix A). Two patterns are particularly notable in these methods: (1) to the extent that standardized tools or scales are used, they tend to be in the areas of language and literacy acquisition and assessment tools and practices; and (2) the majority of health-related studies do not entail original data collection but instead rely on existing records.

It is reasonable to ask why the research methods used in much of the literature on young AI-AN children tend to be those judged as less scientifically valid. Put another way, the question asks: Why are commonly accepted social science or education research practices frequently absent from the information base on young AI-AN children?

Based on materials obtained and reviewed for this synthesis, the following list summarizes the research methods and measures used in studies of young AI-AN children:

Sample sizes tend to be small. Most studies on learning styles, culturally appropriate curricula and practices, language acquisition, and assessment tools have relatively few study participants, ranging from as few as six to as many as a few dozen.
An exception to small sample sizes is found in health-related material regarding American Indian and Alaska Native populations. In the health field, several studies have used large extant databases or retrospective record reviews.
Most research studies identified for this synthesis use convenience or purposive samples. Very few have any sort of control or comparison group.
Studies that focus on reservation-based American Indian populations tend to be tribe specific; that is, most do not involve comparative studies across different tribal populations. Studies that are more general for the AI-AN population tend to draw on urban Indians and do not generally segment findings according to tribal affiliation.
Several studies have used standardized assessment instruments among AI-AN children, but the wide range of tools used (see Appendix B) suggests that there is not a consensus among researchers regarding the most reliable and valid instruments.
Some studies have examined the use of standardized assessment instruments among AI-AN children and have produced mixed findings regarding their appropriateness for the population. Several, for example, have looked at the WISC-R, which is a scale that measures intelligence. One study concluded that the WISC-R has low reliability and predictive validity for Navajo children, another found a range of bias from none to substantial for Papago children, a third showed significantly lower scores for Cherokee and Kiowa children on some subtests, and a fourth determined no significant differences between American Indian and Caucasian children.
AI-AN children tend to score “worse” than national norms on standardized measures. The current state of knowledge does not consistently and definitively explain why this is so. As a result, caution needs to be used in both (1) inferring lower levels of ability among AI-AN children and (2) adopting particular tools to use with this population.

The work conducted to date provides a wealth of rich detail essential to describe the conditions within which AI-AN Head Start programs operate, and they have provided good suggestions that programs may want to try in their efforts to better serve students. Some of the statements in these studies, however, may be too ambitious, perhaps driven by a commitment to enhance the knowledge base and do better for American Indian and Alaska Native people.

To be valid, research does not have to meet a rigid list of certain criteria, such as sample size, experimental design features, or quantitative measures. But it is likely that efforts to improve services for AI-AN children would be better served if the studies that have been (and are being) conducted are reviewed critically and subjected to the question: If the practice recommended in this study were followed, is there evidence to suggest that the investment in the practice would help young American Indian and Alaska Native children achieve better outcomes? Some findings are plausible; others may require more robust research before they are adopted.

Table of Contents | Previous | Next