Deciding on a sampling procedure for a study on understanding teaching and learning relations for minority children in Botswana classrooms.

Abstract

Sampling is an important component of the research process for a study and it hinges on other aspects of a study such as its external validity, or comparability, translatability or totalization. When dealing with qualitative data, which describes or tries to understand social phenomenon, it is non-probabilistic in its representation of the universal world beyond its boundaries. This paper explores sampling procedure in a study on understanding teaching and learning relations for minority children in Botswana classrooms. The sampling procedure followed uses the maximum purposive sampling technique to understand teacher-learner relations ”grounded” in the four different school sites across the country representing a range of language, social and cultural identities of the San. This has been necessary given the diverse composition of the San population in terms of its social ecology, culture, and linguistic make up. The on-going or sequential and progressive process of the study uses comparative case studies from the four selected cases of schools to give a more in-depth understanding and description of teachers and learners interactive classroom relations. However, given the inadequacy of these sampling procedures in understanding of the multilayered, superfluous, relations of teachers and learners coming from different socio-cultural backgrounds in an institutional setting, a further suggestion is made to use ethnomethodology, conversation and discourse analysis in the sample of cases for the study in order to treat each case in the study as distinctive, deserving its own analysis. At the same time the researched will be given more pronounced role to construct reality from their own contexts and not that of the researcher’s.

By

Lucky Daniel Tshireletso

Institute for Educational Research. Faculty of Education,

University of Oslo. Email:

Paper prepared for the First NETREED Conference on Doing Educational Research in Developing Countries- the research agenda and appropriate methods.

Rainbow Gausdal Høifjellshotel near Lillehammer. 7th to 9th January 2002.

Deciding on a sample for a study is one of the crucial stages of the research process and stands to influence generalizability or external validity, which can also be described as comparability, translatability or totalization of the study results. Sampling is defined by Merriam (1998:60) as the selection of a research site, time, people and events in a field research. According to Le Compte and Preissle (1993:57) sampling has to do with representation of individuals and subsets making up the population group from which results can be generalized. However, how to achieve generalizability from study results is also dependent on the nature of data needed for the particular study as well as the research design. When using quantitative data, sampling has implications for probability and statistics whereby the sample subset is calculated to have a high probability or chance of representing the entire population group. There is emphasis on empirical data obtained in an experimental setting and generalization of results is to bring element of prediction future cases as result of proving or disproving the hypotheses of the study. Whereas qualitative data mostly describes, or tries to understand social phenomena it is non-probabilistic in its representation of the universal world beyond its boundaries. Generalization of research results is achieved through comparability, translatability or totalization of the sample to similar instances or cases occurring elsewhere in the social universe. According to Merriam (1998:62) the number of participants in a sample depend on questions being asked, data being gathered, the analysis in progress, the resources available to support the study, and so on. This paper explores sampling procedure in a study on understanding teaching and learning relations for minority children in Botswana classrooms. We shall firstly consider the questions or focal areas of concern for the study, and the kind of information or data required to address the stated questions or concerns, as well as the analytic procedures that maybe necessary during the on-going process of the study.

In understanding teaching and learning relations of minority children in Botswana classrooms the study intends to make an in-depth description of existing teacher education provision in the Botswana setting with particular reference to how it addresses teaching and learning conditions for these children. The study will also make an understanding of concepts such as globalisation, culture, the state, especially as they relate to education provision for San or Basarwa minorities who are the focal population group for the study. The objectives of the study are to:

1)  describe and analyse the socio-cultural context of the San or Basarwa children by describing their home and learning environment.

2)  refine roles and relationships of teachers and learners in the classroom setting, and the roles of the state, colleges of education, the university and local schools in initial teacher preparation.

3)  redefine and recontextualise education (i.e. bringing education back to the context) through roles new of teachers and learners; new roles and possibilities for colleges of education, the university and local schools in initial teacher preparation for minority children and redefine the role of the state in the context of local and global relationships.

From the listed set of objectives for the study, we can then determine the attributes of the data units for the study. According to Le Compte and Preissle (1993:61),data units determine population needed for the study by addressing the problem of where, when and how such units maybe found and they also determine selection and sampling procedures to be followed by the study. In addition to being confined to populations, study units can be artefacts (which include written documents, symbolic forms of interactive communication, and so on), events or contexts. School sites in which Tswana teachers and San or Basarwa learners commonly interact will provide a bounded context in which to describe and understand the relationship. For each of the school cases, roles and interactional relationships of Tswana teachers and San or Basarwa learners in classrooms; the socio-cultural environment and ecological space of San or Basarwa minorities; the initial teacher preparation program at colleges of education, the university and schools’ based initial teacher preparation; and the state policy and role in initial teacher preparation, will be the units of analysis. However, given the wide distribution of schools across the country which largely cater for San or Basarwa children, as well as the heterogeneous nature of the San or Basarwa population particularly in terms of language, social and cultural identity, a maximum purposive sampling strategy will be used in the selection of the case studies to allow for comparison between the different sites. According to Merriam (1998:61) purposive sampling emphasizes on a criterion based selection of information rich cases from which a researcher can discover, understand and gain more insight on issues crucial for the study. Le Compte and Preissle (1993:69) observe that criterion based sampling is used at two levels in ethnographic studies, firstly to provide a set of criteria or list of attributes that the study unit is to possess and secondly to determine as the study progresses new set of phenomena. In applying the maximum purposive sampling technique consideration is made to understand teacher and learner relations in the classroom ”grounded” on wide variation of school sites across the country representing a range of language, social and cultural identities of the San or Basarwa who prefer to link their identity and livelihood closer to the primordial or traditional sense and those who do not. The second, on-going or sequential and progressive process of the study is the comparative case studies which will involve collecting and analysing data from the four selected cases of schools in order to understand and describe influence of socio-cultural background of teachers and learners to their classroom interactive relations and how this can inform initial teacher preparation programs in Botswana. Merriam (1998:40) goes on to observe that the comparative case studies strategy enhances validity and generalizability of research findings because they are grounded, specifying the how, the where and the why of the on-goings of a phenomenon.

However, generalization or universalization of data that are descriptive and have the intention of explaining or understanding socio-cultural relations in particular contexts as is the case with this study is always problematic. This is true in the sense that data or information on socio-cultural relations in the context of a classroom is multi layered and not static. Baszanger and Dodier (1997:11) questions the logic of generalization of ethnographic research from the different approaches it follows. They find monographic totalization that can be reached from an integrative ethnographic approach as lacking in its use of one context to claim universality of culture to other contexts. This is particularly so because conditions of culture differ from one context to the next even within institutions such as schools. From this approach they also question the validity and reliability of the researcher as a neutral instrument of research as he or she shapes the epistemological frame of the study. This is argued from the position that the researcher is likely to represent his or her own knowledge background in a study. They also question the validity of empathetic relations that develop between the researcher and the ”other” as the basis for reconstructing the culture of the ”other”. Karter and Delamont (1996:xi) likens the researcher’s role as more like that one of writing his or her own biography through a research, and with that researcher’s role they strongly question the validity and reliability of such a project. In addition Baszanger and Dodier (1997:11) find it difficult to say if such a relationship can be relied on for a transparent communication of the ”other” self to the researcher. The studying of culture as a hermeneutic process, whereby an individual’s activities and daily relationships with others can be recorded in the form of reflexive narratives, although powerful in articulating the emic perspective is again questioned by Beck (2001:299) who indicates that it is not possible to understand the individual as a separate entity from his or her social context. She also notes Wersch’s interpretation of Vygotsky, observing that cultural traditions of researchers themselves can be a constraint on their research. This highlights the significance of understanding the first the social contexts as a basis for understanding social relations and behaviours of individuals. How then can we best understand multi layered, superfluous, relations of teachers and learners from different socio-cultural backgrounds in an institutional setting? At this point Lahdenpera (2000:206) notes the inadequacy of a single individual approach in the study of multicultural contexts since it studies phenomena from interethnic, intercultural viewpoints where different cultural perspectives meet and complement each other. Miller (2000:26) goes on to suggest ethnomethodology, conversation and discourse analysis as distinct standpoints from which concrete aspects of social life may be seen and analysed. According to him ethnomethodology, conversation and discourse analysis go beyond treating a case study as only interesting and informative when compared and contrasted with other case studies but treats every case as distinctive, deserving its own analysis. At the same time consideration is made of instances of reality construction in each case as involving knowledge about transsituational skills, and issues which to some extent are context free. This implies that for this study generalization will not only be based on a comparative analysis of data, but also by an ethnomethodological analysis of the role and influences of the local and the global in some research settings, analysis of conversations and the discourse of classroom interactive behaviours of teachers and learners from the different socio-cultural backgrounds. It is hoped that the further analysis will enrich the grounded data of each selected case study within its own context.

In a way data in the study will be accounted for in the realist and reflexive forms of ethnography. This position is necessitated for by the epistemological understanding of the socio-cultural phenomenon from two views: firstly as something that can be factually accounted for through observation of interactional behaviour of individuals in a community and institutional setting, and secondly as something that is constructed by individuals continuously at different levels of their social lives and can best be understood by a researcher reflexively. Luttrel (2000:504) found herself using a similar approach in her study of white and black working-class women in the United States, although she observes that she experienced tension between reflexive and realist research during the process of her research work. She goes on to observe that given the dearth of what can be termed as ”good enough” research methods in this kind of studies, one possibility is for the researcher to accept rather than defend the healthy tensions of fieldwork. By doing this she suggests that the researcher needs to accept his or her blind sports, social, emotional and intellectual involvement in the study by noting these down as part of field experiences. Furthermore to bring an element of control on the researcher and his or her influence on the study, Morse (1994:24) recommends for the researcher to use what is described as the audit trail, whereby he or she can use another researcher to follow the cognitive development of a project as it develops. In this particular study, the audit trail used will involve in part a peer review of the study process at the pre data collection stage as well as a reflexive review of data by the research participants during the data collection and field analysis stage.

Finally, although in each case study, data will be in-depth at both factual and reflexive levels, the comparative phase of the study will attempt to synthesize the stories, experiences told and recorded, to give a description of typical patterns or behaviours emerging from the case studies. Morse (1994:24) describes this process as decontextualization of data, as it is removed from the specific and particular instance, of a case study, to the abstract and more generalized description. The next level is the theorizing process whereby abstracted data has to be linked with established theory through falsification and verification, revision and discarding of non-relevant information. At the end of which there is the recontextualization of data in which emerging theory can be applied or generalized to other settings. During the process the researcher has to relate the emerging theory with established knowledge by identifying findings that support established knowledge, or refute it or claim new contributions to existing knowledge. In this way the study findings can be claimed with certainty to have achieved levels of generalizability or comparability.