The unit which is the focus for investigation obviously has an important impact on the results of the investigation, and these resuls are further influenced by the concepts and theories which the investigator employs in formulating and carrying out his work. The choice of units for study and other conceptual and theoretical considerations has further implications, and among these are the ones directly pertinent to area studies and their status and value. A brief discussion of this point will make it clear.

In classic structural-functionalist studies the unit for study, or within which the study is to take place, is patent. The anthropologist sets out to study the Ganda, the Yoruba, the Rjonga, or some other tribe. Just what constitutes a tribe may not be, and need not be, meticulously defined (and with such groups as the Ibo, Chagga, or Balante it will not be sharply delimited) ; but it does provide a guide for the smaller-scale unit (a village, a lineage, or whatever) within which the collection of data is actually carried out. Sometimes data taken from specific kinship or territorial units are used to make generalizations about the tribe as a whole, as when there are studies of witchcraft among the Pogoro or economic organization among the Pangwa, and sometimes the studies concern characteristics of the unit's organization, as when investigations are made of Rjonga lineage structure or Chagga settlement patterns. In all cases the boundaries of investigation are taken as given, and what is variable and to be examined is what goes on within those boundaries.