Some methodological issues in preintervention farming systems research : selecting appropriate techniques for data collection

This paper, based on fieldwork in Ecuador, suggests an alternative approach: different methodologies should be used for different units of analysis. Specifically, regional variation can be tapped by analyzing secondary data, followed by judicious use of informant interviewing, while socioeconomic variation can be measured by random surveys with an appropriate questionnaire design. The paper first considers methods of identifying regional variation, discussing the preparation for and conduct of structured interviewing in subzones which have been identified through prior analysis of secondary data. Survey research techniques are discussed in the second section, which begins by considering the universe from which one might sample and proceeds to the design of questionnaires for specific social strata; this section argues that survey research allows one to identify with precision which social strata exist in a subregion. General methodological issues are analyzed in a concluding section which summarizes the strengths and weaknesses of informant interviewing and survey research, and identifies some implications for staffing and multidisciplinary collaboration. A 26-item bibliography (1922-84) is included.