Questionnaires


Download

The English version of the questionnaires administered by the MDICP can be downloaded in .pdf format below. Contact the project PIs if you would like versions of the questionnaires in the vernacular (Yao, Chichewa and Tumbuka).

 

Questionnaire Development

Final versions of the English-version questionnaire drew on:

(a) qualitative work conducted in the research areas, and

(b) questionnaires used in previous rounds of both the KDICP and MDICP.

The former included semi-structured interviews, focus groups, and/or fieldtesting of a previous version(s) of the questionnaire.

For the MDICP-1, the English-version questionnaire was translated into and administered in the local language (Yao, Chichewa and Tumbuka). The translation was done by supervisors who had first been trained by the PIs with respect to the objectives of the project and the meanings of the questions over the period of several days, and who had constant access to the PIs for clarification of a question's interpretation.

For the MDICP-2 and MDICP-3, the Eglish-version questionnaire was translated into the local language (Yao, Chichewa and Tumbuka), but the interviewers were left free to decide whether to administer it in the local language or in English. The experiences in the first round of the KDICP and the MDICP had led to the conclusion that there was no evidence that local language questionnaires were preferable in these contexts. Interviewers had difficulty reading the local language questionnaires, probably because of the three languages only Chichewa, the national language, is taught in school. The supervisors idicated that it would have been easier for the interviewers to have an English questionnaire, but in training to emphasize understanding the questions and to provide uniformity on any possibly ambiguous wording. More generally, it has long been recognized that the validity of survey research is largely a function of the extent to which respondents interpret the questions in the same way. The traditional approach to this problem in developing world surveys has been to use a standardized translation that, it is assumed, leads to a uniform interpretation of the question by the respondent. It is possible, however, that standardizing the semantic form of the question is not the best way to standardize respondents' interpretations since it ignores both local differences in language (eg. between sites) and also idiosyncratic variation in interpretations (ie. between individuals). Informed, well-trained, and highly motivated interviewers may be able to effect the interpretive standardization more effectively, therefore, if they are given more freedom to translate questions into local idioms. Indeed, if, as was the case in the Kenyan and the Malawian surveys, the interviewers for each site are chosen locally, then trusting interviewers' ability to translate into the micro-local idiom (which the urban supervisor is unlikely to know unless he grew up in that locale) can be a good way to increase the level of interpretive standardization and hence the validity of inter-respondent comparability.

All rounds of the MDICP also collected village-level data, by means of an improved version of the community questionnaire developed for the second round of the KDICP. The supervisors with primary responsibility for a certain village were instructed to ask a number of key informants these questions while walking around the village with the interviewers.

 

 
 

Quick links:

Malawi
Transfers
Survey data
Qualitative Projects
Data quality
People
Papers

last updated January 21, 2007

Please direct comments or questions about this site to sbignami@ssc.upenn.edu