A forgotten part or generally ignored aspect of survey research is the preparation of the research instrument for suitable and efficient data entry and analysis. What ever is available appears to elude most researchers in the social sciences and humanities. The consequence is that these researchers, some of whom are authorities in their fields, frequently fall victim to poorly designed instruments, which can only be used to answer their research questions in the most extraneous manner. After many years of interacting with researchers from several areas of enquiry, and following urging from some of the researchers, I decided to write this paper as a sort of "where there is no doctor" type of paper. The problems discussed are illustrated using actual questionnaire items from a wide range of instruments, which I have been involved in at one stage or another. The main focus is how to turn questionnaire items into suitable data for data entry and analysis. Other issues discussed include understanding data handling in statistical packages, especially in relation to the traditional classifications of variables that most empirical researchers are familiar with. The paper highlights challenges in designing questionnaire instruments that are relevant in investigating the research hypotheses in social and behavioural enquiries.
Read
