Research Instrument: How much is enough!

Postgraduate students in Mass Communication and other Social Sciences in Nigerian universities face a number of challenges in the course of their academic programme. From poor or nonexistent research skills to misplaced expectations of what postgraduate education is all about to system stagnation to misunderstanding between them and their supervisors, the list is endless. It is even more pronounced at the doctoral level. Recently, one of the doctoral students we have been managing since his programme started about three years ago came calling. This time, it was about his instruments of data gathering. The student is conducting a triangulation study and as such, he is using two methods for his data gathering – Survey and IDI (In Depth Interview). As a result, he is using two research instruments – Questionnaire and Interview Guide. We had advised him on what the instruments should contain but his supervisor disagreed and practically dictated to him what the instrument should contain. We had also advised him on how to conduct a pilot test for his study. Normally, one pilot test is enough for a study at every point in time. It should be enough to show how strong or weak the instrument(s) is (are). For this particular doctoral student, his supervisor insisted that he should conduct two pilot tests. He did and each time, the questionnaire was shown to be weak. Having seen the results, the supervisor asked him to make corrections and proceed to gather the main data for his thesis. Having looked through again his Research Questions (No Hypotheses), we advised him on how the questionnaire should look like. Like Professor Ralph Akinfeleye, a professor of Mass Communication, would say in his class, a good machine has no unnecessary parts. Same way, a good research instrument should not contain unnecessary variables. We advised the student as such. Half way through the fieldwork, the supervisor called him back to look through the Questionnaire again. He looked at the Questionnaire and insisted that the student should include additional 10 variables (questions) to the existing 23. To him, 23 questions (variables) are not enough for a doctoral study. This is what we will be looking at here. How much is enough for a postgraduate study? What determines the number of variables in an instrument?
Traditionally, like applied research, academic (basic) research has structure and pattern. Hardly would anyone conduct a basic research without research questions or hypotheses or both to guide. This is what determines the findings one would end up getting. After all, to have answers, one must have questions. It is these research questions or hypotheses that determine the questions a researcher would ask the respondents to answer. Let us assume that a researcher is conducting Survey for an academic study. He or she is expected to have research questions or hypotheses. Let us assume that such fellow has four or five research questions. It is these research questions that will determine the number of questions such fellow will have. At most, the person should be able to extract from each research question four questions (variables) that will be included in the Questionnaire. Where such fellow has four research questions, the number of questions in the Questionnaire would be 16. If need be, relevant demographic variables would be included to make it 20 or thereabout. The caveat –if the need be – indicates that except the inclusion of demographics questions would provide relevant insights to the study, they should be left out. It is, therefore, the number of Research Questions that determines the number of questions (variables) in a Questionnaire and not the programme of study, i.e. Ph.D. or M.Sc.
In Nigeria, there is no doubt that we have poor research culture. Members of the society, including big corporations and governments at all levels, don’t appreciate the importance of research to our society as much as those in the western countries. It always requires persuasion in its highest form to get respondents for a Survey and where they reluctantly oblige, it would be suicidal to bore them with too many questions that could make them abandon the Questionnaire midway. Why don’t you have a Questionnaire that is concise and straight forward but good enough to answer your research questions or hypotheses? After all, that is the purpose of your study.
Early this year (2017), we have analysed a Survey for a doctoral candidate who was investigating the correlation between advertisement of a milk brand and purchase. Because the lead supervisor insisted that the Questionnaire must have at least 50 questions, the candidate was asking irrelevant questions like “Do you have parents”, “Do your parents drink milk”, “Do they watch TV”, “Have they seen the so so advertisement”, among other silly questions. As a result she got silly responses from the respondents and it took her more time to get the silly responses than she should have spent.
Therefore, it is not the programme that determines the number of questions in a Questionnaire, it is the number of Research Questions.
© 2017. Richard Adeyinka Emmanuel

Please follow and like us: