Logical analysis of questions in a questionnaire

Data gathering techniques based on a questionnaire are very popular and abundant research experience in this field has resulted in many directives helpful in the construction of that research tool. From among such directives, relating to various problems, have been selected those concerning the formal aspect of questions. The formal aspect of questions is also of interest for logicians. Using N. Belnap’s logical theory of questions it is possible to derive a series of directives for the construction of formally correct questions. In the first part of the study directives formulated by sociologists and relating to a suitable construction of closed questions as well as directives derived from the logical theory of questions, have been listed and juxtaposed with each other. As a result of this confrontation it is possible to say that, in respect of formal correctness, sociological directives do not bring anything that has not been, so far, taken into account by logical rules. Sociological rules have the form of' instructions which are not considered as absolutely binding and in addition, they are unclear whereas logical rules define rigorously the criteria of formal correctness of questions. It seems, therefore, desirable to have logical rules' included in the set of principles to be followed in constructing a questionnaire. Such rules should be unconditionally observed, in order to secure formal correctness of questions asked from respondents. When using questionnaire-based techniques of data gathering, one should keep in mind that questions contained in the questionnaire are not the only ones involved in the research. Beside them, distinguished may be the researcher’s own questions and the problem questions. Using the rules of correspondence it is possible to pass, from sentences being the answers to questionnaire questions to those constituting the answers to researcher’s own questions; such rules are unilateral or bilateral reductive sentences. Two categories of indirect answers have been distinguished by logicians: (1) partial, and (2) full answers to given questions. The answers given by respondents are most often indirect answers to the researcher’s own question. Whether such an answer is partial or full depends on the structure of the rule of correspondence defining the role of the questionnaire question concerned. If this rule is a bilateral reductive sentence, the answer to the question introduced by this rule, is always an indirect full answer to the researcher’s own question. If the rule of correspondence is a unilateral reductive sentence, two situations are possible: (a) if the answer to the questionnaire question is the one foreseen by the rule of correspondence, it is an indirect full answer, (b) if the answer to the questionnaire question is different from the one foreseen by the rule of correspondence, it is an indirect partial answer to researcher’s own question. Realizing these relationships may help to avoid unjustified conclusions drawn from answers given by respondents. A serious weakness of sociological works stems from the fact that the above-mentioned rules which constitute an essential premise in passing from respondents’ answers, through answers to researcher’s own questions – up to the answers to the problem question – are never revealed