Response Preference as a Function of Instructions in an Unstructured Questionnaire

Abstract

In responding to a questionnaire containing no questions but merely indicating possibilities, people show clear preferences for certain options over others. The present study confirms earlier findings indicating that this response bias is due mainly to preference for positive options over negative or neutral options. Data indicate that instructional content may affect subjects' preferences. Three of 10 items showed such an effect, i.e., a more balanced distribution was reached, leaving order of preference unchanged. On only one item a reversal of preference was obtained by varying the type of response mentioned in the instruction.