The project explores how interactive features in web surveys can influence respondent behavior.
In August 2008, the LISS panel was presented a questionnaire on the effect of including motivational statements and follow-up probes on response length and response quality of the responses to open-ended questions. The questionnaire included four open-ended questions and several closed-ended questions about the current situation in Dutch society. Respondents to the internet survey were randomly assigned to one of the four versions of the open-ended questions. The control version simply displayed the open-ended question; no motivating statement was included with the question and respondents were not branched to a follow-up probe. In the second version the open-ended question included the motivational statement ‘This question is very important to our survey’, but similar to the control version, respondents were not asked a follow-up probe. In the third version respondents were asked the open-ended question and after they submitted their initial response, they were branched to a follow-up probe screen. This screen displayed the respondent’s answer to the initial open-ended question and a follow-up probe (e.g., ‘Is there anything else you would like to add?’). Respondents to the fourth version were also asked the initial open-ended question and then branched to a follow-up probe with a motivational statement. In addition, people who did not answer the initial open-ended question (or who provided an answer like ‘don’t know’ or other answers that were less than four characters) were routed to a nonrespondent version that displayed the text ‘Please provide a response. This question is very important to our survey” and repeated the open-ended question.