Questionnaire Length

Questionnaire length refers to the amount of time it takes a respondent to complete a questionnaire. Survey instruments can vary in length from less than a minute to more than an hour. The length of a questionnaire is important because it can directly affect response rates, survey costs, and data quality. Longer questionnaires result in higher data collection costs and greater respondent burden and may lead to lower response rates and diminished quality of response. However, practical experience and experimental data suggest that below a specifie threshold, questionnaire length bears little relationship to response rate or data quality and has only a minor impact on survey costs.

For telephone and online surveys, although longer questionnaires will result in higher data collection costs, it is generally recognized that questionnaire length is a major factor for these modes only when it exceeds 20 minutes. The field cost of increasing the length of a telephone interview 1 minute beyond the 20-minute mark is 2 to 3 times as high as the cost of adding another minute to a 15-minute telephone interview. This higher cost reflects the lower response rates associated with longer interviews (due to the effects of the increase in respondent burden) and the greater number of interviewer hours needed to achieve the same sample size. The cost of administering a Web survey is similarly increased when the length of interview exceeds 20 minutes. In this case, additional monetary incentives and rewards must be offered to prompt respondents to complete a longer interview. For mail surveys, questionnaire length has a small impact on field costs compared to phone and online surveys. Longer mail questionnaires may require more postage and data entry will be more expensive, but these cost increases are generally not as large as those for other modes of data collection. With in-person interviewing, respondents do not appear to be as affected by the length of the questionnaire as they do by the length of other survey modes.

Questionnaire length also affects data quality. For telephone surveys, long questionnaires get lower response rates than short questionnaires and may therefore be subject to more potential nonresponse bias. The length of a telephone questionnaire affects response rates in two ways. First, when interview length is specified in the respondent introduction or otherwise disclosed by interviewers prior to the first substantive question, the number of refusals tends to be higher for a longer questionnaire than for a shorter one. Second, the more time an interview takes, the greater the chance that a respondent will break off the interview because of disinterest, to attend to personal business, or for some other reason. Refusals and break-offs can be recontacted in order to complete an interview, but this adds time and cost to a survey research project. There is also evidence that both respondents and interviewers can get fatigued toward the end of a long telephone survey, possibly compromising data quality, for example, through an increased tendency to satisfice. Mail survey data quality can also be affected by questionnaire length through lower response rates, but there is no widely accepted threshold beyond which a mail survey's response rate declines significantly.

...