Conference Paper International Workshop on Household Survey Nonresponse
Response patterns in mail panel surveys of young people
12 Sep 2001
This paper will present findings relating to a number of aspects of response and non-response on mail surveys of young people in Scotland. The main questions addressed are the following:
1. What are the characteristics of various categories of respondents and non-respondents?
Specifically, respondents can be divided into groups defined by the number and nature of contacts required to elicit a response, while non-respondents can be divided into those who explicitly refuse, those with whom no contact is made and those for whom the reason for non-response is unknown. Some frame information is available for all sample members and can be used to compare all categories of respondents and non-respondents. This information includes the number and level of school examination passes, gender, type of school attended and region. The various categories of sample members will be compared in terms of these variables and the results related to the emerging theories of the survey response process.
2. To what extent is response behaviour consistent over time?
The data are from panel surveys, so we can compare behaviour at different waves at the individual level. This allows us to assess, for example, to what extent people who respond promptly on one occasion are likely also to respond promptly on another occasion. The findings may be of interest from a theoretical perspective (viz. 'situational' non-response vs. 'pre-disposition' to response) but are also of practical value in helping to predict response rates to panel surveys (based on knowledge of previous response behaviour) and to target response elicitation strategies.
The data come from the Scottish School Leavers Surveys, which are large repeated panel surveys of people aged 16 to 23 in Scotland. Some trends in the response rates to the surveys over the past decade will also be presented as context for the analysis.