Targeted initial letters to longitudinal survey sample members: effects on response rates, response speed, and sample composition

This study draws on two previously-separate strands of methodological research, one on the effect of the content of letters to sample members on response rates, and one on targeted respondent communications. We test targeted appeals in prenotification and invitation letters in a panel survey, based on an experimental design which allows the testing of effects in different contexts. The targeted letter improves response rates for previous wave non-respondents and recent panel entrants, though these effects depend on data collection mode. Response speed is improved for recent panel entrants, particularly in the context of a mixed mode design.