Access Panel und Mixed-Mode Internet Survey

点击翻译按钮获取中文摘要

Access panels have found a remarkable spread in the last years. Therefore an overarching objective is to examine the access-panel approach to survey research for its suitability. Primarily this implies to test if sample surveys drawn from access panel frames make valid inferences possible. There are two major threats to the validity of such inferences: a) selfselection processes and b) mode/response effects.The 1st two project years focused on building-up a large probability-sampling based access panel and a detailed nonresponse analysis, to assess the quality of the recruitment sample and hence the quality of the resulting access-panel sampling frame. Now, the 2nd two years will continue these analyses on response propensity, interviewer effects, survey cooperation and follow-up survey cooperation, and will extend them to a thorough analysis of the ways access panels and genuine panels shape responses to survey questions (effects of panel conditioning). In addition panel attrition (loss of respondents over time) will be analyzed.As most access panels are “online” panels we have to take seriously possible survey-mode and response effects associated with this mode. Therefore a series of seven survey-mode experiments will be carried out to analyze how responses to survey questions are affected or shaped by survey mode, using in particular the telephone mode for comparison purposes.