ESRA 2027 sessions by theme

Back to Overview of Sessions

Session Details

In an era of declining response rates, it is crucial to design a survey that is as lightweight as possible for the participants. One method to reduce survey burden for respondents is to reduce the survey length by relying on methods of planned missingness, in which individual respondents receive only parts of the questionnaire. However, since respondents are not asked all questions, their answers to the omitted items remain unknown, resulting in planned missing values. This planned missingness could stem from a split-questionnaire design, where the assignment to some questions is random, or from other designs, such as adaptive designs, where parts of the planned missingness depend on prior information about the respondent, such as their previous answers. This session is dedicated to such planned missingness designs and to studies answering questions like:

(1) How effective are individually planned missing designs in reducing survey burden and survey nonresponse and nonresponse bias?

(2) How should surveys conducted with this design be handled, given their (often large portions of) planned missingness, in different sets of variables (e.g., metric, ordinal, or multinomial variables) in a survey setting with often limited sample sizes, weak relationships between variables, or a large number of variables? Methods might include imputation, weighting, FIML, or alternative approaches for analyzing the data with planned missingness.

(3) How should a survey conducted with planned missing designs be made accessible and usable to data users?

Contributions addressing these or related questions in survey research in and outside the social sciences are welcome in this session.