Abstract
A family of models for memory effects of answers to retrospective questions is discussed. These models can be used when a respondent is asked how many times he undertook a certain action in a time interval T. The models take into account that the memory effect may depend on the elapsed time since the actions took place and the number of actions that have taken place. The theory is illustrated by examples from the 1981 Dutch Health Survey.

This publication has 0 references indexed in Scilit: