A portable datalogger to evaluate recall-based time-use measures.

Abstract

Self-completed recall diaries have become a common tool in epidemiology and exposure assessment to determine the location and/or activities of subjects during study periods. Unfortunately, little effort has been made to determine the accuracy of such an approach for providing information without significant bias. It is usually assumed that subjects are truthful and complete in their recall. An electronic datalogger has been designed to allow subjects to record location changes as they occur in real time. Subjects carried the datalogger for 3-5 days and completed recall diaries at the end of each day. The concordance between the two records was found to be relatively good for the most commonly visited locations, e.g., bed/bath, workplace, etc. Poorer agreement was found for locations of short or infrequent visits, e.g., cellar or vehicle. The merits and shortcomings of the datalogger approach are discussed. While having external validation for recall diary records is clearly necessary, reliance on subject input to the datalogger presents its own problems related to subject compliance. Electronic approaches which do not require active subject input are recommended to aid in the evaluation of recall diaries.