An analysis of the re-test artefact in longitudinal studies of psychiatric symptoms and personality

It was found that the artefact is unrelated to the time lag between occasions, being equally strong at intervals ranging from 4 to 34 weeks, and consistent with both the 'mechanical responding' and 'social desirability' hypotheses of the re-test artefact.

Abstract

When general population samples are assessed on two occasions with psychiatric symptom or personality measures, a mean change in scores towards less psychopathology is often observed. This re-test artefact is a potential threat to the validity of longitudinal studies. Data from a longitudinal general population study were analysed to discover under what circumstances the re-test artefact occurs. It was found that the artefact is unrelated to the time lag between occasions, being equally strong at intervals ranging from 4 to 34 weeks. However, the artefact did not occur for all measures, but was confined to those assessing negative self-characteristics and administered orally by an interviewer. These findings are consistent with both the 'mechanical responding' and 'social desirability' hypotheses of the re-test artefact.