Reliability and validity of self-reported duration of computer use

The aim of the present study was to investigate the reliability and validity of self-reported duration of computer and mouse use at work. Data of the PROMO project were used, in which an internet questionnaire was administered in about 2000 computer workers. Questions were asked on risk factors in computer work, such as ergonomic and psychosocial factors. Duration of computer use and mouse use at work was reported in 7 categories. To investigate reliability, questionnaires were administered twice, with an intervening period of two weeks. To investigate validity, exposure to computer use was measured using the software program WorkPace (Niche Software Ltd/ErgoDirect) as a gold standard. Test-retest reliability of the self-reported question on the duration of computer use was high: kappa was 0.85 for computer use and 0.75 for mouse use. Regarding validity, the mean computer use durations measured by the software program statistically significantly increased with increasing categories of self-reported computer use. However, standard deviations within a single report category were large. After categorising the software data into categories that were comparable with the self-reported data, kappas were below 0.10. Workers tended to overestimate the duration of computer use with 2.5 hours per day, but this overestimation did not occur systematically. We conclude that the test-retest reliability of the question on duration of computer use was satisfactory, but that the validity was low. It is recommended to use objective ways of exposure assessment, such as software programs, to estimate the duration of computer use.