Verzuimmeting via zelfrapportage en registratie: verschillen tussen de Nationale Enquête Arbeidsomstandigheden en de Nationale Verzuim Statistiek
article
Research to establish (work related) risk factors of sickness absence is usually restricted to self-reported measures of sickness absence. When risk factors and outcome measures are both collected with self-administered questionnaires, however, this may suffer from monomethod bias. In the current research, absence was measured both with self-report (Netherlands Working Conditions Survey), as with registered measures (National Absence Registry), among the same individuals (N = 9.893). On average, it appeared that self-reported absence was significantly higher than registered absence (4,6 vs 3,8%), but both measures also appeared to be consistent (relatively high on absence according to one source was also relatively high according to the other source). Apart from some exceptions, analyses using self-reported absence or registered absence reached the same conclusions regarding risk factors for sickness absence and the question in which subgroups sickness absence is relatively high. Therefore, monomethod bias appears not to be a major problem when analyzing sickness absence and (work related) risk factors of sickness absence, when both are measured with self-administered questionnaires. Self-reported sickness absence in research based on large-scale probability samples therefore appears to be a reliable alternative to registered sickness absence.
Topics
TNO Identifier
573493
Source
TSG: Tijdschrift voor gezondheidswetenschappen, 88(2), pp. 71-78.
Pages
71-78