This article examines the construct validity of the Occupational Stres
s Indicator (OSI) in its Chinese version as well as its criterion-orie
nted validity against self-reported data on absenteeism and accidents,
using a final sample of 1054 workers in the Chinese manufacturing and
power industries. The divergent and convergent validities of five of
the OSI scales-Job satisfaction, Physical ill-health, Mental ill-healt
h, Type A behaviour and Control-were examined by a multimethod-multitr
ait matrix technique. Results indicated that the Chinese OSI had fairl
y good validity in relation to other well-established measures as well
as relevant organizational behaviours.