Measuring Life Events

The investigation of the etiologic role of life events on illness has been hampered by a wide variety of methodological issues. This paper represents an attempt to describe these issues in a systematic way and to test them using data from a large community survey. Each issue (scope of item content, multidimensional structure, confoundedness with dependent variables, objective-subjective scoring, desirability) is outlined and its effect on the relationship between life events and depressive symptoms is examined. The study concludes that even when different ways of evaluating life events are considered, the relatively small relationship to depressive symptoms cannot be improved substantially.