Role stress—a methodological trap?
研究发现，许多角色变量与压力症状之间的中等相关性可能是由于自变量和因变量在操作上相似造成的；通过清晰分离两组变量，角色与压力之间没有关联。
Abstract This study explores the notion that many existing studies which show moderate relationships between role variables and stress symptoms are doing so because of the operational similarity of independent and dependent variables. It was predicted that the removal of such an artefact by clearly separating the two sets of variables would result in a nil relationship. Matched ‘cases’ and ‘controls’ were compared in two different organizations. The independent judgments of occupational health specialists were used to measure stress on a number of agreed dimensions, and role measurements were taken by the researchers in three different ways. On all comparisons made there was a strong trend suggesting no role/stress associations. The implications of the results for role stress studies are discussed.