Whitehead W E
Health Psychol. 1994 Mar;13(2):99-102. doi: 10.1037//0278-6133.13.2.99.
Stress researchers, in attempting to circumvent methodological problems in correlation studies, have resorted to analytical techniques that increase the likelihood of a Type II error. For instance, studies suggesting that stress plays no role in irritable bowel syndrome are at variance with a large body of clinical experience and may be in error. Several measurement issues are discussed, but the most significant concerns the way investigators deal with individual differences in chronic levels of stress and symptoms. J. Suls and colleagues, in this issue, recommend correcting for these individual differences before estimating the relationship between stress and symptoms, but this is illogical because most stressors are chronic--they are related to the social and economic circumstances in which people live. A more important confound, not considered by most investigators, is the contribution of somatization to stress assessment and symptom reports.
压力研究人员在试图规避相关性研究中的方法问题时，采用了一些分析技术，这些技术增加了出现II类错误的可能性。例如，一些研究表明压力在肠易激综合征中不起作用，这与大量临床经验不符，可能是错误的。文中讨论了几个测量问题，但最重要的问题涉及研究人员处理压力和症状慢性水平个体差异的方式。J. 苏尔斯及其同事在本期中建议在估计压力与症状之间的关系之前校正这些个体差异，但这是不合逻辑的，因为大多数压力源是慢性的——它们与人们生活的社会和经济环境相关。一个更重要的混淆因素，大多数研究人员没有考虑到，是躯体化对压力评估和症状报告的影响。
