Watson C G, Detra E, Fox K L, Ewing J W, Gearhart L P, DeMotts J R
Research Service, Department of Veterans Affairs Medical Center, MN 56303, USA.
J Clin Psychol. 1995 Sep;51(5):676-84. doi: 10.1002/1097-4679(199509)51:5<676::aid-jclp2270510513>3.0.co;2-l.
The concurrent validities of five self-report alcoholism measures were compared in a psychiatric hospital population using Diagnostic Interview Schedule (DIS) alcohol dependence module outputs as criterion. The Michigan Alcohol Screening Test was the most accurate identifier of alcohol dependence at some point in the subjects' life, while the CAGE interview best identified alcohol dependence in the previous year. The correlations of a quantity-frequency measure, the Clinical Signs checklist, and the Medical History questionnaire with DIS alcohol dependence module outputs were generally weaker. The data also suggested that cutting scores based on local norms may offer better hit rates in psychiatric settings than those proposed by the instruments' authors, but additional cross-validations are needed to confirm that possibility.
在一家精神病院的人群中，以诊断访谈表（DIS）酒精依赖模块的输出结果为标准，比较了五种自我报告酒精中毒测量方法的同时效度。密歇根酒精筛查测试是在受试者生命中的某个时刻对酒精依赖最准确的识别工具，而CAGE访谈则最能识别前一年的酒精依赖情况。数量频率测量法、临床体征清单和病史问卷与DIS酒精依赖模块输出结果的相关性通常较弱。数据还表明，基于当地规范设定的划界分数在精神科环境中可能比工具作者提出的分数有更高的命中率，但需要更多的交叉验证来证实这种可能性。
