Hovanitz C A, Jordan-Brown L
J Clin Psychol. 1986 Jan;42(1):100-8. doi: 10.1002/1097-4679(198601)42:1<100::aid-jclp2270420115>3.0.co;2-b.
This study was designed to assess the validity of four clinical scales with and without subtle items; and the relative validity of obvious and subtle items. Criteria were questionnaire, mental status, diagnostic, and drug treatment measures; subjects were 73 psychiatric inpatients. Across criteria, the results indicated that the validity of the full scales was not improved by the removal of subtle items; and with the exception of the PD scale, subtle items generally were found to be of similar convergent validity and possibly superior discriminant validity. These findings have implications for psychometric theory and test construction strategies.
本研究旨在评估四个包含和不包含细微条目的临床量表的效度;以及明显条目和细微条目的相对效度。标准为问卷调查、精神状态、诊断和药物治疗措施;受试者为73名精神科住院患者。综合各项标准来看,结果表明,去除细微条目并未提高完整量表的效度;除了人格障碍(PD)量表外,一般发现细微条目具有相似的收敛效度,且可能具有更高的区分效度。这些发现对心理测量理论和测试构建策略具有启示意义。