柯氏性格量表效度概化之統合分析

A Meta-Analysis of KMHQ's Validity Generalization

本研究之目的乃在透過統合分析法（meta-analysis），對「科氏性格量表」過去34年來所累積的研究資料，做客觀、計量性的分析，以瞭解其效度概化的程度。經電腦程式Meta-HC分析後顯示：當以「科氏性格量表」作為人格上的區辨工具時，以在心理症病人的效度最高，其次依序為特殊族群、不同心理症病人、犯人；而當施測於生理症病人與男女之間時，其區辨力稍嫌不足。綜觀本研究的結果顯示：「科氏性格量表」的效度概化程度頗高，區辨力頗為理想，因此是一個值得進一步開發、研究以及推廣使用的人格測驗。

This research aims at objective and statistical analysis on the investigative data of Ko's Mental Health Questionnaire (KMHQ) in the past thirty-four years to get some idea about the extent of KMHQ's validity generalization. This research employs the computer program Meta-HC as the analyzing tool. The research reveals : Among the six independent variables， the weighted average effect size of the psychological-invalids is the highest， meaning that KMHQ's distinguishing capability being the strongest The next three independent variables are special clans， nonpsychological-invalids and criminals. However， the effect size on physical-invalids and sex only reach mid or low degree， showing that KMHQ's distinguishing capability on physical-invalids and sex is not high. To sum up， the utmost conclusion of this research is that， the extent of KMHQ's validity generalization is pretty high， its distinguishing capability is pretty idealized. Therefore， KMHQ is a personality test that deserves to be further developed， studied as well as popularly adopted.