本研究之目的在發展一份「短式BASIC 量表」精神病傾向分量表之工具，期 能以最有效、最經濟，又可達到相同效益的方法對新兵進行心理健康評估，作到 「及早發現，及早預防」，提高心輔人員的實務評估效率，以增進助人工作的效 能。研究對象為海軍新訓中心1,749 名受訓人員，將「BASIC 量表」精神病傾向分 量表的36 題，以多向度Rasch 研究方法精簡為20 題，其中包括「精神分裂症傾 向」、「躁症傾向」、「憂鬱症傾向」及「自殺意念」等四個向度，其受試者分離 信度（person separation reliability）分別是.67、.63、.70、.70，並針對「短式BASIC 量表」精神病傾向分量表提供內容、結構、類推、解釋及本質效度證據。本研 究以實徵資料驗證柯永河等人（2001）根據多年臨床經驗所發展的「BASIC 量 表」題目之有效性診斷，且與全球人口罹患相關疾病之盛行率互相呼應，在臨床實 務上具有貢獻。此外，長短版量表二者之相關為.94，精神分裂症傾向分量表之相 關為.88，躁症傾向分量表之相關為.93，憂鬱症傾向分量表之相關為.89，自殺意 念分量表之相關為.84，均為顯著正相關。可見本研究發展之短式量表足以代替原 始量表，未來為增加受試者之填答意願，確保評估效度，可善用「短式BASIC 量 表」，進行心理健康之評估。

The purpose of this study was to develop “the psychotic tendencies subscale of BASIC Short-Form” (PT-BASIC-SF) as a mental health measurement for military. 1,749 students from Naval Recruit Training Center were recruited. The PT-BASIC-SF included four dimensions: schizophrenic tendencies, manic tendencies, depression tendencies, and suicidal ideation. We cut down 36-item psychotic tendencies subscale to form 20-item in using multidimension Rasch techniques. They were applied to assess model-data fit and to provide the multiple validity evidence of the PT-BASIC-SF. The person separation reliabilities of the measures from four dimensions were .70, .67, .74 and .57, respectively. In this study, the findings supported the study based on many years of clinical experience by Ko et al. (2001) and have a contribution in clinical practice. In addition, the significant positive correlation between length version and short version of schizophrenic tendencies, manic tendencies, depression tendencies and suicidal ideation were .88, .93, .89 and .84. The results indicated that the PT-BASIC-SF could be replaced with the original scale to increase participants’ motivation as a better way to assess mental health.