溝通與互動技巧評估工具---中文版信效度探討

Abstract

研究背景：目前本國缺乏以職能治療為理論依據，並具有良好心理計量特性之標準化社交技巧評估工具。溝通與互動技巧評估工具(the Assessment of Communication and Interaction Skills, ACIS)為一觀察型測驗，具有良好建構效度、信度及區辨力。但尚未使用於本國精神科患者，缺乏中文化版本及其心理計量特性驗證。 研究目的：檢驗中文版溝通與互動技巧評估工具應用於台灣精神科患者之信、效度。 研究方法：本研究的對象為北部和東部共四所醫學中心之日間病房之住院精神病患者，診斷根據DSM-IV標準診斷為精神分裂症及情感型疾患、教育程度具有國小畢業程度以上。藥酒癮患者，由於牽涉到腦傷影響，因此本研究不予收集。總共收案101位。個案經說明解釋研究內容及目的、同意並簽署同意書後，進行簡短式智能測驗(MMSE)，測驗得分大於24分以上者，繼續進行觀察與相關量表測驗。本研究使用ACIS-C評估個案參與合作性團體活動及標準化會談時，其社交與互動技巧之表現。以此檢驗ACIS-C之建構效度、區辨效度與再測信度。ACIS-C測驗結果並進一步與職能治療綜合評量表之社交行為、魏氏智力兩分測驗進行相關性分析，以驗證同時效度。研究資料以WINSTEP 3.56版本進行羅序模式分析，並以SPSS11.5版統計軟體進行描述性及推論性分析。 研究結果：101位個案平均年齡為38.56歲。55.4%為女性，44.6%為男性。診斷77.2%為精神分裂症，憂鬱症和躁鬱症佔少數。婚姻狀況未婚居多(70.3%)。教育程度以高中到大學居多。97%個案無全職工作。羅序分析結果顯示ACIS-C題項，除第七題外，其餘項項目符合羅序模式假設為單一建構效度。ICC=.93，而加權kappa值介於.365~1間，顯示再測信度良好。項目分離係數為.96，顯示項目內在一致性佳。測驗難易度可涵蓋所有個案不同程度的社交技巧。個案分離係數為4.64，可將個案分成至少6種不同程度之能力。項目難易度順序，最簡單為詢問，最困難為建立關係。 討論：中文版溝通與互動技巧評估工具經過嚴謹的翻譯過程，經過專家意見修正，建立中文版，適用在本國精神科患者以驗證其信效度，結果顯示中文版溝通與互動技巧評估工具具有適當之信效度，並可區分不同程度之個案能力，可供臨床治療師評估個案社交技巧問題。第七題發音不符合羅序模式假設，可能是因為藥物副作用影響個案口腔動作。臨床設計訓練活動，建議可先從肢體動作臉部表情開始，循序漸進地改善個案之社交技巧缺失。本研究發現可能因慢性化過程殘餘症狀趨於穩定，因此無法顯現台大症狀量表和憂鬱量表與ACIS-C之相關。未來研究應建立施測者間信度及增加團體觀察機會。

Background：There is a lack of standardized social skill assessments which are developed based on occupational therapy theories with good psychometric properties in Taiwan. The Assessment of Communication and Interaction Skills (ACIS) is an observational rating scale with acceptable construct validity, discriminative validity and reliability. However, the psychometric properties of the Chinese version are unclear. Objective：The purpose of this study is to examine the reliability and validity of the Assessment Communication and Interaction Skills---Chinese version (ACIS-C) with psychiatric patients in Taiwan. Methods：The subjects were recruited from 3 day care wards in northern and eastern Taiwan. The minimal educational level is elementary school. Exclusion criteria for the subjects included diagnoses of organic brain injury and substance abuse. We recruited 101 patients. After explaining the process and purpose of the study to patients and obtaining their informed consent, each subject was tested on the Mini-Mental State Examination (MMSE). Only patients whose scores on MMSE were equal or larger than 24 were recruited. After securing the permission of the patient and ward staff, the researcher observed subjects’ performance during group session and a simulated social situation. The ACIS-C was used to assess the communication and interaction skills of the patients. Construct and discriminative validity as well as test-retest reliability were examined through Rasch measurement model and related inferential statistics. Results: The mean age of the subjects are 38.6 years (SD=11.5years). 55.4% are female and 44.6% are male. The majority of the subjects are diagnosed as schizophrenia (77.2%). 70.3% of the patients were unmarried. Most of the educational level is at college or above. 97% of the subjects were unemployed. The results of the Rasch analysis showed that the items of the ACIS-C form a unidimensional construct except for one item (item 7). The test retest reliability is good (ICC=.93). The internal consistency of the scale is .96. The ACIS-C is a sensitive scale which could be used to differentiate social skills among psychiatric patients. The easiest item is ”ask” and the hardest item is”relate”. Discussion：The ACIS was translated into Chinese through careful two stage translation process. The ACIS-C has acceptable construct validity, discriminative validity and test-retest reliability. The reason for misfit item could be attributed to side effect of psychotropic medications. Therapists can utilize the level of difficulties of ACIS-C items from the study to design matching training programs. For example, a therapist could start the training focus from easier item such as body movement, then progress to harder items such as relates and respects. The results of the study also showed that the severity of symptoms did not influence the patients’ social skills. Future study needs to emphasize inter-rater and intra-rater reliability of the ACIS-C as well as ACIS-C’s validity in different context.