觸控式電腦化螢幕上肢訓練系統之建構效度與活動難易度分析

觸控式電腦化螢幕上肢訓練系統是一套特別針對中樞神經損傷患者合併上肢動作障礙而設計的上肢功能復健設備，本系統具有任務導向及動作結果立即回饋等諸多特點，可補償Wii、Kinect等遊戲軟體之使用限制。本系統活動之設計與分類是依據研發者的專業知識與經驗來編排，並未有實證資料支持，因此本研究之目的除驗證上肢功能訓練活動內容的建構效度，同時也透過羅序分析方式將活動難易度排列，我們預期活動難度可涵蓋不同能力之個案。共60位單側偏癱腦中風受試者以患側上肢隨機進行本系統的第一大類及第二大類活動，總計16個項目。每個項目選擇最簡單的活動進行1分鐘，並自電腦化上肢訓練系統記錄1分鐘內可正確完成的動作次數。資料分析方式使用因素分析驗證此二類別上肢訓練活動之建構效度。羅序模式分析以Winsteps 3.75版將此二類之活動難易度排列，發展個案能力與活動難易度的配對圖。研究結果顯示，在類別一的11項上肢動作及肩關節穩定度活動中，皆具有一個共同的主要因素，因素解釋變異性達66.72%，我們將此因素命名為上肢動作。在類別二的5項上肢伸手活動中，也只具有一個共同的主要因素，因素解釋變異性達87.04%，我們將此因素命名為上肢伸手。類別一的項目難度階層中，最簡單的活動是項目7區塊填滿，最困難的是項目6抬手平撥。類別二中最簡單的活動是項目11固定伸手指物，最困難的是項目13來回目標向肩。本研究結論是本系統所提供的上肢訓練活動具有良好的建構效度，顯示本系統之活動設計與編排符合其分類方式。此外，本系統之上肢功能訓練活動的難度階層涵蓋各種不同能力之個案，因此可做為臨床使用者選擇活動時的參考。

關鍵字

腦中風；虛擬實境；上肢復健；羅序分析

並列摘要

The computerized touch-screen system is a set of upper limb motor rehabilitation equipment and specific to central nervous system injury patients with upper limb motor impairment. The system has a task-oriented and action results immediately feedback and many other features, it can compensate for limitation of game software like as Wii and Kinect. The activities design and classification of system is based on the expertise and experience of developers to arrange, did not have the empirical data to support, so the purpose of this study is not only to verify the construct validity of the activities, but also through the way of Rasch analysis to arrange item difficulty, we expected activities can be difficult to cover different abilities of cases. The sixty hemiplegia stroke subjects used hemiplegic side of the upper limb to randomly execute the category I and II of the system, total are sixteen items. Each item selected the most simple activity for one minute, and recorded the correct number of execution within one minute from the system. Data analysis used of factor analysis to verify the construct validity of these two category activities. The Rasch model analysis arranged the item difficulty of these two category activities with Winsteps 3.75 version, and developed the person ability and item difficulty map. The results showed that the 11 items of arm movement and shoulder stability in the category I, all have a common main factor that the factor variance of explained is 66.72%, we named this factor arm movements. In the category II, 5 items of arm reaching only have a common main factor that the factor variance of explained is 87.04%, we named this factor arm reaching. In the item difficulty hierarchy of the category I, the easiest activity is item 7 Blocks Painted, and the most difficult is item 6 Raising Hand and Horizontal Move. In the category II the easiest activity is item 11 Reach to Fixed Target, and the most difficult is the item 13 Back and Forth Target to Shoulder. In conclusion, the activities provided by the system have good construct validity, display that the activity design and layout is conformed with the classification of the system. In addition, the item difficulty hierarchy of the system covered a variety of cases of different abilities, so it can be used as a reference when the clinical user to select activities.