本研究是以感覺訊息處理與感覺統合理論為基礎，發展一套的本土化標準化評量工具，提供評量者全面、快速且容易解釋的評量工具。本研究進行兩個連續性的研究，研究一是預試以建構感覺統合功能評量表，針對126 位中部地區幼稚園及國小之一般兒童進行預試施測，以確定量表題目，形成正式量表。研究二是了解分量表之鑑別力、建立信度、驗證理論，以台灣北、中、南三區3～10歲兒童為研究對象，共發出1,600份問卷，回收八成，有效問卷為1,000份（62.5%）。預試題目是以傳統題目分析與WINSTEPS 分析；問卷資料以皮爾森積差相關、測驗訊息量、探索性及驗證性因素分析等方法進行統計分析。研究結果如下：1. 「感覺統合功能評量表」具有七項構念：姿勢動作、兩側協調動作順序、感覺區辨、感覺調適、感覺搜尋、注意力與活動量、行為與情緒等。2. 所編量表由七個向度，共98 題組成，有良好的信度、效度，應具有運用的價值。3. 七個分量表訊息量較大的範圍不同。4. 從模式基本適配度、內在結構適配度、整體適配度三項分析而言，「感覺統合功能評量表」之三階因素模式結構與實證的資料是適配的。

The foundation of this research is based on two main stems: Sensory information processing and sensory integration theory. Main aim and objective is to develop a standardized assessment tool and testify theory. Such a tool is expected to provide all users,with expedient and easily understandable final results.This research was segmented into two phases of serial investigative study: 1. The establishment of an sensory integration function assessment scale; main sample basis includes a total of 126 kindergartens and elementary schools children (demographics:
Central Taiwan). Pre-testing rounds were held during this phase, in order to finalize questionnaire content, also to formatting the assessment scale. Traditional item analysis and WINSTEPS analysis were used to analyze data. 2. During this phase, it was intended to identity
differentiate validity of the scale, in order to achieve necessary reliability.Questionnaire respondents targeted were children between the age of 3 to 10 years old (demographics: Northern, Central, Southern Taiwan). A total of 1,600 questionnaires were sent out, 80% of which was returned; effective response ratio is 62.5% (with 1,000 effective copies). Test information, exploration and confirmation factor
analysis were used to analyze data. Research results obtained are as stated hereunder:1.Sensory integration functions were investigated from seven different sub-scale, such as: postural movement, bilateral coordination and motor sequencing, sensory discrimination,sensory modulation, sensory seeking, attention and activity level, behavior and
emotion, etc. 2. Sensory integration functions have seven sub-scale, a total of 98 test items. There was strong support for firm reliability and validity, as well as possessing realistic applicable value. 3. There were range differentiations detected concerning test information amongst the 7 sub-scale. 4. From analytical perspectives of “preliminary fit
criteria”, fit of internal structure of model”, and “overall model fit”, research data fit triorder models of factors structure based on sensory integration functions.