本研究旨在探討「特殊教育學報」、「特殊教育研究學刊」和「特殊教育與復健學報」等三份國內主要之特殊教育學術研究期刊所刊證之研究報告，其統計方法的使用情形。研究結果指出，在三份期刊裡，計有380篇研究報告，其中356篇是以量化的方式分析其研究資料。在全部1104次的統計方法中，主要以描述性統計法使用的次數最多，其次以積差相關、變異數分析、事後比較及t考驗等所用的比率在百分之三十以上。使用次數最少的依次是線性結構模式、C統計法、重複量數和群集分析等。就統計學難易程度來分，初等統計法使用的次數最高，佔55%；中等統計次之，佔34.9%；高等統計使用次數最少，佔約10%。個別期刊間的比較是相似性多於相異性，其中以「特殊教育研究學刊」使用較多如多變項變異數分析、群集分析、典型相關、區別分析等複雜、高深的高等統計方法。本研究邁議將來可再進一步就特殊教育領域的學術研究報告所使用的統計方法是否恰當進行研究。

　　　　　The purpose of this study was to study the statistical methods employed in three major journals of special education in Taiwan: Journal of Special Education, Bulletin of Special Education, and Bulletin of Special Education and Rehabilitation. Totally 356 articles were analyzed. The results indicated that the descriptive statistics were the most frequently used statistical techniques (54.5%), followed by Pearson correlation (49.4%), ANOVA (44.4%), post hoc comparisons (31.2%) and t-test (30.9%). The least frequently used methods were structural equation modeling, C statistic, ANOVA with repeated measurement and cluster analysis. In terms of level of difficulty, 55% of the techniques were basic, 34.9% intermediate and 10% advanced. When comparisons were made between three research journals, the results indicated that Bulletin of Special Education used more advanced statistical techniques such as multivariate, cluster analysis, canonical analysis and discriminate analysis than two other journals. Future research is suggested to investigate whether the statistical methods used by researchers in speical education are appropriately employed.