本研究目的在於以SAS語言設計Erlebacher(1977)法之基本與應用電腦 程式，該分析法可應用於偵�alJ實驗設計是否干擾實驗處理。一般而言，運動心理學及行為科學的實驗設計之基本型態大致可分為組間(between-)及組內(within-subjects)兩大類，假若使用不同類型的設計而得到不同的實驗結果， 設計類型即干擾了處理效果，可能因而導致謬誤的結論。一九七七年，Erlebacher將實驗設計(組間與組內)以一因子(factor)處理，而實驗處理視 為另一因子，發展出該法之統計參數期望值與F值之誤差項。但是其計算繁複費時，電腦程式化方能使其普及使用。而今該電腦程式尚未問世，基於SAS具備F分配機率功能(probabilityfunction of F distribution)並為其所獨 有，並備有基本程式語言指令及樣本統計功能，故以其書寫本程式。程式之設計依三個步驟流程進行:一、撰寫程式架構，二、依程式架構撰寫程式，三、按程式段落逐步測試直至最後。程式架構大致可分四部分:←)設定輸出格式，ω設定資料步驟，(;)設定程序步驟，�明結束SAS程式。其中最主要的資料步驟大致可分鳥五部分:1.程式基本參數之輸入，2.細格(cell)資 料基本統計運算，3. Erlebacher法運算， 4. Erlebacher法各效果F值之機率 (p)算取， 5.原始資料之輸入(程式內資料檔時)。當資料從外輸入時，其輸 入指令位於1典2之間並刪除第5部份。程式之語法錯誤，由執行過程記錄檔(logfile)搶視糾正;非語法性錯誤，由執行輸出檔(lst file)對照預期結果糾正。程式經三步驟流程進行而完成，基本程式輸進Grice和Hunter(1964)資 料槍驗;應用程式以Lueft(1966)資料驗證'所得結果與預期符合無誤。基 本程式為2處理水，準及20個細格觀察數，應用至異於此二參數情況時，僅需在對應此二參數地方增修程式即可。

The purpose of the experiment was to design the basic and applied SAS programs for the Erlebacher (1977) method which can be used to detect the
confounding of the design with treatment. Thebetween- and within-subjects are the two basic types of the experimental designs in the sport psychology and behavior science researches. Theresults of experiments may be different due to different designs， and lead to a false conclusion. In 1977， by controlling the design as a factor and the treatment as another， Erlebacher developed the expected values and error terms for the parameters of such a factorial experiment. Thesignificant interaction indicates the design confounding. Since the calculation for the Erlebacher analysis is complicated， the computer programming of the method is necessary. Amongdifferent languages， SAS processes sample statistical functions， especially the probability function of the F distribution which are appropriate for writing the program of the Erlebacher analysis. The design procedures of the program follow three steps as: I.Outline the program structure， II. Writing the program according to the structure， III. Test the program step by step until the end. The structure of the program can generally be divided as four portions: (a) the output format， (b) data step， (c) procedure step， (d) end SAS program. Amongthose， the data step is the major part and can be identified as five sections as: (a) the input of the basic parameters， (b) basic statistical calculations of the cells， (c) the calculation of the Erlebacher method， (d) the probabilities of Fs，(e)the input of raw data from inside program.The correction for the error of the statement rule in the program was according to the "log" file. The error of the non-statement rule was judged from the "1st" file when it (the output file) did not match the planed result. After getting through the three procedure steps， the program was finished and tested without error.The basic program is 2 treatment levels and 20 cell size. As applying to different number(s) of the treatment levels or/and cell sizes， it is needed only amending the corresponded parameters.