李克尺度及語意差異尺度之尺度類型在行銷研究應用的本質上為順序尺度，然而 ，實證分析上卻往往將其視為區間尺度處理，此種處理方式即意謂著：順序尺度的等距假設 。本研究主要目的即在探討此一假設對驗證性因素分析模式估計的影響。研究上採用蒙地 卡羅模擬的方式進行實驗，操弄因子包括樣本大小及峰態。根據模擬分析的結果顯示：各種 峰態分佈下，尺度等距假設之模式參數估計會產顯著的偏誤，模式評估亦容易發生型一的錯 誤。這些發現除了顯示過去應用驗證性因素模式的研究在虛無假設拒絕上之不當歸因外，亦 突顯文獻上一些看法的不適切性。 同時，本研究也發現峰態低 (但不偏 ) 的情形下，卡方 值並不見得易於過高，這與過去的發現結果不同。

　　　　　The Likert scale and the semantic differential scale used on the applicatiion of marketing research are ordinal. However, in empirical studies, these two scales are usually assumed as interval. The purpose of this paper investigates what influences this assumption brings about to the model of confirmatory factor analysis (CFA). An experiment(varying:sample sizes and kurtosis) is employed based on the Monte Carlo simulation. The results show that the interval assumption of ordinal scale significantly produces the biases on the parameter estimation and the type one errors of the overall test. These findings falsify the justifications of previous analysis. Furthermore, unlike what has been pointed out in the past studies, it is suggested that the chi-square values are not necessarily higer than the critical ones under some conditions.