銷售預測向來被認為是攸關企業成功與否的關鍵因素之一。近年來由於企業誰出新產品的腳步遠遠超過了以往的水率，且產品生命週期縮短，因此創新擴散模型的研究又再次地引起了學術界廣泛的注意。然而過去三十年間所發展出的模型數目之多，已經使得實證研究的工作遠遠地落後模型發展的速度。研究者應該要有系統地檢驗各種擴散模型，找出具最佳配適度 ( Model Fit ) 的模型。本研究將針對20種不同的彈性擴散模型進行實證研究，再以後設分析法分析比較各模型的配適能力。研究結果有二：1.如應用三參數水準的彈性擴散模型來進行產品銷售量的預測即可獲致良好的模型配適度，如果繼續增加參數水準的話，則應採用五參數水率下，研究的完整模型才能獲得配適度的顯著提昇。而在三參數水準中，最佳的模型唯有修正後的NUI模型一種，亦即研究設計中的模型9。2.產品別因素的確對彈性擴散模型間的配適度造成顯著的差異，這項研究結果與Sultan, Farley與Lehmann ( 1990 )所做的實證研究結果一致。由於產品不同會造成模型配適度間的顯著差異，因此可判定彈性擴散模型無法將產品特性列入考量，適用於所有產品的創新擴散預測上，因此將產品特性列入擴散模型中有其必要性。

Sales forecast has always been considered as one of the key success factors of a company. In recent years, the studies of the Innovative Diffusion Model has regained the researcher's attention in the academic circle, due to the ever- fast speed of new product development and the shortening of product life cycle. As a result of numerous models developed in the past thirty years, the path of empirical research has been far behind the progress of the model development. Therefore, one of the major tasks for researchesr was to examine the Innovative Diffusion Models, and find out the model most-fitted. An empirical research on twenty innovative diffusion models were conducted through comparisons in this sutdy. The major findings of the study were: (1) the degree of model-fit could be improved by using three parameters; however. The levels of fitness would be significantly increased when five parameters were adopted. A good model by using three parameters could only be attained through modification, as represented by Model 9-NUI Model. (2).The individual differences among products did play an important role in deciding the fitness of a model. This finding was consistent with the results of the empirical research by Sultan, Farley and Lehmann ( 1990 ). As a result, it was essential that the characteristics of a product be incorporated into the model for a better result of sales forecast to be attained.