|
Previous diffusion model researches have focused on the durable goods by adoption the annual data for the prediction purpose. In this study we attempt to explore if monthly date can be used to predict the traditional Bass model, as well as to explore if there exists a certain relationship between monthly data and annual data while predicting Bass mode. We use five different models to test some hypotheses and use NLS to evaluate the model's parameters. The research results indicate that: (1) monthly data do have the interference of periodicity factor, for the products of refrigerator, color TV and video recorder. (2)generally the monthly data have better prediction than annual data.
|