Abstract
A quantitative method for analyzing the characteristics of communication networks in an organization is described, in which a framework model is proposed to unifying the learning curve models that were discriminated in terms of obstruction probability in the past study. Moreover the time to process the subject and the quantity of errors for the subject by using the estimates of the learning effects obtained from the proposed models are estimated. The high multiple correlation coefficients and estimates of the parameters in meeting the actual data are obtained through our analyses, and the validity of the models us verified.
