Principal component analysis is the study of how a small number of uncorrelated principal components indicate the internal structure of multiple variables. The purpose is to simplify the data, reduce the dimension of redundancies, and make the new variable uncorrelated. So it is widely utilized in multivariate statistical analysis. These years, however, some people apply the principal component analysis to determine the weight of evaluation index system, even include the method in the textbook. This paper will clarify the unscientific method of valuing the weight in four aspects.