期刊名称：International Journal of Statistics and Applications
印刷版ISSN：2168-5193
电子版ISSN：2168-5215
出版年度：2019
卷号：9
期号：5
页码：160-169
DOI：10.5923/j.statistics.20190905.05
语种：English
出版社：Scientific & Academic Publishing Co.
摘要：The paper provides an account of the principal components regression (PCR) and uses some examples from the literature to illustrate the following: (1) the importance of PCR in the presence of multicollinearity; (2) some cautions on its correct implementation in SPSS, as some researchers use it improperly; (3) the use of the correct formulas, in accordance with the choice of scaling the variables; (4) the choice of principal components to be dropped; (5) the conditions for the PCR to outperform ordinary least squares, in the minimum mean-square-error sense; and (6) the robustness of the estimates to substantial changes in the sample.
关键词：Multicollinearity; Principal Components; MSE; SPSS
