The Use and Interpretation of Logistic Regression in Higher Education Journals: 1988-1999.

摘要: This article examines the use and interpretation of logistic regression in three leading higher education research journals from 1988 to 1999. The were selected because their emphasis on research, relevance issues, broad coverage topics, reputable editorial policies. term “logistic regression” encompasses logit modeling, probit tobit modeling significance tests estimates. A total 52 articles identified as using regression. Our review uncovered an increasingly sophisticated for a wide range topics. At same time, there continues be confusion over terminology. sample sizes used did not always achieve desired level stability parameters estimated. Discussion results terms delta-Ps marginal probabilities was cautionary, according definitions. is concluded with recommendations journal editors researchers formulating appropriate policies practice applying versatile technique communicating its readers research.