The purpose of this exploratory study is to examine the determinants of alternative forms of academic misconduct among undergraduate students in a Korean university. An economic model of college cheating is estimated using logistic regression. The study differs from other studies in that it frames the issue of academic dishonesty in the context of recent developments to internationalize higher education in Korea, highlighting certain unique aspects of the Korean educational system that are undergoing rapid change and have largely remained unexamined in the literature. The presence of a native English-speaking foreign instructor in class, English-taught classes, and class levels are significant predictors of in-class cheating and such types of academic misconduct as seeking outside help and plagiarism. GPA, class size, cheating tolerance, and study-abroad experience are significant in some types of illicit academic behavior but not in others. Gender and being a foreign student turn out to be weak predictors.

No abstract

Sept. 2009) estimated a logistic model to predict the likelihood of enrollment by admitted students in a small, private college. Given the natural link between the enrollment and persistence decisions of high school graduates as part of an overall college-going behavior, it seems natural to extend the analysis. Can the same admissions data be used to predict first-semester success? The predictors for enrollment and persistence may not be the same for different-size colleges or universities considering that, compared to their larger-sized and betterendowed counterparts, small private colleges may lack adequate laboratory or sports facilities, have fewer course offerings, meager marketing budgets, and fewer feeder high school networks that could affect college choice. This research concludes that the predictors of first-semester success for the small, private college studied here are high school GPA, an offer of two scholarships, being a second choice school, and indecision on major at admission.Studies of first-year academic performance usually rely on single-equation estimation methods or separate, single-equation regressions (Johnson Research in Higher Education, 49, 2008; Wolniak and Engberg Research in Higher Education, 51, 2010). Since this study uses admissions data with many zeroes for the dependent variable to predict first-semester success, it differs by postulating that enrollment and persistence decisions come from two data generating processes-one for zeroes and the other for positive outcomes. Hence, it employs a two-part (two-equation), mixed model with a common formulation for both parts. Assuming the covariates x are exogenous and the

No abstract

No abstract