The Course Placement Service of the American College Testing Program (ACT) uses logistic regression to model the relationships between outcomes in standard college courses and placement test scores. The logistic regression results are used together with empirical test score data to obtain estimates of validity statistics (e.g., proportion of…

After a review of the disadvantages of linear models for estimating the probability of academic success from previous school records and admission test results, the use of a probit model is proposed. The model is illustrated with admissions data from the Ecole des Hautes Etudes Commerciales in Montreal. (Author/BW)

The authors show how the work of Nobel Laureates in economics can enhance student understanding and bring them up to date on topics such as probability, uncertainty and decision theory, hypothesis testing, regression to the mean, instrumental variable techniques, discrete choice modeling, and time-series analysis. (Contains 2 notes.)

The model presented used available data to predict whether or not a student will drop out at some time during his or her college career. The model successfully identified students who would or would not drop out approximately 80% of the time. Logistic regression analysis was employed to predict chances of attrition for matriculating freshmen soon…

This article uses data from the National Longitudinal Survey of Youth 1979-1998 to understand the differences between those who exited the welfare system voluntarily (leavers) and those who stayed (non-leavers), 1990-1996. The purpose of this research was to examine the differences between the groups in their personal and family characteristics,…

By using a competing risks model, survival analysis methods can be extended to predict which of several mutually exclusive outcomes students will choose based on predictor variables, thereby ascertaining if the profile of risk differs across groups. The paper begins with a brief introduction to logistic regression and some of the basic concepts of…

This paper provides a method for analyzing data consisting of event sequences and covariate observations associated with Markov chains. The objective is to use the covariate data to explain differences between individuals in the transition probability matrices characterizing their sequential data. (TJH)

Two studies that compare the use of analysis of patterns in time (APT) (formerly non-metric temporal path analysis) and the linear models approach (LMA) are reported. In the first study, scores on the College Entrance Examination Board French Achievement Test (CEEBFAT), number of years of study of high school French (YHSF), and course achievement…

A two-parameter logistic regression model for personnel selection is proposed. The model was tested with a database of 84,808 military enlistees. The probability of job success was related directly to trait levels, addressing such topics as selection, validity generalization, employee classification, selection bias, and utility-based fair…

The purpose of this study is to examine step-wise admission to two medical programs with regard to selection effects. By using data from a large-scale longitudinal project, the influence of background variables on regular and step-wise admission is estimated. Characteristics of both applicants and those admitted are investigated using descriptive…

Cox (1972) proposed a discrete-time survival model that is somewhat analogous to the proportional hazards model for continuous time. Efron (1988) showed that this model can be estimated using ordinary logistic regression software, and Singer and Willett (1993) provided a detailed illustration of a particularly flexible form of the model that…

The current study attempts to differentiate effects of phonotactic probability (i.e. the likelihood of occurrence of a sound sequence), neighbourhood density (i.e. the number of phonologically similar words), word frequency, and word length on expressive vocabulary development by young children. Naturalistic conversational samples for three…

Three versions of a probabilistic model adapted from the theory of information retrieval--a binary version, a version using the full value of the data, and a version using principal components--were tested and applied to data available from application forms to predict graduate school performance of library school students. (EM)

This report analyzes the relationship of Advanced Placement Program® (AP®) teacher practices and student performance on AP Biology and AP U.S. History Exams. Using a national survey of AP teachers, the study developed four models for each subject with public school teachers only and both public and nonpublic school teachers, using two standards of…

This study compared essay scores from paper-based and computer-based versions of a writing test for prospective teachers. Scores for essays in the paper-based version averaged nearly half a standard deviation higher than those in the computer-based version, after applying a statistical control for demographic differences between the groups of…