Maximum likelihood estimation in a model with interval data: a comment and extension

The purpose of this article is, first, to extend Poon et al. 's (1993) maximum likelihood estimation (MLE) of the correlation coefficient based on interval data to the regression case. Secondly, this paper shows how the traditional method of collecting interval data with the intervals chosen by the researcher can be easily modified to avoid the problems discussed by Poon et al. (1993). The MLE for this modification to the regression problem is presented. Finally, all the methods discussed in this paper are used to estimate the effects of grade point average and gender on student perceptions of the percentage of their classmates who have cheated on at least one exam in college.     

Female athletic participation and income: evidence from a latent class model

This paper introduces and applies an EM algorithm for the maximum-likelihood estimation of a latent class version of the grouped-data regression model. This new model is applied to examine the effects of college athletic participation of females on incomes. No evidence for an “athlete” effect in the case of females has been found in the previous work by Long and Caudill [12], Henderson et al. [10], and Caudill and Long [5]. Our study is the first to find evidence of a lower wage for female athletes. This effect is present in a regime characterizing 42% of the sample. Further analysis indicates that female athletes in many otherwise low-paying jobs actually get paid less than non-athletes.     

Estimating class-specific parametric models using finite mixtures: an application to a hedonic model of wine prices

Hedonic price models are commonly used in the study of markets for various goods, most notably those for wine, art, and jewelry. These models were developed to estimate implicit prices of product attributes within a given product class, where in the case of some goods, such as wine, substantial product differentiation exists. To address this issue, recent research on wine prices employs local polynomial regression clustering (LPRC) for estimating regression models under class uncertainty. This study demonstrates that a superior empirical approach – estimation of a mixture model – is applicable to a hedonic model of wine prices, provided only that the dependent variable in the model is rescaled. The present study also catalogues several of the advantages over LPRC modeling of estimating mixture models.     

Pooling choices or categories in multinomial logit models

The estimation of multinomial logit models today is routine. With this increased use has also come a need for testing. A test to determine whether choices can be combined is important. This paper presents a likelihood ratio test for combining choices in multinomial logit models. The use of the test is demonstrated with a simple example.     

Quantitative models for optimal rest period scheduling

Mathematical and simulation models studying work-rest schedules in a production process are developed to arrive at an optimal rest policy to maximize work output per unit time. A CSMP simulation study is used to test the sensitivity of the results for different specific cases.     

High-risk drinking among college fraternity members: a national perspective

This survey, with its 85% response rate, provides an extensive profile of drinking behaviors and predictors of drinking among 3,406 members of one national college fraternity, distributed across 98 chapters in 32 states. Multiple indexes of alcohol consumption measured frequency, quantity, estimated blood alcohol concentration levels (BACs), and related problems. Among all members, 97% were drinkers, 86% binge drinkers, and 64% frequent binge drinkers. On the basis of self-reports concerning the 4 weeks preceding the time of survey, the authors found that members drank on an average of 10.5 days and consumed an average of 81 drinks. Drinkers had an average BAC of 0.10, reaching at least 0.08 on an average of 6 days. These fraternity members appear to be heavier drinkers than previously studied fraternity samples, perhaps because they were more representative and forthright. All 6 preselected demographic attributes of members and 2 chapter characteristics were significantly related to the drinking behaviors and levels of risk, identifying possible targets for preventive interventions.     

Stadium size, ticket allotments and home field advantage in college football

This study builds on a conceptual framework that explores of the relationship between relative fan attendance, stadium size and home field advantage in college football. We extend the conceptual model by providing empirical estimates of the impact that various ticket allotment arrangements and home stadium size have on the probability of winning in college football. The Auburn–Alabama rivalry (the “Iron Bowl”) provides a unique case given that the series has been played both at a neutral site with an equal division of tickets, as well as on a home-and-home basis with an uneven ticket split and nonuniform stadium capacities.     

More on testing the normality assumptionin the Tobit Model

In a recent volume of this journal, Holden [Testing the normality assumption in the Tobit Model, J. Appl. Stat. 31 (2004) pp. 521–532] presents Monte Carlo evidence comparing several tests for departures from normality in the Tobit Model. This study adds to the work of Holden by considering another test, and several information criteria, for detecting departures from normality in the Tobit Model. The test given here is a modified likelihood ratio statistic based on a partially adaptive estimator of the Censored Regression Model using the approach of Caudill [A partially adaptive estimator for the Censored Regression Model based on a mixture of normal distributions, Working Paper, Department of Economics, Auburn University, 2007]. The information criteria examined include the Akaike’s Information Criterion (AIC), the Consistent AIC (CAIC), the Bayesian information criterion (BIC), and the Akaike’s BIC (ABIC). In terms of fewest ‘rejections’ of a true null, the best performance is exhibited by the CAIC and the BIC, although, like some of the statistics examined by Holden, there are computational difficulties with each.     

How robust is the rationality assumption in economics? A statistical test based on student grade distributions

This note re-addresses the question of the rationality of economic actors in a new setting. We examine the distribution of final course grades for a “Principles of Economics Class” in an effort to draw conclusions about rational student work effort. Based on the concepts of the opportunity cost of time, the relative reward for earning various grades (based on university codes for determining grade points) and the concept of rationality, we expect to find a greater percentage of grades at the lower end of the distribution for each grade level. We find statistical evidence that there are more students at the lower end of each grade interval than at the upper end. This result is consistent with what economists have long conjectured: students are rational in the allocation of study time.     

Heterogeneous skewness in binary choice models: predicting outcomes in the men's NCAA basketball tournament

Several authors have recently explored the estimation of binary choice models based on asymmetric error structures. One such family of skewed models is based on the exponential generalized beta type 2 (EGB2). One model in this family is the skewed logit. Recently, McDonald (1996, 2000) extended the work on the EGB2 family of skewed models to permit heterogeneity in the scale parameter. The aim of this paper is to extend the skewed logit model to allow for heterogeneity in the skewness parameter. By this we mean that, in the model developed, here the skewness parameter is permitted to vary from observation to observation by making it a function of exogenous variables. To demonstrate the usefulness of our model, we examine the issue of the predictive ability of sports seedings. We find that we are able to obtain better probability predictions using the skewed logit model with heterogeneous skewness than can be obtained with logit, probit, or skewed logit.