The Texaco racial discrimination case and shareholder wealth

We analyze the wealth ejfects of the Texaco racial discrimination lawsuit both on the shareholders of Texaco and its major U.S. competitors. Employing a comprehensive data set which included every case docket entry and every Wall Street Journal article on the case as an experimental stimulus, our findings suggest that the overall cost of the case to Texaco shareholders exceeded $500 million, that Texaco's tribulations had little, if any, impact on the share prices of its major competitors, and that Wall Street Journal coverage of the case was highly correlated with significant changes in Texaco stock prices. This last finding provides significant support for Hite 's suppostion that newspaper editors “key ” on ex post stock price changes in selecting the events to be covered in the next day's edition. The authors are grateful to Kee Chung for helpful comments on earlier drafts and also acknowlege the help-ful assistance of the staff of the law library at the Cecil C. Humphreys School of Law at The University of Memphis.     

A Comparison of Partially Adaptive and Reweighted Least Squares Estimation

The small sample performance of least median of squares, reweighted least squares, least squares, least absolute deviations, and three partially adaptive estimators are compared using Monte Carlo simulations. Two data problems are addressed in the paper: (1) data generated from non-normal error distributions and (2) contaminated data. Breakdown plots are used to investigate the sensitivity of partially adaptive estimators to data contamination relative to RLS. One partially adaptive estimator performs especially well when the errors are skewed, while another partially adaptive estimator and RLS perform particularly well when the errors are extremely leptokur-totic. In comparison with RLS, partially adaptive estimators are only moderately effective in resisting data contamination; however, they outperform least squares and least absolute deviation estimators.     

Exploring multivariate data graphically:a brief review with examples

Many methods for obtaining graphical representations of multivariate data have been suggested over the last decade, but it is only relatively recently that software has become available which makes use of these procedures routine. In this paper, a number of examples are used to illustrate the application of graphical methods for the exploration of multivariate data. The diagrams produced may prove useful in consultations with clients and in examining the results from more formal methods.     

Sources of drug information among adolescent students

A sample of 1023 eighth and tenth grade students in small to medium-sized central Texas school districts was assessed to determine the amount of information they receive from ten sources about six categories of drugs. The amount of information males reported receiving about each drug category was significantly greater than what females reported, and the amount of information that eighth graders reported receiving about each drug category was significantly greater than what tenth graders reported. Television was the primary source of drug information for all categories of drugs except inhalants, for which friends and television were equally important sources. Parents and printed media (magazines or newspapers) were of secondary importance, followed by friends and teachers. Adolescents were less likely to receive drug-related information from experience, siblings, church, doctors, and police. The reliance on the mass media for drug information in smaller school districts is a pattern which has been previously observed in larger urban districts. This consistency suggests that mass media approaches to drug education are likely to be as effective in rural areas and smaller towns as they are among urban adolescents. Implications for television programming are discussed.     

Book Review: Beyond Earth Day: Fulfilling the Promise. Gaylord Nelson with Susan Campbell and Paul Wozniak. Madison,WI: The University of Wisconsin Press, 2002: 160 pp.

Population and Environment -     

ON THE INDEPENDENCE OF THE SAMPLE MEAN AND TRANSLATION-INVARIANT STATISTICS FOR MATRIX NORMAL DISTRIBUTIONS

We present an explicit characterization of the joint dependency structure of an n×p matrix normal random matrix such that the p-dimensional sample mean vector is independent of all translation invariant statistics.     

QUESTION WORDING AND PARTISANSHIP: CHANGE AND CONTINUITY IN PARTY LOYALTIES DURING THE 1992 ELECTION CAMPAIGN

Gallup macropartisanship varies more over time than aggregatemeasures of partisanship employing the standard Michigan SurveyResearch Center (SRC) party identification measure, but previousanalyses do not provide direct evidence about why Gallup macropartisanshipis more variable. Although these differences could result fromthe short-term focus of the Gallup party affiliation question,aggregate-level analyses cannot test the effects of questionwording directly. Between March and October 1992, we conducteda series of question-wording experiments, employing six statewidecomputer-assisted telephone interview (CATI) surveys of Michiganadults, including a four-wave panel study. Our analyses stronglysuggest that the Gallup measure responds more to short-termpolitical conditions and clearly demonstrate that the Gallupmeasure is less stable over time. These individual-level resultshelp explain why Gallup macropartisanship varies more over timethan aggregate measures of partisanship employing the standardSRC measure and raise questions about the degree to which onecan generalize from analyses using the Gallup data to the researchliterature on party identification.     

Bayesian analysis of directed graphs data with applications to social networks

Summary.  A fully Bayesian analysis of directed graphs, with particular emphasis on applica- tions in social networks, is explored. The model is capable of incorporating the effects of covariates, within and between block ties and multiple responses. Inference is straightforward by using software that is based on Markov chain Monte Carlo methods. Examples are provided which highlight the variety of data sets that can be entertained and the ease with which they can be analysed.     

D-efficient window experimental designs

In the paper we present a new method of calculating sampling intervals, so-called windows, allowing an experimenter some flexibility in timing the sample collection, while a minimum required design efficiency for parameter estimation is assured. The method is based on the Equivalence Theorem for D-optimality what makes the length of each window related to the parameter sensitivities. An example of calculating the windows in a pharmacokinetic study is presented. Some other methods of calculating efficient sampling windows are briefly discussed.