Computer tools for statistical quality control

Since the 1920' sanduntil recently, numerical computation has beena limiting factor in the application of statistical methods to the improvement of product quality. This restriction is be in geliminated by the in troduction of computer tools for statistical quality control.

Both novice and expertusers of statistical methods can ben-efitsubstantially from the availability o intergrated soft ware for computing, graphics, and data management. Theuse of such soft ware in SQC training program senables the student to focus on the under standing of statistical techniques, rather than the irmechanicaldetails. In productionen vironments, properly designedinterfaces facilitated at aent ryand access to statistical soft ware by plant personnel, with out requiring know ledge of a computer language. The sesame tool scan be used by management to retrieve in for mationandob tain summaries and display sofcriti-caldata gatheredover different period softime. Finally, computer tool sprovide the applied statistician with agreater range of advanced methods, includinganalytical and graphical extensions of the traditional She whart control chart.     

Following the Rules? Candidate Strategy in Presidential Primaries*

Objectives. The competitiveness of the 2008 presidential primaries in both the Republican and Democratic parties has prompted a reconsideration of the role of delegate‐selection rules in influencing the strategic behavior of presidential candidates. Using advertising and candidate state‐visit data from the 2004 and 2008 presidential nominating campaigns, we reexamine the strategies presidential candidates use when competing for the nomination of their party. Methods. Using the Wisconsin Advertising Project Data from 2004 and 2008, we estimate several multiple regressions designed to analyze the factors predicting candidate visits and advertising. Results. We find that, to a large extent, the rules of the game help predict where candidates allocate their political advertising and campaign stops; candidates consider whether a contest is a primary or caucus, they pay attention to how many delegates are at stake, and they consider whether a state's delegate‐allocation method is largely proportional or winner take all. Yet we also find some differences in how the rules influence frontrunners and long‐shot candidates, and we discover how other factors, including a candidate's access to financial resources, influence the allocation of ads and visits. Conclusion. Our findings offer some of the first empirical evidence for the idea that a state's delegate‐allocation method influences candidates' resource‐allocation behavior. That these rules matter at all is somewhat of a surprise given that the delegate‐allocation methods used by states have become more homogenous within each party.     

Record values and related statistics - a review

In this brief survey, we discuss major developments of the past decade in the study of record values, record times, inter-record times and some related statistics from a series of observations     

Some properties and a characterization of trivariate and multivariate binomial distributions

By considering a trivariate binomial distribution, the regression equations are obtained and a set of necessary and sufficient conditions are given for the regression to be linear. Under the limiting conditions, the expressions approach those of trivariate Poisson distribution derived by Mahamunulu (1967). A characterization of the trivariate binomial distribution is also established based on the distribution of the sum of two trivariate random vectors. The results are shown to extend to the multidimensional case in a natural way.     

Testing Goodness of Fit of Parametric AFT and PH Models with Residuals

In this article, we consider the problems of testing the goodness of fit of the parametric accelerated failure time model and the Cox proportional hazards model. We consider omnibus test statistics based on residuals. The statistical distributions of Kolmogorov, Cramer-von Mises–Smirnov, and Anderson–Darling statistics are all investigated by means of Monte Carlo simulations. Type-I, Type-II, and independent random censoring situations are all considered in this study. A Monte Carlo power study has also been carried out for these tests to distinguish between various baseline models, which reveals that the Anderson–Darling test performs better than the others.