Mary M. Whiteside

A nonparametric stochastic procedure for FMS evaluation

Abstract Evaluating alternative manufacturing technologies in the presence of multiple performance measures is often a difficult task for the decision maker. It is for this reason that justification and evaluation of flexible manufacturing systems has been receiving significant attention in the manufacturing circles. This paper proposes an innovative framework, which is based on the combined application of data envelopment analysis and nonparametric statistical procedures, for the selection of flexible manufacturing systems. The strengths of this methodology are that it incorporates variability measures in the performance of alternative systems, provides decision maker with effective alternative choices by identifying homogeneous groups of systems, and presents graphic aids for better interpretation of results. The methodology is illustrated through its application on a previously reported data set.

Maintaining distributed relationships

Electronic communication works best when it increases interaction and collaboration through a variety of media.

Comparing the Effectiveness of Traditional and Active Learning Methods in Business Statistics: Convergence to the Mean

This research shows that active learning is not universally effective and, in fact, may inhibit learning for certain types of students. The results of this study show that as increased levels of active learning are utilized, student test scores decrease for those with a high grade point average. In contrast, test scores increase as active learning is introduced for students in the lower level grade point average group. Every student involved in the experiment is taught three topics, each one by a different teaching method. Students take a test following each learning session to assess comprehension. The experiment involves more than 300 business statistics students in seven class sections. Method topic combinations are randomly assigned to class sections so that each student in every class section is exposed to all three experimental teaching methods. The effect of method on student score is not consistent across grade point average. Performance of students at three different grade point average levels tended to converge around the overall mean when learning was obtained in an active learning environment. The effects of the teaching method on score do not depend on other student characteristics analyzed (i.e. gender, learning style, or ethnicity). A linear mixed model is used in the analysis of results.

Reverse Regression, Collinearity, and Employment Discrimination

Apparently contradictory results between direct and reverse regression in employment-discrimination data analysis are a manifestation of collinearity in the data. An easily implemented guideline that alerts the analyst to the presence of contaminating collinearity is illustrated with employment data from Title VII litigation.

A comparison of some nonparametric tests for scale

This paper presents simulated power functions for the sum of squared ranks, Wilcoxon Savages T and locally most powerful tests for the two-sample scale problem using small samples drawn from gamma distributions. Also, pitman efficiencies of each of these tests relative to the best test test are compared for gamma distributionsd. In addition, the sum of squared ranks is shown to be prferable to a composite test due to Woinsky in some situations.

A Better Estimate of the Number of Valid Signatures on a Petition

The 2006 Texas gubernatorial race with independent candidates Kinky Friedman and Carole Keeton Strayhorn, the 2004 attempts of Ralph Nader to gain ballot access as an independent candidate for President of the United States, and the recall of California Governor Gray Davis in 2003 are among many examples that illustrate the importance of validating signatures on a petition. Signatures may be invalid for several reasons, including not a registered voter and replication. The statistical problem is interesting because replicated signatures must be estimated differently than other invalid signatures. This article presents a new nonlinear estimator that is unbiased with smaller standard error in cases considered and suggests an innovation to current practice: first estimate the number of replicated signatures, then the number of valid signatures.

A new look at life and job satisfaction with sunbelt MBA's

Work and extra-work correlates of life and job satisfaction for a sample of MBA graduates (N=1495) are compared to those for a probability sample from the general public using the Rice, Near, Hunt model for analysis. For both groups, the unique relationship between life and job satisfaction is quite weak, however, predictability of job satisfaction is significantly different for MBAs.1

Asymptotic efficiency of the sum of squared ranks test for discrete distributions

Asymptotic (Pitman) efficiencies of the sum of squared ranks test relative to the best test are presented when the underlying distributions are Poisson. binomial, uniform discrete, and negative binomial when the average scores and midrank methods for resolving ties are used. These results are then compared to efficiencies of other nonparametric tests for these specific discrete distributions.

The sampling distribution of the W estimator of the number of valid signatures on a petition

Abstract A new non linear estimator, W, for the number of valid, unique signatures on a petition has been shown better, for the cases enumerated and with certain restrictions, than a popular Goodman-type statistic, G. This article extends those results with relaxed conditions by developing the exact probability mass function and mean of W and a close approximation of the variance (Var(W)). If the proportion of valid signatures among unique and duplicated signatures is the same, then Var(W) is approximately a function of the means and variances of the two sample statistics. Using the delta method, we estimate Var(W), with the resulting approximation shown to be good, even when the condition of equal proportions does not hold. We compare W to G and establish which estimator is preferred for different intervals of the design parameters. Data from a Washington State petition illustrate the findings.

Modified chance constrained linear programming under uncertainty

This paper considers the sensitivity of chance constrained linear programming solutions where the coefficients of the left-hand side of a constraint function are estimated from a sample using multiple linear regression. The modified nonlinear constraint provides considerable assurance that the true, but unknown, stochastic linear constraint will be satisfied at a given level of probability for the conditions of the simulation herein. Ordinary least squares and least absolute value regression criteria are considered along with normal, uniform and double exponential distributions of error.