How the Reflection of Linear Correlation in Odds Ratios Depends on the Cut-Off Points

When an OR is calculated based on data measured on continuous scales, the magnitude of the OR is dependent on the cut-offs chosen for dichotomizing both dependent and independent variables even when the relationship is strictly linear. Cuts away from the median on either or both dependent and independent variables increase the expected OR. This increase is quite substantial when the cut-off is made at the extremes. Illustrations of the consequences of cut-off on OR for populations with varying values of linear correlation are provided in simulated and real data. Potential circumstances and motivations for such dichotomizations are discussed.     

The generalized symmetric eigenproblem in multivariate statistical models

The solution of the generalized symmetric eigenproblem Ax = λBx is required in many multivariate statistical models, viz. canonical correlation, discriminant analysis, multivariate linear model, limited information maximum likelihoods. The problem can be solved by two efficient numerical algorithms: Cholesky decomposition or singular value decomposition. Practical considerations for implementation are also discussed.     

Nonparametric multiple testing procedures

Assume that we have n_i independent observations from each of k independent populations. Each population has the same distribution except for a translation parameter. We are interested in specific pairwise differences of the parameters in various settings, such as treatment vs. control, change point or all pairwise differences. We propose new multiple testing procedures for the pairwise differences. The new procedures are based on ranks and they have desirable practical properties not shared by existing procedures. These include tests that satisfy the interval property. Furthermore, the test method provides an interval that serves as an estimate of the difference in the parameters of interest.     

Maximizing competence through professional development: increasing disability knowledge among One-Stop Career Center staff

The Workforce Investment Act of 1998 (USA) mandates that partners in the One-Stop Career Center system be prepared to serve a diverse customer base. Effective service delivery depends in part on a focus on human resources and professional development. This article presents innovative strategies for One-Stop Career Center staff training related to serving customers with disabilities. Findings from case study research conducted in several One-Stops across the country revealed that staff struggled with both knowledge and attitudes around disability issues. To address these concerns, local leaders developed practices that provided opportunities to gain practical skills and put acquired knowledge to use. These included a formalized curriculum focused on disability issues; informal support and consultation from a disability specialist; and exposure and learning through internships for students with disabilities. Implications are offered to stimulate thinking and creativity in local One-Stops regarding the most effective ways to facilitate staff learning and, in turn, improve services for customers with disabilities.     

A Revised Economic Analysis of Restrictions on the Use of Cell Phones While Driving

Evidence that cell phone use while driving increases the risk of being involved in a motor vehicle crash has led policymakers to consider prohibitions on this practice. However, while restrictions would reduce property loss, injuries, and fatalities, consumers would lose the convenience of using these devices while driving. Quantifying the risks and benefits associated with cell phone use while driving is complicated by substantial uncertainty in the estimates of several important inputs, including the extent to which cell phone use increases a driver's risk of being involved in a crash, the amount of time drivers spend using cell phones (and hence their aggregate contribution to crashes, injuries, and fatalities), and the incremental value to users of being able to make calls while driving. Two prominent studies that have investigated cell phone use while driving have concluded that the practice should not be banned. One finds that the benefits of calls made while driving substantially exceed their costs while the other finds that other interventions could reduce motor vehicle injuries and fatalities (measured in terms of quality adjusted life years) at a lower cost. Another issue is that cell phone use imposes increased (involuntary) risks on other roadway users. This article revises the assumptions used in the two previous analyses to make them consistent and updates them using recent data. The result is a best estimate of zero for the net benefit of cell phone use while driving, a finding that differs substantially from the previous study. Our revised cost-effectiveness estimate for cell phone use while driving moves in the other direction, finding that the cost per quality adjusted life year increases modestly compared to the previous estimate. Both estimates are very uncertain.     

Second order asymptotic and non-asymptotic optimality properties of combined tests

Let p independent test statistics be available to test a null hypothesis concerned with the same parameter. The p are assumed to be similar tests. Asymptotic and non-asymptotic optimality properties of combined tests are studied. The asymptotic study centers around two notions. The first is Bahadur efficiency. The second is based on a notion of second order comparisons. The non-asymptotic study is concerned with admissibility questions. Most of the popular combining methods are considered along with a method not studied in the past. Among the results are the following: Assume each of the p statistics has the same Bahadur slope. Then the combined test based on the sum of normal transforms, is asymptotically best among all tests studied, by virtue of second order considerations. Most of the popular combined tests are inadmissible for testing the noncentrality parameter of chi-square, t, and F distributions. For chi-square a combined test is offered which is admissible, asymptotically optimal (first order), asymptotically optimal (second order) among all tests studied, and for which critical values are obtainable in special cases. Extensions of the basic model are given.