Publication Type: Conference Paper/Unpublished Manuscript
Abstract: How do we motivate the concept of error in hypothesis testing? How can we help statistics students understand the meaning of Type I and Type II errors, the likelihood of each error for a given test, and the connection between that likelihood and computing p-values or selecting decision criteria in classical tests? Often the probabilities α and β are simply mysterious Greek symbols whose definitions aren’t meaningful because students haven’t considered how errors can occur during testing or how they might be controlled. The challenge of teaching about error is now easier because of ready access to probability simulations that take advantage of improved random number features on graphing calculators.

In this talk, I shall demonstrate how I develop the concept of error in an applied statistics course via a calculator exploration that uses repeated simulations of a test with different decision criteria and both true and false null hypotheses. I’ll provide a worksheet that helps students discover the types of errors, assess the probability associated with each, and improve their understanding of this critical feature of hypothesis testing.

Publication Type: Conference Paper/Unpublished Manuscript
Abstract: This paper explores the conditions under which the World Bank chooses targeted aid projects as compared to nationwide projects and also the conditions under which the Bank chooses project lending as compared to programmatic lending.

Publication Type: Conference Paper/Unpublished Manuscript
Abstract: Through analyzing the Recidivism of Prisoners Released in 1994: United States dataset, this study examines the survival time distribution of recidivists for different types of offenses by the type of release from prison. The relationships between demographic factors, crime history, and the severity of punishment and the hazard of recidivism are examined and compared for different types of releases. Findings indicate the hazard of rearrest is higher for unconditionally released inmates than the risk of rearrest for conditionally released inmates. While gender and types of offenses attribute to differences in conditionally released inmates, they are not associated with the differences of rearrest rates in unconditionally released inmates. The severity of punishment measured by sentence length is significantly related to the hazard of recidivism; however, other two measures of severity of punishment--time served and the percentage of sentence served before release are not associated with the hazard of recidivism for both conditionally and unconditionally released inmates.

Publication Type: Conference Paper/Unpublished Manuscript
Abstract: Abstract:
The tempered ultradistribution space of Roumieu type for the spaceHμ,ν is defined, which
is a subspace of the Hausd¨orff locally convex topological linear space. Further, results are
obtained for the multipliers and operators on the tempered ultradistribution spaces for
the distributional Hankel-type transformation spaces.

Publication Type: Abstract
Abstract: Type II diabetes mellitus, also known as non-insulin dependent diabetes mellitus (NIDDM), is an insidious and protracted major health concern in the United States. Risk factors include age, obesity, genetic predisposition and physical inactivity. The purpose of this instructional laboratory was to give Pharm D. students practical experience with animal handling, blood drawing and the use of the glucometer to understand the pathophysiologic principles in NIDDM. These principles were demonstrated in the laboratory by using obese and lean ZDF rats, a genetic strain predisposed to NIDDM. Upon housing, obese female ZDF rats were fed a high fat diet for 4-6 weeks, which induces NIDDM in these animals. Animals were fasted for 16 hours before lab and given a bolus dose of glucose (2ml/kg -60% glucose solution) by gavage. Blood taken from locally anesthetized tails of dosed rats were tested for glucose at 0, 30, 60, 90 and 120 minutes using commercially available glucometers. Students plotted blood glucose levels of each rat over 120 minutes and observed higher glucose levels in obese ZDF rats at all time points, when compared to lean controls. Glucose levels of lean rats returned to baseline by 120 minutes. Glucose levels of obese rats failed to return to pre-glucose administration levels in this time frame. Some preparation and expertise in handling of animals was required, but as a visual demonstration of NIDDM, this laboratory proved to be an excellent teaching tool for our students to explain this major health issue.