Thomas E. Bradstreet, Ph.D.
Associate Director, Scientific Staff
E-mail Address: Thomas_Bradstreet@merck.com
Phone: (484) 344-2542
Address: Clinical Pharmacology Statistics
Merck Research Labs, BL 3-2
10 Sentry Parkway Blue Bell, PA 19422

Title: Sampling Plans and Combinatorial Theorems for a Stopping Rule of Order k
Duration: 1 hour seminar
Audience: Graduate (and Undergraduate) Students, Faculty, in Statistics/Biostatistics
Abstract: Consider generalized inverse binomial sampling which stops with the first occurrence of k, k>=2, successes in-a-row. Let there be s, s>=1, strings of successes and failures generated. We present sampling plans, combinatorics, and a probability structure for this process. Path counts in the sampling plan form Pascal Triangles of order (k,s). comparison is made to the k=1 case.
Honorarium: Not necessary

Title: Illustrating the Neyman-Pearson Lemma with a Stopping Rule of Order k
Duration: 1 hour seminar
Audience: Graduate (and Undergraduate) Students, Faculty, in Statistics/Biostatistics
Abstract: Consider generalized inverse binomial sampling which stops with the first occurrence of k, k>=2, successes in-a-row. A curved exponential family results where dim{minimal sufficient statistic}=2 but dim{parameter space}=1. Power comparisons of the N-P MP test with two competing tests show that the N-P test is MP but not UMP. Comparisons made to k=1 case.
Honorarium: Not necessary

Title: Conditional Two Sample Permutation Tests for a Stopping Rule of Order k
Duration: 1 hour seminar
Audience: Graduate (and Undergraduate) Students, Faculty, in Statistics/Biostatistics
Abstract: Consider generalized inverse binomial smapling which stops with the first coccurence of k, k>=2, successes in-a-row. We compare power of two permutation tests. The first considers differences in proportions, the second differences in lenghs of strings of successses and failures. Each conditioned on a two dimensional statistics sufficient for a one dimensional parameter.
Honorarium: Not necessary

Title: Examples of Preclinical and Clinical Trials Used in Drug Develpment
Duration: 3 hour seminar (can be shortened or developed further)
Audience: Graduate (and Undergraduate) Students, Faculty, in Statistics/Biostatistics
Abstract: We present some study designs and subject matter issues from animal and human studies conducted in pharmaceutical research. Examples include teratology, tumerigenicity, cytotoxicity, does response, bioequivalence, alcohol interaction, hypertension, SHF, GERD, and ophthalmics. Some data and studies are stored at www.villanova.edu/~tshort/Bradstreet/
Honorarium: Not necessary

Title: Preparing for the Interview: Student Preparation for Finding a Job in Industry
Duration: 60 - 90 miutes, can be shortened
Audience: Graduate and Undergraduate students in Statistics/Biostatistics
Abstract: We present guidance on where to find a job, interview prepration, and expected resume content. Emphasis is on academic and professional activities. For example: Industrial work experience; computer skills; communication skills; networking, expected education; cover letter and resume; telephone interviews, standard interview questions; and visibility.
Honorarium: Not necessary