Sponsoring Section/Society: WNAR
Session Slot: 2:00- 3:50 Sunday
Estimated Audience Size: xx-xxx
AudioVisual Request: xxx
Session Title: Biostatistical Applications of Multi-Level Models
Theme Session: No
Applied Session: Yes
Session Organizer: Witte, John Case Western Reserve University
Address: Department of Epidemiology and Biostatistics 2500 MetroHealth Drive Cleveland, OH 44109-1998
Phone: (216) 778-8523
Fax: (216) 778-3280
Email: witte@darwin.cwru.edu
Session Timing: 110 minutes total (Sorry about format):
First Speaker - 30 minutes Second Speaker - 30 minutes Third Speaker - 30 minutes Discussant - Floor Discusion - 15 minutes
Session Chair: Witte, John Case Western Reserve University
Address: Department of Epidemiology and Biostatistics 2500 MetroHealth Drive Cleveland, OH 44109-1998
Phone: (216) 778-8523
Fax: (216) 778-3280
Email: witte@darwin.cwru.edu
1. Getting It Right? Several Analyses Of a Pediatric Pain Repeated Measures Data Set
Weiss, Robert, UCLA Biostatistics
Address: Department of Biostatistics UCLA School of Public Health Los Angeles, CA 90095-1772
Phone: (310) 206-9626
Fax: (310) 267-2113
Email: rob@rem.ph.ucla.edu
Abstract: I discuss statistical modeling issues in the context of 9 different analyses of a repeated measures data set collected to study the effects of pain coping style (CS) and a treatment. Time pressure usually demands that Statisticians analyze data quickly, resulting in the use of models and methods that are simple and easily available in standard software. Occasionally we have time to take one data set thoroughly apart so as to make an attempt at getting the model completely correct, or, in this case, to try out several different analyses. This talk sketches a portion of several analyses of a Pediatric pain repeated measures data set. The data consists of up to 4 repeated pain tolerance trials on 64 kids. Pain tolerance is assessed by the length of time the children can keep an arm immersed in cold water. Some of the problems with the data set are (a) some missing data, (b) censoring of large responses, (c) repeated measures, (d) non-constant variance, (e) skewness and (f) a lack of overwhelming statistical significance.
2. Assessment of Gene-Environment Interactions with Hierarchical Models
Aragaki, Corinne, Fred Hutchinson Cancer Research Center
Address: Quantitative Genetic Epidemiology Metropolitan West Building, Room #805 Seattle WA 98104
Phone: 206-667-7256
Fax:
Email: caragaki@cclink.fhcrc.org
Abstract: Assessment of gene-environment interactions with hierarchical models is proposed. The properties of the estimates are studied in finite samples with a limited simulation study. This method is illustrated with an application to diet-NAT2* interaction estimation in a colorectal polyp data set.
3. Radon Dose Imputation in the Colorado Plateau Miners Chort Using a Hierarchical Model for the Temporal and Geographical Distribution of True Exposure
Stram, Daniel, Univ. of Southern California
Address: Suite 220, 1540 Alcazar Street Los Angeles, CA 90033
Phone: 213-342-1817
Fax: 213-342-2349
Email: stram@rcf.usc.edu
Huberman, Mark, Univ. of Southern California
Abstract: This talk will describe measurement error correction of analyses of the Colorado Uranium Miners Cohort lung cancer data based upon a hierarchical model for the geographic and temporal distribution of radiation dose in the Colorado Plateau from 1950-1969. The implications of the hierarchical model's dose imputations in a realistic model for lung cancer mortality will be addressed.
List of speakers who are nonmembers: None