# Instructor Class Description

Time Schedule:

Norbert David Yanez Iii
BIOST 513
Seattle Campus

### Medical Biometry III

Analysis of categorical data including two sample methods, sets of 2 x 2 tables, R x C tables, and logistic regression. Classification and discrimination techniques. Survival analysis including product limit estimates and the Cox proportional hazards model. Prerequisite: BIOST 512 or permission of instructor. Offered: Sp.

Class description

This course introduces the principles and methods of statistical inference for categorical data and survival data. The major topics covered are: contingency table methods; logistic regression; Kaplan-Meier and log-rank methods; and Cox regression.

Student learning goals

Perform standard tests of homogeneity and trend with data from 2xC tables and compute odds ratio estimates and confidence intervals from 2x2 and stratified 2x2 tables.

Perform logistic regression to estimate model coefficients and odds ratios (adjusted), compute and interpret coefficient and odds ratio confidence intervals, and test hypotheses that one or more coefficients in the logistic regression model are zero.

Compute and interpret the product limit (Kaplan-Meier) estimate of survival and associated confidence intervals.

Perform and interpret the log-rank test for differences between survival curves with right censored survival data.

Perform Cox regression to estimate proportional hazards model coefficients, interpret coefficient estimates and confidence intervals, and test hypotheses that one or more coefficients in the regression model are zero.

Interpret and critique the results of application of these statistical techniques as found in the health sciences literature.

General method of instruction

Lectures with discussion sections.

Recommended preparation

Upon entering this course the student is expected to be able to: (1) use a statistical package (STATA) to enter data and calculate summary statistics, (2) compute and interpret confidence intervals for one-sample and 2-sample situations, (3) perform "multivariable" linear regression, interpret model coefficients, compute confidence intervals for model coefficients, and test hypotheses using partial F-tests and Wald tests, (4) perform 2-sample hypothesis tests for binary data (i.e., proportions), (5) compute and interpret confidence intervals for one-sample and two-sample situations.