620-270 Applied Statistics

Students may only gain credit for one of 620-270, 620-272 or 620-370.

Passing 620-270 precludes subsequent credit for 620-152 or 620-160.

Students who have completed 620-371 or 620-372 may not enrol in this subject for credit.

One of 620-131, 620-152 or 620-160.

This subject demonstrates the importance of statistical methods for interpreting data, the role of exploratory and formal data analysis and the importance of experimental design. Students should learn to examine data to determine underlying structures, formulate statistical models for a range of practical situations and check the assumptions of the model in specific situations. They should also learn to use the computer to carry out standard statistical analyses and to express conclusions in scientifically useful terms.

Introduction to statistical inference topics include estimation; confidence intervals; hypothesis testing including the power of tests; and determination of sample size using the width of confidence intervals and power. Correlation and regression topics include assumptions; method of least squares; interpretation; hypothesis testing; confidence and prediction intervals; residuals; regression diagnostics; transformations; collinearity; model selection; and polynomial regression. Analysis of variance models (one-way and two-way with equal numbers of observations per cell) topics include model; assumptions; estimation and hypothesis testing; interaction and its interpretation; transformations; residuals; and diagnostics. Design of experiments topics include randomisation; replication; blocking; standard designs including completely randomised, randomised block and Latin square designs; factorial experiments: analysis; interpretation; and introduction to confounding. Analysis of covariance topics include detailed treatment of the case with one factor and one covariate; and extension to more complex situations. Contingency tables topics include tests for association; odds ratios, including confidence intervals; and introduction to loglinear models.

Up to 50 pages of written assignments due during the semester (25%); a 3-hour written examination in the examination period (75%).