37495 Statistical Design and Models for Evaluation Studies6cp; 4hpw: 2hpw (lecture), 2hpw (computer lab)
Requisite(s): ((37252 Regression and Linear Models AND 37161 Probability and Random Variables) OR 36103 Statistical Thinking for Data Science )
Traditional randomised control trials remains the ‘gold-standard’ for evaluation studies but there is increasing use of alternative model-based approaches. Building on the students’ knowledge of regression models, this subject explores the use of basic RCTs where appropriate, time series approaches useful for evaluating interventions such as the Sydney ‘Lock-Out Laws’, propensity score adjustment, and two-stage regression approaches. Where possible, the subject draws on real-world case-studies to illuminate the various designs and approaches. The subject also explores concepts of statistical design relating to internal verses external validity.
Autumn semester, City campus
Detailed subject description.