Learning outcomes

A student that has completed this course should be able (independently and cooperatively) to:

- Discuss and use selected designs. Write randomisation protocols.

- Specify the linear model following the randomisation-distribution.

- Perform a randomized experiment and analyse the data.

- Write and present reports

- Define and explain key concepts. State and use important results; prove selected results.

Content

Concepts: Code of Ethics of the International Statistical Institute. Experimental and observational units. Replication versus subsampling. Random effects. Contrasts, covariate.

Designs: Completely randomized block, row-column designs (e.g. "Latin squares"), cross-over, and factorial. Matching. Selecting replication-numbers to satisfy error-requirements (power) given effect-size. Propensity scores from logistic regression.

Statistical-analysis reflecting the randomized assignment of treatments to experimental units. Mean-unbiased estimators, analysis of variance (anova), analysis of covariance (ancova), multiple-comparisons and hypothesis testing.

Instruction

Lectures, seminars and computer-assisted work.

Assessment

Written examination at the end of the courses. Compulsory assignments.