This course is designed to provide a detailed understanding of the use of multilevel regression modeling for data analysis, as well as to provide a basic familiarity with non-parametric approaches and Bayesian modeling. Concepts and techniques are demonstrated using the statistical platform R, which is open source (weblink http://www.r-project.org/) and runs under most operating systems. Learning is reinforced through weekly assignments that involve working with different types of data.

20 hours, 2 hours per week

Coursework based assessment (100%)

Aims

■ To introduce students to basic techniques involved in organizing and processing complex datasets.

■ To provide a non-technical introduction to nonparametric and robust techniques to improve on parametric statistics and detect outliers.

■ To provide a basic understanding of the regression framework, including how to express study design through regression.

■ To provide an understanding of multilevel regression models and their use in experimental research.

■ To provide a basic familiarity with Bayesian approaches to modelling data;

■ To train students to use the statistical programming language R.

By the end of this course students will be able to:

■ use R to organize, restructure, and visualise complex datasets;

■ explain the basic ideas behind resampling and robust statistics and their relation to classic parametric techniques;

■ make predictions from a multiple regression equation and explain the interpretation of parameter estimates;

■ express various study designs within a multilevel regression framework;

■ compute basic quantities within a Bayesian framework.

Students must submit at least 75% by weight of the components (including examinations) of the course's summative assessment.