Statistical Methods in Mental Health

This module is intended to provide students with a thorough grounding in the principles and practicalities of basic statistical analysis. A large part of the teaching will be computer based, with students learning to use STATA statistical software to analyse practice data sets.

The module is optional for students on the MSc in Clinical Mental Health Sciences, but they are strongly advised to take it if contemplating a project with substantial statistical analysis involved, unless they already have very strong statistical skills. It is compulsory for students on the MSc in Mental Health Sciences Research (unless they can either demonstrate that they have fully covered the ground and learnt the relevant practical skills in a previous Programme, or that another available statistical module better fits their needs).

Module Leaders

Ms Rebecca Jones

Rebecca Jones is a statistician who began her career as a psychology graduate. She moved from the London School of Hygiene and Tropical Medicine to UCL in 2014, and divides her time between statistical support for the Division of Psychiatry MScs and working with Professor Michael King on the analysis of large epidemiological data sets.

Module Contents

The core teaching will be a series of practical workshops, taught at computers, in which students will first discuss with the statistician leading this module the principles of the statistical tests to be covered, then students will learn how to use STATA to carry out these tests on practice data sets and will learn how to interpret and report the output. Descriptive statistics, univariate analyses, regression analyses and power calculation will be taught in this way. A final session will involve introducing students to some more complex statistical techniques that they will not be expected to carry out but whose basic principles it is desirable for them to grasp. Following the main seminar series there will be a set of further problem classes which students can attend to practice the techniques learnt further. A series of Moodle exercises accompanying the seminars will involve further practice on using the statistical techniques and also on interpreting reports of analyses.

Learning Outcomes

The intended learning outcomes are:

  • To enable students to read papers that report statistical data using straightforward univariate and multivariate analyses with understanding and to be able to comment on their quality.
  • To enable students to enter and clean data in STATA statistical software and to carry out appropriate descriptive analyses and report and comment on these.
  • To enable students to select and carry out appropriate univariate tests, including Chi squared and Fisher's exact tests, t tests, one-way analysis of variance, and correlations, and to enable them to report and comment on these.
  • To familiarise students with the principles of regression analyses and enable them to carry out appropriately and report on multiple and logistic regressions using STATA software.
  • To enable students to carry out and interpret simple power analyses, such as for a clinical trial, using STATA software.
  • To familiarise students with the main principles of further techniques including factor analyses, the analysis of clustered data, multilevel modelling and survival analysis, allowing them to read papers based on these techniques with understanding.