ESRC Research Methods Programme

Introduction to the problem of missing values (missing responses, missing explanatory variables, bias, precision)
Taxonomy of missing data (MCAR, MAR, NMAR, ignorable/non-ignorable)
Consequences of ignoring missing data
Main approaches to adjustment and correction for missing values
Introduction to imputation (deterministic/random, single/multiple)
Introduction to weighting
Consultant's forum

Pre-requisites:

Comfortable with basic probability, conditional probability, Bayes' Theorum and basic statistics (variance, standard errors, t-tests)
Familiarity with linear regression

The course will involve a combination of lectures, interactive group work and practical computing using R, Stata and WinBUGS. Prior knowledge of these packages is not required.

Each workshop is limited to 20 people and everyone who books will be asked to complete a short questionnaire to allow the organisers to refine the programme to reflect participants’ interests.

The workshop is sponsored by the ESRC Research Methods Programme and is free to participants. However, no expenses can be paid in respect of travel and accommodation.

index