Multilevel Modelling Using Stata II - Moving beyond two-level continuous response models
Application deadline: TBC
This course is only open to social science PhD students from the Bristol, Bath, Exeter, Plymouth or UWE arms of the South West Doctoral Training Partnership (SWDTP).
To apply for a place, please email firstname.lastname@example.org and give your name, department and institution. Please only apply for a place if you satisfy the course prerequisites and if you can attend both days.
Course tutor: Dr George Leckie
Date: 23-24 May 2017
Time: 10:00 - 17:00
Location: Small Computer Room 1.4n, School of Geographical Sciences, University Road, Bristol, BS8 1SS.
Description: This ‘second course’ in multilevel modelling covers a range of intermediate topics surrounding the analysis of continuous and binary responses (dependent or outcome variables) when the data are clustered or hierarchical.
On Day 1 we move beyond standard two-level multlilevel models to consider three-level models and cross-classified and multiple membership models for when the data are not strictly hierarchical. We demonstrate the consequences of ignoring these complexities when one naively fits two-level models in these complex settings. These modelling extensions are appropriate, for example, when individuals are simultaneously nested within different contexts such as schools and neighbourhoods, and where their membership of these contexts may change over time due to residential mobility.
On Day 2 we consider multilevel logistic regression for binary responses. We start with a review of conventional logistic regression before proceding to extend this model to multilevel settings. We consider both cross-sectional and longitudinal data examples. For example the binary voting behaviour of individuals within US states (i.e., Republican vs. Democrats), and repeated binary health outcomes of subjects within randomised control trials.
10:00 - 12:00 Session 1 - Three-level models - variance partition coefficient (VPC) & intraclass correlation coefficient (ICC), adding level-1, level-2 and level-3 covariates, consequences of ignoring clustering at level-3, consequences of ignoring clustering at level-2, treating the level-3 random-effects as fixed-effects dummy variables, ...
12:00 - 12:45 Lunch
12:45 - 14:45 Session 2 - Cross-classified models - classification notation and diagrams, interaction classifications, consequences of ignoring cross-classified structure, VPC and ICC, including level-1 and level-2 covariates, constrained hierarchical formulation, ...
14:45 - 15:00 Break
15:00 - 17:00 Session 3 - Multiple membership models - classification notation and diagrams, alternative weighting schemes, consequences of ignoring multiple membership structure, VPC and ICC, contextual effects, constrained hierarchical formulation, ...
10:00 - 12:00 Session 4 - Review of single-level binary response models - linear probability model, logistic regresssion model, probit regression model, generalized linear model formulation, latent response formulation, log-odds, odds, odds ratios, probabilities, ...
12:00 - 12:45 Lunch
12:45 - 14:45 Session 5 - Binary response models I - variance-components model, estimation methods, VPC and ICC, random-intercept model with covariates, random-slope model, contextual effects, cross-level interactions, ...
14:45 - 15:00 Break
15:00 - 17:00 Session 6 - Binary response models II - adaptive quadrature, population-averaged (marginal) vs. cluster-specific (conditional) inference, marginal models, ...
The course will also describe popular multilevel modelling resources which participants can use to support their learning after the course (useful web sites, online course, software, discussion boards, email lists, books, ...).
Each new methodological development will be illustrated with applications to social science data sets. The course will consist of an approximately 2:1mix of lectures and computer practicals using the Stata statistical software package (especially the -mixed- and –melogit- commands). On completion of this course, participants should be able to apply multilevel models to their own data using Stata.
Prerequisites: Participants must be familiar with estimating and interpreting two-level multilevel models for continuous responses to the level of knowledge obtained by completing Module 5 of the LEMMA online course (including the writing and interpretation of model equations, hypothesis testing and model selection, and the use and interpretation of variance-components models, random-coefficients, cross-level interactions).
Participants should ideally be familiar with single-level logistic regression for binary responses to the level of knowledge obtained by completing Module 6 of the LEMMA online course (including the understanding of odds, odds ratios, and probabilities). However, a brief review of single-level logistic regression will be given before we introduce multilevel modelling of binary outcomes.
Previous experience of Stata would also be highly beneficial. Participants who have not used Stata before are strongly encouraged to familiarise themselves with Stata before the course by reading and, ideally, working through the Stata practical which accompanies Module 5 of the LEMMA online course.
Optional background reading: Chapter 2 of Hox (2011) provides a review of two-level multilevel modelling for continuous response variables.
- Hox, J. (2011). Multilevel analysis: Techniques and applications, 2nd edition. Lawrence Erlbaum Associates: Mahwah, NJ.
Other details: Participants are expected to attend both days of the course. Please bring a memory stick so that you can take the course electronic materials away with you. While there is no dedicated time to analysing your own data, you are welcome to apply what you have learnt on the course to your own data during the lunch and coffee breaks should you wish to do so. The instructor will be around for some of this time and is happy to discuss your research with you and in particular to answer questions relating to carrying out multilevel modelling in your own research.
Refreshments: As this is a free course, lunch and coffee are not provided. However, there are various cafes close to the computer lab where you can purchase refreshments.
Contact: Please email George Leckie (email@example.com) if you have any queries about the course.