Multilevel Modelling
Presenters: Dr. Damon Berridge and Dr. Mick Green
The aim of this course is to highlight the problems that occur when the hierarchical nature of many social surveys is ignored in routine statistical analysis. The classical variance components ANOVA model will be described and extended to the multilevel model. Estimation techniques such as REML will be outlined. The analysis of such models will be illustrated using appropriate software.
Cost
- Lancaster University staff and postgraduates - £50
- External participant from an academic institution - £270
- External participant from a non academic institution - £390
The course fees include all supporting documentation, refreshments and lunches.
Topics
The nature of heterogeneity and intraclass correlation. Random effects in ANOVA models, variance components and REML. The problem of bias in naïve statistical methods. Multilevel models and random parameter models. Estimation methods. Illustrations of statistical inference in statistical software such as GLLAMM.
Learning
Students will learn through the application of concepts and techniques covered in the course to real substantive issues.
Successful students will be able to:
- understand the issue of heterogeneity in social studies
- carry out advanced statistical procedures
- apply theoretical concepts
- identify and solve problems
- analyse data and interpret statistical output
