REML and multi-level models

Course details

Who Should Attend Practical experience of ordinary analysis of variance is necessary, but there will be no need for any complicated maths. The lectures will be interspersed with practicals to introduce you to real-life data sets and illustrate the methods. The practicals also give you the opportunity to discuss your own problems and investigations with the presenter.
Duration / Date 1 day
Overview REML provides several important types of analysis, with application areas that include biology, medicine, industry and finance. It can be used to analyse models with several types of error variation (often called “multi-level models”), as well to fit models to correlated data like repeated measurements. GenStat has a very powerful set of REML facilities, that are nevertheless very straightforward and easy to use.
Learning Objectives The course is designed to introduce you to REML in GenStat, and give you the underlying knowledge and confidence to use it correctly and effectively. It shows how the REML menus guide you through even very complicated analyses, and also explains the REML commands so that you can program any non-standard analyses that you need.
Methods of Evaluation Subsequent to instructor facilitated sessions, participants will be expected to complete example analyses and exercises unaided.
Training Methods Instructor facilitated, interactive computer sessions.
Contents The Course is in 4 sessions:

  1. Analysis of variance Similarities and differences between REML and ordinary ANOVA, unbalanced designs, variance components, residual plots, means plots and predictions.
  2. Meta analysis Combined analysis of several data sets to provide estimates of treatment effects that use all the available information.
  3. Spatial analysis Modelling covariances between effects of a random term, assessing different covariance models, advantages over conventional blocking for experiments with many treatments
  4. Repeated measurements Covariance models for time effects, random-coefficient regression.
Location Details here

Next steps

To discuss this course email training, or contact us on +44(0)1442 450230

Contact Us

For further information about this course or on-site training, please email the training team or call them on +44 (0)1442-450230.