ISSN 2043-8087
Journal of Experimental Psychopathology
 Volume 2, Issue 2, 271-293, 2011
Abstract  E-mail Abstract   Article Purchase as PDF 
Return to Issue List 
A Primer on Using Multilevel Models in Clinical and Experimental Psychopathology Research

Authors
Andy P. Field (a), Daniel B. Wright (b)
(a) School of Psychology, University of Sussex
(b) Psychology Department, Florida International University

Volume 2, Issue 2, 2011, Pages 271-293
DOI: http://dx.doi.org/10.5127/jep.013711

Abstract
A Multilevel model is a statistical tool for analysing data that has a hierarchical data structure (in other words, data are nested within contexts). This paper describes what a multilevel model is, how it is described mathematically, the advantages of using this data analysis technique, and some practical issues to consider. We then move on to describe the application of multilevel models using two scenarios pertinent to researchers interested in clinical trial analysis and experimental psychopathology research. We describe how to use the software R to run these analyses.

Table of Contents
A Primer on Using Multilevel Models in Clinical and Experimental Psychopathology Research
I Already Know How to Analyse My Data, Why Should I Learn Something New?
  Overcoming the Independence of Errors Assumption
  Homogeneity of Regression Slopes
  Missing Data
What is a Multilevel Model?
  Fixed and Random Coefficients
    The random intercept model
    The random slope model
  Building a multilevel model
Some Practical Issues
  Assessing the Fit and Comparing Models
  Types of covariance structures
  Assumptions
  Sample Size and Power
  Centring Variables
Doing Multilevel Models
  A Very Brief Guide to R
  Example 1: Clinical Trial Data
    Initializing the package and importing the data
    Fitting a baseline model
    Fitting a random intercepts model
    Fitting a model with random intercepts and slopes
  Example 2: Reaction Time Data
    Importing the data
    Fitting a baseline model
    Modelling covariance structures
    Including the time series variable
    Fitting a random slopes model
Summary
Further Reading
References

Correspondence to
Prof. Andy P. Field, Child Anxiety Theory and Treatment Laboratory (CATTLab), School of Psychology, University of Sussex, Falmer, Brighton, East Sussex, BN1 9QH.

Keywords
Multilevel models, clinical data, psychopathology

Dates
Received 17 Dec 2010; Revised 21 Jan 2011; Accepted 21 Jan 2011; In Press 5 May 2011







Bookmark and Share

Related articles by AUTHORS
None Found

Related articles by KEYWORDS
None Found




© Copyright 2009-2016 Textrum Ltd . All rights reserved. Published in the UK. - Contact Us Advertise | Privacy Policy | Terms of Use