Modelling correlated data: Multilevel models and generalized estimating equations and their use with data from research in developmental disabilities.
MLMs map personal change in IDD data; GEEs guard against model mistakes.
01Research in Context
What this study did
Dimitrios et al. (2018) wrote a how-to guide for researchers who work with clustered IDD data. They compare two statistical tools: multilevel models (MLMs) and generalized estimating equations (GEEs).
The paper lists when each tool works best and what can go wrong. No new data were collected; the goal was to stop bad stats in disability research.
What they found
MLMs are the pick when you want to chart each person’s growth over time. GEEs are safer if you might have the wrong model shape.
Pick the wrong tool and you can mis-count standard errors, which leads to false-positive results.
How this fits with other research
Raab et al. (2018) put the MLM advice into action. They used a multilevel model in an RCT with toddlers who had developmental delays. The model split child effects from practitioner effects, just as Dimitrios recommended.
DeHart et al. (2019) echo the theme for single-subject designs. They show that mixed-effects modeling keeps each client’s own trend while still giving p-values, lining up with Dimitrios’s push for person-level stats.
Georgiades et al. (2014) and T-Tsai et al. (2014) both used latent growth methods years earlier. Dimitrios updates the field by saying MLMs now do the same job more easily for IDD data.
Why it matters
If you run or read IDD research, this paper is your stats checklist. Use MLMs when you track change across time points or nest kids in classrooms. Use GEEs when the shape of change is messy or unknown. Share the guide with grad students or journal clubs so future studies give clean, believable results.
Want CEUs on This Topic?
The ABA Clubhouse has 60+ free CEUs — live every Wednesday. Ethics, supervision & clinical topics.
Join Free →Open your last data file and check if you used the right model—switch to MLM if tracking growth, GEE if shape is unclear.
02At a glance
03Original abstract
BACKGROUND: The use of Multilevel Models (MLM) and Generalized Estimating Equations (GEE) for analysing clustered data in the field of intellectual and developmental disability (IDD) research is still limited. METHOD: We present some important features of MLMs and GEEs: main function, assumptions, model specification and estimators, sample size and power. We provide an overview of the ways MLMs and GEEs have been used in IDD research. RESULTS: While MLMs and GEEs are both appropriate for longitudinal and/or clustered data, they differ in the assumptions they impose on the data, and the inferences made. Estimators in MLMs require appropriate model specification, while GEEs are more resilient to misspecification at the expense of model complexity. Studies on sample size seem to suggest that Level 1 coefficients are robust to small samples/clusters, with any higher-level coefficients less so. MLMs have been used more frequently than GEEs in IDD research, especially for fitting developmental trajectories. CONCLUSIONS: Clustered data from research in the IDD field can be analysed flexibly using MLMs and GEEs. These models would be more widely used if journals required the inclusion of technical specification detail, simulation studies examined power for IDD study characteristics, and researchers developed core skills during basic studies.
Research in developmental disabilities, 2018 · doi:10.1016/j.ridd.2018.04.010