People can change: Measuring individual variability in rehabilitation science

Lauren Terhorst, Shannon B. Juengst, Kelly B. Beck, Saul Shiffman

Research output: Contribution to journalArticlepeer-review

3 Scopus citations


Purpose/Objective: The purpose of this report is to provide rehabilitation researchers with an explanation of multilevel item response theory (MLIRT), specifically applied to data collected using ecological momentary assessment (EMA) methods. Design: This is a didactic brief report of a statistical method. The advantages of the method are illustrated using examples from the literature or clinical experience, and potential implications for rehabilitation science are highlighted. Results: Multilevel item response theory can be a used in rehabilitation research to develop items that characterize within-individual changes in disability, measure disability with instruments that have not yet been used in an EMA framework, examine temporal patterns of behaviors or symptoms that are associated with disability, or create adaptive tests based on individual performance. Conclusions: MLIRT methods are underutilized in rehabilitation research despite their unique advantages. Psychometric properties of instruments used to measure change over time should be evaluated at the within-individual level. Additionally, MLIRT offers opportunity to investigate temporal variability of behaviors or symptoms, and can potentially reduce participant burden when used in adaptive testing.

Original languageEnglish (US)
Pages (from-to)468-473
Number of pages6
JournalRehabilitation Psychology
Issue number3
StatePublished - Aug 1 2018


  • Item response theory
  • Measurement
  • Multilevel analysis
  • Rehabilitation research
  • Statistical models

ASJC Scopus subject areas

  • Physical Therapy, Sports Therapy and Rehabilitation
  • Rehabilitation
  • Clinical Psychology
  • Psychiatry and Mental health


Dive into the research topics of 'People can change: Measuring individual variability in rehabilitation science'. Together they form a unique fingerprint.

Cite this