WEBINAR (Sponsored by the Analytic Special Interest Group of the TBI Model Systems)
Reliability and Error in Measurement Instruments developed with Classical Test Theory and Item Response Theory
Presenter: Alan J. Kozlowski, Ph.D., B.Sc (PT)
Date/Time: April 22, 2014; 4:00 PM EDT
The webinar has concluded. Slides from the presentation may be downloaded from the TBI National Data and Statistical Center website.(PDF)
Analyses conducted using data from the Model Systems National Databases allow researchers and clinicians to examine patterns of change in constructs such as physical and cognitive functioning using scores from the motor and cognitive subscales of the FIM™ Instrument, psychological functioning using the Patient Health Questionnaire (PHQ-9), or a variety of other outcome measures. However, the original versions of these instruments were developed using Classical Test Theory (CTT) methods. One limitation of CTT-based instruments is that the scores can be interpreted only for the scale or subscale summary scores; individual items cannot be interpreted at a single point in time or for change over time. Another limitation of many multi-item CTT-based instruments, whether patient-reported or clinician-rated, is that the resulting scores are ordinal in nature, even though numerically they have the appearance of interval-level scaling. Together, these and other design issues contribute error to measurement, and consequently influence the reliability of the instrument. As we look across time, measurement reliability becomes more important.
In this webinar, the first of three in the series, the presenters will discuss the aspects of test development using CTT and a contemporary approach called Item Response Theory (IRT) that relate to measuring and interpreting change over time.
Alan J. Kozlowski, Ph.D., B.Sc (PT) is an Assistant Professor and Director of the Bionics Research Program in the Department of Rehabilitation Medicine in the Icahn School of Medicine at Mount Sinai. He received his B.Sc. in physical therapy in 1991 and his PhD in Rehabilitation Sciences in 2010, both from the University of British Columbia. He practiced as a physical therapist and rehabilitation manager in work disability prevention before pursing graduate studies. Dr. Kozlowski completed his postdoctoral fellowship at the Center for Rehabilitation Outcomes Research at Rehabilitation Institute of Chicago and the Center for Healthcare Studies at Northwestern University, in which he constructed individual growth models for FIM Instrument scores for persons with spinal cord injuries and traumatic brain injuries. He has instructed courses in physical therapy clinical skills, measurement in clinical practice (online), and research methods.
We at NARIC are posting this for information purposes. We are not involved in the planning or presentation of this webinar.