Sheng Luo1, Yuanyuan Liu1, Jeanne A Teresi2,3, Glenn T Stebbins4, Christopher G Goetz4. 1. Department of Biostatistics, University of Texas Health Science Center, School of Public Health, Houston, Texas, USA. 2. Columbia University Stroud Center at New York State Psychiatric Institute, New York, NY, USA. 3. Research Division, Hebrew Home at Riverdale, Bronx, New York, USA. 4. Department of Neurological Sciences, Rush University Medical Center, Chicago, Illinois, USA.
Abstract
OBJECTIVE: Test if differential item functioning due to gender, age, race/ethnicity, or education impacts Unified Dyskinesia Rating Scale scores. BACKGROUND: Testing rating scales for differential item functioning is a core validation step. If differential item functioning exists, interpretation of item scores must consider secondary influences on dyskinesia ratings. METHODS: Using Unified Dyskinesia Rating Scale translation databases (N = 3,132), we tested uniform and nonuniform differential item functioning. We required confirmation by two independent methods and considered differential item functioning pertinent if McFadden pseudo R2 magnitude statistics exceeded negligible ratings. RESULTS: No age, race/ethnicity, or education nonuniform differential item functioning was identified. Gender nonuniform differential item functioning occurred for 2 items, both with negligible magnitude. Gender, race, and education uniform differential item functioning was observed for multiple items, all with negligible magnitude. CONCLUSIONS: The Unified Dyskinesia Rating Scale items effectively capture dyskinesia severity without pertinent gender, age, race/ ethnicity, or education influence.
OBJECTIVE: Test if differential item functioning due to gender, age, race/ethnicity, or education impacts Unified Dyskinesia Rating Scale scores. BACKGROUND: Testing rating scales for differential item functioning is a core validation step. If differential item functioning exists, interpretation of item scores must consider secondary influences on dyskinesia ratings. METHODS: Using Unified Dyskinesia Rating Scale translation databases (N = 3,132), we tested uniform and nonuniform differential item functioning. We required confirmation by two independent methods and considered differential item functioning pertinent if McFadden pseudo R2 magnitude statistics exceeded negligible ratings. RESULTS: No age, race/ethnicity, or education nonuniform differential item functioning was identified. Gender nonuniform differential item functioning occurred for 2 items, both with negligible magnitude. Gender, race, and education uniform differential item functioning was observed for multiple items, all with negligible magnitude. CONCLUSIONS: The Unified Dyskinesia Rating Scale items effectively capture dyskinesia severity without pertinent gender, age, race/ ethnicity, or education influence.
Authors: Christopher G Goetz; Yuanyuan Liu; Glenn T Stebbins; Lu Wang; Barbara C Tilley; Jeanne A Teresi; Douglas Merkitch; Sheng Luo Journal: Mov Disord Date: 2016-11-07 Impact factor: 10.338
Authors: Carlo Colosimo; Pablo Martínez-Martín; Giovanni Fabbrini; Robert A Hauser; Marcelo Merello; Janis Miyasaki; Werner Poewe; Cristina Sampaio; Olivier Rascol; Glenn T Stebbins; Anette Schrag; Christopher G Goetz Journal: Mov Disord Date: 2010-07-15 Impact factor: 10.338