RESEARCH REPORT POSTER DISPLAY

APPLICATION OF RASCH ANALYSIS TO THE KOREAN VERSION OF THE GROSS MOTOR FUNCTION MEASURE. Park S¹, Velozo C¹, Yi C², Kwon H³, Kwon O², Ahn D⁴, Jeon H²; ¹University of Florida, Gainesville, USA. ²Yonsei University, Wonju, Korea (South). ³Daegu University, Daegu, Korea(South). ⁴Inje University, Kimhae, Korea(South)

PURPOSE: The purpose of this study was to investigate the item-level psychometrics of a Korean translation of the Gross Motor Function Measure (GMFM) using Rasch analysis. RELEVANCE: The translated GMFM is the most widely used outcome measure in Korea. Unfortunately, the translation is based on an earlier version of the GMFM and there is little evidence that the translated instrument shows good psychometric characteristics. PARTICIPANTS: The participants were 206 Korean children with cerebral palsy diagnosed by physician. Their mean age was 4.9 years (range: 8 months to 14.5 years). METHODS: The Korean GMGM user’s manual was re-translated to reflect the most recent changes to the English version. The final translation was based on unanimous agreement of 29 Korean physical and occupational therapists on both the manual text and item definitions. GMFM data for 206 children with cerebral palsy were then collected by the 29 therapists across 11 outpatient rehabilitation facilities in Korea between August 2004 and April 2005. ANALYSIS: The Winsteps software program was used to assess whether the Korean version of the GMFM fit the Rasch measurement model. RESULTS: The re-translated version of the GMFM demonstrated good internal construct validity showing a person separation reliability of 0.99 (analogous to Cronbach’s alpha). The GMFM has a person separation index was 12.7 which indicates that the instrument separates the sample into 17.3 statistically significant strata. Ninety-six percent (85/88) items fit the Rasch model. The items, “Supine: Lift head 45 degrees,” “Supine: Flexes right hip and knee through full range,” “Supine: Flexes left hip and knee through full range” showed high infit statistics. In general, the order of item difficulty of the remaining 85 items showed a logical item difficulty hierarchy with the “lying and rolling” items being the easiest and the “jumping and hopping” items being the most difficult. CONCLUSIONS: The Korean version of the GMFM shows outstanding item-level psychometric qualities. IMPLICATIONS: This study suggests that the translated GMFM produces valid measures of gross motor ability in children with cerebral palsy and furthermore appears to be sensitive in differentiating children with different gross motor abilities. KEYWORDS: Cerebral palsy, Rehabilitation, Treatment Outcome. FUNDING ACKNOWLEDGEMENTS: This study was unfunded. CONTACT: soyeonparkpt@gmail.com