Objective: To analyse in patients with Parkinson's disease the psychometric performance of the Activities-specific Balance Confidence scale (ABC) and its 3 short versions, using both a classical test theory approach and Rasch analysis. Methods: A sample of 217 patients with Parkinson's disease was assessed by ABC and the 3 short versions: Berg Balance Scale, Fear of Falling Measure, and the Unified Parkinson's Disease Rating Scale. Results: Cronbach's α was 0.95 in ABC, and ranged from 0.88 to 0.90 in its short versions. At Rasch analysis the 11 original rating categories were collapsed to 5 levels. In ABC 15 out of 16 items fitted the Rasch model. The ABC showed high correlation only with Fear of Falling Measure (r = 0.85), and excellent correlation with its short versions (r > 0.93). The 3 short versions showed a limited range of item difficulty estimates, low reliability levels, floor effect, a mismatch between mean item difficulty and mean ability of the patients. Conclusion: In patients with Parkinson's disease: (i) the ABC has an adequate unidimensionality; (ii) the selection of its items is satisfactory, although there is room for some minor refinement; (iii) the 0-10 rating scale should be simplified, and a 5-level response format seems able to improve the measurement qualities of the scale; (iv) the 3 short versions showed some psychometric limitations in comparison with the ABC.
Rasch validation of the Activities-specific Balance Confidence Scale and its short versions in patients with Parkinson's disease
Ferriero G
2014-01-01
Abstract
Objective: To analyse in patients with Parkinson's disease the psychometric performance of the Activities-specific Balance Confidence scale (ABC) and its 3 short versions, using both a classical test theory approach and Rasch analysis. Methods: A sample of 217 patients with Parkinson's disease was assessed by ABC and the 3 short versions: Berg Balance Scale, Fear of Falling Measure, and the Unified Parkinson's Disease Rating Scale. Results: Cronbach's α was 0.95 in ABC, and ranged from 0.88 to 0.90 in its short versions. At Rasch analysis the 11 original rating categories were collapsed to 5 levels. In ABC 15 out of 16 items fitted the Rasch model. The ABC showed high correlation only with Fear of Falling Measure (r = 0.85), and excellent correlation with its short versions (r > 0.93). The 3 short versions showed a limited range of item difficulty estimates, low reliability levels, floor effect, a mismatch between mean item difficulty and mean ability of the patients. Conclusion: In patients with Parkinson's disease: (i) the ABC has an adequate unidimensionality; (ii) the selection of its items is satisfactory, although there is room for some minor refinement; (iii) the 0-10 rating scale should be simplified, and a 5-level response format seems able to improve the measurement qualities of the scale; (iv) the 3 short versions showed some psychometric limitations in comparison with the ABC.| File | Dimensione | Formato | |
|---|---|---|---|
|
2014_JRM.pdf
non disponibili
Tipologia:
Versione Editoriale (PDF)
Licenza:
Copyright dell'editore
Dimensione
575.43 kB
Formato
Adobe PDF
|
575.43 kB | Adobe PDF | Visualizza/Apri Richiedi una copia |
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.



