![]() |
CiteULike | ![]() |
rfieo's CiteULike | ![]() |
![]() |
|
![]() |
Register | ![]() |
Log in | ![]() |
Applying psychometric criteria to functional assessment in medical rehabilitation: III. Construct validity and predicting level of care.by: R. F. Harvey, B. Silverstein, M. A. Venzon, K. M. Kilgore, W. P. Fisher, M. Steiner, J. P. Harley
Archives of physical medicine and rehabilitation, Vol. 73, No. 10. (October 1992), pp. 887-892.
|
Reviews
[Write a review of this article]
Notes for this articlerehab, rasch
Find related articles from these CiteULike users
Find related articles with these CiteULike tags
Posting History
AbstractA discriminant analysis was performed using Rasch ability estimates derived from four Patient Evaluation and Conference System (PECS) subscales, to distinguish among the functional independence and impairment profiles exhibited by patients admitted into any of three levels of medical rehabilitation delineated by the National Association of Rehabilitation Facilities (NARF): (1) inpatient hospital, (2) atypical nursing home (intermediate), or (3) day program. Two discriminant functions accounted for 91% and 9% of the between-group variance, respectively. Cross-validated classification of patients into one of the three levels of care, based on discriminant function scores, produced 75% correct classification; a 66% improvement over the percentage of correct classification likely by chance alone. Results support the construct validity of the PECS subscales and indicate they may be useful in validating clinically-based admission decisions among three of the levels of care promulgated by NARF.
BibTeX record
RIS record