Statistical Evaluation Criteria for Outcome Measures

Statistical Evaluation Criteria for Outcome Measures
Reliability
Chronbach’s α or split-half statistics
Excellent ≥ 0.80
Adequate 0.70 – 0.79
Poor < 0.70
Test-retest or Inter-rater reliability (ICC or kappa statistics)
Excellent ≥ 0.75
Adequate 0.40 – 0.74
Poor < 0.40
Validity
Construct/convergent and concurrent correlations
Excellent ≥ 0.60
Adequate 0.31 – 0.59
Poor ≤ 0.30
Receiver Operating Characteristic (ROC) analysis – Area under the curve (AUC)
Excellent ≥ 0.90
Adequate 0.70 – 0.89
Poor < 0.70
Responsiveness
Sensitivity to change (standardized effect sizes)
Small < 0.5
Moderate 0.5 – 0.79
Large ≥ 0.8
Floor/ceiling effects
Excellent No floor/ceiling effects
Adequate Floor and ceiling effects ? 20%
Poor > 20% of patients attain either the minimum or maximum score

*Based on tables presented in Salter, Jutai & Teasell, 2005 and
in Andresen, 2000.

Andresen, E. M. (2000). Criteria for assessing the tools of
disability outcomes research. Archives of Physical Medicine and
Rehabilitation
, 81, 12(2), S15-S20.

Salter, K., Jutai, J., Foley, N., Teasell, R. (2005). Outcome
Measures in Stroke Rehabilitation. http://www.ebrsr.com/modules/module21.pdf
(accessed October 5, 2006).

Landis, J. R., Koch, G. G. (1977). The measurement of observer
agreement for categorical data. Biometrics, 33, 159-174.