Coefficient of variation as a measure of subject effort

Arch Phys Med Rehabil. 1995 Jun;76(6):516-20. doi: 10.1016/s0003-9993(95)80504-4.


This study describes average coefficients of variation (CVs) for various isometric strength tests in an impaired population, determines if differences exist in CVs between the various tests or between persons with different diagnoses, and determines the correlations between the CVs of differing tests. Survey data were collected from a consecutive sample of 270 functional capacity evaluations conducted on previously injured patients at an outpatient rehabilitation center. Patients were grouped by diagnosis. Three trials on each of up to eight isometric strength tests (two pushing, two pulling, and four lifting) were performed. CVs were calculated for each test for each subject. Mean CVs were then calculated for each subject for all tests completed. Mean CVs were also calculated for each test for all subjects who performed the test. Mean CVs on the various tests ranged from 6.1% to 12.0%. There was a significant difference in means between tests (p < .0001). Mean CVs between diagnostic groups were not different. No evidence was found to indicate that the CV can be used independently to determine sincerity of subject effort. CVs will vary depending on the specific test. CVs do not necessarily vary because of differing diagnoses. Judgments of sincerity of effort should not be made when based solely on the CV.

MeSH terms

  • Adult
  • Analysis of Variance
  • Biomechanical Phenomena
  • Data Interpretation, Statistical
  • Female
  • Humans
  • Isometric Contraction
  • Male
  • Middle Aged
  • Patient Compliance*
  • Physical Exertion*
  • Physical Therapy Modalities / methods
  • Reproducibility of Results*