Clinical measurement property | Rating method | |
---|---|---|
Inter-rater reliability | A measure of the consistency for the scores obtained between two rates that are measuring the same subject. | |
Test-retest reliability | A property of which the stability of the result using the same outcome measure with the same group of subjects in a repeated test. | |
Internal consistency | A reliability measure; ensuring all measurement items measure the same context. | |
 | Kappaa | A preferred statistical measure for rating the inter-rate reliability; cut-off value: ≥ 0.70– acceptable [10, 16, 17]. |
Cronbach’s αb | A statistical measure for rating the internal consistency; cut-off value: ≥ 0.70– acceptable [10, 18]. | |
ICCc | A statistical measure for rating the reliability; cut-off value: ≥ 0.70– acceptable [10]. | |
Concurrent validity | Â | A type of criterion validity; examining the consistency of the score obtained by the testing instrument compared with a gold standard. |
Predictive validity | Â | A type of criterion validity; measuring a correlation between the testing score and future events. |
Construct validity | rd | An essential measurement validity which assesses whether a testing instrument measures what it was made to measure by comparing hypothetical and non-criterion parameters. |
A statistical measure for rating the correlation of linear relationship between two variables; cut-off values: > 0.75 - excellent; 0.50 to 0.75 - moderate; 0.25 to 0.49 - mild; < 0.25 – weak [14]. | ||
Diagnostic accuracy | ROCe | An ability to discriminate the population with targeted health condition. |
A statistical measure for expressing how well the measure can distinguish between two populations; cut-off values: 0.9 to 1.0 - excellent; 0.8 to 0.89 - very good; 0.7 to 0.79 - good; 0.6 to 0.69 - sufficient; 0.5 to 0.59 - insufficient; < 0.5 - no use [15]. | ||
Responsiveness to change | An ability to detect the changes in health outcomes following the intervention. | |
 | MDCf | A statistical measure for rating the response to changes which is not due to the measurement error. |