Table 3 Measurement properties of outcome measures
From: Clinical outcome measures in dementia with Lewy bodies trials: critique and recommendations
Domain | Measurement property | Definition | Methods of assessment |
|---|---|---|---|
Reliability | Â | The extent to which an OMI is free from random error | Â |
Reproducibility | The proportion of the total variance in the measurement due to true differences between patients. Includes Inter-rater rand test–retest reliability | Kappa coefficient | |
Interclass correlation coefficient | |||
Internal consistency | The degree of interrelatedness among the items | Cronbach’s alpha | |
Item-total correlations | |||
Homogeneity coefficient | |||
Measurement error | The systematic and random error of a patient’s score that is not attributed to true changes in the construct to be measured | Standard error of measurement | |
Bland–Altman method | |||
Validity | Â | The extent to which an OMI the construct that it is supposed to measure | Â |
Content validity | The degree to which the OMI covers the important parts of the construct to be measured | Lynn’s Content Validity Assessment from an expert panel | |
Construct validity | The degree to which the scores of an OMI are consistent with the expected scores, based on the existing knowledge of the construct. Includes structural, hypothesis-testing, and cross-cultural validity | Structural validity: Factor analysis | |
Hypothesis-testing: Convergent, divergent, and internal validity compared to another validated measure | |||
Cross-cultural: measurement invariance in different populations | |||
Criterion validity | The degree to which the scores of an OMI are an adequate reflection of the gold standard (if available) | Criterion-outcome measure agreement or correlation coefficient | |
Responsiveness | Responsiveness | The ability to detect change over time in the construct to be measured | Effect size |
Standardized response mean | |||
Responsiveness statistic |