Reliabilities and Error Rates of the Three Competency Measures
| Estimate | Description | Scales | Competency Measure | ||
|---|---|---|---|---|---|
| CAST-MR | McCAT-CA | ECST-R | |||
| Reliability | Inter-rater reliability establishes the reproducibility of scores when evaluated by two or more experts at the same time; lower estimates equal larger errors. | Competency | 0.90* | 0.83 | 0.93 |
| Feigning | NA | NA | 0.996 | ||
| SEM | Standard error of measurement (SEM) measures the likely variability in the accurate measurement of a single score; larger scores equal greater errors. | CompetencyFeigning | 1.15† | 1.28 | 1.28 |
| Feigning | NA | NA | 0.18 | ||
| 95% CI | 95% Confidence interval establishes the range of scores possible for a single score that is likely to occur most of the time (i.e., 95 of 100 times); larger scores equal greater errors. | CompetencyFeigning | 2.25 | 2.51 | 2.51 |
| Feigning | NA | NA | 0.35 | ||
* Inter-rater reliabilities were reported only as percentages; this correlation represents test-retest reliability for nonforensic cases.
† Based on unweighted scale averages (M = 10.14, SD = 3.63) for four small subsamples of competency cases (n = 58).