Table 3.
Standard error of measurement (SEM) for depression tests calibrated to the PROMIS metric
Test | Mean (SD) | Median | Interquartile range | Range |
---|---|---|---|---|
PHQ-9, PROMIS metric | 4.9 (1.4) | 4.6 | 3.5–6.6 | 2.8–8.3 |
PROMIS short form, PROMIS metric | 3.5 (1.8) | 2.6 | 1.8–5.5 | 1.5–7.5 |
PROMIS-simulated CAT, 9 items | 2.7 (1.4) | 2.0 | 1.7–3.4 | 1.4–5.2 |
PROMIS-simulated CAT, 8 items | 2.8 (1.3) | 2.2 | 1.8–3.4 | 1.5–5.2 |
PROMIS-simulated CAT, 3 items | 3.7 (1.3) | 3.2 | 2.7–5.6 | 2.3–5.6 |
Note that the data presented in the table refer to means and standard deviations of standard errors of measurement; smaller is better. The PROMIS metric uses a SD of 10; on the unadjusted IRT logit scale, then, a mean SEM of 4.9 corresponds to a mean SEM of 0.49 standard theta units