WURS-25 validity and reliability
At the same cutoff, the WURS-25 separated ADHD from healthy comparison participants more clearly than it separated ADHD from depression. The comparison group matters.
What did the original study measure?
| Study group | Participants | Result |
|---|---|---|
| Adults with ADHD | 81 | 86% reached the threshold (sensitivity). |
| Healthy comparison adults | 100 | 99% scored below it (specificity). |
| Adults with unipolar depression | 70 | 81% scored below it (specificity). |
The last two rows explain why a claim like “99% accurate” would mislead a reader. Specificity changes with the people included in the comparison. Ward, Wender & Reimherr (1993).
What did a larger study add?
Brevik and colleagues compared 646 adults with diagnosed ADHD and 908 population controls in Norway. The Norwegian WURS had high internal consistency (Cronbach?s alpha 0.967). The WURS had an area under the receiver operating characteristic curve of 0.956 (95% confidence interval 0.946–0.965). This measures how well scores ranked members of the two groups across thresholds. It cannot give an individual their probability of ADHD. Brevik et al. (2020).
What remains uncertain?
Reliability concerns the consistency of measurement. Validity concerns what inferences the scores support. A consistent recollection can still be incomplete. Neither a high internal consistency estimate nor good separation in a selected sample verifies what happened in one person’s childhood. When citing an estimate, include the version, language, threshold and comparison population.
Differential assessment in psychiatric outpatients
Kouros and colleagues studied 121 psychiatric outpatients aged 18–25 in Sweden, including patients assessed for ADHD, bipolar disorder and borderline personality disorder. At a cutoff of 39, sensitivity was 0.88 and specificity 0.70. The area under the ROC curve was 0.87 (95% CI 0.80–0.94). This sample addresses a different clinical question from a comparison with healthy volunteers. Kouros et al. (2018), Swedish psychiatric outpatient validation.
| Estimate | Interpretation in the cited sample |
|---|---|
| Internal consistency | An alpha of 0.967 in the Norwegian study describes how closely items covaried during one administration. It cannot establish the accuracy of a memory. |
| Discrimination | In Norway, the WURS AUC was 0.956, compared with 0.904 for the ASRS and 0.964 for the combination. These comparisons describe classification across candidate cutoffs in that cohort. |
| Stability across time | Lundervold and colleagues repeated the WURS-25 after about seven years. Current ADHD symptom burden influenced reports about the same childhood period. |
Brevik et al. (2020), Norwegian WURS and ASRS validation · Lundervold et al., longitudinal WURS-25 reliability study.
Using these findings in a report or review
State who completed the form, how the comparison participants were recruited and how ADHD status was established. Give the threshold for sensitivity and specificity. Give the sample and confidence interval for an AUC. A result from adult cases and population controls should be identified as such when discussing an outpatient service where most patients already have substantial difficulties.
References
Frequently asked questions
Can the WURS-25 track treatment progress?
Its time frame is childhood. Discuss a measure of current symptoms and functioning if the aim is to monitor change during treatment.