Answer timings classify insomnia-questionnaire labels in a student cohort
Source check: Source checked. Not independent scientific confirmation.
Finding: Learned classifiers used the time taken to answer seven questionnaire items to distinguish concurrent insomnia-symptom score labels. Repeated internally balanced evaluations reported an AUROC of 0.824 for raw timings. This is agreement with a self-report scale, not an insomnia diagnosis or a forecast.
Technology: Logistic regression; multilayer perceptron
Publication: Peer-reviewed paper · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.
- AI contribution
- Logistic regression and a multilayer perceptron learned a mapping from seven response durations to the concurrent ISI score threshold; model interpretation assessed derived timing features.
- Test and result
- A 2,729-student screening cohort supplied questionnaire timings. Ten random balanced samples each contained141 symptom-positive and141 symptom-negative participants, evaluated with10-fold cross-validation. Raw-timing logistic regression and MLP each reported AUROC0.824.
- Limitation / still unknown
- Reference labels are ISI scores of at least8, not independent clinical interviews. The same questionnaire supplied timing features and reference scores. Model comparison and first-fold tuning do not establish a fully nested evaluation; preprocessing and feature selection containment are incomplete. The source reports inconsistent final participant denominators, and balanced resampling does not establish population predictive values. The single-university setting and uncontrolled device/environment effects limit generalization; prospective utility is unestablished.
Evidence sources and annotation context
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
Evidence sources: https://pmc.ncbi.nlm.nih.gov/articles/PMC11261114/
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: not linked in this catalog; this does not mean no paper exists.
Review date: 21 Nov 2026. Review scope: Original primary front;Sections2.1–2.7,3.4–3.5 and4;current primary bibliographic status metadata.
No follow-up recorded; this is not a claim that no follow-up exists. Follow-up review scope and date are not recorded. No source-supported relationships recorded.
Editorial assessment: Impact 1/5 · Validation 2/5 · Novelty 2/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.