Participant-Aware Model Validation for Repeated-Measures Data: Comparative Cross-Validation Study
Abdolamir Karbalaie, Farhad Abtahi, Charlotte K Häger, Gang Luo, Lobes Herdiman, Valentina Palama, Yihan Hu · 2026
Read Participant-Aware Model Validation for Repeated-Measures Data: Comparative Cross-Validation Study by Abdolamir Karbalaie, Farhad Abtahi, Charlotte K Häger, Gang Luo, Lobes Herdiman, Valentina Palama, Yihan Hu (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/pmc-13176815.txt.
Description
Abdolamir Karbalaie and co-authors warn that common validation habits inflate model scores when each person gives many trials. Their twenty twenty-six study deals with repeated-measures data, standard in biomechanics and digital health. If trials from one person land in both training and test sets, the model has in effect seen the answer.
The case study predicts fear of reinjury after knee ligament surgery from hop-test mechanics, with hundreds of trials from seventy-two people. Four validation schemes were compared across ten classifiers. Plain ten-fold splitting overrated accuracy badly against a leave-one-person-out scheme. Group and nested schemes gave steadier, more honest scores. The authors urge participant-aware validation for clinical trust.
Catalog record
| CREATOR | Abdolamir Karbalaie, Farhad Abtahi, Charlotte K Häger, Gang Luo, Lobes Herdiman, Valentina Palama, Yihan Hu |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 62 KB |
| VERSION | 1.0 |
| LICENSE | CC BY 4.0 |
| RIGHTS | CC BY |
| SOURCE | Europe PMC |