Beyond classification metrics: a psychometric-aware benchmark for data augmentation in imbalanced student mental health surveys
Chen Shao, Shengnan Qiao, Ang Li, Shipeng Liu, Yanglong Chen, Yuzhe Tan, Zhibo Liang, Xuanming Si · 2026
Read Beyond classification metrics: a psychometric-aware benchmark for data augmentation in imbalanced student mental health surveys by Chen Shao, Shengnan Qiao, Ang Li, Shipeng Liu, Yanglong Chen, Yuzhe Tan, Zhibo Liang, Xuanming Si (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/PMC13310988.txt.
Description
Chen Shao and co-authors benchmark fake data methods for student mental health surveys in a 2026 study. Machine screening of survey answers can aid clinicians, yet urgent cases are rare in data, so models miss them. Fake added data may help, but whether fakes keep the survey meaning intact was untested.
Eight methods were tried on three public sets, from a small mood scale set to two large student sets. Overall score gains did not split methods clearly. A new generator that guards factor shape and item links won most on the small set, while simple oversampling led on the large set with neural nets. Both accuracy and meaning must be graded.
Catalog record
| CREATOR | Chen Shao, Shengnan Qiao, Ang Li, Shipeng Liu, Yanglong Chen, Yuzhe Tan, Zhibo Liang, Xuanming Si |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 113742 |
| LICENSE | https://creativecommons.org/licenses/by/4.0 |
| RIGHTS | CC BY |
| SOURCE | Europe PMC |