A penalized heteroskedastic ordered probit model for DIF (measurement invariance) testing of single-item assessments in cross-cultural research
R. Noah Padgett · 2026
Read A penalized heteroskedastic ordered probit model for DIF (measurement invariance) testing of single-item assessments in cross-cultural research by R. Noah Padgett (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/arxiv-2601-18889.txt.
Description
R. Noah Padgett builds a way to test if a lone quiz item acts in a fair way across lands and groups. Such a test is called a check for item bias. It asks if folk with the same true skill, but from unlike groups, pass the item at the same rate. Past tests need more than one item to work. With a lone item there is no mate item to set the skill scale, and no trait to lean on as a stand-in.
Padgett adds a fine for wild swings to a kind of ordered score model that can take spread in its counts. The fine keeps the model from chase of noise in thin groups. A sound lone-item test could serve short polls and health checks that have just one item per trait. Those are in wide use across lands, so fair checks count.
Catalog record
| CREATOR | R. Noah Padgett |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 25 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2601-18889