RACE/ETHNICITY AND THE MEASUREMENT OF COGNITION IN NSHAP: RECOMMENDATIONS FOR ROBUSTNESS
Bibliographic record
Abstract
Abstract In this study, we interrogate measurement of cognition by race, in order to move towards a less-biased and more-inclusive set of measures for capturing cognitive change and decline in older adulthood. We use data from Round 2 (N=3377) and Round 3 (N=4777) of the National Social Life Health and Aging Project (NSHAP), and examine the study’s version of the Montreal Cognitive Assessment (MoCA). We employ exploratory factor analyses to explore configural invariance by racial/ethnic group (Non-Hispanic White, Non-Hispanic Black, Hispanic, All else), and then log-likelihood tests of scalar and metric invariance. Using modification indexes we identify items that seem robust to bias by race. We test the predictive validity of the full (18-item) and short (7-item) scales using self-reported dementia diagnosis, Instrumental Activities of Daily Living (IADLs), proxy reports of dementia death, and National Death Index (NDI) reports of dementia death. We found that 7 measures, out of the 18 used in NSHAP’s MoCA, formed a scale that was more robust to racial bias. The shortened form predicted consequential outcomes equally well, compared to NSHAP’s full MoCA. The short form was also highly correlated with the full form, and displayed lower test-retest correlation between Round 2 and 3. Although sophisticated structural equation modeling techniques could be useful for assuaging measurement invariance by race in NSHAP, the shortened form provides a quick way for researchers to carry out robustness checks – to see if the disparities and associations they document are ‘real’ or the product of artifactual bias.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.541 | 0.787 |
| Meta-epidemiology (narrow) | 0.005 | 0.004 |
| Meta-epidemiology (broad) | 0.008 | 0.027 |
| Bibliometrics | 0.010 | 0.017 |
| Science and technology studies | 0.005 | 0.012 |
| Scholarly communication | 0.011 | 0.013 |
| Open science | 0.014 | 0.011 |
| Research integrity | 0.008 | 0.011 |
| Insufficient payload (model declined to judge) | 0.006 | 0.002 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".