Is linear regression sufficient for normative psychological assessment? A comparison between OLS, GAM, and GAMLSS across seven public reference datasets
- 1NeuronUP S.L
Abstract
BackgroundNormative Z-scores are essential for interpreting individual cognitive performance relative to healthy reference populations. Traditional norms often compare individuals within predefined groups, such as age bands. This is useful and easy to interpret, but it can lose information because variables such as age change continuously and because age, sex, and education may contribute jointly to performance. Regression-based norms estimate the expected score for each individual from their specific covariates and then express the difference between the observed and expected score as a normative Z-score. Therefore, the key methodological question is not simply whether a more complex model can be fitted, but whether that added complexity changes the Z-score in a meaningful way.
MethodsWe implemented a comparative framework and fitted ordinary least squares (OLS), generalized additive models (GAM), and generalized additive models for location, scale, and shape (GAMLSS) to the seven public reference datasets distributed in the NormData R package. The datasets span processing speed, verbal memory, verbal fluency, academic achievement, personality, and anxiety. For each dataset we compared each regression model with the classic subgroup norm, and then compared the regression models with one another. Metrics included Pearson r, residual RMSE, and the percentage of participants whose model-based Z-score agreed with the classic Z-score.
ResultsAll three models reproduced the classic Z-scores closely (r = 0.964-1.000). OLS and GAM produced nearly identical Z-scores throughout (r = 0.996-1.000), indicating that the observed age trajectories were smooth enough for a quadratic term to capture. GAMLSS diverged from OLS mainly in datasets with non-constant residual variance (Fluency, sigma-ratio = 1.79; TMAS, sigma-ratio = 1.50), where modeling the scale parameter recovered dispersion differences that the classic subgroup method encodes implicitly.
ConclusionsAcross seven heterogeneous reference datasets, OLS provides a sufficient normative Z-score computation when the sample size is large enough and residual variance is approximately constant. GAM is useful when the mean trajectory is strongly nonlinear; GAMLSS is useful when residual variance changes across covariate profiles. The practical message is deliberately simple: add model complexity only when the data show what that complexity contributes to the individual Z-score.
Related
- The Hungarian version of the oxford cognitive screen (OCS-HU): Normative data and comparison of mild stroke to neurologically healthy adults.
- From implicit assumptions to explicit ethics: assessment of normative frameworks in digitalization recommendations in rheumatology
- Establishment of age- and sex-stratified normative data for the Brief Assessment of Cognition (BAC) app in healthy Japanese individuals.
- Can fluoxetine's diminished efficacy in pediatric depression be explained by study sites, baseline severity, age, and psychological interventions? A secondary exploratory meta-regression
- Exercise for tension-type headache: protocol for a systematic review and meta-analysis with meta-regression and assessment of exercise reporting completeness.
BibTeX
@misc{gomez2026linear,
title = {{Is linear regression sufficient for normative psychological assessment? A comparison between OLS, GAM, and GAMLSS across seven public reference datasets}},
author = {M. Arbizu Gomez and P. Rodriguez Royo and E. Maltseva and C. Sastre Barrios and J. M. Corada and C. Ortea Suarez and I. Fernandez de Pierola and M. Rios Lago and J. M. Cortes},
howpublished = {medRxiv},
year = {2026},
month = sep,
doi = {10.64898/2026.09.16.26363213},
url = {https://www.medrxiv.org/content/10.64898/2026.09.16.26363213v1},
note = {CC BY-NC},
}
RIS
TY - UNPB
TI - Is linear regression sufficient for normative psychological assessment? A comparison between OLS, GAM, and GAMLSS across seven public reference datasets
AU - M. Arbizu Gomez
AU - P. Rodriguez Royo
AU - E. Maltseva
AU - C. Sastre Barrios
AU - J. M. Corada
AU - C. Ortea Suarez
AU - I. Fernandez de Pierola
AU - M. Rios Lago
AU - J. M. Cortes
PY - 2026
DA - 2026/09/16
PB - medRxiv
DO - 10.64898/2026.09.16.26363213
UR - https://www.medrxiv.org/content/10.64898/2026.09.16.26363213v1
AB - BackgroundNormative Z-scores are essential for interpreting individual cognitive performance relative to healthy reference populations. Traditional norms often compare individuals within predefined groups, such as age bands. This is useful and easy to interpret, but it can lose information because variables such as age change continuously and because age, sex, and education may contribute jointly to performance. Regression-based norms estimate the expected score for each individual from their specific covariates and then express the difference between the observed and expected score as a normative Z-score. Therefore, the key methodological question is not simply whether a more complex model can be fitted, but whether that added complexity changes the Z-score in a meaningful way. MethodsWe implemented a comparative framework and fitted ordinary least squares (OLS), generalized additive models (GAM), and generalized additive models for location, scale, and shape (GAMLSS) to the seven public reference datasets distributed in the NormData R package. The datasets span processing speed, verbal memory, verbal fluency, academic achievement, personality, and anxiety. For each dataset we compared each regression model with the classic subgroup norm, and then compared the regression models with one another. Metrics included Pearson r, residual RMSE, and the percentage of participants whose model-based Z-score agreed with the classic Z-score. ResultsAll three models reproduced the classic Z-scores closely (r = 0.964-1.000). OLS and GAM produced nearly identical Z-scores throughout (r = 0.996-1.000), indicating that the observed age trajectories were smooth enough for a quadratic term to capture. GAMLSS diverged from OLS mainly in datasets with non-constant residual variance (Fluency, sigma-ratio = 1.79; TMAS, sigma-ratio = 1.50), where modeling the scale parameter recovered dispersion differences that the classic subgroup method encodes implicitly. ConclusionsAcross seven heterogeneous reference datasets, OLS provides a sufficient normative Z-score computation when the sample size is large enough and residual variance is approximately constant. GAM is useful when the mean trajectory is strongly nonlinear; GAMLSS is useful when residual variance changes across covariate profiles. The practical message is deliberately simple: add model complexity only when the data show what that complexity contributes to the individual Z-score.
ER -