Separation-like irregularity and sample size optimism in high-discrimination logistic prediction models

Q: What does this research mean for the field?

Closed-form minimum sample size criteria for developing logistic prediction models, such as the Riley framework, underestimate the required sample size when anticipated discrimination is high. Novelty: ClaimNovelty.CONTRADICTORY. Consensus alignment: ConsensusAlignment.CHALLENGES_CONSENSUS.

Ye Liang; Louis Shuo Wang; Jiguang Yu; Xin Zan

doi:10.1371/journal.pone.0342286

August 24, 2026PLoS ONEOpen Access

Separation-like irregularity and sample size optimism in high-discrimination logistic prediction models

View Full Paper

Ask AI

Bookmark

Share

Key Points

Assess whether closed-form sample size formulas, such as the Riley framework, underestimate the sample size required to develop stable logistic prediction models when anticipated discrimination is high.
Conducted a Monte Carlo simulation study comparing formula-recommended sample sizes against empirical sample sizes required for out-of-sample calibration-slope stability.
Evaluated scenarios fixing parameter dimension at p = 10 while varying predictor distributions (normal, skewed, binary), signal density (dense vs. sparse), event prevalence (5%, 10%, 20%), and target discrimination (AUC 0.70 to 0.90).
Closed-form sample size criteria systematically underestimate sample size requirements as target discrimination increases toward higher AUC thresholds.
Empirical sample sizes required to achieve stable out-of-sample calibration slopes exceed formula recommendations due to separation-like irregularity in high-discrimination settings.

No abstract yet

We only have this paper's title and publication details.

Read it at the source

Implication

Monte Carlo simulation study demonstrates formula-based sample size calculations underestimate requirements in high-discrimination logistic models, indicating larger cohorts are necessary for...

Discussion

No takes yet. Share an insight, caveat, or question.

Cite This Study

Liang et al. (2026) studied this question.

synapsesocial.com/papers/6a8ebb54451774b83f3b4beb https://doi.org/10.1371/journal.pone.0342286

View Full Paper

Ask AI

Bookmark

Share

Separation-like irregularity and sample size optimism in high-discrimination logistic prediction models

Key Points

No abstract yet

Implication

Discussion

Cite This Study

Separation-like irregularity and sample size optimism in high-discrimination logistic prediction models

Key Points

No abstract yet

Implication

Discussion

Cite This Study

Keep reading

More related papers

More from PLoS ONE

Keep reading

More related papers

More from PLoS ONE