🤖 AI Summary
This study addresses the challenge of effectively discriminating among non-nested survival models with a cure fraction when their baseline distributions differ. To this end, we propose J-type and MJ-type score tests that extend the Vuong test to the cure survival modeling framework. By incorporating information from competing models into the null log-likelihood, our approach assesses model redundancy and constructs unsigned statistics suitable for multi-model comparison. Integrating restricted maximum likelihood estimation, parametric bootstrap, and Kullback–Leibler divergence estimation, the proposed methodology supports both global hypothesis testing and model selection. Notably, the MJ statistic is capable of identifying whether at least one candidate model is correctly specified. Compared to existing approaches, this framework offers greater theoretical rigor while maintaining strong discriminative power and practical utility.
📝 Abstract
We propose specification tests for discriminating among non-nested parametric survival models with a cure fraction, focusing on models that differ only in their baseline distributions. The proposed approach augments the null log-likelihood with information from competing models and applies a score test to assess whether the additional information is redundant. Because the test relies only on restricted maximum likelihood estimates, it avoids fitting augmented models. For two competing models, the score statistic reduces to a quadratic form in the sample mean of the individual log-likelihood differences. We show that its signed square root coincides with Vuong's test statistic, although our framework differs in three important respects: it tests the specific null hypothesis that a given model is the true data-generating process, it uses an unsigned statistic that extends naturally to $M \ge 2$ competing models, and it estimates the Kullback-Leibler bias by parametric bootstrap. The resulting MJ statistic combines the individual J tests to assess the global null hypothesis that at least one candidate model is correctly specified, while also providing a model-selection criterion.