University of Amsterdam, Amsterdam, Netherlands.
University of Tasmania, Hobart, Australia.
Behav Res Methods. 2020 Apr;52(2):918-937. doi: 10.3758/s13428-019-01290-6.
Over the last decade, the Bayesian estimation of evidence-accumulation models has gained popularity, largely due to the advantages afforded by the Bayesian hierarchical framework. Despite recent advances in the Bayesian estimation of evidence-accumulation models, model comparison continues to rely on suboptimal procedures, such as posterior parameter inference and model selection criteria known to favor overly complex models. In this paper, we advocate model comparison for evidence-accumulation models based on the Bayes factor obtained via Warp-III bridge sampling. We demonstrate, using the linear ballistic accumulator (LBA), that Warp-III sampling provides a powerful and flexible approach that can be applied to both nested and non-nested model comparisons, even in complex and high-dimensional hierarchical instantiations of the LBA. We provide an easy-to-use software implementation of the Warp-III sampler and outline a series of recommendations aimed at facilitating the use of Warp-III sampling in practical applications.
在过去的十年中，贝叶斯证据积累模型的估计方法得到了广泛的应用，这主要得益于贝叶斯层次框架所提供的优势。尽管在贝叶斯证据积累模型的估计方面取得了一些新进展，但模型比较仍然依赖于次优的方法，例如后验参数推断和倾向于过度复杂模型的模型选择标准。在本文中，我们提倡基于 Warp-III 桥抽样获得的贝叶斯因子进行证据积累模型的比较。我们使用线性弹道积累器 (LBA) 证明了 Warp-III 抽样提供了一种强大而灵活的方法，可应用于嵌套和非嵌套模型比较，即使在 LBA 的复杂和高维层次实例中也是如此。我们提供了一个易于使用的 Warp-III 抽样器的软件实现，并概述了一系列建议，旨在促进 Warp-III 抽样在实际应用中的使用。
