分层反应时间模型的贝叶斯估计和最大似然估计
Bayesian and maximum likelihood estimation of hierarchical response time models.
作者信息
Farrell Simon, Ludwig Casimir J H
机构信息
University of Bristol, Bristol, England.
出版信息
Psychon Bull Rev. 2008 Dec;15(6):1209-17. doi: 10.3758/PBR.15.6.1209.
Abstract
Hierarchical (or multilevel) statistical models have become increasingly popular in psychology in the last few years. In this article, we consider the application of multilevel modeling to the ex-Gaussian, a popular model of response times. We compare single-level and hierarchical methods for estimation of the parameters of ex-Gaussian distributions. In addition, for each approach, we compare maximum likelihood estimation with Bayesian estimation. A set of simulations and analyses of parameter recovery show that although all methods perform adequately well, hierarchical methods are better able to recover the parameters of the ex-Gaussian, by reducing variability in the recovered parameters. At each level, little overall difference was observed between the maximum likelihood and Bayesian methods.
摘要
在过去几年中，分层（或多级）统计模型在心理学领域越来越受欢迎。在本文中，我们考虑将多级建模应用于前高斯分布，这是一种常用的反应时间模型。我们比较了单级和分层方法来估计前高斯分布的参数。此外，对于每种方法，我们还比较了最大似然估计和贝叶斯估计。一组参数恢复的模拟和分析表明，尽管所有方法的表现都相当不错，但分层方法通过减少恢复参数的变异性，能更好地恢复前高斯分布的参数。在每个层面上，最大似然法和贝叶斯法之间总体差异不大。
相似文献
1
Bayesian and maximum likelihood estimation of hierarchical response time models.分层反应时间模型的贝叶斯估计和最大似然估计
Psychon Bull Rev. 2008 Dec;15(6):1209-17. doi: 10.3758/PBR.15.6.1209.
2
Multilevel modeling of single-case data: A comparison of maximum likelihood and Bayesian estimation.多水平模型的单案例数据：最大似然和贝叶斯估计的比较。
Psychol Methods. 2017 Dec;22(4):760-778. doi: 10.1037/met0000136. Epub 2017 Mar 30.
3
Empirical Bayes Gaussian likelihood estimation of exposure distributions from pooled samples in human biomonitoring.基于经验贝叶斯高斯似然估计的人类生物监测中混合样本暴露分布研究
Stat Med. 2014 Dec 10;33(28):4999-5014. doi: 10.1002/sim.6304. Epub 2014 Sep 12.
4
Fitting the psychometric function.拟合心理测量函数。
Percept Psychophys. 1999 Jan;61(1):87-106. doi: 10.3758/bf03211951.
5
A multilevel model for hierarchical, repeated, and overdispersed time-to-event outcomes and its estimation strategies.用于分层、重复和过度分散的事件发生时间结局的多级模型及其估计策略。
J Biopharm Stat. 2013;23(6):1420-34. doi: 10.1080/10543406.2013.834914.
6
Likelihood-based parameter estimation and comparison of dynamical cognitive models.基于似然性的动态认知模型参数估计与比较
Psychol Rev. 2017 Jul;124(4):505-524. doi: 10.1037/rev0000068. Epub 2017 Apr 27.
7
On Latent Trait Estimation in Multidimensional Compensatory Item Response Models.多维补偿性项目反应模型中的潜在特质估计
Psychometrika. 2015 Jun;80(2):428-49. doi: 10.1007/s11336-013-9399-0. Epub 2014 Mar 7.
8
Hierarchical approximate Bayesian computation.分层近似贝叶斯计算。
Psychometrika. 2014 Apr;79(2):185-209. doi: 10.1007/s11336-013-9381-x. Epub 2013 Dec 3.
9
Classical and Bayesian inference in neuroimaging: applications.神经影像学中的经典推理与贝叶斯推理：应用
Neuroimage. 2002 Jun;16(2):484-512. doi: 10.1006/nimg.2002.1091.
10
Maximum-likelihood versus maximum a posteriori parameter estimation of physiological system models: the C-peptide impulse response case study.生理系统模型的最大似然与最大后验参数估计：C肽脉冲响应案例研究
IEEE Trans Biomed Eng. 2000 Jun;47(6):801-11. doi: 10.1109/10.844232.
引用本文的文献
1
Dissociable roles for the striatal cholinergic system in different flexibility contexts.纹状体胆碱能系统在不同灵活性背景下的可分离作用。
IBRO Neurosci Rep. 2022 Apr 1;12:260-270. doi: 10.1016/j.ibneur.2022.03.007. eCollection 2022 Jun.
2
Applying hierarchical bayesian modeling to experimental psychopathology data: An introduction and tutorial.应用分层贝叶斯模型于实验心理病理学数据：介绍与教程。
J Abnorm Psychol. 2021 Nov;130(8):923-936. doi: 10.1037/abn0000707.
3
Unveiling the neuromechanical mechanisms underlying the synergistic interactions in human sensorimotor system.揭示人类感觉运动系统协同作用的神经力学机制。
Sci Rep. 2021 Jan 8;11(1):203. doi: 10.1038/s41598-020-80420-z.
4
Computing Bayes factors for evidence-accumulation models using Warp-III bridge sampling.使用 Warp-III 桥采样计算证据积累模型的贝叶斯因子。
Behav Res Methods. 2020 Apr;52(2):918-937. doi: 10.3758/s13428-019-01290-6.
5
Perceptual Decision-Making: Biases in Post-Error Reaction Times Explained by Attractor Network Dynamics.感知决策：吸引子网络动力学解释了后错误反应时间中的偏差。
J Neurosci. 2019 Jan 30;39(5):833-853. doi: 10.1523/JNEUROSCI.1015-18.2018. Epub 2018 Nov 30.
6
On the importance of avoiding shortcuts in applying cognitive models to hierarchical data.论避免在将认知模型应用于层次数据时走捷径的重要性。
Behav Res Methods. 2018 Aug;50(4):1614-1631. doi: 10.3758/s13428-018-1054-3.
7
2D-EM clustering approach for high-dimensional data through folding feature vectors.通过折叠特征向量的 2D-EM 聚类方法对高维数据进行聚类。
BMC Bioinformatics. 2017 Dec 28;18(Suppl 16):547. doi: 10.1186/s12859-017-1970-8.
8
Divisive hierarchical maximum likelihood clustering.分裂式层次最大似然聚类。
BMC Bioinformatics. 2017 Dec 28;18(Suppl 16):546. doi: 10.1186/s12859-017-1965-5.
9
Bayesian inference for psychology, part III: Parameter estimation in nonstandard models.贝叶斯推断在心理学中的应用，第三部分：非标准模型中的参数估计。
Psychon Bull Rev. 2018 Feb;25(1):77-101. doi: 10.3758/s13423-017-1394-5.
10
Failures of cognitive control or attention? The case of stop-signal deficits in schizophrenia.认知控制还是注意力缺陷？精神分裂症中停止信号缺陷的案例。
Atten Percept Psychophys. 2017 May;79(4):1078-1086. doi: 10.3758/s13414-017-1287-8.
本文引用的文献
1
Analyzing the dynamics of free recall: An integrative review of the empirical literature.分析自由回忆的动态：实证文献的综合评述。
Psychon Bull Rev. 1994 Mar;1(1):89-106. doi: 10.3758/BF03200763.
2
Three case studies in the Bayesian analysis of cognitive models.认知模型贝叶斯分析中的三个案例研究。
Psychon Bull Rev. 2008 Feb;15(1):1-15. doi: 10.3758/pbr.15.1.1.
3
Diffusion model analysis with MATLAB: a DMAT primer.使用MATLAB进行扩散模型分析：DMAT入门指南。
Behav Res Methods. 2008 Feb;40(1):61-72. doi: 10.3758/brm.40.1.61.
4
Short-term recognition memory for serial order and timing.对序列顺序和时间的短期识别记忆。
Mem Cognit. 2007 Oct;35(7):1724-34. doi: 10.3758/bf03193505.
5
Individual differences in components of reaction time distributions and their relations to working memory and intelligence.反应时间分布各成分的个体差异及其与工作记忆和智力的关系。
J Exp Psychol Gen. 2007 Aug;136(3):414-29. doi: 10.1037/0096-3445.136.3.414.
6
Multilevel models for the experimental psychologist: foundations and illustrative examples.实验心理学家的多层模型：基础与示例
Behav Res Methods. 2007 Feb;39(1):101-17. doi: 10.3758/bf03192848.
7
Analyzing individual differences in sentence processing performance using multilevel models.
Behav Res Methods. 2007 Feb;39(1):31-8. doi: 10.3758/bf03192841.
8
An EZ-diffusion model for response time and accuracy.一种用于反应时间和准确性的EZ扩散模型。
Psychon Bull Rev. 2007 Feb;14(1):3-22. doi: 10.3758/bf03194023.
9
Switch and maintenance of task set in schizophrenia.精神分裂症中任务集的切换与维持
Schizophr Res. 2006 Jun;84(2-3):345-58. doi: 10.1016/j.schres.2006.01.022. Epub 2006 Mar 24.
10
An introduction to Bayesian hierarchical models with an application in the theory of signal detection.贝叶斯层次模型介绍及其在信号检测理论中的应用
Psychon Bull Rev. 2005 Aug;12(4):573-604. doi: 10.3758/bf03196750.
Zotero 插件