Predicting working memory failure: A subjective bayesian approach to model selection

Bradley P. Carlin, Robert E. Kass, F. Javier Lerch, Brian R. Huguenard

Research output: Contribution to journalArticlepeer-review

12 Scopus citations


We use Bayes factors to compare two alternative characterizations of human working memory load in their ability to predict errors in database query-writing tasks. The first measures working memory load by the number of different features each task contains, while the second attempts instead to measure the complexity of the task by giving more weight to features that require more mental time for their correct execution. We reanalyze data from a previously conducted experiment using two logistic regression models with random subject effects nested within an experimental condition factor. The two models have alternative covariates based on the alternative measures of working memory load. We construct prior distributions based on our subjective knowledge gleaned from related experiments, providing details of the elicitation process. We examine sensitivity of our results to the effects of prior misspecification and case deletion. Asymptotic approximations are used throughout to facilitate computations. Finally, we comment on the strengths and limitations of the approach in light of our experience.

Original languageEnglish (US)
Pages (from-to)319-327
Number of pages9
JournalJournal of the American Statistical Association
Issue number418
StatePublished - Jun 1992

Bibliographical note

Funding Information:
* Bradley P. Carlin is Assistant Professor, Division of Biostatistics, School of Public Health, University of Minnesota, Minneapolis, MN 55455. Robert E. Kass is Associate Professor, Department of Statistics, Carnegie Mellon University, Pittsburgh, PA 15213. F. Javier Lerch is Assistant Professor and Brian R. Huguenard is Doctoral Student, Graduate School of Industrial Administration, Carnegie Mellon University, Pittsburgh, PA 15213. The work of the first-named author was supported in part by National Science Foundation Grant DMS 88-05676, and that of the second-named author was supported in part by National Science Foundation Grants DMS 87-05646, DMS 88-05676, and DMS 90-05858, and National Institutes of Health Grant R01-CA54852-01. The authors thank John B. Smelcer for permitting us access to the data used in this study; Alan Genz for assistance with computation; and the editor, associate editor, and three conscientious referees for numerous helpful comments that led to improvements in the presentation. This researchwas done whilethe first-named author wasvisiting the Department of Statistics at Carnegie Mellon University.


  • Bayes factor
  • Case deletion
  • Laplace method
  • Model choice
  • Prior elicitation
  • Prior sensitivity


Dive into the research topics of 'Predicting working memory failure: A subjective bayesian approach to model selection'. Together they form a unique fingerprint.

Cite this