Advertisement

Informed priors

  • Douglas A. Schroeder
Chapter
Part of the Springer Series in Accounting Scholarship book series (KLAS, volume 5)

Abstract

When building an empirical model we typically attempt to include our understanding of the phenomenon as part of the model. This commonly describes both classical and Bayesian analyses (usually with locally uninformed priors). However, what analysis can we undertake if we have no data (new evidence) on which to apply our model. The above modeling strategy leaves us in a quandary.With no new data, we are not (necessarily) in a state of complete ignorance and this setting suggests the folly of ignoring our background knowledge in standard data analysis. If our model building strategy adequately reflects our state of knowledge plus the new data, we expect inferences from the standard approach described above to match Bayesian inference based on our informed priors plus the new data.

Keywords

Background Knowledge Maximum Entropy Earning Management Data Generate Process Incentive Payment 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

Copyright information

© Springer Science+Business Media, LLC 2010

Authors and Affiliations

  1. 1.The Ohio State UniversityColumbusUSA

Personalised recommendations