Premium
An elicitation method for multiple linear regression models
Author(s) -
Garthwaite Paul H.,
Dickey James M.
Publication year - 1991
Publication title -
journal of behavioral decision making
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.136
H-Index - 76
eISSN - 1099-0771
pISSN - 0894-3257
DOI - 10.1002/bdm.3960040103
Subject(s) - computer science , weighting , bayesian linear regression , task (project management) , bayesian multivariate linear regression , linear regression , multivariate statistics , bayesian probability , parameterized complexity , regression , proper linear model , conjugate prior , regression analysis , artificial intelligence , statistics , machine learning , prior probability , bayesian inference , mathematics , algorithm , medicine , management , economics , radiology
This paper describes a method of quantifying subjective opinion about a normal linear regression model. Opinion about the regression coefficients and experimental error is elicited and modeled by a multivariate probability distribution (a Bayesian conjugate prior distribution). The distribution model is richly parameterized and various assessment tasks are used to estimate its parameters. These tasks include the revision of opinion in the light of hypothetical data, the assessment of credible intervals, and a task commonly performed in cue‐weighting experiments. A new assessment task is also introduced. In addition, implementation of the method in an interactive computer program is described and the method is illustrated with a practical example.