What question did this study set out to answer?

To develop a Bayesian framework for Box-Cox power transformations in linear models using an outcome-independent prior distribution.

January 1, 1981Biometrika

A Bayesian approach to transformations to normality

View Full Paper

Ask AI

Bookmark

Authors

LPLuis R. PericchiUniversity of Puerto Rico System

Discussion

Member takes

Implication

Theoretical analysis demonstrates formal equivalence between Bayesian inference and likelihood methods using outcome-independent priors, highlighting unified estimation of data transformations.

Key Points

To develop a Bayesian framework for Box-Cox power transformations in linear models using an outcome-independent prior distribution.
Analyzed transformation of observations within linear models assuming normal error distributions as formulated by Box and Cox (1964).
Derived and applied an alternative noninformative, unnormed prior distribution that avoids dependence on observed outcomes.
Achieved a formal mathematical identity between likelihood-based and Bayesian inference for estimating optimal normalizing transformations.
Demonstrated exact Bayesian and likelihood equivalence when assessing homoscedasticity and model additivity under the transformation, with extensions to related estimation problems.

Cite This Study

Luis R. Pericchi (1981) studied this question.

synapsesocial.com/papers/6a088a8b280cd4e998e8c7f7 https://doi.org/10.1093/biomet/68.1.35

View Full Paper

Ask AI

Bookmark

Implication

Theoretical analysis demonstrates formal equivalence between Bayesian inference and likelihood methods using outcome-independent priors, highlighting unified estimation of data transformations.

A Bayesian approach to transformations to normality

Discussion

Member takes

Implication

Key Points

Cite This Study

A Bayesian approach to transformations to normality

Discussion

Member takes

Implication

Key Points

Cite This Study

Also Consider

Also Consider