Computer Science
6 years ago

New method accurately computes Bayes factors from data with missing values.

Psychological Methods
Computing Bayes factors from data with missing values.
Herbert Hoijtink, Xin Gu, Joris Mulder, Yves Rosseel

Paper Summary

Related papers

Theme background image
Mathematics
4 years ago

New method simplifies model selection in Bayesian statistics without prior specification.

Regularization Methods
Missing Data
Multicollinearity
Theme background image
Computer Science
6 years ago

New method accurately computes Bayes factors from data with missing values.

Secure Communication
Time Series Forecasting
Missing Data
1 view
Theme background image
Mathematics
8 years ago

New method revolutionizes comparison of lognormal variances in Bayesian hypothesis testing.

Skew Distributions
Multicollinearity
Experimental Design
1 view

The article shows how to calculate Bayes factors when some data is missing. By using multiple imputations to fill in the missing values, a more accurate Bayes factor can be obtained. This method is important because it ensures that the Bayes factor is based only on the information in the observed data. The researchers found that using a single imputation for missing data leads to very inaccurate Bayes factors. The approach outlined in the article can be applied using R packages for multiple imputation along with specific Bayes factor packages.