On this page you will find news about missing data and applied biostatistical methods, referrals to other interesting blogs or articles.

Calibration belt

With package givitiR a calibration belt can be plotted to assess the calibration of dichotomous outcome models. This is assumed a better procedure compared to the shortcomings of the Hosmer and Lemeshow test, that may become significant in large datasets, although predicted and observed probabilities agree well.

Package naniar

Package naniar provides tidy ways to summarise, visualise, and manipulate missing data.

Package missCompare

You can compare different missing data imputation algorithms with package missCompare.