MI_boot function

Bootstrap validation in Multiply Imputed datasets

MI_boot Bootstrapping in each (original) Multiply Imputed dataset for internal validation. Called by function psfmi_perform.

MI_boot(pobj, p.crit, nboot, direction)

Arguments

• pobj: An object of class pmods (pooled models), produced by a previous call to psfmi_lr.
• p.crit: A numerical scalar. P-value selection criterium used for backward or forward selection during in the bootstrap samples. When set at 1, validation is done without variable selection.
• nboot: The number of bootstrap resamples, default is 10.
• direction: The direction of predictor selection, "BW" is for backward selection and "FW" for forward selection.

See Also

psfmi_perform

Author(s)

Martijn Heymans, 2020