plrtest.mdyplFit: Penalized likelihood ratio test for '"mdyplFit"' objects

In brglm2: Bias Reduction in Generalized Linear Models

Penalized likelihood ratio test for "mdyplFit" objects

Description

Computes the Diaconis-Ylvisaker prior penalized likelihood ratio test statistic or its adjusted version using high-dimensionality correction under proportional asymptotics. Associated p-values are also computed using a chi squared distribution.

Usage

## S3 method for class 'mdyplFit'
plrtest(object1, object2, hd_correction = FALSE, ...)

Arguments

object1	a "mdyplFit" object
object2	a "mdyplFit" object
hd_correction	if FALSE (default), then the summary corresponding to standard asymptotics is computed. If TRUE then the high-dimensionality corrections in Sterzinger & Kosmidis (2024) are employed to updates estimates, estimated standard errors, z-statistics. See Details.
...	further arguments to be passed to methods. Currently not used.

Details

Both object1 and object2 should have been fitted using the mdyplFit() method for glm(), and the same shrinkage parameter alpha; see mdyplFit() and mdyplControl() for setting alpha.

If hd_correction = TRUE then the deviance and the associated p-value are adjusted using a high-dimensionality correction under proportional asymptotics as in Sterzinger & Kosmidis (2024); see summary.mdyplFit().

Author(s)

Ioannis Kosmidis ⁠[aut, cre]⁠ [e-mail]

References

Sterzinger P, Kosmidis I (2024). Diaconis-Ylvisaker prior penalized likelihood for p/n \to \kappa \in (0,1) logistic regression. arXiv:2311.07419v2, https://arxiv.org/abs/2311.07419.

See Also

mdyplFit(), summary.mdyplFit(), mdypl_control()

brglm2 documentation built on Aug. 29, 2025, 5:25 p.m.