Skip to content

Multi-category Models and Weights #3

@jluchman

Description

@jluchman

Hi Claudio,

I would like to request (and/or offer support helping to develop if desired) an extension of the methods to incorporate polr in MASS as well as multinom in nnet. I believe both would be fairly straightforward extensions of the glm based methods currently available as both suggested models accept logLik and the fit metrics applied to glm should apply.

Additionally, I would like to request/support the inclusion of observation weights or case weights for models that accept them such as lm, glm, polr, and multinom. This suggestion might be more complex than I think as I am somewhat newer to R programming and am not quite sure how the weights are used by each method yet by comparison to weights as used by functions in the package survey. That is, this request is to obtain functionality similar to that offered by the relaimpo package in incorporating complex survey weights.

To more fully disclose some of my motives here, I am hoping to extend the functionality I have discussed in two publications to the R environment (see below).

- joe

Luchman, J. N. (2014). Relative importance analysis with multicategory dependent variables: An extension and review of best practices. Organizational Research Methods, 17(4), 452-471.

Luchman, J. N. (2015). Determining subgroup difference importance with complex survey designs: an application of weighted dominance analysis. Survey Practice, 8(5), 2835.

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions