Skip to contents

Returns a table describing the modeling, reporting, import, diagnostic, and export features that are currently supported, experimental, or planned in mlmr.

Usage

mlm_supported_models()

Details

This helper is intended for transparent documentation and teaching. It gives users a compact way to see the current production path, advanced workflows, and areas that require extra statistical judgment or future development.

Value

A data frame with columns Area, Status, Scope, and User_responsibility.

Examples

mlm_supported_models()
#>                      Area                               Status
#> 1          Model families                            Supported
#> 2       Nested structures                            Supported
#> 3  Three-level structures                            Supported
#> 4  Crossed random effects                         Experimental
#> 5     Multiple membership                              Planned
#> 6       Repeated measures Supported through grouping structure
#> 7          Random effects                            Supported
#> 8               Centering                            Supported
#> 9            Interactions                            Supported
#> 10 Categorical predictors        Supported through R contrasts
#> 11          Uploaded data                            Supported
#> 12            Diagnostics                            Supported
#> 13      Reporting exports                            Supported
#> 14     Software citations                            Supported
#>                                                                                                                                                       Scope
#> 1  Gaussian linear mixed models are the primary production path; binomial, Poisson, negative binomial, and Gamma GLMMs are available as advanced workflows.
#> 2                                                                                             Two-level nested models are supported in the app and backend.
#> 3                                                     Three-level nested models are supported through additional grouping factors and random-effect blocks.
#> 4                                         Separate lme4 random-effect blocks can be specified; users should verify the design and interpretation carefully.
#> 5                                                                          Weighted membership models are outside the current lme4-backed production scope.
#> 6                               Longitudinal models can be specified when occasions, persons, and higher-level units are represented by grouping variables.
#> 7                                                   Random intercepts, random slopes, and correlated or independent random-effect structures are supported.
#> 8                                                      No centering, grand-mean centering, and cluster-mean centering are supported for numeric predictors.
#> 9                                                                                 Fixed interactions and cross-level interaction-style terms are supported.
#> 10                                              Factors and character variables are modeled using R's contrast system; dummy-coding summaries are reported.
#> 11                                                       CSV, TSV/TXT, Excel, SPSS, SAS, and Stata files are supported when optional readers are installed.
#> 12                                     Convergence messages, singular-fit checks, gradients, overdispersion for GLMMs, and variance summaries are reported.
#> 13                                                           APA tables, raw LaTeX, equations, Quarto-ready reports, and reproducible R code are supported.
#> 14                                                      R/package version tables, APA software statements, and optional papaja citation code are supported.
#>                                                                                       User_responsibility
#> 1                                                  Confirm distributional assumptions and link functions.
#> 2                                                Confirm the grouping IDs correctly represent the design.
#> 3                                  Confirm enough units exist at each higher level for stable estimation.
#> 4                                       Confirm the design is truly crossed and not a miscoded hierarchy.
#> 5                           Use specialized methods outside mlmr for weighted multiple-membership models.
#> 6                Confirm time is coded appropriately and that autocorrelation assumptions are acceptable.
#> 7  Inspect convergence, singular fits, and whether the random-effects structure is supported by the data.
#> 8                                                       Match centering choices to the research question.
#> 9                             Interpret lower-order effects conditionally when interactions are included.
#> 10                                           Confirm reference categories and contrasts before reporting.
#> 11                                Check variable types, missing data, and imported labels before fitting.
#> 12                      Treat diagnostics as evidence for revision, not as automatic pass/fail decisions.
#> 13                                    Review all manuscript text, labels, and notation before submission.
#> 14                                      Confirm package citations meet the target journal or style guide.