Probing, Plotting, and Interpreting Multilevel Interaction Effects R package mlmoderator version 0.2.1 Provides a unified workflow for probing, plotting, and assessing the robustness of cross-level interaction effects in two-level mixed-effects models fitted with 'lme4' (Bates et al., 2015) <doi:10.18637/jss.v067.i01>. Implements simple slopes analysis following Aiken and West (1991, ISBN:9780761907121), Johnson-Neyman intervals following Johnson and Fay (1950) <doi:10.1007/BF02288864> and Bauer and Curran (2005) <doi:10.1207/s15327906mbr4003_5>, and grand- or group-mean centering as described in Enders and Tofighi (2007) <doi:10.1037/1082-989X.12.2.121>. Includes a slope variance decomposition that separates fixed-effect uncertainty from random-slope variance (tau11), a contour surface plot of predicted outcomes over the full predictor-by-moderator space, and robustness diagnostics comprising intraclass correlation coefficient shift analysis and leave-one-cluster-out (LOCO) stability checks. Designed for researchers in education, psychology, biostatistics, epidemiology, organizational science, and other fields where outcomes are clustered within higher-level units. Software Depends: R (>= 4.1.0) Imports: lme4, ggplot2, stats, utils, grDevices, rlang Suggests: testthat (>= 3.0.0), dplyr, tibble, knitr, rmarkdown, patchwork Subir Hait <haitsubi@msu.edu> Comprehensive R Archive Network (CRAN) Subir Hait [aut, cre] (ORCID: <https://orcid.org/0009-0004-9871-9677>) MIT + file LICENSE (https://CRAN.R-project.org/package=mlmoderator/LICENSE) 2026-04-03 application/tgz https://CRAN.R-project.org/package=mlmoderator doi:10.32614/CRAN.package.mlmoderator en-US