Want to share your content on R-bloggers? click here if you have a blog, or here if you don't.
Frequentist and Bayesian statistics are sometimes regarded as fundamentally different philosophies. Indeed, can both methods qualify as philosophies, or is one of them just a pointless ritual? Is frequentist statistics about \(p\) values only? Are frequentist estimates diametrically opposed to Bayesian posterior distributions? Are confidence intervals and credible intervals irreconcilable? Will R crash if lme4
and brms
are simultaneously loaded? If only we could fit frequentist and Bayesian models to the same data and plot the results together, we might get a glimpse into these puzzles.
All the analyses shown below can be reproduced using the materials at https://osf.io/gt5uf. The combination of the frequentist and the Bayesian estimates in the same plot is achieved using the following custom function from Bernabeu (2022).
Visualising frequentist and Bayesian estimates in one plot
Both frequentist and Bayesian statistics offer the options of hypothesis testing and parameter estimation (Cumming, 2014; Kruschke & Liddell, 2018; Rouder et al., 2018; Schmalz et al., 2022; Tendeiro & Kiers, 2019, 2022; van Ravenzwaaij & Wagenmakers, 2022). In the statistical analyses conducted by Bernabeu (2022), hypothesis testing was performed within the frequentist framework, whereas parameter estimation was performed within both the frequentist and the Bayesian frameworks (for other examples of the estimation approach, see Milek et al., 2018; Pregla et al., 2021; Rodríguez-Ferreiro et al., 2020).
Now it’s time to consider in earnest:
Is frequentist statistics about \(p\) values only? Are frequentist estimates diametrically opposed to Bayesian posterior distributions? Are confidence intervals and credible intervals irreconcilable? Will R crash if
lme4
andbrms
are simultaneously loaded?
Session info
If you encounter any blockers while reproducing the above analyses using the materials at https://osf.io/gt5uf, my current session info may be useful. For instance, the legend on the plot may not show if the latest versions of the ggplot2
and the tidyverse
packages are used. Instead, ggplot2 3.3.5
and tidyverse 1.3.1
should be installed using install_version('tidyverse', '1.3.1')
and install_version('tidyverse', '1.3.1')
.
sessionInfo() ## R version 4.2.2 (2022-10-31 ucrt) ## Platform: x86_64-w64-mingw32/x64 (64-bit) ## Running under: Windows 10 x64 (build 19045) ## ## Matrix products: default ## ## locale: ## [1] LC_COLLATE=English_United States.utf8 ## [2] LC_CTYPE=English_United States.utf8 ## [3] LC_MONETARY=English_United States.utf8 ## [4] LC_NUMERIC=C ## [5] LC_TIME=English_United States.utf8 ## ## attached base packages: ## [1] stats graphics grDevices utils datasets methods base ## ## other attached packages: ## [1] ggtext_0.1.2 Cairo_1.6-0 forcats_0.5.2 ## [4] stringr_1.5.0 dplyr_1.0.10 purrr_1.0.0 ## [7] readr_2.1.3 tidyr_1.2.1 tibble_3.1.8 ## [10] ggplot2_3.3.5 tidyverse_1.3.1 knitr_1.41 ## [13] xaringanExtra_0.7.0 ## ## loaded via a namespace (and not attached): ## [1] Rcpp_1.0.9 lubridate_1.9.0 assertthat_0.2.1 digest_0.6.31 ## [5] utf8_1.2.2 plyr_1.8.8 R6_2.5.1 cellranger_1.1.0 ## [9] ggridges_0.5.4 backports_1.4.1 reprex_2.0.2 evaluate_0.19 ## [13] highr_0.10 httr_1.4.4 blogdown_1.16 pillar_1.8.1 ## [17] rlang_1.0.6 uuid_1.1-0 readxl_1.4.1 rstudioapi_0.14 ## [21] jquerylib_0.1.4 rmarkdown_2.19 labeling_0.4.2 gridtext_0.1.5 ## [25] munsell_0.5.0 broom_1.0.2 compiler_4.2.2 modelr_0.1.10 ## [29] xfun_0.36 pkgconfig_2.0.3 htmltools_0.5.4 tidyselect_1.2.0 ## [33] bookdown_0.31 fansi_1.0.3 crayon_1.5.2 tzdb_0.3.0 ## [37] dbplyr_2.2.1 withr_2.5.0 commonmark_1.8.1 grid_4.2.2 ## [41] jsonlite_1.8.4 gtable_0.3.1 lifecycle_1.0.3 DBI_1.1.3 ## [45] magrittr_2.0.3 scales_1.2.1 cli_3.4.1 stringi_1.7.8 ## [49] cachem_1.0.6 farver_2.1.1 fs_1.5.2 xml2_1.3.3 ## [53] bslib_0.4.2 ellipsis_0.3.2 generics_0.1.3 vctrs_0.5.1 ## [57] tools_4.2.2 glue_1.6.2 markdown_1.4 hms_1.1.2 ## [61] fastmap_1.1.0 yaml_2.3.6 timechange_0.1.1 colorspace_2.0-3 ## [65] rvest_1.0.3 haven_2.5.1 sass_0.4.4
References
Bernabeu, P. (2022). Language and sensorimotor simulation in conceptual processing: Multilevel analysis and statistical power. Lancaster University. https://doi.org/10.17635/lancaster/thesis/1795
Cumming, G. (2014). The new statistics: Why and how. Psychological Science, 25(1), 7–29. https://doi.org/10.1177/0956797613504966
Kruschke, J. K., & Liddell, T. M. (2018). The Bayesian New Statistics: Hypothesis testing, estimation, meta-analysis, and power analysis from a Bayesian perspective. Psychonomic Bulletin & Review, 25(1), 178–206.
Milek, A., Butler, E. A., Tackman, A. M., Kaplan, D. M., Raison, C. L., Sbarra, D. A., Vazire, S., & Mehl, M. R. (2018). “Eavesdropping on happiness” revisited: A pooled, multisample replication of the association between life satisfaction and observed daily conversation quantity and quality. Psychological Science, 29(9), 1451–1462. https://doi.org/10.1177/0956797618774252
Pregla, D., Lissón, P., Vasishth, S., Burchert, F., & Stadie, N. (2021). Variability in sentence comprehension in aphasia in German. Brain and Language, 222, 105008. https://doi.org/10.1016/j.bandl.2021.105008
Rodríguez-Ferreiro, J., Aguilera, M., & Davies, R. (2020). Semantic priming and schizotypal personality: Reassessing the link between thought disorder and enhanced spreading of semantic activation. PeerJ, 8, e9511. https://doi.org/10.7717/peerj.9511
Rouder, J. N., Haaf, J. M., & Vandekerckhove, J. (2018). Bayesian inference for psychology, part IV: Parameter estimation and Bayes factors. Psychonomic Bulletin & Review, 25(1), 102–113. https://doi.org/10.3758/s13423-017-1420-7
Schmalz, X., Biurrun Manresa, J., & Zhang, L. (2021). What is a Bayes factor? Psychological Methods. https://doi.org/10.1037/met0000421
Tendeiro, J. N., & Kiers, H. A. L. (2019). A review of issues about null hypothesis Bayesian testing. Psychological Methods, 24(6), 774–795. https://doi.org/10.1037/met0000221
Tendeiro, J. N., & Kiers, H. A. L. (2022). On the white, the black, and the many shades of gray in between: Our reply to van Ravenzwaaij and Wagenmakers (2021). Psychological Methods, 27(3), 466–475. https://doi.org/10.1037/met0000505
van Ravenzwaaij, D., & Wagenmakers, E.-J. (2022). Advantages masquerading as “issues” in Bayesian hypothesis testing: A commentary on Tendeiro and Kiers (2019). Psychological Methods, 27(3), 451–465. https://doi.org/10.1037/met0000415
R-bloggers.com offers daily e-mail updates about R news and tutorials about learning R and many other topics. Click here if you're looking to post or find an R/data-science job.
Want to share your content on R-bloggers? click here if you have a blog, or here if you don't.