statistics

Who's through with convergence warnings? A list of papers that used ‘allFit’ to check ‘coefficients’ across ‘optimizers’

How to break down colour variable in sjPlot::plot_model into equally-sized bins

Whereas the direction of main effects can be interpreted from the sign of the estimate, the interpretation of interaction effects often requires plots. This task is facilitated by the R package sjPlot. For instance, using the plot_model function, I plotted the interaction between two continuous variables. library(lme4) #> Loading required package: Matrix library(sjPlot) #> Learn more about sjPlot with 'browseVignettes("sjPlot")'. library(ggplot2) theme_set(theme_sjplot()) # Create data partially based on code by Ben Bolker # from https://stackoverflow.

How to map more informative values onto fill argument of sjPlot::plot_model

Whereas the direction of main effects can be interpreted from the sign of the estimate, the interpretation of interaction effects often requires plots. This task is facilitated by the R package sjPlot. For instance, using the plot_model function, I plotted the interaction between a continuous variable and a categorical variable. The categorical variable was passed to the fill argument of plot_model. library(lme4) #> Loading required package: Matrix library(sjPlot) #> Install package "strengejacke" from GitHub (`devtools::install_github("strengejacke/strengejacke")`) to load all sj-packages at once!

How to visually assess the convergence of a mixed-effects model by plotting various optimizers

To assess whether convergence warnings render the results invalid, or on the contrary, the results can be deemed valid in spite of the warnings, Bates et al. (2023) suggest refitting models affected by convergence warnings with a variety of optimizers. The authors argue that, if the different optimizers produce practically-equivalent results, the results are valid. The allFit function from the ‘lme4’ package allows the refitting of models using a number of optimizers.

Covariates are necessary to validate the variables of interest and to prevent bogus theories

The need for covariates—or nuisance variables—in statistical analyses is twofold. The first reason is purely statistical and the second reason is academic. First, the use of covariates is often necessary when the variable(s) of interest in a study may be connected to, and affected by, some satellite variables (Bottini et al., 2022; Elze et al., 2017; Sassenhagen & Alday, 2016). This complex scenario is the most common one due to the multivariate, dynamic, interactive nature of the real world.

Cannot open plots created with brms::mcmc_plot due to lack of discrete_range function

I would like to ask for advice regarding some plots that were created using brms::mcmc_plot(), and cannot be opened in R now. The plots were created last year using brms 2.17.0, and were saved in RDS objects. The problem I have is that I cannot open the plots in R now because I get an error related to a missing function. I would be very grateful if someone could please advise me if they can think of a possible reason or solution.

A table of results for Bayesian mixed-effects models: Grouping variables and specifying random slopes

Here I share the format applied to tables presenting the results of Bayesian models in Bernabeu (2022). The sample table presents a mixed-effects model that was fitted using the R package 'brms' (Bürkner et al., 2022).

A table of results for frequentist mixed-effects models: Grouping variables and specifying random slopes

Here I share the format applied to tables presenting the results of frequentist models in Bernabeu (2022). The sample table presents a mixed-effects model that was fitted using the R package 'lmerTest' (Kuznetsova et al., 2022).

Why can't we be friends? Plotting frequentist (lmerTest) and Bayesian (brms) mixed-effects models

Frequentist and Bayesian statistics are sometimes regarded as fundamentally different philosophies. Indeed, can both qualify as philosophies or is one of them just a pointless ritual? Is frequentist statistics only about $p$ values? Are frequentist estimates diametrically opposed to Bayesian posterior distributions? Are confidence intervals and credible intervals irreconcilable? Will R crash if lmerTest and brms are simultaneously loaded?

Bayesian workflow: Prior determination, predictive checks and sensitivity analyses

This post presents a run-through of a Bayesian workflow in R. The content is *closely* based on Bernabeu (2022), which was in turn based on lots of other references, also cited here.

Language and vision in conceptual processing: Multilevel analysis and statistical power

Research has suggested that conceptual processing depends on both language-based and vision-based information. We tested this interplay at three levels of the experimental structure: individuals, words and tasks. To this end, we drew on three …

Language and sensorimotor simulation in conceptual processing: Multilevel analysis and statistical power

Research has suggested that conceptual processing depends on both language-based and sensorimotor information. In this thesis, I investigate the nature of these systems and their interplay at three levels of the experimental structure---namely, …

Mixed-effects models in R, and a new tool for data simulation

In this talk, I will look over the rationale for LMEMs, and demonstrate how to fit them in R (Brauer & Curtin, 2018; Luke, 2017). Challenges will also be covered. For instance, when using the widely-accepted 'maximal' approach, based on fitting all possible random effects for each fixed effect, models sometimes fail to find a solution, or 'convergence'. Advice for the problem of nonconvergence will be demonstrated, based on the progressive lightening of the random effects structure (Singman & Kellen, 2017; for an alternative approach, especially with small samples, see Matuschek et al., 2017). At the end, on a different note, I will present a web application that facilitates data simulation for research and teaching (Bernabeu & Lynott, 2020).

Naive principal component analysis in R

Principal Component Analysis (PCA) is a technique used to find the core components that underlie different variables. It comes in very useful whenever doubts arise about the true origin of three or more variables. There are two main methods for performing a PCA: naive or less naive. In the naive method, you first check some conditions in your data which will determine the essentials of the analysis. In the less-naive method, you set those yourself based on whatever prior information or purposes you had. The 'naive' approach is characterized by a first stage that checks whether the PCA should actually be performed with your current variables, or if some should be removed. The variables that are accepted are taken to a second stage which identifies the number of principal components that seem to underlie your set of variables.

At Greg, 8 am

The single dependent variable, RT, was accompanied by other variables which could be analyzed as independent variables. These included Group, Trial Number, and a within-subjects Condition. What had to be done first off, in order to take the usual table? The trials!