bayes4psy -- an Open Source R Package for Bayesian Statistics in Psychology

Erik \v{S}trumbelj; Grega Repov\v{s}; Jure Dem\v{s}ar

arxiv: 1907.01952 · v1 · pith:7UWVMNSRnew · submitted 2019-07-03 · 📊 stat.AP · cs.MS· stat.ME

bayes4psy -- an Open Source R Package for Bayesian Statistics in Psychology

Jure Dem\v{s}ar , Grega Repov\v{s} , Erik \v{S}trumbelj This is my paper

Pith reviewed 2026-05-25 09:32 UTC · model grok-4.3

classification 📊 stat.AP cs.MSstat.ME

keywords Bayesian statisticsR packagepsychologyreaction timest-testbootstrappingdata analysisteaching tool

0 comments

The pith

The bayes4psy R package supplies Bayesian models tailored to psychology experiment data.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper presents an open-source R package to make Bayesian statistical analysis available to psychologists who generate distinctive data sets. Standard Bayesian tools often demand technical skills in probabilistic programming that are not part of typical psychology training, limiting their use. The package supplies ready models for t-tests, bootstrapping, reaction times, success rates, and colors, together with diagnostic, analytic, and visualization tools. A sympathetic reader would view this as removing a practical obstacle so researchers and students can apply Bayesian methods directly to their experiments.

Core claim

The bayes4psy package offers a collection of Bayesian models and methods for data analysis that arises from psychology experiments and serves as a teaching tool for Bayesian statistics in psychology. It contains Bayesian t-tests and bootstrapping along with models for reaction times, success rates, and colors, and supplies all diagnostic, analytic, and visualization tools needed for the modern Bayesian data analysis workflow.

What carries the argument

The bayes4psy R package, which implements and packages Bayesian models for common psychology data types along with workflow tools.

If this is right

Psychologists can analyze reaction time and success rate data with Bayesian methods without writing custom probabilistic code.
The package can serve as a direct teaching resource for introducing Bayesian workflows in psychology courses.
Diagnostic and visualization functions reduce the chance of misapplying Bayesian procedures in routine analyses.
Bootstrapping and t-test functions become immediately available for psychology-specific experimental designs.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

Wider availability of such models could increase the proportion of psychology papers that report full posterior distributions rather than point estimates.
The package structure suggests a template for building similar domain-specific Bayesian toolkits in other experimental sciences.
Users could test whether the supplied models recover known parameter values on simulated psychology data generated from the same likelihoods.

Load-bearing premise

The specific models supplied in the package are correctly implemented, numerically stable, and statistically appropriate for the data structures that arise in psychology experiments.

What would settle it

Applying the package models to standard psychology benchmark data sets and obtaining results that deviate substantially from independent Bayesian implementations or from known frequentist benchmarks on the same data.

read the original abstract

Research in psychology generates interesting data sets and unique statistical modelling tasks. However, these tasks, while important, are often very specific, so appropriate statistical models and methods cannot be found in accessible Bayesian tools. As a result, the use of Bayesian methods is limited to those that have the technical and statistical fundamentals that are required for probabilistic programming. Such knowledge is not part of the typical psychology curriculum and is a difficult obstacle for psychology students and researchers to overcome. The goal of the bayes4psy package is to bridge this gap and offer a collection of models and methods to be used for data analysis that arises from psychology experiments and as a teaching tool for Bayesian statistics in psychology. The package contains Bayesian t-test and bootstrapping and models for analyzing reaction times, success rates, and colors. It also provides all the diagnostic, analytic and visualization tools for the modern Bayesian data analysis workflow.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Desk Editor's Note private letter to a colleague

bayes4psy bundles standard Bayesian models for reaction times, success rates, and colors into an R package aimed at psychology users who skip probabilistic programming.

read the letter

The paper describes an R package that puts Bayesian t-tests, bootstrapping, and models for reaction times, success rates, and color data into one accessible tool, complete with diagnostics and plots. The curation for psychology experiments is the main practical point, and it targets a real audience that lacks the background for general tools like Stan or brms. If the implementations are straightforward wrappers around established methods, the package could lower the barrier for both analysis and teaching without requiring users to write their own models. The open-source release is a plus because anyone can inspect the code directly. The manuscript stays descriptive and does not include validation runs, simulation checks, or comparisons against other packages on real or synthetic data. That leaves the reliability of the specific models for psychology data structures as an assumption rather than a demonstrated result. The claim itself is narrow enough that the absence of benchmarks does not break it, but a software paper would be stronger with at least a couple of worked examples that match known frequentist outcomes. This is for psychologists or methods instructors who want ready-made Bayesian options rather than for readers hunting new statistical theory. It is worth sending to peer review because the target use case is common and the contribution is a usable piece of software even if the write-up needs more on verification.

Referee Report

1 major / 0 minor

Summary. The manuscript describes the bayes4psy R package, which provides a collection of Bayesian models and workflow tools for psychology experiment data, including t-tests, bootstrapping, reaction time models, success rate models, color models, and associated diagnostic, analytic, and visualization tools. The package is presented as both a practical analysis tool and a teaching aid to make Bayesian methods accessible to psychology researchers and students without requiring advanced probabilistic programming expertise.

Significance. If the supplied models prove correct, numerically stable, and appropriate for typical psychology data structures, the package would address a genuine accessibility gap and could increase adoption of Bayesian methods in the field while serving an educational role. The open-source release is a strength, as it permits direct inspection, reproduction, and community extension of the code.

major comments (1)

[Abstract] Abstract: the claim that the package supplies models 'to be used for data analysis' is not supported by any implementation details, validation benchmarks, or correctness checks. Without such evidence it is impossible to assess whether the models are statistically appropriate or numerically reliable for the data structures arising in psychology experiments.

Simulated Author's Rebuttal

1 responses · 0 unresolved

We thank the referee for their feedback. We address the single major comment below.

read point-by-point responses

Referee: [Abstract] Abstract: the claim that the package supplies models 'to be used for data analysis' is not supported by any implementation details, validation benchmarks, or correctness checks. Without such evidence it is impossible to assess whether the models are statistically appropriate or numerically reliable for the data structures arising in psychology experiments.

Authors: We agree that the manuscript as submitted does not include explicit implementation details, validation benchmarks, or correctness checks to support the abstract's claim. The open-source nature of the package permits code inspection, but this is insufficient for readers to evaluate statistical appropriateness or numerical reliability. In the revised version we will add a dedicated validation section that reports simulation studies on typical psychology data structures, comparisons against known frequentist results, and checks for numerical stability and convergence diagnostics. revision: yes

Circularity Check

0 steps flagged

No significant circularity: software package description with no derivations

full rationale

The paper presents an open-source R package (bayes4psy) containing pre-built Bayesian models for psychology data (t-tests, reaction times, success rates, colors) plus workflow tools. No mathematical derivations, parameter fits, predictions, or uniqueness theorems are claimed or derived. The central claim is descriptive (the package exists and supplies these models), which is externally verifiable by code inspection rather than by internal reduction to inputs. No self-citation chains, ansatzes, or fitted-input-as-prediction patterns appear. This is a normal non-circular software paper.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

The work introduces no new mathematical entities or free parameters; it packages existing Bayesian modeling techniques for a new user group.

axioms (1)

domain assumption Bayesian hierarchical models are suitable for typical psychology experiment data structures
The package is built on this premise without providing new justification.

pith-pipeline@v0.9.0 · 5697 in / 1146 out tokens · 39262 ms · 2026-05-25T09:32:30.588293+00:00 · methodology

bayes4psy -- an Open Source R Package for Bayesian Statistics in Psychology

Core claim

What carries the argument

If this is right

Where Pith is reading between the lines

Load-bearing premise

What would settle it

discussion (0)