CGMoreh
- 3 skills
- 0 followers
- 1 day ago last updated
- ▌ Bayes Workflow R · cgmoreh bundleRun a Bayesian analysis in R with brms, rstanarm or cmdstanr on Stan, as an iterative workflow rather than a single fit. Covers choosing priors and checking what they imply, prior predictive simulation, MCMC failure read as a modelling problem, posterior predictive checks that name the next model, leave-one-out model comparison with pointwise attribution, prior and likelihood sensitivity by power-scaling, design calibration by simulation at the real sample size, and what weights() in a brm() call does and does not adjust for. Use when fitting, expanding, diagnosing or troubleshooting a model in any of those packages; when Stan warns about divergent transitions, treedepth, low effective sample size or non-convergence; or when a small sample needs a defence against the objection that the priors drove the result and the evidence for that defence has to be generated. For causal estimands and target populations use bayes-estimands-r; for writing the analysis up use bayes-reporting-r.
- ▌ Bayes Estimands R · cgmoreh bundleDefine and estimate the quantity a Bayesian analysis is actually about, rather than reading it off a coefficient table. Covers stating the estimand before the model, the difference between sample and population average treatment effects, poststratifying posterior draws to a target population, computing causal contrasts by simulating each counterfactual assignment, direct and indirect effects through a mediator, and checking whether a model expansion has inflated the uncertainty of the estimand while coefficients still look stable. Use when a Bayesian model is fitted or planned and the question concerns a treatment or exposure effect; when results must generalise beyond the sample; when posterior estimates from a survey sample must represent a target population whose composition differs from the sample; when a mediator sits between exposure and outcome; or when asked what a brms coefficient means on the response scale. For fitting and checking the model itself, use bayes-workflow-r.
- ▌ Bayes Reporting R · cgmoreh bundleWrite up a Bayesian analysis for a journal, so the workflow behind the estimates is visible in the paper rather than buried in an appendix. Covers the methods section stage by stage, wording for priors and their justification, reporting posteriors without significance thresholds, what belongs in the main text against the supplement, building a workflow appendix from the analysis log, checking a draft's numbers against its output, and the results section in the book's own interpretive register. Use when drafting or revising the methods or results section of a paper reporting Bayesian estimates; when a reviewer challenges the priors or the choice of a Bayesian analysis and the defence goes into a response memo or robustness paragraph; when converting an analysis notebook into a manuscript; when checking that a draft's numbers trace to the output behind them; or when deciding which checks to report. For running the analysis use bayes- workflow-r; for the quantity reported use bayes-estimands-r.