fact_checkyour constructs, quantified

Media & mass communication scales, Bayesian-ready

Choose a theoretical construct. Each entry gives you its canonical measurement, a defensible reference prior drawn from the published literature, and preset sample data, then runs the exact Bayesian analysis for the effect you're studying. Change the numbers and it all recomputes.

What is a "reference prior" here? A weakly informative or literature-grounded starting distribution, following Jeffreys (1961) and the modern default-prior guidance of Gelman, Jakulin, Pittau, & Su (2008). It is a transparent starting point to be sensitivity checked, not a hidden assumption. Every prior used here is disclosed in the pink "Your prior" card below.

Select a construct to load its theory, citation, and a Bayesian analysis.

science Your prior (disclosed)

campaign Your sample data

Prior Posterior
Posterior mean
95% equal-tail CI
Posterior SD
Naive proportion
P(θ > 0.5 | data)