Prior Elicitation Calculator
Translates your assumptions about a binary event rate into a Beta prior distribution. You can specify a median and an interval, an effective prior sample size and mean, or historical data with discounting, and the tool fits the matching Beta prior for use in the other Bayesian design calculators.
How it works
Three modes are supported: quantile matching (fit a Beta to a specified median and interval), effective-sample-size based (choose a prior mean and how many patients' worth of information the prior represents), and historical (build a discounted power prior from past data). The tool reports the fitted Beta parameters, its effective prior sample size, and a prior-predictive check of the outcomes the prior implies.
When to use it
- You need to encode expert judgment or prior evidence as a Beta prior for a proportion.
- You want to control how much weight the prior carries via an effective sample size.
- You want a prior-predictive check to see whether your prior implies plausible data before using it downstream.
Assumptions & limitations
- A prior encodes assumptions; it is not an objectively correct distribution, and different reasonable inputs produce different priors — sensitivity analysis across specifications is recommended.
- The effective prior sample size (α + β for a Beta prior) describes how much information this prior contributes under the Beta–Binomial model; it is not a universal information measure.
- The fitted interval summarizes prior belief, not the frequentist coverage of a confidence interval.
- Elicitation currently targets a Beta prior for a binary proportion; the quality of downstream sample-size and operating-characteristic results depends directly on the prior you specify.
For the full methodology, derivation, and worked examples, read the complete guide.