Statistical framework · BEGINNER LESSON
Bayesian
Bayesian statistics is a way to update a belief or probability after new data are observed. It combines prior knowledge with the likelihood of the data to produce a posterior probability or posterior distribution.
What it is and what it measures
Bayesian statistics is a framework for making probabilistic statements about a hypothesis, parameter, or treatment effect after data are observed. It starts with a prior, meaning what is believed before the new data, and then updates that belief using the likelihood of the observed data to produce a posterior. The posterior is the main output: it tells us the probability of different parameter values or hypotheses after combining earlier knowledge with the new evidence. In practice, this can be used to estimate the probability of a treatment benefit, a probability distribution for an effect, or a credible interval, which is a range of values that the posterior assigns high probability to.
[1] Bayesian statistics: a primer for perioperative medicine clinicians [2] Beyond the blank page: Frequentist and Bayesian ... - PMC [3] Bayesian analyses for research on Alzheimer's disease ...
How the update works
The core idea is Bayes’ theorem, a rule for updating uncertainty when evidence arrives. The prior represents knowledge from earlier studies, expert judgment, theory, or sometimes a deliberately weak or noninformative starting point. The likelihood describes how probable the observed data are for different possible parameter values. Combining prior and likelihood produces the posterior distribution, which reflects both earlier information and the new data. This differs from frequentist analysis, which uses only the observed dataset and does not formally include prior knowledge.
[1] Bayesian statistics: a primer for perioperative medicine clinicians [2] Beyond the blank page: Frequentist and Bayesian ... - PMC [3] Bayesian analyses for research on Alzheimer's disease ...
Why people use it
Bayesian methods are useful when existing knowledge matters, such as in clinical research, small samples, or complex models. They can borrow strength from earlier evidence, which may avoid extreme estimates and improve stability. They also give outputs that are easy to interpret as probabilities, rather than only test results based on long-run error rates. Some versions can also include practical thresholds, such as the region of practical equivalence, which marks values small enough to be treated as essentially no meaningful difference.
[1] Bayesian statistics: a primer for perioperative medicine clinicians [2] Beyond the blank page: Frequentist and Bayesian ... - PMC [3] Bayesian analyses for research on Alzheimer's disease ...
Limits and cautions
Bayesian analysis has real costs. Results can depend strongly on the chosen prior, especially when data are limited, so different reasonable priors may lead to different conclusions. Many Bayesian models also need heavy computation, often using Markov chain Monte Carlo, which can take time and requires checking that the algorithm has converged. Because the ideas of prior, posterior, and probability updating are unfamiliar to many readers, the method can be harder to explain and may raise questions about reproducibility if the prior choice is not well justified.
[1] Bayesian statistics: a primer for perioperative medicine clinicians [2] Beyond the blank page: Frequentist and Bayesian ... - PMC [3] Bayesian analyses for research on Alzheimer's disease ...
SEE THE IDEA
Bayesian update from belief to posterior
Starting belief before seeing the new data, based on earlier evidence, expert judgment, or weak assumptions.
How well different parameter values would explain the observed data.
Updated belief after combining the prior with the likelihood.
Used to describe probabilities of effects, credible intervals, or practical thresholds.
Prior and likelihood combine to produce the posterior, which is then used for interpretation.
A Bayesian analysis updates earlier knowledge with observed data to form a new probability statement.
TRY IT YOURSELF
Test your understanding
Choose an answer, then check the explanation.
PUT IT TOGETHER
Match each concept to its meaning
Drag a label onto its description, or select a label and then select a description. This also works with a keyboard.
4 pairs to match
Sources and further reading
This general lesson was generated from public source material. It explains a concept; it does not interpret a private paper or provide medical advice.