The beta-binomial model updates uncertainty about an unknown success probability by combining a Beta prior with binomial observations. Its conjugate structure gives an exact posterior through parameter addition. section 7.4.2
The posterior is Beta(3, 27) after updating the prior with two defects and eight non-defects. section 7.4.2
Estimating a defect rate
A prior defect rate of 5% and a sample with 2 defects among 10 parts yield a posterior mean of 10%, balancing prior and sample evidence. section 7.4.2
