Not only did she have a perfect score, but she "adamantly" corrected a mistake in the control group. That must have some impact on the Bayesian estimate too. Her correction is worth something, and her confidence in providing that correction is worth something.