Steps to reproduce:
1. Review of grader report. 4 quizzes taken by 5 students. There are two groups (group A - 2 students, group B - 3 students). Preferences set to show percentages of non-empty grades, to 2dp.
2. Review of the grader report for the whole group gives the correct results for the average.
3. Review of the grader report for each group gives incorrect results for the group averages (I cannot understand how these are calculated?)
Screen shots are attached for the whole group and the two sub-groups showing the errors in calculation of the group averages.