Grade Distribution Normality Check

Seeing the student grade distributions of the 6 large residential courses, the team is tempted to draft recommendations for instructors and report to them what particular aspects could be addressed to improve students' academic learning outcome. However, before they launch statistical tests, they need to verify if the student grades data approximately follows normal distribution, a sufficient condition rendering the design of statistical models valid for those courses. You suggest that a QQ-plot is a great method to determine how similar a distribution is to another.

• Make a 3 * 2 figure (again, 6 subplots) so that for each course you have a QQ plot using the student grade samples versus the normal distribution with the same mean and standard deviation
• You need to use a legend on each plot to specify the corresponding course name and number of students involved. For example, you can draw a legend and specify "STATS 250, n=5000" to indicate that you are analyzing STATS 250 course with 5000 enrolled students records being used for analysis
• For each QQ-plot, use annotate tool inside the graph to circle observations which are 2 standard deviations outside from the QQ-line (a straight line showing the theoretical values for different quantiles under normal distribution)
• Write a couple of sentence about the figure discussing the courses and whether they seem to be normally distributed.

Hint: You may find using fig = plt.figure() and fig.add_subplot() functions helpful to create subplots. You don't have to use these functions though.

def grade_normality(): return None grade_normality()