The points-per-game statistics from the 1993 NBA season were analysed for basketball players who went to college in four particular ACC schools: Duke, North Carolina. North Carolina State. and Georgia...

The points-per-game statistics from the 1993 NBA season were analysed for basketball players who went to college in four particular ACC schools: Duke, North Carolina. North Carolina State. and Georgia Tech. We want to find out if scoring is different for the players from different schools. Can this be analysed with a parametric procedure? Why or why not? The classical F-test that assumes normality of the populations yields F = 0.41 and H_O
is not rejected. What about the nonparametric procedure?