From time to time I will adjust a set of scores. I promise never to do this on final course grades, but I may do it on quizzes, exams and other graded activities. When I make an adjustment, with the help of my computer, I create a situation in which the highest score in the class (whatever that score may be) is transformed into 100, and the average (mean) score for the class is transformed into 80. This means that the best students will receive As, while the class average will be a B-. I will perform a transformation on a set of scores only if the scores, on the whole, are much higher (resulting in grade inflation) or much lower (resulting in unhappy students) than I feel they should be, or when the assignment is such that no standard exists for setting grades other than comparing students with each other.

To adjust a set of scores I do the following:

1. 

I determine the mean () for the class.

2.

I determine the standard deviation (s) for the class.

3.

I then compute a z-score for each student by taking that student's score, subtracting the mean and dividing the result by the standard deviation. (For any set of z-scores the standard deviation will be 1 and the mean will be 0).

4.

I divide 20 by the highest z-score in the class in order to obtain the standard deviation for the curve and then multiply each z-score by the dividend and add 80. The resulting curved score is a modified T-score.