A better norm-referenced grading using the standard deviation criterion.

BACKGROUND

The commonly used norm-referenced grading assigns grades to rank-ordered students in fixed percentiles. It has the disadvantage of ignoring the actual distance of scores among students.

PURPOSES

A simple norm-referenced grading via standard deviation is suggested for routine educational grading.

METHODS

The number of standard deviation of a student's score from the class mean was used as the common yardstick to measure achievement level. Cumulative probability of a normal distribution was referenced to help decide the amount of students included within a grade. RESULTS of the foremost 12 students from a medical examination were used for illustrating this grading method.

RESULTS

Grading by standard deviation seemed to produce better cutoffs in allocating an appropriate grade to students more according to their differential achievements and had less chance in creating arbitrary cutoffs in between two similarly scored students than grading by fixed percentile.

CONCLUSIONS

Grading by standard deviation has more advantages and is more flexible than grading by fixed percentile for norm-referenced grading.