I read in the Trenton Times Nov. 8 that Diane Metcalf-Leggette ’13 had sued the University for violating the Americans with Disabilities Act because it refused to grant her “100 percent extra time on classroom examinations.”
Apparently the University had granted some accommodations (her exams are limited to one per day, she can take exams in a private area without distractions, and she can take breaks during them). But she contended that this was insufficient to compensate for her disabilities.
“How much is enough?” is not merely a critical existential question, it is also an important practical one. Providing accommodations in exams for students with disabilities is not only good institutional manners; it is required by law.
One solution is to give all students unlimited time. This is impractical as well as unsuitable for many situations. A practical alternative is a variation of the method of low-dose extrapolation common in drug research, but it requires a data-gathering effort by the University.
Specifically, an exam should be given to students without disabilities who are divided randomly into groups that receive, say, one hour, two hours, three hours, etc. Then their average scores are plotted as a function of the amount of time they received. This allows us to connect time with score, and thence to adjust the scores to what they would have been with unlimited time. Next, students with disabilities are provided unlimited time (if they request it). When we then use the adjusted scores, the field is level.
A study by Educational Testing Service on the SAT found that allowing extra time on the verbal portion yielded no improvement in score, and so extra time for disabled examinees could be allocated freely. But extra time on the quantitative section had a significant positive effect, and so some adjustment ought to be made.
For further details, see Chapter 7 in my recent book, Picturing the Uncertain World, Princeton University Press.