第11章
On the day of the test, let students know when feedback will be provided. The longer students have to wait
for correct answers, the less interested they become in checking their work. Provide immediate feedback if
you can. Here are two suggestions for accomplishing this:
. Students record their answers on both the test and the answer sheet. They hand in the answer
sheet and keep the test, scoring it themselves from a key or keys you post immediately after the
test. The computerized test bank makes it possible for you to provide feedback that includes
both correct answers and page references from the text.
. Create an exam that can be completed in 30 minutes (25–30 multiple-choice questions) and
devote the rest of a 50-minute class to feedback. Ask students to jot down their answers on their
own paper after they complete the test. Then collect the students’ answer sheets, while students
retain their notes of their answers and the test. Review the test items, allowing students to ask
questions about perceived ambiguous items. At the end of the class, collect the tests and the
students’ notes. With this procedure, you only have to return the graded answer sheets in order
for students to receive maximum feedback.
How can I avoid hassles about test items?
First, select and write items carefully, being sure that you agree that the keyed answer is the best answer.
Then make your key carefully. Examine the printout of scores before posting it, and check any item for
ambiguity or incorrect keying if half or more of the students miss it. If you make a mistake in your choice of
items or on the key, correct it graciously.
Do not spend class time arguing about test items. Have students submit their case in writing if they feel that
their choice for an item is better than the keyed answer.
DECISIONS ABOUT EVALUATION
Should I use numbers or letters to grade students’ work?
Any activity that is to contribute to students’ grades should be assigned a numerical score so an
unambiguous point system can be used for determining final grades. Some students have difficulty
understanding that an A on a minor assignment is not weighted as heavily as a D on the final exam.
If you use large numbers, particularly for a large class, the range of final totals for students is larger and the
cutoff between letter grades appears reasonable.
How will I assign final grades?
There are two traditional methods for assigning grades, the criterion-based system and the normative-based
or on-the-curve system.
In criterion-based grading, criteria are published in the syllabus. For example, if a student gets 92 percent to
100 percent of the maximum number of points, the student can be assured of an A. This method is preferred
by the majority of instructors because it is easy for students to understand and accommodates variability in
the performance of different classes.
xix
In normative-based grading it is predetermined that you will, for example, give an A to the top 10 percent of
students, a B to students in the next 20 percent, and so forth. This method has the advantage of not
requiring you to commit yourself to absolute criteria, but it has the disadvantage of causing some students
to be uncertain about where they stand during the term. Also, normative-based grading assumes that your
class is a random sample of students who take introductory courses at your school and does not take into
account that some classes as a whole are better than others. Identical student performance could earn an A
in one class and a B in another when the normative-based system is used.
Should I use standard scores?