Date: Tue, 14 Feb 2006 02:11:58 -0500
From: "Dr. Brenda Ponsford" <bponsford@neo.rr.com>

Dr. Morris has some valuable cautions.

As one who uses test bank questions, I have a few suggestions:

1. Do not use the "select randomly" feature.  Instead, carefully read each question you select.  Pay careful attention to how the material is presented in the textbook including variations in marketing jargon.  Check for typos, etc.
2. Always have a textbook page reference for every question used. This saves headaches from lecture development, question quibbles with students, and perhaps, the rejection of the question for all students in calculating a curve after the exam is given.
3. Because people interpret mutliple choice questions differently, I developed a practice that dates to a time when I had many ESL and international students which posed interpretation problems. This method can also save a lot of wasted time in picky arguments.  I adopted the common practice of choice that we use on essay questions such as "answer any 5 from the 7"  to multiple choice.   I have the students answer any 50 from the 55 questions and just skip their rejects on the scan-tron sheet for ease of grading.  In my example, the grade is how many right out of 50, then multiple by 2% for each question.  If a question is peculiar to the student (back in the 80s, a Soviet student was unfamiliar with the "Cola Wars"), the student uses British English, etc., then the student can skip out "bad" questions.   Only count the scantron answers to avoid cheaters who intentionally bubble in choice one on the scantron while circling their back-up choice on the exam booklet and then claim innocent transcription errors.

Your ordinary student will view you as merciful but they still have to answer 50 questions.  Remind them to proof the scantrons for accurate transcription.

Hope this helps,
B. Ponsfor