We discuss attempts to change the attitude of students towards an Electricity and Magnetism class taken by Engineering majors in the required Introductory Physics sequence. Class size is large (150 to 200 students) leading to little personal interaction. Testing has, in the past, been based on multiple choice problems which may be graded efficiently. We have demonstrate that students do not solve these from the physics but rather develop strategies for selection of answers based on memorized

more »

... es and formulae. The two popular texts at this level encourage memorization by listing formulae to be remembered and by supplying training questions which are easily related to the formula lists rather than to the text. The overall student attitude is that they should approach the course by memorizing operational formulae. Performance is poor with up to 30% failing to attain a satisfactory grade. We have been attempting to change the student attitude towards the course and to improve performance. As a first step we have examined the multiple choice testing format to determine whether it is a reliable measure of performance. We select questions which have been previously used under the multiple choice format and where there is a historical record of student success. The same questions are now asked in a format where a numerical answer is required and only a completely correct answer is acceptable. Under the new format the correct answer is achieved only half as frequently as under the previous multiple choice format. Clearly the multiple choice format substantially overestimates the student's ability to achieve a correct result. We conclude that this format should be abandoned. After the student's initial shock at the change to question format has worn off we find that students adapt and performance improves.