Abstract:

As teachers we deliver new knowledge to our students and we always want to know if it was learned and how well it was learned. There are certain indicators if students comprehended it or not. For example, the interaction with students during the class, their participation in different class discussions. Such discussions can show if class as a whole understands the material but can't check whether each student in the class does: not all students actively participate for different reasons, and those who do often rely on other students' replies and comments.

The ultimate indicator if our work as instructors is well done is when a former student tells you that at another class that used the knowledge gained in your lecture (for example, in higher level math class or in physics), he was well prepared (or not prepared). Or when your fellow instructor of a different class tells you about his experience with your former students math preparation.

Unfortunately, we have to give grades to our students during the course, not few months/years later, we need to measure knowledge we delivered and they received. We test them throughout the course and at the end of it. We all have our philosophy about testing, some test often, some test once or twice during a semester, some require students to rely only on their memory and analytic skills during a test, some allow students to use their books and notes. The discussion what is the better way of testing, pros and cons of closed book vs open book did not start yesterday, it is old and have hundreds of supporters on each side. During the last couple of years I was experimenting with both types of testing at different levels of math classes and in this article I will summurize my findings.