It is a widespread idea that CATs is a very efficient way of assessing people's language abilities.

Here's a rough description of how it works:

Each student is given an initial item of average difficulty. The students who give a correct answer are then presented with a more difficoult item. Those who give a wrong answer will have an easier item to deal with. The computer adapts to the students' different language levels, raising or lowering the difficulty, until an evaluation is possible. This kind of evaluation is based on statistical analysis (scary!! ;D).

What do you think of this kind of testing?

What are the pros and cons?

Are oral interviews a form of Adaptive Testing?