Exam administration in large is fundamentally flawed because it
1) tests mostly for memory capability rather than analytical ability, and
2) tests for performance under extreme pressure.
Both are very bad metrics for predicting the academic ability of a person because in the real-world these conditions are rarely present, BOCTAOE. And sure, memory is somewhat of an indicator for capability but not without the other (and besides, we now have Google to compensate.) Einstein said that “education is what remains after one has forgotten everything he learned in school.” In other words, education = understanding, not memory.
I’m a poor test taker, so you might question my motivation for changing the system although I honestly feel that what I propose is fair for all and a much better indicator for capability than current methods.
All exams should be
1) open material (anything paper allowed except for the one in front of the person sitting next to you,) and
2) loosely bounded by time (say, x3 the present settings.)
This will allow for
1) hard questions based on understanding rather than memory, and
2) elimination of the pressure element that is a differentiation factor that has very little to do with real life requirements.
The only down side to this is that it requires more work from the people who write the exams, which is the very reason why it will never be implemented. I had two teachers at UC Santa Cruz who implemented the above; they were the best educators I have ever had (if you are curious who they are, email me.)