Having spent a decade of my life in for-profit education, specifically standardized test preparation (Sylvan, Kaplan, College Network, Grockit, Veritas Prep), I can say that there is more data that could be collected from tests. For example, instead of looking who got the "right" answer, you could look at which wrong answers people selected and why. For example, students who chose "A" probably have experience with microscopes, perhaps don't read instructions carefully.
More interestingly, you could learn a lot about a student's thought process by analyzing the time spent on a particular question, or whether they chose another answer choice before settling on one, or whether It's possible that the computer adaptive exams (GRE, GMAT, NCLEX, etc.) take this into account because the computer can collect this data, but paper and pencil tests simply cannot.
A student who chose answer choice "A" after 5 seconds is probably careless or overconfident, whereas a student who chose "A" after a minute of waffling between "A" and "C" possibly lacks confidence.
I think standardized tests could become a lot more useful if we could collect more data from them.
More interestingly, you could learn a lot about a student's thought process by analyzing the time spent on a particular question, or whether they chose another answer choice before settling on one, or whether It's possible that the computer adaptive exams (GRE, GMAT, NCLEX, etc.) take this into account because the computer can collect this data, but paper and pencil tests simply cannot.
A student who chose answer choice "A" after 5 seconds is probably careless or overconfident, whereas a student who chose "A" after a minute of waffling between "A" and "C" possibly lacks confidence.
I think standardized tests could become a lot more useful if we could collect more data from them.