Analysis of proctor marking accuracy in a computer-aided personalized system of instruction course.

The team looked at a college course that uses CAPSI. CAPSI lets students move at their own speed and pass small tests to advance.

Two proctors graded every test. The study asked: does the second proctor catch mistakes the first one missed?

Double marking caught far more errors. Accuracy jumped when the second proctor looked at clearly wrong answers.

The boost was biggest for items that were obviously incorrect, not for partly right ones.

Hawley et al. (2004) reviewed rate-building studies and found little extra gain once practice and praise were held steady. Their warning to keep fluency claims modest pairs well with CAPSI’s tight feedback loop: accurate scores matter more than fast scores.

Cameron et al. (1996) showed that matching reading difficulty to the learner gives the best generalization. CAPSI uses the same idea: each test must match the unit just studied, and accurate marking keeps the match honest.

Jackson et al. (2025) found college students own plenty of assistive tech yet still under-use it. Likewise, CAPSI gives instant computer feedback, but human proctors still need to stay sharp; tech alone is not enough.

If you run self-paced courses, have two staff grade the same key tests. Swap proctors each week so fresh eyes catch slips. You will keep student progress honest and save time on later re-grades.