The test closed on its own and the summary screen said 22 attempted, 18 not visited. You knew how to do at least half of those 18. Nothing in the test was beyond you, and you still scored like someone who had not studied. The feeling afterwards is that you were unlucky with a hard paper, which is the wrong diagnosis and the reason it repeats in the next drive.
A machine-scored screening test: 40 questions in 45 minutes, all sections in one timer, questions presented one per screen with a Next button. First or second attempt of a placement season. The student prepared by solving topics, not by rehearsing a clock, and had never once timed a full section before test day. Linked module: 6.2 (Online assessment prep).
Do the arithmetic the test screen was doing all along. 45 minutes is 2,700 seconds, and 2,700 divided by 40 is 67.5 seconds per question. That is the whole budget, before you take off anything for reading instructions and waiting for pages to load.
Now price the mistake. The attempt log showed 9 minutes on question 14, a probability sum that was eventually abandoned. Nine minutes is 540 seconds, and 540 divided by 67.5 is exactly 8. One question consumed eight questions' worth of clock and returned nothing. The 18 unvisited questions were not lost to difficulty; they were lost to one refusal to let go.
Two mechanisms did the damage together, and both are ordinary. The first is strict order: answering question 1, then 2, then 3 means the questions you can do fastest sit behind the questions you cannot do at all, and the ones at the end are the ones you never price. The second is the sunk-cost feeling — after four minutes on a sum you feel you must finish or the four minutes are wasted. They are already wasted. The next minute is a fresh decision.
Note what is not the cause. The paper was not unusually hard, the student was not slow at arithmetic, and no section was unfair. The failure is a scheduling failure, and scheduling is a skill with a method, which is why this is fixable in a week rather than a semester.
Three passes, and two caps you obey by the clock rather than by feel.
The cap is enforced by a rule, not by willpower: when the cap hits, you mark your best current option, flag the question, and move on that second. If you are already in the situation with ten minutes gone and twenty questions untouched, stop solving and run pass 1 immediately across everything you have not seen. Cheap marks at the end of a paper are still marks.
Never take an untimed practice set again. Every round gets a visible timer and a per-question log, because a topic you can do in four minutes and not in sixty seconds is a topic you cannot do on this test. Rehearse full sections, not single topics, so that the skip-and-return motion becomes automatic. Keep one number in your head for every test you sit: seconds per question. If a student cannot state that number for a test they just took, they were not running a plan.
Q: "You had a low score on the online assessment. What happened?" The weak answer blames the paper. The strong answer is arithmetic: "The budget was 67.5 seconds per question, I spent 540 seconds on one probability sum, which was eight questions' worth, and 18 questions went unattempted. I now run three passes with a 45-second and a 100-second cap, and I log seconds per question on every practice set." That answer describes an engineer diagnosing a resource problem, and it lands very differently from bad luck.