The 40-Minute Reading Section and the Clock-Watcher
A candidate sits the EF SET reading section. The section is adaptive: item difficulty shifts based on whether the previous answers were correct. There are roughly 40 minutes for the reading block, and the candidate has 22 items remaining with 14 minutes left on the clock. On her last five items she got three right and two wrong, and the two wrong ones were both long multi-paragraph passages she guessed on after about 90 seconds each. She reports that the short grammar items take her about 20 seconds and she rarely misses them, but the long passages take 3-4 minutes when she actually reads them. She is now anxious and is considering clicking quickly through the remaining long passages to 'bank' the short items, since she believes short items are worth the same as long ones.
1. Is the candidate's plan to speed-click through long passages and bank the short items a good strategy? Explain the reasoning an adaptive test rewards.
Read the worked answer
No - the plan is built on a false premise.
Two things matter here:
Adaptive scoring is not a flat point-per-item system in the way she assumes. On an adaptive reading test, the difficulty of upcoming items is driven by her recent accuracy. Blowing through the long passages produces wrong answers, which lowers the difficulty of subsequent items. That may feel safer, but it caps the ceiling of what she can demonstrate. If the test can report a higher band, she needs to keep getting harder items right, not retreat to easy ones.
The time math doesn't support the plan either. 14 minutes / 22 items = ~38 seconds per item average. Her short items cost 20s; her real long-passage cost is 3-4 min. Even if she banks every short item, the long ones still exist and she cannot skip all of them without a large accuracy hit.
Better play: triage by cost, not by length. Skim the first sentence of each long passage, decide in ~15 seconds whether the question is answerable from a targeted read (e.g., a detail or vocabulary-in-context item) versus a whole-passage inference. Do the targeted ones properly and make a fast, reasoned guess on the whole-passage inference items rather than guessing blindly. Never spend 90 seconds and then guess - either commit to reading or commit to a fast guess.
2. She has 14 minutes and 22 items. Sketch a time budget and state the one assumption your budget depends on.
Read the worked answer
Budget sketch (38s average):
- Short grammar/vocab items: ~20s each. If ~12 of the 22 are short, that's ~4 min.
- Remaining ~10 long items: ~10 min left → ~60s each.
- Reserve ~1 min buffer for the final items and to avoid a rushed last click.
So: 12 short × 20s = 4 min; 10 long × 60s = 10 min; total 14 min with zero buffer - which is too tight. Realistic version: push short items to 15s (3 min), give long items 55s each (~9 min), leaving ~2 min buffer.
The assumption: that the mix is roughly 12 short / 10 long. She should verify this in the first 60 seconds by glancing at item types rather than assuming. If the mix is heavier on long passages, she must accept more fast guesses on whole-passage inference items and protect accuracy on the targeted-detail ones.