Reconstruct the decision
For each uncertain item, record where you looked, what evidence you selected, how you interpreted it, and why the chosen option defeated the alternatives. An error may begin with navigation to the wrong section, a missed reference, an unchecked condition, a paraphrase gap, or an inference that exceeded the text. Choose the earliest failure. If the correct answer was a guess, review it as an unstable success rather than treating it as mastered.
- Situation
- A learner finds the right workshop row but misses a general notice requiring advance registration.
- In practice
- Primary category: cross-document condition check, not vocabulary. The table search succeeded; integration with the notice failed.
- Why it works
- The correction should train combined rules rather than faster scanning of the same row.
Try this
- Log location, evidence, interpretation, and option decision.
- Mark correct guesses and slow answers for review.
- Assign one primary error category at the earliest break.
Retest the process on a new document
A useful correction changes behaviour on unfamiliar material. Navigation errors call for heading-prediction drills. Reference errors need substitution practice. Condition errors need profile-to-table matching. Paraphrase errors need claim rewriting, while inference errors need scope and certainty comparisons. After two or three focused exercises, test the same process in a new passage and compare accuracy, time, and confidence. Improvement means the process became more reliable, not that one answer was memorized.
- Situation
- Three viewpoint errors involve confusing concessions with final recommendations.
- In practice
- Correction: build claim/reason/concession/action grids for two new debates, then answer recommendation questions without the grid.
- Why it works
- A transfer task tests whether the distinction now survives outside the original wording.
Try this
- Count category frequency across multiple sets.
- Choose one correction with a visible behaviour change.
- Retest using new content and the same evidence demand.
Tactics to keep
- Review every low-confidence answer before comfortable correct answers.
- Track time by question type, not passage total alone.
- Save one evidence sentence and one distractor distortion per error.
- Retest without keeping the answer key visible.
Common pitfalls
- Calling all errors vocabulary problems.
- Reviewing only wrong answers and ignoring guesses.
- Repeating the same passage until familiarity replaces navigation.
Practice prompts
Prompt 1
Review one complete reading set and its timing notes.
Create an evidence-failure table for wrong, guessed, and unusually slow questions.
Check yourself
- Categories name a process.
- Timing evidence is included.
- A dominant pattern is visible.
Prompt 2
Select the dominant error category from your table.
Design two focused exercises and one unfamiliar transfer check.
Check yourself
- Exercises isolate the cause.
- The transfer uses new text.
- Success includes accuracy and time.
