Selective trial-test review: build an error log you can use

What should an error-log entry contain?

Keep the task reference, the step you got wrong, a corrected explanation and a new example to try later. The entry should tell you what to do when you reopen it.

A screenshot can preserve the question, but it does not explain your mistake. Add a short note in your own words before moving on. Keep any access and ownership requirements for the original material in mind.

Name the incorrect operation

Original entry: “I reduced $80 by 20%, then added 20% to the reduced price and expected $80 again.” The correction is that the second percentage has a new base: $80 × 0.8 = $64, then $64 × 1.2 = $76.80.

A useful retest changes the original price and the percentages. Repeating the same answer immediately would mostly check memory of the reviewed question.

Group by the correction you need

“Maths” is too broad to choose tomorrow’s work. “Percentage base changed” is specific. For reading, “I treated an example as the conclusion” is more useful than “comprehension mistake”.

Do not create dozens of categories you will never use. Start with a few recurring actions and refine them when the entries show a real distinction. The log should make practice easier to select.

Give each entry a later independent check

At the next review, cover the corrected solution and attempt the changed example. Mark whether you could explain the method, rather than only whether the final number matches.

After targeted review, use a new mixed task to see whether the correction survives a less obvious context. Retain the original attempt so the assisted explanation does not replace your first result.

Try it yourself

Rewrite the entry “Careless maths error” so it gives a specific next action for the percentage example.