Beyond hypercorrection: remembering corrective feedback for low-confidence errors.

Area of Science:

• Cognitive Psychology
• Educational Psychology
• Learning Sciences

Background:

• Effective error correction is crucial for learning.
• The hypercorrection effect suggests high-confidence errors are better remembered than low-confidence errors.
• Prior research primarily focused on high-confidence errors.

Purpose of the Study:

• To investigate if low-confidence errors can be corrected and retained.
• To explore factors influencing the correction of low-confidence errors.
• To examine the role of confidence, surprise, and feedback memory in error correction.

Main Methods:

• Participants completed a multiple-choice test with confidence ratings.
• Immediate corrective feedback was provided.
• A cued-recall test assessed error correction after a short delay.
• Two experiments were conducted using control and trick questions.

Main Results:

• The typical hypercorrection effect was observed for control questions.
• Low-confidence errors on trick questions were corrected as effectively as high-confidence errors.
• Memory for feedback and original responses significantly predicted error correction, more so than confidence or surprise.

Conclusions:

• The hypercorrection effect is not universal and depends on question type.
• Effortful processing and problem-solving for low-confidence errors on trick questions may enhance feedback memory.
• Memory for corrective feedback is a critical factor in successful error correction, even for low-confidence errors.