Dunning-Kruger: the gap between prediction and performance

No comments

Evidence says students struggle to predict their performance at all ages. David Read suggests how to help students be more aware of their progress for improved study

A pawn before a mirror, reflected as a king

Source: Vasileios Karafillidis / Shutterstock.com

The Dunning–Kruger (D–K) effect refers to someone’s ability to predict their performance on a task. In particular, lower-performing individuals tend to be unable to recognise weaknesses and take corrective action. This leads to unrealistically high predictions of their own performance. Higher-performing participants report more accurate predictions.

The D–K effect has appeared in several chemistry education studies in recent years. In a recent study, researchers looked at the D–K effect at different levels of chemistry degree programmes.

The D–K effect has appeared in several chemistry education studies in recent years. In a recent study, researchers looked at the D–K effect at different levels of chemistry degree programmes.¹

The authors speculated the D–K effect might diminish in more senior years, since lower performing students are less likely to progress to more advanced courses. They also thought that the perceived difficulty of chemistry might lead lower-performing students to be more accurate in their predictions.

The researchers stapled a short survey to exam papers for students to record predictions of the marks they expected to achieve. They collected data from around 3000 students.

As in previous studies, the highest performing students (those achieving above 80% in the exams) showed a tendency to underpredict their mark. Those achieving lower marks tended to overpredict their mark, with the level of overprediction increasing as attainment decreased.

To put this in perspective, first year students scoring less than 50% on the exam predicted a mark of 69% on average – but their actual average mark was just under 37%, a difference of over 32%.

The researchers observed similar patterns for all the courses in the study. The data shows that the D–K effect persists through senior levels of study despite the attrition of weaker students in earlier years. And it remains in chemistry at lower levels of attainment, regardless of the perceived difficulty of the subject.

Students unaware of their own poor preparation for assessments are detrimentally affected. It prevents them from taking steps to improve and it even compromises success for students who are putting a lot of effort into their studies.

Teaching tips

So, what can teachers do to help? Explicitly teaching metacognitive study methods will support students in overcoming the Dunning–Kruger effect by making their study more effective. Introduce them to Bloom’s taxonomy, show that they are working at lower levels and define activities that constitute higher order thinking. This paper outlines a study plan that follows a preview-attend-review-study-assess schedule:

So, what can teachers do to help? Explicitly teaching metacognitive study methods will support students in overcoming the Dunning–Kruger effect by making their study more effective. Introduce them to Bloom’s taxonomy, show that they are working at lower levels and define activities that constitute higher order thinking. Elzbieta Cook and colleagues outline a study plan that follows a preview-attend-review-study-assess schedule in their 2013 article:²

Preview – teachers provide guidance on how students can prepare for future sessions, perhaps through guided reading or video support material.
Attend – students should ask questions based on their preview of the material, as well as completing tasks as directed.
Review – very soon afterwards, students should read and annotate their notes, identifying any further questions they have.
Study – revisit material soon after the lesson – devise (and answer) questions such as ‘why?’, ‘how?’, ‘what if?’ etc.
Assess – perform regular reality checks, asking, ‘Are my study methods effective?’, ‘Could I teach this material to another student?’

This approach exposes students to material three times within a short period, supporting rapid consolidation of material. Incorporating regular testing provides opportunities for students to refine their predicting abilities and monitor their progress. This helps them to check the impact of new study methods and builds their confidence.

Some students may have achieved high grades in their previous studies using ineffective study strategies, like memorisation, and may not be receptive to different approaches. Give these students opportunities to take tests and make predictions. An unexpected, disappointing result can be a valuable reality check, opening the door for students to adapt their study practices. This will improve both their attainment and their ability to predict it.