Does student confidence on multiple-choice question assessments provide useful information?

Area of Science:

• Dental Education
• Medical Education Assessment
• Psychometrics in Education

Background:

• Multiple-choice question (MCQ) assessments typically provide only a percentage correct score.
• Student competence is inferred from scores, neglecting confidence and clinical impact of errors.
• Limited feedback hinders identification of student overconfidence or underconfidence.

Purpose of the Study:

• To evaluate dental students' confidence in their incorrect responses.
• To determine how confidence is influenced by the clinical impact of incorrect answers, question type, and gender.
• To differentiate between uninformed (incorrect, not confident) and misinformed (incorrect, confident) responses.

Main Methods:

• Exploratory, cross-sectional study with 104 Year 3 dental students.
• Students completed 20 MCQs on implant dentistry, indicating correctness and confidence for each answer.
• Incorrect responses were categorized by faculty as benign, inappropriate, or potentially harmful; question type (factual/complex) and gender were recorded.

Main Results:

• Misinformed responses (incorrect but confident) were more common than uninformed ones (22% vs. 8%).
• Misinformed responses were more frequent with complex questions than factual ones (p < 0.05).
• Contrary to expectations, student confidence did not decrease as the potential clinical harm of incorrect answers increased.

Conclusions:

• Assessing student confidence alongside accuracy effectively identifies misinformed responses.
• This approach can inform targeted remediation strategies for dental students.
• Making calibration errors (confidence vs. accuracy) visible is valuable for professional development feedback.