Related Experiment Video
Updated: Mar 26, 2026
04:46
'Boden Food Plate': Novel Interactive Web-based Method for the Assessment of Dietary Intake
Published on: September 18, 2018
7.9K
Long-menu questions in computer-based assessments: a retrospective observational study
Bernard Cerutti1, Katherine Blondon2, Annick Galetto3
1Unit of Research and Development in Medical Education, Faculty of Medicine, University Geneva, 1 Rue Michel Servet, Geneva, 1211, Switzerland. bernard.cerutti@unige.ch.
BMC Medical Education
|February 11, 2016
Summary
Long-menu questions in pediatric assessments demonstrate strong psychometric properties, proving easier and more discriminative than Type A and Pick N formats. These findings suggest their value in enhancing computer-based medical education.
Area of Science:
- Medical Education
- Psychometrics
- Pediatric Assessment
Background:
- Computer-based assessments in pediatrics utilize sequential clinical case information.
- Common question formats include Type A (single answer), Pick N, and Long-menu.
- Long-menu questions involve a progressive narrowing of answer choices based on initial student input.
Purpose of the Study:
- To analyze and compare the psychometric properties of Long-menu questions against Type A and Pick N formats.
- To evaluate the difficulty and discrimination indices of different question types in pediatric computer-based assessments.
Main Methods:
- A retrospective analysis of pediatric exam data from 2009-2015 was conducted.
- Difficulty level and discrimination index of Long-menu, Type A, and Pick N questions were compared.
- Multiple-way analyses of variance were employed for statistical comparison.
Main Results:
- The analysis included 558 items across 13 exam sessions with 855 students.
- Long-menu questions (38%) were found to be significantly easier than Type A questions.
- Long-menu questions exhibited a higher discrimination index compared to both Type A and Pick N questions.
Conclusions:
- Long-menu questions possess favorable psychometric properties compared to Type A and Pick N formats.
- Their design offers greater variety and reduces the cueing effect, potentially mirroring real-world clinical practice.
- Further research is recommended to confirm these findings and explore their application in computer-based assessments.
