Show simple item record

AuthorKassab, Salah E
AuthorFida, Mariam
AuthorRadwan, Ahmed
AuthorHassan, Adla B
AuthorAbu-Hijleh, Marwan
AuthorO'Connor, Brian P
Available date2021-09-01T10:03:25Z
Publication Date2016
Publication NameMedical education
ResourceScopus
URIhttp://dx.doi.org/10.1111/medu.13054
URIhttp://hdl.handle.net/10576/22432
AbstractContext In problem-based learning (PBL), students construct concept maps that integrate different concepts related to the PBL case and are guided by the learning needs generated in small-group tutorials. Although an instrument to measure students’ concept maps in PBL programmes has been developed, the psychometric properties of this instrument have not yet been assessed. Objectives This study evaluated the generalisability of and sources of variance in medical students’ concept map assessment scores in a PBL context. Methods Medical students (Year 4, n = 116) were asked to construct three integrated concept maps in which the content domain of each map was to be focused on a PBL clinical case. Concept maps were independently evaluated by four raters based on five criteria: valid selection of concepts; hierarchical arrangement of concepts; degree of integration; relationship to the context of the problem, and degree of student creativity. Generalisability theory was used to compute the reliability of the concept map scores. Results The dependability coefficient, which indicates the reliability of scores across the measured facets for making absolute decisions, was 0.814. Students’ concept map scores (universe scores) accounted for the largest proportion of total variance (47%) across all score comparisons. Rater differences accounted for 10% of total variance, and the student × rater interaction accounted for 25% of total variance. The variance attributable to differences in the content domain of the maps was negligible (2%). The remaining 16% of the variance reflected unexplained sources of error. Results from the D study suggested that a dependability level of 0.80 can be achieved by using three raters who each score two concept map domains, or by using five raters who each score only one concept map domain. Conclusions This study demonstrated that concept mapping assessment scores of medical students in PBL have high reliability. Results suggested that greater improvements in dependability might be made by increasing the number of raters rather than by increasing the number of concept map domains.
Languageen
PublisherBlackwell Publishing Ltd
SubjectBahrain
clinical decision making
concept formation
curriculum
education
evaluation study
human
learning
medical education
medical student
problem based learning
procedures
psychology
Bahrain
Clinical Decision-Making
Concept Formation
Curriculum
Education, Medical, Undergraduate
Educational Measurement
Humans
Learning
Problem-Based Learning
Students, Medical
TitleGeneralisability theory analyses of concept mapping assessment scores in a problem-based medical curriculum
TypeArticle
Pagination730-737
Issue Number7
Volume Number50
dc.accessType Abstract Only


Files in this item

FilesSizeFormatView

There are no files associated with this item.

This item appears in the following Collection(s)

Show simple item record