Validity and Reliability of Rubrics for Assessment of Preclinical Orthodontic Practices in Dental Students.

Background: Objective assessment of preclinical dental practices remains challenging, as it requires consistency, transparency, and standardization in the evaluation of practical competencies. Rubrics have been proposed as useful tools for improving the quality of assessment; however, evidence regarding their psychometric properties in preclinical orthodontic education remains limited. Aim: This study aimed to analyze the reliability, internal structure, and academic usefulness of assessment rubrics applied to preclinical orthodontic practice in undergraduate dental education. Methods: A cross-sectional psychometric retrospective study was conducted with 175 third-year dentistry students enrolled in Orthodontics I at the European University of Madrid. Two analytical rubrics designed to assess wire-bending and cephalometry practices were evaluated. Reliability was assessed using Cronbach's alpha, construct validity using exploratory factor analysis, and associations with academic performance by correlation analyses. Results: The wire-bending rubric showed moderate internal consistency (α = 0.61), whereas the cephalometry rubric demonstrated adequate reliability (α = 0.71). Exploratory factor analysis yielded evidence consistent with a one-dimensional structure for both instruments. Cephalometry performance was positively associated with theoretical examination scores and seminar performance, whereas no significant associations were observed for wire-bending performance. Students reported generally positive perceptions of rubric use, although lower post-assessment ratings were observed in the wire-bending activity. Conclusions: These findings provide preliminary evidence supporting the reliability and educational usefulness of rubric-based assessment in preclinical orthodontic education. These instruments may contribute to greater transparency and consistency in student assessment. Further studies involving multiple evaluators, multicenter samples, longitudinal designs, and psychometric approaches specifically developed for ordinal data are required.