Documenting medical students' use of self-explanations: tool development and initial validity evidence.

Introduction: Self-explanation (SE), an individual learning strategy for the development of clinical reasoning skills, has been implemented in undergraduate medical curricula. A tool for documenting students' appropriate use of SE is needed to ensure benefit on learning. The objective of this project was to develop such a tool and report on evidence of its validity. Methods: We used DeVellis's eight steps to develop the tool. Assessors applied the tool to 85 audio-recorded SEs. We calculated students' mean number of inferences (biomedical, clinical, monitoring) and case elements used when self-explaining to document validity evidence of content. We used interrater agreement, with intraclass correlation coefficients, to document validity evidence of response processes. Results: Three assessors documented the students' use of SE with the tool. They found means of 13.33 to 17.76 biomedical inferences, 16.27 to 27.04 clinical inferences, 2.03 to 3.10 monitoring statements, and listed 21.77 to 26.87 case elements. Interrater reliability varied from 0.53 to 0.96. Discussion: We developed the tool using the principles underlying SE. The way students used SE aligned with our expectations. Assessors used the tool in a consistent way. The tool could document students' use of SE in experimental or educational contexts.