Peer-to-peer dialogue can enhance students’ understanding of mathematics by stimulating active processing and articulation of knowledge. However, this type of interaction also places demands on working memory, which may hinder learning if cognitive load becomes excessive. To optimize classroom dialogue, it is important to distinguish between different types of cognitive load: intrinsic load (IL), extraneous load (EL), and germane load (GL). Existing self-report instruments do not account for the distinct cognitive demands associated with students’ roles as listeners or explainers. This study aimed to develop and validate a questionnaire to measure IL, EL, and GL separately for both listening and explaining roles during peer-to-peer dialogue in secondary mathematics classrooms. The development process involved a literature review, analysis of existing instruments, adaptation for adolescent learners, and integration of mathematical dialogue characteristics. The resulting instrument consists of 18 items, 9 for each role. To validate the instrument, two studies were conducted using peer instruction in Dutch secondary school classes (n = 65 and n = 32; ages 15-17). Principal component analysis confirmed a three-factor structure aligned with the three types of cognitive load for both roles. The results suggest that the questionnaire is a promising tool for measuring differentiated cognitive load during classroom dialogue. It may inform instructional design aimed at balancing cognitive demand and supporting effective peer interaction in mathematics education. ... Read more