Developing a Multidimensional Rating Scale for Content-Visual Proving in Primary School

Assessing primary-school children’s content-visual proofs requires instruments that capture not only the level of justification reached but also which mathematical relationships were used, which construct underlies the argument, and how far children’s own representational tools support generality. Building on Balacheff’s (1988) proof stages, we report the development of a multidimensional rating scale that combines an established justification-level dimension with two further dimensions: a task-specific construct dimension grounded in the Abstraction in Context framework, and a representational-tool dimension. The scale is being developed for a planned longitudinal study on the development of content-visual proving from grade 4 to grade 6 in Switzerland and Germany. We illustrate the scale with examples of sums of even and odd numbers as well as consecutive numbers and report initial results from piloting the scale on 71 authentic children’s solutions, including interrater reliability.