Students’ mathematical digital competence in assessment: Validity considerations in the instrumental use of spreadsheet tools

Abstract in English

This thesis explores the opportunities and challenges of incorporating digital tools into summative assessments in mathematics. While technological developments have led to an increasing integration of digital tools into mathematics teaching and learning, there is a lack of understanding of how specific tools can be responsibly incorporated into assessments without compromising their validity. Accordingly, the aim of this thesis is to investigate and discuss what is assessed and what key considerations emerge when integrating digital tools into summative digital assessment of mathematics. Focusing on spreadsheet tools, the thesis employs an instrumental theoretical approach and the concept of instrumental genesis to analyze students’ techniques and mental schemes. Michael Kane’s argument-based validation framework is used to evaluate the plausibility of validity claims. The four studies included draw on both numerical and non-numerical data. Study I examines test developers’ perspectives on digital mathematics assessments, suggesting that digital test formats can enhance validity by enabling new constructs, while simultaneously introducing validity threats, particularly in the presentation of mathematical reasoning. Study II examines students’ interactions with spreadsheet tools, showing that prior tool knowledge cannot be assumed and that opportunities for practice are vital for assessment validity. Study III uses PISA 2022 data to analyze the technical and conceptual demands of spreadsheet tool use, demonstrating that technical skills alone are insufficient; productive tool use also requires conceptual understanding of both the tool use and related mathematics. Study IV, also using PISA 2022 data and applying explanatory item-response modeling, demonstrates that spreadsheet tool items are generally more challenging than items without such functionality and that students’ familiarity with spreadsheets may reduce the likelihood of a correct response. Overall, the thesis concludes that assessment validity depends on the alignment between learning and assessment, including opportunities to develop both technical and conceptual aspects of digital tool use. Achieving a relevant balance between tool authenticity and accessibility emerges as a central design challenge for valid digital mathematics assessment.