Developing a Conceptual Cognitive GPS Matrix for Representational Learning and Assessment in STEM Subjects

Jean Claude Uwayezu
Rwanda Polytechnic-Gishari College, Rwanda; Kigali Independent University (ULK), Rwanda
ORCID: https://orcid.org/0000-0002-5706-4545
Email: jcuwayezu2025@gmail.com

Abstract: Representational competence (RC) is a cornerstone in developing 21st century skills, especially for effective scientific communication. However, research on representational competencies in science, technology, engineering, and mathematics (STEM) lacks a unified conceptual framework for a systematic and hierarchical development of these competencies. This hinders the efficiency of assessment for learning. However, the revised Bloom’s taxonomy (RBT) can offer a good hierarchical structure to model representational competency progression. Thus, employing the antecedents, decisions, outcomes (ADO) model, the present study systematically mapped perceptual, conceptual, and meta-representational competencies to six (6) RBT cognitive levels to establish a conceptual framework for STEM subjects’ learning and assessment designs. It involved the extraction of competency descriptors from synthesized research publications, the identification of cognitive process verbs, and analysis of exemplar STEM studies. The results include a 6×3 cognitive GPS matrix with 18 discrete and assessable competency-level intersections, a competency development trajectory map, and a competency integration framework at both lower-order and higher-order thinking levels. The framework’s elements are also exemplified in different empirical STEM studies. Further, the framework has uncovered opportunities for using emerging technologies like intelligent tutoring systems (ITS), machine learning (ML), and Generative AI, for enhancing representational learning in STEM subjects. The developed conceptual framework significantly adds value to existing RC frameworks in STEM, as it presents a clear and hierarchical progression of representational competency from novice to expert. It is a valuable conceptual tool for STEM educators and researchers, particularly in the areas of representational learning and instructional planning.