Abstract
Introduction Developing pupils’ digital competence is a core goal of primary education, yet many schools lack scalable approaches that combine authentic tasks with consistent scaffolding. This study tested whether student-facilitated, scaffolded project-based learning (PBL) improves pupils’ DigComp-aligned digital competence compared with standard instruction. Methods A quasi-experimental, non-equivalent control group pretest–posttest design was conducted in one mainstream public primary school in Kazakhstan with Grades 2–4 pupils ( N = 124; experimental n = 62; control n = 62). The intervention comprised six scaffolded PBL modules delivered over one academic term (16 weeks) and supported by 24 trained teacher-education undergraduates acting as classroom facilitators under teacher and researcher supervision. Outcomes included a DigComp 2.1-informed performance diagnostic across four domains (information literacy; communication/collaboration; content creation; digital safety), a 5-point pupil self-assessment of digital confidence, and structured observations (Kendall’s W = 0.84). Between-group differences were tested using independent-samples t-tests (Welch’s t where appropriate) and reported with Cohen’s d. Results Groups were comparable at pretest on the overall index (2.47 ± 0.52 vs. 2.49 ± 0.55; p = 0.857). At posttest, the experimental group outperformed the control group on overall digital competence (3.85 ± 0.44 vs. 3.08 ± 0.51; t = 7.11; p < 0.001; d = 1.10) and across domains ( d = 0.52–1.34). Digital confidence increased more strongly in the experimental group (Δ = 1.42 vs. 0.54; t = 6.95; p < 0.001; d = 1.25). Discussion Student-facilitated, scaffolded PBL was associated with meaningful gains in primary pupils’ digital competence and confidence over one term. The findings suggest a feasible university–school partnership model for supporting early-grade digital skills development, although evidence is currently limited to a single-school context.