From Guided Hardware–Software Experiments to an Autonomous Robot Competition: A Six-Year Repeated Cross-Sectional Mixed-Methods Analysis of Student Evaluations

Introductory engineering courses must balance guidance with opportunities for independent design. This retrospective repeated-cross-sectional mixed-methods study examined perceptions of a two-semester course progressing from guided Arduino-based activities to an autonomous robot competition. It integrated 87 group-level evaluation records from 2020–2025 with 415 archived comment units and their thematic assignments. The common-period semester comparison (2021–2023) did not reach statistical significance (p = 0.126; rank-biserial correlation = 0.24). A secondary full-period comparison favored Semester 1 (p = 0.010), but year coverage and overall-item wording differed. Scores varied across years, H(5) = 15.18, p = 0.010, although no Holm-adjusted pairwise contrast was significant. Feedback addressed practical activity, instructor support, workload, and organization. Only lecture–laboratory alignment differed between semesters after Holm adjustment, including in the common-period analysis. Original-score comparisons by theme presence detected no Holm-significant differences; their confidence intervals allowed effects in either direction. A conflicting source score was examined by substitution and exclusion. The findings describe perceptions and recorded feedback, subject to changing measures, aggregated records, and unestablished intercoder reliability. They identify course-design priorities but do not demonstrate learning gains, development of autonomy, or an independent competition effect.