From Mimicry to Mastery: Teaching Coding to Mixed-Background Learners in Data-Intensive Courses

Abstract Book of the 5th International Education Conference

Year: 2026

[PDF]

From Mimicry to Mastery: Teaching Coding to Mixed-Background Learners in Data-Intensive Courses

Dr. Vladimir Shapiro

ABSTRACT:

Purpose. Many students entering higher education courses in Statistics, Analytics, Machine Learning, Artificial Intelligence, and other data-intensive disciplines (including Bioinformatics, Computational Finance, and Digital Humanities) arrive with diverse coding backgrounds, ranging from none to fluency in languages other than the course language. A conventional approach would either make programming a prerequisite or teach a language from scratch, both of which consume time and risk disengaging non-technical learners before they engage with the discipline. This paper presents an integrated pedagogical approach for heterogeneous cohorts, refined across 16 sections over seven years. Methodology. The approach combines three components: (1) in-class tutorials showing each week’s methods in code, which students mimic and extend to solve assignments; (2) a weekly instructor-curated “Hall of Fame” of student snippets — a bank of peer-worked examples with commentary on novelty and quality; (3) bonus incentives for creative coding, formatting, and visualization. The design draws on Use–Modify–Create (UMC) and Predict–Run–Investigate–Modify–Make (PRIMM) — established learning progressions for novice programmers, worked-example theory, and self-efficacy research. Findings. First, solution quality is subjectively indistinguishable between students with and without prior coding experience from the very first assignment, with no warm-up. Second, students report higher satisfaction than with language-first instruction. Third, an earned-membership effect: once featured, students work to defend that recognition in subsequent weeks — a pattern not cleanly accommodated by established constructs of motivation or self-efficacy. Conclusion. The approach offers a transferable, theoretically grounded model for teaching coding-as-a-tool in data-intensive disciplines and warrants controlled, multi-instructor empirical evaluation.

Keywords: Coding Pedagogy; Earned-Membership Effect; Mixed-Background Learners Peer-Worked Examples; Use–Modify–Create