Browser-based interactive coding: an experimental evaluation of the Art-Vire web environment for an introductory Verilog course

Abstract: Problem. In engineering education, there is a significant gap between the complexity of digital device design using FPGAs and the cognitive accessibility of introductory training. Traditional CAD systems (Quartus) require a lengthy synthesis and firmware flashing cycle, which creates a high cognitive load for students at the initial learning stage, reduces the frequency of code experimentation, and hinders the formation of stable practical skills. Aim. To experimentally verify whether the Art-Vire environment, with instant feedback and multimedia principles, provides statistically significant im-provements in objective Verilog learning outcomes compared to the traditional FPGA cycle. Methods. A quasi-experimental study was conducted with control and experimental groups (n = 12 each). Participants – 24 third-year stu-dents without Verilog experience but with basic knowledge of digital logic – were randomly assigned to groups. Group homogeneity was confirmed by pre-testing (p > 0.05, Mann–Whitney U-test). The experimental group (Art-Vire, instant compilation) and the control group (Quartus + Cyclone IV, 3–15 min synthesis and firmware flashing cycle) completed five identical tasks of increasing complexity. The following parameters were recorded: task completion time, proportion of completed projects, final test score, and number of experimental code modifications. Statistical analysis involved non-parametric Mann–Whitney U and Pearson's chi-squared χ2 tests, with α = 0.05. Results. The reduction in task completion time in the experimental group was 87.0 % (from 187.5 to 24.3 min; U = 0, p < 0.001, r = 0.81). The proportion of students successfully completing all five projects increased from 41.7 % to 91.7 % (χ2 = 6.75, p = 0.009, φ = 0.53). Final test scores in the experimental group were statistically significantly higher (26.4 ± 2.1 vs. 21.7 ± 3.8; U = 38.5, p = 0.008, r = 0.54). The number of creative code modifications increased from 16.7 % to 75.0 % (χ2 = 8.31, p = 0.004, φ = 0.59). Conclusions. The Art-Vire environment provides statistically significant reductions in task completion time, increased project comple-tion rates, improved test scores, and enhanced student creativity compared to the traditional FPGA cycle.