Background and Context: To educate students about programming plagiarism and collusion, we introduced an approach that automatically reports how similar a submitted program is to others. However, as most students receive similar feedback, those who engage in plagiarism and collusion might feel inadequately warned. Objective: When students are likely to be engaging in plagiarism or collusion, we would like the system to apply enough pressure on them to make them reconsider their actions. Method: This study proposes a variation of the approach, which is less likely to report coincidental similarity. The variation was compared with its predecessor via quasi-experiments with 202 computing students. Findings: Students with the new approach are slightly more aware of programming plagiarism and collusion than those with the previous approach with a reduction in cases of such misconduct. Implications: There is another way to automatically educate students about programming plagiarism and collusion with appropriate pressure.