Improving the Cross-Comparison of Educational Achievement and Evaluation

The achievements of engineering students are evaluated using particular grading scales. Arithmetic mean is usually employed for grade aggregation. We show that this measure renders correct results only for equivalent scales as defined in the paper. If grades from non-equivalent scales are aggregated, rank “errors” as well as “absurd” averaging may happen, as we originally observed in practice. Decision-making based on arithmetic mean aggregation of grades may be true, false, or fuzzy, according to our analysis. We also argue for the choice of grading scales in conducting cross-comparison of students’ achievements. Our analysis also has relevance for regular grading and scaling methods, which we tested on engineering students with excellent student feedback. The grading technique proposed in this paper is particularly suited to engineering courses and it appears fit for promoting higher teaching and evaluation standards, paralleled by increased interest and higher competition levels for all student categories.