Beyond surface learning: Grade 12 performance on traditional versus reformulated Lebanese biology assessments

Abstract

The predictable nature of the Baccalaureate Biology Exam in Lebanon may lead to "teaching to the test" and an emphasis on developing format-specific proficiency instead of transferable conceptual understanding. Lebanese research has not provided a direct assessment of whether high performance on the traditional Baccalaureate biology exam reflects true understanding or proficiency in responding to the exam's format. This study provides direct performance evidence by examining whether students who perform well on the traditional exam can apply their knowledge to alternative formats with differing cognitive demands. Using a quasi-experimental within-subjects design, 40 Grade 12 students from three Lebanese schools completed Test A (the traditional exam) and Test B (an alternate version with multiple choice, application-based, and analytical tasks). Students performed significantly better on Test A (M = 85.12%, SD = 8.46%) than on Test B (M = 64.75%, SD = 12.12%), with a 20.38-point difference, t(39) = 9.692, p < 0.000001, and a very large effect size (d = 1.532). School-level analysis showed significant variation in performance gaps, F(2,37) = 11.430, p = 0.0001. The correlation between Test A and Test B was low (r = 0.203, p = 0.210), suggesting that the two versions assessed different skills despite the same content. Overall, the findings suggest that performance on the traditional Baccalaureate biology exam may reflect familiarity with the exam format rather than strong knowledge transfer, and that the performance gap varies significantly across schools.