Richard R. Hake analyzed gain/change scores in physics education, comparing "Interactive Engagement" (IE) courses with "Traditional" (T) courses. He defined normalized average gain ⟨g⟩ as the ratio of actual average gain ⟨G⟩ to the maximum possible gain, using pre and post test scores from the Force Concept Inventory. He categorized courses based on ⟨g⟩ values: High-g (>0.7), Medium-g (0.7–0.3), and Low-g (<0.3). His analysis showed that all 14 T courses had low ⟨g⟩ (⟨⟨g⟩⟩ = 0.23 ± 0.04), while 85% of 48 IE courses had medium ⟨g⟩ (⟨⟨g⟩⟩ = 0.48 ± 0.14). No course had high ⟨g⟩. He concluded that ⟨g⟩ is a reliable measure of course effectiveness in promoting conceptual understanding, as it showed a low correlation with pretest scores, unlike posttest scores or average gain. He calculated an effect size (d) of 2.78 for IE vs T courses, comparable to other studies. He noted that studies on cooperative learning often lack pretesting and may underestimate its effectiveness. Hake's work highlights the importance of normalized gain as a method for analyzing pre/post test results, despite its limited use in education research. His findings suggest that IE methods significantly enhance conceptual understanding compared to traditional methods.Richard R. Hake analyzed gain/change scores in physics education, comparing "Interactive Engagement" (IE) courses with "Traditional" (T) courses. He defined normalized average gain ⟨g⟩ as the ratio of actual average gain ⟨G⟩ to the maximum possible gain, using pre and post test scores from the Force Concept Inventory. He categorized courses based on ⟨g⟩ values: High-g (>0.7), Medium-g (0.7–0.3), and Low-g (<0.3). His analysis showed that all 14 T courses had low ⟨g⟩ (⟨⟨g⟩⟩ = 0.23 ± 0.04), while 85% of 48 IE courses had medium ⟨g⟩ (⟨⟨g⟩⟩ = 0.48 ± 0.14). No course had high ⟨g⟩. He concluded that ⟨g⟩ is a reliable measure of course effectiveness in promoting conceptual understanding, as it showed a low correlation with pretest scores, unlike posttest scores or average gain. He calculated an effect size (d) of 2.78 for IE vs T courses, comparable to other studies. He noted that studies on cooperative learning often lack pretesting and may underestimate its effectiveness. Hake's work highlights the importance of normalized gain as a method for analyzing pre/post test results, despite its limited use in education research. His findings suggest that IE methods significantly enhance conceptual understanding compared to traditional methods.