Correct responses to Newton's Third Law questions rose while forces were taught, peaked at the end of that instruction and fell back toward chance by the end of the quarter¶

Claim · All claims
Evidence · 1 study · associational r2 · q2
Currency · newest study 2010 · checked against its sources 2026-10-20 (automated, 1 of 1 entry) · not reviewed by a person

Subclaims¶

q2 i? In Mechanics, response rises from near chance before instruction to a maximum in week 6 (the last week of force instruction and the exam), then rapidly drops to near chance by quarter's end. → Clark 2010

Evidence¶

Clark 2010¶

Clark, J. W., Sayre, E. C., and Franklin, S. V. (2010). Fluctuations in Student Understanding of Newton's 3rd Law. arXiv preprint. https://arxiv.org/abs/1009.0260 doi:10.48550/arxiv.1009.0260

q2 · i? · associational · r2

Between-student response curve for Mechanics (Fig. 3), averaged across fall and winter sections. Before instruction responses hover around the chance line of 20%; the curve peaks in week 6 and then "rapidly drops" toward chance. Different students answered in different weeks, and the week-6 peak coincides with both the end of force instruction and the exam, so the curve tracks a rise and fall alongside instruction without isolating instruction as its cause.

"Instruction on forces begins in week 4, and student performance begins to rise, culminating with a maximum performance in week 6. Not coincidentally, week 6 is also the last week of instruction on forces, and includes the examination. After instruction, the response rapidly drops, with two of the questions ending just above the chance line at the end of the quarter."

Discussion¶

Related Claims¶

• Direct instruction's advantages do not persist over the longer term — the general claim this finding bears on, with the studies behind it
• Instruction on scalar electric potential produces a pronounced dip in correct vector-based Newton's Third Law responses in E&M — related
• Students enter E&M with a higher initial Newton's Third Law response (66%) than they exited Mechanics, most likely due to a winnowing effect — related
• The interference dip disappears once instruction returns to vector-based topics such as magnetic fields — related
• Current instruction affects previously learned knowledge (interference), calling into question the accuracy of any single assessment — a broader claim this one bears on
• In Experiment 1, unlearning of the initial category and re-learning of the target category proceeded in parallel rather than unlearning preceding re-learning. — related