Open Collections

Kraft, Aaron M.

University of British Columbia

In this thesis I investigate how students experienced a one-week lab in the first-year curriculum: the pendulum lab. In this lab, students are tasked with comparing the period of a pendulum at 10 and 20 degrees. A high-quality measurement should show that the periods are distinguishable, contradicting the model that many students learn in high-school or first year physics. The design intention is that this model breakdown forces students to engage in scientific inquiry as they grapple with unexpected results. I ran an intervention to assess whether asking students to generate a hypothesis impacted their experience in the lab. I saw no evidence that the intervention impacted their experience. However, these data ended up uncovering a whole host of interesting results. First, I found that many students do not report inconclusive results. Students employ a modified statistical test called a t-Score to compare the periods in this lab. When a students t-Score is “in-tension” (inconclusive), they tend to avoid reporting that their comparison is inconclusive. I attribute this to a discomfort with inconclusive results because of a misconception that they must report a definite answer in the lab. I also uncovered that students treat the t-score as a categorical measure instead of a continuous one. This is analogous to scientists erroneously treating anything with a p-value above .05 as a statistically insignificant result, no matter how close it is to the cutoff. In response to these results, I designed a second intervention that removed inconclusive results and stressed that the t-score in continuous. The data I collected indicate that this intervention was generally successful, although it had unintended consequences. Second, I found that most students do not actually experience the intended model breakdown because so few of them knew about the period equation before the lab. I discuss whether providing the equation, and forcing the model fail, would be beneficial for students. But I ultimately conclude that the risk of confirmation bias, which I uncover in a subset of students, outweighs the benefits of forcing the model fail.

Text

2024-09-17

Attribution-NonCommercial-NoDerivatives 4.0 International

http://hdl.handle.net/2429/89258

Physics

University of British Columbia

UBCV

http://creativecommons.org/licenses/by-nc-nd/4.0/