In this thesis we investigate how students experienced the pendulum lab, a one-week lab in the first year lab curricula. This lab course is designed to teach scientific critical thinking through an iterative process of collecting data, making a quantitative comparison and reflecting on that comparison. In the pendulum lab, students are asked to compare the periods of a pen[1]dulum at 10 and 20 degrees. The quantitative comparison is performed using a modified statistical test that we refer to as the t-score. The lab is designed such that a reasonable measurement of the periods shows that they are different. But this result contradicts the equation for the period of a pendulum that students learn in first year physics courses because of an approximation. Previous work has shown that students who see the lab as a model confirmation activity struggle with labs like this.
Our study set out to see if asking students to hypothesize would hinder their ability to interpret their data correctly because of confirmation bias. But instead, we found that students’ main struggles were because of issues with inconclusive results. Confirmation bias did play a role in some of our data, but the impact was limited. We also found that only half of the students were surprised by their results in this lab. Most students did not know about the equation for the period of a pendulum if they took this lab in the Fall semester. However, students who took the lab in the Spring semester did know about the pendulum equation and showed indications of confirmation bias. We also found that only half of the students in the lab performed a high enough quality measurement to uncover the difference, which was far below our expectations.
We proposed and implemented an intervention to deal with inconclusive results by redesigning how students interpret their t-scores. We also added more time to the lab so more students could get high quality results. These changes were not extremely successful but provided useful information for future improvements to the lab.
Add to Calendar
2024-06-25T10:00:002024-06-25T12:00:00The Pendulum Lab: Understanding Common Experiences and Pitfalls in a Lab With an Intentional Model FailureEvent Information:
Abstract:
In this thesis we investigate how students experienced the pendulum lab, a one-week lab in the first year lab curricula. This lab course is designed to teach scientific critical thinking through an iterative process of collecting data, making a quantitative comparison and reflecting on that comparison. In the pendulum lab, students are asked to compare the periods of a pen[1]dulum at 10 and 20 degrees. The quantitative comparison is performed using a modified statistical test that we refer to as the t-score. The lab is designed such that a reasonable measurement of the periods shows that they are different. But this result contradicts the equation for the period of a pendulum that students learn in first year physics courses because of an approximation. Previous work has shown that students who see the lab as a model confirmation activity struggle with labs like this.
Our study set out to see if asking students to hypothesize would hinder their ability to interpret their data correctly because of confirmation bias. But instead, we found that students’ main struggles were because of issues with inconclusive results. Confirmation bias did play a role in some of our data, but the impact was limited. We also found that only half of the students were surprised by their results in this lab. Most students did not know about the equation for the period of a pendulum if they took this lab in the Fall semester. However, students who took the lab in the Spring semester did know about the pendulum equation and showed indications of confirmation bias. We also found that only half of the students in the lab performed a high enough quality measurement to uncover the difference, which was far below our expectations.
We proposed and implemented an intervention to deal with inconclusive results by redesigning how students interpret their t-scores. We also added more time to the lab so more students could get high quality results. These changes were not extremely successful but provided useful information for future improvements to the lab.Event Location:
Zoom : https://ubc.zoom.us/j/2140943545?pwd=RGdIb0swbmRxM0QrWEtWejY2VGpVUT09, Meeting ID: 214 094 3545 , Passcode: 876743