Investigating student ability to draw conclusions from measurement data

In this study, we surveyed students from a calculus-based and an algebra-based introductory physics course at a liberal arts college about their ability to draw conclusions from measurement data. Both courses are taught in a studio mode and use the Workshop Physics curriculum. The survey questions were adapted from Kok et al. (2019), who found that an increase in the number of decimal places hinders students' ability to compare data sets. We administered the survey online before and after instruction on measurement uncertainty. On the survey, students considered two experiments that differ only by one setup. Students were first asked to make predictions about the experimental outcomes as to whether or not the outcomes agree, and then were given data to analyze and draw conclusions. The survey had two versions where the measurement data had either two or four decimal places. We used the framework of point and set paradigms to characterize student reasoning. The set paradigm emphasizes a measurement distribution rather than a single measurement; it is considered more expert-like. The results show that students tended to switch from a correct to an incorrect answer after analyzing the data. The number of decimal places did not seem to correlate with the switch in student answers. We also found that after instruction on measurement uncertainty, student reasoning tended to shift from the point paradigm toward set paradigm as many students included the standard deviation or standard deviation of the mean in the analysis. However, many students did not seem to recognize how the uncertainty could inform the conclusion as to whether or not the data sets agree; most students appeared to rely only on the comparison of the means. We discuss implications for instruction as well as future research areas.