Traditional laboratory exercises are often intended to provide hands-on experience in science to students, reinforce the rules and relationships presented in lecture, and convey a sense that science is cool. Other potential goals, such as clarifying the scientific method and improving student understanding of measurement, error and assumptions, are poorly framed and largely neglected during instruction. Even primary goals are not always made clear, which frustrates students and promotes negative attitudes toward the laboratory, or worse, science in general. This project is developing three different instructional strategies for the laboratory component of an introductory physics course: a strongly guided, content-focused strategy; a question-driven, minimally guided strategy; and a question-driven, scientific ability-focused strategy. Each of these strategies is being developed with assessable student learning outcomes in mind, to ensure that we can evaluate the success of each strategy. Formative assessment of student efforts to reach these goals involves rubric-based evaluation of individual laboratories and laboratory-focused examination questions. Summative assessment incorporates the final exam, FCI or CSEM, and CLASS. This assessment is informing future iterations of laboratory instruction and provides significant insight about strategies for laboratory-based education in the science education community.

Broader impact: This project advances understanding on how different levels of guidance in content and scientific abilities impact student learning, epistemological beliefs, and laboratory skills. As laboratory instructors can have very different goals for their courses, the project is assessing several different outcomes to determine how instructors can design their laboratories to match these goals. Furthermore, the project is gathering additional data on student behavior in laboratories - through videotapes and interviews - that can be used to structure laboratories and modify how students interact with each other, the instructor, and the laboratory equipment. The quantitative and qualitative data allow for extensive collaborations with other researchers and can advance understanding on designing, implementing, and assessing laboratories.

Intellectual Merit: The project is deeply embedded in current research in STEM education. Recent research on student learning in the laboratory setting suggests that the instructional strategy is crucially important. The STEM community is just beginning to develop a cohesive framework for understanding these factors, and this study plays a pivotal role in its development. The laboratory component of an introductory science course is often the only exposure that nonmajors get on what physicists do and how they think. Without effective laboratory instruction, students may leave the course with unclear views about science and be less knowledgeable consumers of scientific research beyond the classroom. By helping instructors develop laboratories that not only impact future scientists, but help develop the scientific content knowledge and abilities of the general public, the project can lead to a more scientifically literate society, able to make reasoned decisions involving scientific data.

Project Report

The introductory physics laboratory, sometimes treated as an appendage to the primary physics course, actually plays a very important role in physics education. We explore this dimension of student learning and engagement by investigating students’ exhibition of scientific reasoning and data handling skills in laboratory activities designed with different degrees and means of scaffolding inquiry. Our goal is to compare each of these approaches to one another and to compare both approaches to other institutions where comparably structured inquiry-based laboratory activities are implemented. The comparative analysis of students from two sequences of laboratory activities, one that is more exploratory and another that is more explicitly scaffolded, indicates that there are differences between the two groups, although the differences are slight. We measure almost no statistically-meaningful differences between the two sequences across many of the abilities assessed. However, students from each sequence differ strongly in their ability to minimize uncertainty, which may reveal information about the roots of the other, slight differences. The freedom afforded to students in the exploratory sequence may allow them to design and conduct experiments with which they feel the most comfortable engaging. The scaffolded activities require students to occasionally address questions that challenge their expectations about what a solution might look like. Minimizing uncertainty, for instance, is easier to think about when an experiment results in a quantitative value that can be averaged over multiple trials. In an exploratory setting, students may choose experiments that fit this mold. On the other hand, if the scaffolding dictates that experiment must be qualitative, considering and minimizing uncertainty are not necessarily so familiar to students. In other words, more challenge may come with more scaffolding. Overall, less scaffolding is associated with slightly improved exhibition of scientific reasoning abilities. This may result from students engaging more critically throughout the activities, or it may result from students choosing the path about which they are more comfortable and confident. The latter cause doesn’t necessarily mean that the exploratory approach is weaker; confidence and self-efficacy are strongly associated with positive learning outcomes. Practically speaking, as the differences in all facets of assessment are quite small, there is no evidence to suggest that either sequence is better or worse than the other according to the metrics assessed here. Others report that both of these approaches promote stronger exhibition of scientific reasoning skills and time spent sense-making than traditional laboratories (Etkina et al., 2010; Lippmann, 2003); therefore, our findings suggest that both approaches are viable alternatives for laboratory instruction. Some skills, such as the ability to evaluate the consistency of different representations, the ability to design a reliable experiment and the ability to make a reasonable prediction based on a hypothesis, seem to saturate early in the semester and remain saturated throughout. Other skills either plateau or decline throughout the semester. Still others increase throughout the semester but do not saturate. These observations suggest that, regarding many of the scientific reasoning abilities of interest, five bi-weekly laboratory sessions are not sufficient for acquisition. Etkina and colleagues report that only after eight weekly sessions in which students conduct scaffolded design laboratories do they begin to achieve mastery of scientific abilities (Etkina et al., 2008). We cannot determine whether the amount of time, number of meetings, frequency of meetings, or the combination of these factors accounts for their observation. Nonetheless, our findings suggest that fewer meetings at a lower frequency, in spite of differences in student population, are not as effective as activities conducted elsewhere. More opportunities for students to engage in laboratory activities allow for (1) longer duration of scaffolding, (2) more varied scaffolding and (3) more distributed scaffolding; in spite of the fact that only a fraction of the scientific reasoning abilities were incorporated for assessment, students were clearly overwhelmed by the information. Ultimately, we find that students who engage in two different sequences of laboratory activities – exploratory and scaffolded – both demonstrate similarly enhanced scientific reasoning skills; the students who participate in the former sequence slightly outperform those in the latter, though perhaps the difference in performance only occur while the groups are engaging in different activities, as it disappears when the two groups engage in identical final laboratory activities. Regardless of group, some important skills are either not exhibited or retained by many of the students, suggesting that five bi-weekly meetings are not sufficient for acquisition and demonstration of these abilities.

Agency
National Science Foundation (NSF)
Institute
Division of Undergraduate Education (DUE)
Type
Standard Grant (Standard)
Application #
0942044
Program Officer
Duncan E. McBride
Project Start
Project End
Budget Start
2010-07-01
Budget End
2012-06-30
Support Year
Fiscal Year
2009
Total Cost
$200,000
Indirect Cost
Name
Harvard University
Department
Type
DUNS #
City
Cambridge
State
MA
Country
United States
Zip Code
02138