New research and new opportunities within geoscience and across STEM

Like many scientists, educators, and researchers, I belong to several professional organizations and receive several publications and emails from these organizations. Often, they pile up on my desk until I have a spare hour, and then I skim through them and set aside articles I want to read in more detail. One of the journals that piles up quickly is Science—as a weekly, it is hard to stay on top of, but I often flip through it while I am eating lunch.

The issues of Science that hang out the longest on my desk are those that include the Education Forum, an approximately monthly section that was first published in 2006 and highlights “the science of education” (Alberts, 2009). Although the consideration of education research as a science may be obvious to readers of an education research journal such as this one, it was a big step for a premier scientific research journal. Articles in the Education Forum often address topics that are of particular interest to research-intensive readers, such as the merits of training research mentors (Pfund, Maidl Pribbenow, Branchaw, Miller Lauffer, & Handelsman, 2006) and evidence for the effectiveness of active-learning pedagogies (Miller, Pfund, Pribbenow, & Handelsman, 2008).

The March 30, 2018, Education Forum article has been open on my desk for over a month now, with the compelling title “Anatomy of STEM Teaching in North American Universities” (Stains et al., 2018). The authors reported on the results of their observations of teaching in more than 2000 classes across the STEM disciplines at 25 institutions. They used the Classroom Observation Protocol for Undergraduate STEM (COPUS; Smith, Jones, Gilbert, & Wieman, 2013), in which observers record every two minutes what the instructor is doing and what the students are doing, choosing from defined behaviors for each. COPUS is a useful tool for describing what is going on in the classroom for several reasons: It does not rely on self-reporting by instructors or evaluations from students, it is straightforward to train observers, and it has a high interrater reliability, which allows for the kinds of large-scale comparisons made in this study.

From the results of their 2008 observations in 709 courses (each course was observed multiple times), the authors defined instructional profiles based on combinations of four instructor behaviors and four student behaviors. Didactic profiles consist of 80% or more class time devoted to instructors lecturing and students listening. Interactive lecture profiles still focus on instructors lecturing, but include more time spent in group activities. Student-centered profiles spend a majority of class time in group work strategies that have emerged as effective pedagogical practices from discipline-based educational research (National Research Council, 2012, 2015). The authors then used these profiles to explore their results further based on class size, the physical layout of the classroom, the course level, and the STEM discipline of the course.

Figure 1 shows the results of their study broken down by discipline. Observations in geology courses comprise a relatively small proportion of the total observations (6%, or about 120 observations), but there are some distinct differences between geology and the other disciplines. About 30% of observations in geology were student-centered, which is higher by 10% to 20% than all the other STEM disciplines except math. Slightly more than 50% of observations in geology were didactic, which is similar to biology and computer science and significantly less than chemistry, engineering, and physics.

Figure 1. Instructional style as determined by analysis of COPUS data, broken down by STEM discipline. From “Anatomy of STEM Teaching in North American Universities,” by M. Stains, J. Harshman, M. K. Barker, S. V. Chasteen, R. Cole, S. E. DeChenne-Peters, ... A. M. Young, 2018, Science, 359(6383), 1468–1470, doi:10.1126/science.aap8892. Reprinted with permission from AAAS.

What does this mean for our discipline? It would be easy to look at that figure and say we are doing pretty well as a discipline at promoting student-centered, evidence-based pedagogies in our courses. That quick observation takes away some of the power of this study, however, which has both strengths and limitations that provide opportunities for our community in geoscience education and education research.

The primary strength of this research, as in most large-scale studies, is the scope of the work across disciplines, universities, and course levels. As the authors note, “This report provides specific baseline data for comparison for determining the impact of educational interventions, for professional development facilitators to inform the design of their programs, and for faculty when they receive COPUS data” (p. 1470). Baseline data provide a snapshot of how the world looks today, and they can be a powerful tool to imagine the world of tomorrow: They can be used to set specific goals for teaching reform across a department or college (such as, “reduce didactic instruction to less than a third of all geoscience classes” or, “increase the use of interactive lecture instruction by 50%”) as well as for individuals who are working to improve their teaching.

However, those baseline data have some significant limitations. All of the observations took place at doctoral-granting universities, for example, but how representative is that of our entire community? How does it compare to the teaching that takes place at two-year colleges, in which about a third of undergraduate students in the United States are enrolled (National Center for Education Statistics, 2018) and where the geoscience community has strengthened in recent years through the efforts of programs like SAGE 2YC (2017)? How does it compare to teaching at institutions from which the majority of future K–12 teachers receive their degrees? The door has been opened wide for a comparative study across institution types.

Although the strength of the COPUS lies in its relative objectivity, as all it records is the amount of time spent engaged in certain behaviors, this is also a limitation. The authors noted that COPUS does not capture information about the quality of the behaviors or the student learning that takes place as a result; nor is it adapted to fully online or hybrid courses, which are growing across all disciplines. Research that connects the instructional profiles to student learning, persistence to degree, and other measures of success within college or within the discipline could build on this work to strong effect. The development of a protocol that can be used where there is no classroom—whether that is online or in the field—would also allow for a better understanding of the complete anatomy of STEM teaching.

COPUS is not the only observation protocol, of course, and this study is not the only large-scale observation of STEM instruction or geoscience instruction. Within our community, the Classroom Observation Project emerged to better understand geoscience teaching practices nationwide and to validate self-reported results about teaching (SERC, 2017). The project recently published the results of its work: Teasdale and colleagues (2017) used the Reformed Teaching Observation Protocol (RTOP) in 204 geoscience courses across a variety of institution types and similarly classified observations into three groups: teacher-centered, transitional, and student-centered. They supplemented these observations (only one per instructor) with surveys completed by the instructors to give a more longitudinal view of the course. In their results, 30% of observations were in the teacher-centered category (similar in style to the didactic profile), 45% were transitional (similar to the interactive lecture profile), and 25% were student-centered (similar to the student-centered profile defined by Stains et al., 2018). They observed that multiple different types of teaching strategies can lead to a student-centered classroom—a result that was repeated by the authors of the COPUS study—and the nature of the paired RTOP and survey data allowed for documentation of those different strategies.

Research in areas like the nature of STEM teaching benefit from both within-discipline and cross-discipline studies like the ones described here. In their editorial in this journal describing the goals of establishing an alliance of discipline-based education researchers across the STEM disciplines, Shipley, McConnell, McNeal, Petcovic, and John (2017) describe the research opportunities and potential for transformative change that can evolve from such transdisciplinary efforts. Simply being aware of education research in other disciplines is one of the first steps toward becoming part of this broader community. I know I am not the only one whose desk and inbox pile up with unread journals, and I know that it can be challenging to explore new research in other fields, as well. In future issues of JGE, look for more summaries of research articles from other disciplines to help you take that step.