Physicist Tim McKay has taught enough introductory physics courses to know what many university students think about them: They are difficult. You will get a lower grade in them than in your other courses. And worst of all: If you don’t do well, then you probably weren’t meant to study science after all.
Studies have shown that those who face the worst consequences from this mentality are those who are already less likely to be found in many STEM fields: women, underrepresented minorities and students from low-income backgrounds.
In higher education, this is no secret. But creating the cultural shift needed to understand the core of these issues requires effort and collaboration across institutions, not to mention buy-in from the top down.
Luckily, McKay, who has been a professor at the University of Michigan since 1995, is up for the challenge. And he’s equipped with the not-so-secret tool that every physicist turns to for answers: data.
After cutting his teeth early in his career on the Sloan Digital Sky Survey—a collaboration of hundreds of scientists who studied the universe using telescope data—he understood the value of an experiment that brought together top interdisciplinary minds to find new insights in huge datasets.
For the past decade, he has used those tools to study inequities in education. Now, as head of the Sloan Equity and Inclusion in STEM Introductory Courses (SEISMIC) project, he has brought together 180 people from 10 public research universities to understand what makes STEM courses inequitable among students.
Not only that, the group is working to create concrete solutions that shift the way these courses are taught and to create a new way for professors and university leadership to think about education. In a year where education systems across the country are being disrupted and examined, members say that perhaps now is the time to create real change.
“We have to change the culture,” says McKay, who is now an associate dean for undergraduate education at the University of Michigan. “I’d really like for students to take intro to science courses and come out feeling like they had real success, like they were set up to learn the deep roots of the field, rather than feeling like they got through by the skin of their teeth and didn’t understand anything.”
Shifting away from STEM ‘grade penalties’
For many students at large research universities, intro-to-STEM courses involve large lecture halls filled with hundreds of students, all watching a single professor write notes and explain the basics of the field at the front of the room—a scenario that has played out the same way for decades.
For some students, this works fine. As an undergraduate at Temple University in the late 1980s, McKay himself found his way into physics through courses just like these. What inspired him was his enthusiastic, magnetic instructor, Jack Crow, who got him excited about the consistency and logic of physics.
When McKay first became a professor, he tried to emulate that enthusiasm, thinking it was enough to inspire his students. “But I found that teaching is a craft that you learn how to do, and you learn through evidence, not instinct,” he says. “Enthusiasm is not a terrible approach, but it’s far from the most effective.”
He began experimenting in his courses, giving students electronic clickers to answer questions during class. He found that it engaged students who would otherwise passively sit in the back and listen. When he became director of the honors program within Michigan’s College of Literature, Science and the Arts in 2008, he realized he had no idea what sort of pedagogy was working well across disciplines. To find out, he turned to data.
He knew that students tended to get lower grades in intro-to-STEM courses, but he wanted to examine how those grades fared against the average grade they received in all of their other classes. In other words, what were the grade penalties associated with those courses? And were those penalties the same for different groups of students?
First taking a look at penalties for male versus female students, he found that in intro to physics courses, the grade point penalty for male students was .3, while the penalty for female students was .6.
Why was this? Female students tended to outperform male students across disciplines at Michigan, except in the large introductory science and math courses. McKay wanted to know: Was this just happening at Michigan? He began recruiting other universities to ask the same question, and they found that the patterns were the same everywhere.
“These inequities were systemic,” he says. “It’s shameful to discover that you’ve been teaching a course that has this outcome. Once I knew it was everywhere, I thought: I have to do something about this.”
Understanding the problem of inequity
With funding from the Sloan Foundation, McKay created SEISMIC. Since 2019, the group has divided into working groups focused on four topics: measurements, structures, experiments and constructs.
Because of McKay’s findings at Michigan, one of their first targets was the gender gap in intro-to-STEM courses, says Sehoya Cotner, a biology education professor at the University of Minnesota and one of the leaders of the experiments working group.
In large into-to-STEM lecture courses, high-stakes, timed exams are often the evaluation tool professors use to gauge whether students understand the concepts. Cotner’s research—and research at other universities—have shown that female students often perform worse on these types of tests compared to male students.
Scientists have investigated several factors that could contribute to this gap, such as stereotype threat: performance-diminishing anxiety related to a fear of confirming a negative stereotype about a group of which you are a member. Female students are more likely to report experiencing test anxiety than male students, but the cause of this gap likely involves myriad factors and will need more research, Cotner says.
Still, the group is working to find alternative ways to evaluate students in introductory STEM courses other than timed tests. These evaluations could involve tasks more akin to authentic physics research, such as writing out an analysis or working on a group project.
They’re also looking at other interventions to deal with student anxiety, such as early discussions about how challenges in these courses are normal, temporary and surmountable.
Though SEISMIC began by looking at gender, participants are also looking at how these interventions could close gaps between other groups. At the University of Minnesota, Cotner found that “belongingness” discussions lowered the performance gap in an Introduction to Chemistry course between students from groups that are underrepresented at the institution (students who identified as Hispanic/Lantinx, Black, Native American or Pacific Islander) and students who identified as white or Asian.
The SEISMIC group has begun to expand their focus—looking at how gender, race, income and first-generation status all affect students. They want to take a nuanced approach, and early results show that students at the intersections of these different minority groups stand to lose the most from grade penalties.
“We are working beyond gender and looking at what works for certain subjects and in certain situations,” Cotner says. “We want to contextualize it and give our colleagues in STEM information that they can hang their hats on. It’s not fuzzy or theoretical—we want to create concrete, actionable information that will lower barriers and reduce inequities.”
Shifting the course from ‘prior opportunity’
For the past 25 years—first as a chemistry instructor and now as assistant vice provost for educational effectiveness at University of California, Davis—SEISMIC member Marco Molinaro has been gathering data about inequities among students from different backgrounds and neighborhoods.
“Especially in introductory courses,” he says. “They are the Achilles’ heels of STEM education. When you have students that are first-generation, low-income and underrepresented minorities, they can be lost at three times the rate of white students who have benefitted from greater prior opportunities to learn the material. Even if they stay in the class, they are often a full course grade behind other students.”
As one of the leaders of SEISMIC’s structures working group, he is helping to find the right data and approach to communicate and effect change at the structural level. The group is looking at datasets including regional opportunity indexes, census information and estimated income potential. They want to see how students’ backgrounds—and the opportunities or lack of opportunities those backgrounds afforded them—relate to how they perform at the university level. In other words, how their “prior opportunity” level affects future opportunities.
“We are looking at how to change the discussion,” he says. “What is really happening is that their prior opportunity seems to be continuing once they get to college. Now we are looking to find instances of introductory STEM courses where students outperform their prior opportunity level and see how we can replicate that success across institutions.”
Part of SEISMIC’s goal is to not only find the right data, but communicate it in a way that will get even the most stalwart professors and leaders to change their thinking.
For example, when McKay talks to other STEM professors about gender inequities in an intro-to-STEM course, many of them dismiss the idea that the data reveal a problem with the course itself. “Their first instinct is to explain it away.”
Instead, he often first presents the data as coming from students that took the same course but were taught in different classes by different instructors. Professors are always interested in correcting inequities between these classes, McKay says.
Molinaro’s group is now working to create a how-to manual for institutions to examine their own behaviors and structural inequities in STEM courses. “If we can start to answer the question ‘What do we do now?’, that’s where I think we’ll have success,” he says.
Seizing the moment to shift the culture
While SEISMIC was launched in 2019, members say they are seizing this moment—in which educators, students and parents are reckoning with how to work with students during a pandemic—to make their case.
“The pandemic has forced everyone to reconsider and perhaps change the basic structures of their courses,” McKay says. “We know that the pandemic has hit Black [and other ethnic minority] communities much harder than others, and we fear that all of the disruption anticipated for this year will make things worse. It’s an opportunity for us to move forward with this colossal shift that is needed. We can find new modes of evaluations that instructors like and that will help solve equity issues.”
For Cotner, who lives in the Minneapolis-St. Paul area, where the police killing of George Floyd and resulting unrest has brought the issue of systemic racism and social inequities to the forefront, a project like SEISMIC is one way she feels like she can effect real change.
“I can give money, I can go to rallies, but I can also use my training and funding to try to do this one tiny part, which is make science education more equitable,” she says. “We want more people of color to become principal investigators and leaders in science. It’s what we can do in our little part of the world.”