Virtual learning is nowhere near its pandemic peak, but it has entered the K-12 education mainstream and remains an option for many students. That is why it is important for educators and policymakers to understand that remote learning can be an environment primed to draw out teachers’ unconscious biases, according to a new study.
When asked to evaluate identical student math work in a simulated virtual Zoom classroom, teachers were more likely to recommend that Black students get tested for special education than white students and that boys get tested for gifted programs more than girls, the researchers found.
These are consequential decisions that can alter the trajectory of students’ academic careers and lives, said Yasemin Copur-Gencturk, an associate professor of education at the University of Southern California and the lead researcher on the study.
“We know that implicit bias usually happens or occurs when there is little information about a student,” she said. “Teachers are dealing with tons of things, and they are human. Everyone has some subconscious biases. The goal of the research is to increase the awareness, if teachers know it, they won’t do it.”
For the study, teachers were shown the photos, names, and math work of 12 students in what looked like a Zoom classroom. The teachers were asked to evaluate the students’ work, which remained the same for every teacher, but the student images—which were a mix of Black and white students—and names attached to the work were randomly assigned and different for each teacher.
Education Week spoke with Copur-Gencturk about when teachers—or any human, for that matter—are most at risk for subconscious bias and why virtual learning is an environment where it’s easier for implicit biases to influence decisions.
This telephone interview was edited for brevity and clarity.
Did teachers’ implicit biases surface in different ways depending upon the school?
We found [that] the racial profile of the schools in which teachers were working had an impact on their evaluations of both students’ ability and their recommendations for gifted and individualized programs.
So, Black students were assumed to have lower math ability and they were recommended for special education programs at a higher rate than white students in schools with higher concentrations of Black students.
We also found that gender differences were more pronounced for gifted education recommendations in schools with a higher concentration of white students—meaning that white girls are discriminated against more for being recommended for gifted education programs.
Those findings feel a little counterintuitive
It sounds a little bit counterintuitive, but at the same time, unfortunately, it shows that when teachers work with students from stereotyped groups, they seem to generalize their experiences to all students from that group.
So I think we really need to find a way to overcome these implicit biases. We need to come up with good professional development programs to help teachers increase their awareness, and to reduce the ambiguity. Because that’s something that we keep noticing, we observe bias when there is limited information about students. And this is when teachers may unconsciously draw from social information about the students and may unintentionally apply race and gender stereotypes in their assessments.
So, the solution I think is more about finding more-effective ways for teachers to get to know their students and be aware that they should not over-generalize their prior experiences with students from particular groups to all students from that group.
What do you consider the most important finding from this study?
I want people to pay attention to the fact that virtual learning environments are becoming a norm, a new form of teaching, and unfortunately the biases that we see in in-person teaching are also appearing in online classrooms. And we need to do something, because virtual learning environments are providing limited information for teachers about their students.
And we know that when there is limited information, people rely on their biases more.
What should a good virtual learning environment look like?
One where teachers have more information on their kids.
What we have right now is that students log in, sometimes there is no turn-on camera policies. Most of the time kids are joining online environments and teachers are asking questions and mimicking the in-person teaching scenarios.
I sometimes see students try to share their ideas, but it’s just a replication of in-person teaching with limited interaction—that’s how I see teaching is done in an online learning environment. Less accountability on whether kids are paying attention to the class, or whether they are learning. Less formative assessment.
An ideal way of teaching, I know people might not like this idea, but I believe that turning cameras on could be a solution, because we can have virtual backgrounds, so that [students] don’t have to worry about what is going on in their home setting.
More check-ins. Each chance for teachers to ask their students how their days are going, so they can feel some connection. Breakout rooms are not used often, but teachers can use breakout rooms so that kids can connect with peers and their classmates as well.
Teachers don’t have to lecture all the time. Kids, especially in the upper grade levels, can learn how to share their work.
So, it’s really about getting to know each student better. Then teachers are less likely to fall back on implicit biases?
Exactly. That is one component of it. The second component is that by doing that, they can also create a connection. The more connection they create with their students, the more information they have on their students, the less likely they will draw on their unconscious biases.
What are the limitations of the study?
The strength of the study is also the limitation of the study. We did an experimental study so we could rule out all of the alternative hypotheses. But at the same time, these students are not teachers’ [actual] students, and maybe teachers show different types of bias with their own students—we don’t know that part.
