Long-running efforts to increase academic achievement for girls in middle- and high-school math seemed to be bearing fruit by 2019. However, pandemic educational disruption reversed this trend, with both NAEP and TIMSS test results in 2023 showing a far sharper decline in math test scores for girls than for boys. Additionally, female students remained underrepresented in upper-level STEM courses, STEM majors, and among those pursuing careers in STEM fields at both points in time. Researchers interested in continuing the pre-Covid rise for girls recently tested a two-pronged approach aimed to improve math performance: tutoring and female role modeling.
The research team focused on ninth grade students taking Algebra I in five high schools in an unnamed urban school district in New England during the 2021-22 school year. The district intended to offer tutoring fully in-person and solely during the school day via a special math intervention block. However, a lack of available local tutors ended up requiring virtual sessions at three of the schools, including after school hours. Tutoring groups ranged in size from one to three students per tutor. The analytic sample included 422 students (half of them girls) and 23 tutors (39 percent female). The researchers randomly assigned students to either same-gender or opposite-gender tutors. Seventy-six percent of students were Hispanic, 45 percent were white, and 33 percent were Black. Among the tutors, 43 percent were Black, 34 white, and about 2 percent Asian. The vast majority of students (89 percent) were low-income, 18 percent had Individualized Education Plans, and 37 percent were classified as having Limited English Proficiency. Among the data gathered for all students: math course grades, school attendance, and overall achievement data for the prior school year; tutoring attendance and two skills assessment scores at the beginning and end of the tutoring program; and an end-of-year student survey administered by the tutoring provider.
Girls paired with female tutors reported significantly higher STEM interest (0.73 standard deviations) compared to those assigned to male tutors and were 3.9 percentage points more likely to pass their Algebra I course with a C- or better. The effects appear stronger for girls working with tutors in-person as opposed to virtually, and during the school day as opposed to after school. However, there was no impact of tutor gender on end-of-program assessment scores or on students’ attendance patterns.[1] The researchers do note that ongoing Covid concerns contributed to inconsistent school and tutoring attendance, including an overall high rate of chronic absenteeism at all five buildings.
Interestingly, the researchers don’t land on any huge takeaways or suggest broad policy prescriptions based on these findings. In fact, in an interview with EdWeek, research team member Carly Robinson said that matching students and tutors based on any singular factor like gender or race could actually have “potential negative impact” on outcomes. Educator quality, experience, training, and previous success cannot be overlooked. Instead, she encourages schools to make sure that all students can see people like themselves within the ranks of strong and successful adult role models around them. “Ensuring that your tutor base or educator base is more diverse and represents the students who they’re working with,” she said, “would allow for that sort of exposure and role modeling,” especially in STEM areas. For girls, “being exposed to a female tutor who represents the pathway that they could take is impactful.”
SOURCE: Joshua Bleiberg et al., “The Impact of Tutor Gender Match on Girls’ STEM Interest, Engagement, and Performance,” retrieved from Annenberg Institute at Brown University (May 2025).
[1] While data are provided for the boys in the sample in appendices, there is no written analysis or discussion of them or comparisons made to the girls’ outcomes.
