A disappointing test result can do more than lower a student’s average. It can change how strongly that student feels they belong in the classroom, and that altered sense of belonging may then become associated with what happens in the next high-stakes assessment.
That is the pattern reported in a study of introductory biology students at a two-year college in the United States. Rather than treating belonging as a fixed characteristic that students bring into a course, the researchers tracked it across the semester and examined how it interacted with academic performance. The results point to a feedback process that became visible only after students had received meaningful assessment feedback.
The midterm appears to be an important turning point
The study followed students enrolled in 16 sections of Introductory Biology 1 at a two-year college in the Mountain West region during spring 2022. Seven instructors taught the course, with seven sections delivered online and nine in person. Of 426 enrolled students, 310 consented to participate. The main early-semester analyses included 308 students, while 240 completed the late-semester survey and contributed to analyses involving measures collected near the end of the course.
Students completed surveys during the first two weeks of the semester and again during the final two weeks before the cumulative final examination. The measures captured sense of belonging, belonging uncertainty and academic mindset. The researchers also used a 15-question biology pre-assessment to estimate incoming academic preparation and analysed students’ midterm and final examination performance.
The sequencing matters. Early-semester belonging did not significantly predict average midterm performance. Academic preparation did: each additional point on the 15-point pre-assessment was associated with a 0.08-standard-deviation increase in average midterm performance after the other variables in the model were considered.
After the midterm, however, the relationship changed direction. A one-standard-deviation increase in average midterm performance predicted a 0.31-point increase in late-semester sense of belonging on the study’s six-point scale. It also predicted a 0.32-point reduction in belonging uncertainty. Both associations were statistically significant at p<0.001, and the researchers characterised the effects as medium in size.
Late-semester belonging then became associated with the cumulative final exam. After accounting for academic preparation, a one-point increase in late-semester sense of belonging predicted a 0.23-standard-deviation increase in final-exam performance. Greater belonging uncertainty moved in the opposite direction: a one-point increase predicted a 0.12-standard-deviation decrease in final-exam performance.
This does not establish that a low midterm mark directly causes a student to feel excluded, or that increasing belonging alone will produce higher marks. The research is observational and the models cannot eliminate every alternative explanation. What it does show is a time-specific pattern consistent with a recursive process: performance feedback was associated with subsequent perceptions of belonging, and those later perceptions were in turn associated with later performance.
Mindset added another, smaller layer
The researchers also examined academic mindset, distinguishing between entity beliefs, which treat intellectual ability as relatively fixed, and incremental beliefs, which treat ability as more developable.
Fixed beliefs showed the more persistent negative relationship with performance. A one-point increase in early-semester entity mindset predicted a 0.25-standard-deviation reduction in average midterm performance. Incremental mindset moved positively at midterm: a one-point increase predicted a 0.20-standard-deviation improvement. The mindset variables added only a small amount of explanatory power, however, with the models explaining about 10% of the variance in midterm performance.
By the final examination, entity mindset remained significant. A one-point increase was associated with a 0.23-standard-deviation reduction in final performance, while incremental mindset was no longer a significant predictor. In an exploratory model that considered belonging and mindset together, early-semester sense of belonging and entity mindset each retained a significant association with final performance after academic preparation was taken into account.
The modest effect sizes are important. Incoming academic preparation remained the most consistent predictor across the analyses, and much of the variation in performance was left unexplained. Belonging and mindset therefore should not be treated as substitutes for teaching quality, curriculum design, academic support or students’ prior educational opportunities.
Why access-oriented institutions matter
The setting makes the study particularly useful. Two-year colleges often enrol students whose routes through higher education are less conventional than those represented in research at selective four-year universities. In this sample, 65% of participants were classified as moderately nontraditional, 73% were women and more than half of students in the course were taking introductory biology during their first semester of college.
That context has parallels beyond the United States. Universities, universities of technology, colleges and other broad-access institutions increasingly serve mature students, first-generation students, working adults and people balancing study with family responsibilities. In these environments, an assessment can carry psychological meaning as well as academic information. A poor result may be interpreted not simply as evidence that a topic needs more work, but as evidence about whether a student is suited to the field at all.
The study therefore raises a practical question for lecturers: what happens immediately after students receive their first consequential marks? The findings do not justify artificially reassuring students or lowering academic standards. They do suggest that feedback, opportunities to recover, transparent explanations of effective study strategies and classroom practices that make participation feel legitimate may be especially important at the point when students begin interpreting their performance as information about whether they fit.
The results need careful boundaries
The research was conducted in one introductory biology course at one U.S. two-year institution, so the coefficients should not be assumed to generalise to every discipline or education system. Instructors wrote their own examinations, although scores were standardised within instructor groups to improve comparability.
Attrition is another limitation. Of the 308 students included in early-semester analyses, 240 completed the late-semester survey. Students who did not complete that later survey had significantly lower final-exam scores, even though their initial academic preparation and early belonging measures did not significantly differ from those who remained. The late-semester results may consequently underrepresent lower-performing students.
The sample also limited detailed subgroup analysis. Small numbers in some gender categories could not support reliable regression estimates, and the researchers note that characteristics such as socioeconomic status, race, first-generation status and intersecting identities could shape belonging in ways not captured by the models.
Even with those qualifications, the temporal pattern is informative. Students did not simply enter the course with a level of belonging that predicted everything that followed. Their sense of fit appeared responsive to academic experience. The first substantial performance signal was associated with how they felt later, and those later perceptions carried information about final performance beyond prior preparation.
For institutions concerned about retention in gateway STEM courses, that shifts attention from belonging as an abstract student trait to belonging as something that can evolve during ordinary academic events. The moments surrounding assessment and feedback may be where academic and psychological trajectories begin to reinforce one another.
Source Information
Study Title: Social belonging and academic mindset affect student performance in an introductory biology course at a two-year college
Authors: Hector L. Torres, Vanessa Johnson-Ojeda, Melissa E. Hardy and Regina F. Frey
Journal: Frontiers in Education
Year: 2026
DOI: 10.3389/feduc.2026.1825440
