STEM identity is the degree to which a person sees themselves as a “science person,” “math person,” or someone who belongs in a science, technology, engineering, or mathematics field. It goes beyond skill or grades. Two students can earn the same A in chemistry, but only the one who thinks “I’m a scientist” has a strong STEM identity. Researchers treat it as a core sense of self, shaped by both internal confidence and external validation, and it turns out to be one of the strongest predictors of whether someone actually stays in a STEM career path.
The Three Pillars of STEM Identity
The most widely cited framework for understanding STEM identity comes from researchers Heather Carlone and Angela Johnson, who broke it into three interlocking components: competence, performance, and recognition.
- Competence is feeling confident that you have, or can acquire, the knowledge needed to understand STEM concepts. It’s not about being the smartest person in the room. It’s about believing you can figure things out.
- Performance is confidence in your ability to demonstrate those skills publicly, whether that means presenting lab results, solving problems on a whiteboard, or explaining a concept to someone else.
- Recognition is the piece that depends on other people. It means that peers, teachers, mentors, and the broader scientific community see you as someone who belongs. Without recognition, competence and performance often aren’t enough to sustain a person’s sense that they’re “a STEM person.”
All three pillars reinforce each other. A student who feels competent but never receives recognition from instructors may start doubting whether they truly belong. A student who gets recognition but struggles with performance may feel like an imposter. The strongest STEM identities tend to be supported on all three fronts simultaneously.
How STEM Identity Develops
STEM identity isn’t something you’re born with. It builds over time through a mix of psychological factors and social experiences. The three most consistent predictors researchers have identified are interest, self-efficacy, and recognition.
Interest is the starting point. Students who are genuinely curious about how things work, who voluntarily tinker or read about science, are more likely to begin seeing themselves as future scientists or engineers. Self-efficacy, the belief that you can succeed at STEM tasks, turns that curiosity into persistence. Students who trust their own ability to handle challenging material push through setbacks rather than interpreting difficulty as proof they don’t belong.
Recognition acts as the social glue. When teachers call on a student for their insights, when parents celebrate a science fair project, when peers treat someone as the go-to person for math help, those moments accumulate into an identity. Research consistently shows that recognition from teachers, peers, and family members is one of the strongest external forces in affirming a person’s place in STEM. Students who transfer between colleges also show that social adjustment at the new institution positively predicts how strong their STEM identity becomes, reinforcing that identity is as much about environment as ability.
Why It Matters for Staying in STEM
Nearly half of students who enter college intending to pursue a STEM degree never finish one. Between 2003 and 2009, about 20% of beginning STEM majors left school entirely, and another 28% switched to a non-STEM field. That’s a staggering attrition rate, and STEM identity is one of the key factors that separates those who persist from those who leave.
Students with a strong sense of themselves as “STEM people” are more resilient when courses get difficult, more likely to seek out research opportunities, and more likely to envision a future in the field. Programs designed to strengthen that identity, like research-based first-year seminars in biology, have shown measurable effects. In one study, all six student participants who completed such a seminar said they planned to remain in their biology major afterward, and broader data on first-year seminars shows higher retention rates and GPAs through all four years of college.
The flip side is equally telling. When students lack STEM identity, even strong academic performance may not be enough to keep them. They may do well on exams but still feel like outsiders, and that disconnect eventually pushes them toward fields where they feel more at home.
How Race, Gender, and Income Shape the Experience
STEM identity doesn’t develop in a vacuum. A person’s race, gender, and socioeconomic background all shape how easily that identity forms, because those factors determine how much recognition and belonging the surrounding environment provides.
In U.S. higher education, White, upper-income men have historically been treated as the “default” STEM student. Students who don’t fit that profile often face additional friction. A student might experience belonging along one axis of their identity, like being a man, while experiencing systemic barriers along others, like being Black and low-income. These overlapping identities create unique experiences that can’t be understood by looking at any single factor in isolation.
Qualitative research captures this vividly. One Black woman in engineering described being “Black first, then a woman,” yet the microaggressions she encountered were simultaneously about race and gender, targeting racialized beauty standards. A low-income Black man at a predominantly White institution described sensing “something” inhibiting his inclusion in student organizations but couldn’t pinpoint whether it was race, income, or both. That ambiguity itself becomes a barrier, creating a constant low-grade questioning of whether you truly fit.
Income plays a particularly powerful role. Higher-income students have easier access to unpaid internships, study abroad, and other high-impact experiences that reinforce STEM identity. Low-income students often can’t participate in those opportunities, not because of ability, but because of logistics. Research has found that financial security can partially compensate for other factors that negatively affect belonging, like being a student of color in STEM. Conversely, low income amplifies nearly every other disadvantage.
Even broad demographic categories can be misleading. Asian American students are often grouped into an “overrepresented majority” in STEM education data, but that label conceals enormous variation. A Pakistani-born, first-generation college student from an immigrant family may check the “Asian” box on a survey and be statistically categorized as overrepresented, while never once feeling represented in their actual STEM environment.
Building STEM Identity in Young Students
Because STEM identity starts forming early, educators have studied what works in K-12 settings. A systematic review of practices that support STEM identity in young children identified several evidence-based approaches: play-based learning, group activities, inquiry-based projects, teacher scaffolding, and creating learning environments that feel socially relevant to students’ lives.
Play-based learning stands out as especially effective for younger children. When kids experiment, build, and problem-solve through play, they start associating STEM activities with something they naturally do, not something imposed on them. Group activities help by giving students a social context for their STEM engagement, letting them see themselves as part of a community. Inquiry-based activities, where students ask their own questions and design ways to answer them, build the kind of autonomy that feeds both competence and performance.
One nuanced finding involves gender alignment. Using same-gendered characters in STEM materials, like a female scientist protagonist in a story, can help girls see themselves reflected in the field. Teacher scaffolding, where an instructor provides just enough support to help a student succeed without doing the work for them, builds genuine self-efficacy rather than dependence.
How Researchers Measure It
STEM identity might sound abstract, but researchers have developed validated tools to measure it. The Science Identity (SciID) Scale, designed for high school students, breaks identity into two dimensions drawn from broader identity theory: exploration and commitment. Exploration captures how much a student is actively investigating what STEM means to them, trying on the identity. Commitment captures how firmly they’ve adopted it as part of who they are.
Simpler tools also exist. Some researchers use a single-item measure, essentially asking students to rate how strongly they identify as a STEM person. While less nuanced, these quick assessments have proven surprisingly reliable for tracking trends across large groups. The combination of detailed scales and single-item measures gives educators and researchers flexible ways to gauge whether interventions are working.