A lack of visible role models can create a barrier to entry for underrepresented groups

Why is diversity in STEM important?

As we know, diversity in the workplace can improve employee morale, attract top talent and contribute to economic growth. However, we continue to see STEM (science, technology, engineering and mathematics) related roles struggling to achieve this balance.

Interestingly, research suggests that diverse problem solvers perform better than a non-diverse group even if they have high-performing members.

A diverse team can process information carefully, ask more questions and bring fresh perspectives to problem-solving which generates new ideas and fosters creativity.

While we are seeing slow improvements in the male to female ratio, STEM roles are still largely male-dominated even though a diverse team is more likely to perform better. The question is ‘why’?

Were you surprised that artificial intelligence (AI) image generators are challenging stereotypes?

Access Education have conducted research into the top professions where AI is least likely to represent women in key STEM roles, and the ones where AI is challenging stereotypes.

The results of the research were very interesting, especially in the top 10 professions where AI overrepresented women, with robotics engineering coming out on top.

It was also fascinating to see that, contrary to reality, the AI image split between male and female figures are relatively balanced in roles such as surgeon, computer specialist or electrical engineer.

This is important because a lack of visible role models has proven to create a barrier to entry for underrepresented groups in these professions.

Although the AI image generator underrepresented women for roles such as pharmacist, clinical researcher, biotechnologist and game developer, the differences are comparatively smaller.

AI models are ultimately trained and programmable, so it’s not inconceivable that more conscious efforts are placed for certain prompts to promote greater diversity in these visualisations, considering the multiple facets of our identities.

How do stereotypes impact STEM?

As the AI image generators demonstrated, there is clear bias about the male to female ratio in STEM and women are largely underrepresented in certain roles. This lack of visibility can deter women from participating or showing an interest in these subjects at school, especially if there is no visual representation to tell them otherwise.

This links to my own research from the University of Reading which finds that students decide whether or not to pursue STEM disciplines based on stereotyped views of what makes a typical student.

In biology, women are equally represented but students think this subject is the easiest and it’s unclear whether it’s perceived as easy because of the female representation. Students in employment, those with disabilities or caring responsibilities also believe taking on tough subjects wouldn't fit in with their lifestyle.

Students' opinion on themselves, their coursemates and their subjects reveal how wider societal stereotypes impact who can succeed in the four subjects we included in the study: biology, engineering, mathematics, and physics.

Language is also important when describing subjects, for example, mathematics was associated with  more masculine words compared to biology. This can again deter women from pursuing mathematics subjects as they may feel like they don’t fit in.

To widen participation in STEM from underrepresented groups, we need an attitude change, whether that’s in real life or AI. Image plays an important role in who pursues STEM so diversifying images to be more inclusive is essential.

How can schools engage a diverse range of students in STEM subjects?

Of course, schools play a key role in engaging young people from different backgrounds in STEM and there are many strategies they can implement to widen participation.

Schools can provide opportunities for extracurricular and informal STEM teaching, from technology to mathematics, and the spectrum of sciences. This way students have the chance to develop new skills, and show an interest in STEM without any academic pressure.

As educators, we all have a duty to break down perceptions of who can study STEM. This includes challenging the narrow but popular view that certain STEM subjects are difficult, leading students to think they’re not for ‘people like me.’ This is often due to identity mismatch including the belief that certain subjects are only for the very ‘clever’ or ‘talented.’

Teachers can scaffold with students about the different and diverse characteristics and skills that have value across STEM subjects, which could help break down some of these barriers for less experienced learners. They should unpack and construct their desirable and realistic version of the ideal student in STEM, to avoid stereotyped views and increase young people’s confidence in pursuing STEM subjects.

It's important for schools to recognise that representation matters, from the staff to the student composition of STEM subjects. Having mentors, role models, alumni or previous STEM students from underrepresented backgrounds who can share their stories and journeys is invaluable. It’s important that these figures are relatable to help increase interest and participation.