Quartz questions whether role models are useful


There’s so, so much wrong with this article:

Firstly, the “examples” of bad role model projects are nothing of the sort.
Secondly it ignores the research that shows that role models are important.
Thirdly, it seems to think there’s only one type of intervention can be done, when what we need is a patchwork of different ones.
Fourthly, using Lovelace and Curie as anti-role models is hugely disingenuous. Curie especially got a lot of shit both during her life and afterwards because of her romantic life (after her husband’s death), and so gets criticised for both being “too perfect” and “not perfect enough”.It’s lazy to call Curie a bad role model, as if a) she is, (clue: she isn’t) and b) as if Lovelace and Curie are the only role models out there
Fifthly, Evidence on the usefulness of talking to girls specifically about the problems of sexism can hugely backfire if not done carefully. There’s evidence that focusing on women’s exclusion puts women off engaging with STEM.

The research is here:


Has anyone got a moment to take a look at it with a critical eye?



I’ll give the paper a look over.

1 Like


You’re a star!



I read the article not the research as could not access it.

My interpretation was that we need more articles like this as addressing sexism head on rather that skirting around the big issues is a sound strategy. While role models are good for some people at different stages, if people do not relate to them personally, they are not roles models. Science does do a good job of pushing girls away, or rather the sexist culture which pervades pushes girls away. The campaigns which are cited the hack a hairdryer etc just show how out of touch organisations are which commission these very un-creative messages.



The paper is accessible for me, though one does have to click on a picture of Albert Einstein!

If girls/women cannot find a role model to relate to, that means we don’t have enough different sorts of role models, not that role models don’t help. That’s not a failing of role models, it’s a failing of plurality, and it means we need to work harder on creating more varied types of role models. There is evidence that role models help, and a single paper saying they don’t have an impact is not sufficient to throw role models out as a tool in our much broader toolbox, as the Quartz article seems to want to do.

With “addressing sexism head on”, the problem there is that it can backfire. There’s some evidence that talking about the nature of sexism, such as explaining how unconscious bias works, can help girls to understand that the problems they face in STEM are not their fault and that can be positive. However, if it’s not done right, and if it’s just a case of telling girls that sexism exists, that actually puts them off STEM.

So what annoys me about this Quartz piece is that it throws out a perfectly good tool – role models – and says we should replace it with something that’s very easy to get wrong – talking about sexism directly. That’s absurd.

Yes, there’s sexism in STEM. That’s why Ada Lovelace Day and all the other dozens, if not hundreds of campaigns exist. But there is no silver bullet. We need a multiplicity of approaches because we need to reach a wide variety of types of people, and not everyone will react the same to any given approach.




This link from Catalyst suggests five things companies can do. if more women are retained in STEM industries then by default their will not just be more role models but a friendlier more inclusive culture which values social diversity for what it can offer business and academia.

I will try and access the article.

But I would make the point that sexism is everywhere not just in STEM so by tackling it head on boys and girls can learn what it is and how they can contribute to social change.

There is other research which suggests that for many girls role models are much closer to home, like mothers and other family members.

And yes there are some great and no so great campaigns but often they are single issue and not always open to collective and collaborative working. Sexism is sexism and it is systemic in society, if we can focus on this then logic would say that all sectors would be more inclusive.



There are plenty of things that companies and academia can do, yes. Loads and loads of things. As you point out there’s a bit of a chicken-egg problem, because the more women we get into STEM the easier it will be for the women who come after them. Although some of the women who are currently in senior positions aren’t helping other women up the ladder (perhaps due to Stockholm Syndrome, ie sympathising too much with the patriarchal system that they have risen within), so it’s a complex problem without as single solution.

And yes, of course sexism is everywhere. Ada Lovelace Day just focuses on STEM because it’s easier to deal with a subset of problems rather than all the problems all at once.

I’m certainly not suggesting that tackling it “head on” is one way to do it, but again, we have to recognise that doing so comes with risks if we get it wrong. Tackling it head on can result in empowerment or disempowerment, depending on how it’s done. We certainly shouldn’t relegate other tactics to the bottom of the pile based on one study, especially not in favour of something that’s so easy to get wrong.

And yes, role models for girls and women usually are closer to home, but we have to ask why. I would posit that part of the problem is that there just aren’t a huge number of positive female role models in popular culture for girls and women to relate to, so they have to turn to those nearby. There is no female Brian Cox or Dara O’Brien or Richard Attenborough. Some women are coming through, such as Alice Roberts or Maggie Aderin-Pocock, but they simply don’t have the name recognition that the men do because they are on TV so much less. You currently have to search hard to find female role models, whereas our culture is just swimming with males ones.

There are some great campaigns and some not so great campaigns, you’re quite right. And yes, sexism is systemic. But what we need is a plurality of campaigns, different people taking different approaches on the same problem, because it’s a really, really, really big complicated problem and there is no one way to tackle it. I wish there was - it would make life much easier.



This is an interesting take on the Quartz article and harks back to the whole ‘femmephobia’ thing - the idea that you can be a scientist or a woman (interested in things like fashion, beauty, etc) but not both.

we need to be careful that in doing so we aren’t inadvertently condescending to women ourselves by proclaiming that stereotypically girly things are stupid or silly or vapid. Although “all women love pink” is a sexist stereotype that we should fight against, we have to be careful not to erase the fact that some women do love the color pink or love wearing lipstick and flowers and other stereotypically feminine things and many of those women also happen to be scientists.



The Skepchick writer loses me as soon as she calls that Quartz piece ‘great’… that just gets my back up. It’s not a great piece at all, it’s lazy and ignorant. Grrr.

I will try to read that Skepchick piece properly though, as I do agree with her basic thesis. But again, it all comes down to plurality. We need women in STEM who are feminine pinkophiles, but we also need women who are goths, boardroom professional types, tomboys, and everything else in between. And the big fucking issue here that’s staring everyone in the face is that we simply don’t have enough role models or coverage for that plurality to exist. So instead of saying “Let’s just do more cool shit”, people are arguing over which sort of cool shit they thing appeals to most people. Better, maybe, to say “Do the cool shit you’re into, however you fancy doing it.”



fair point on the ‘great’ :smile:

It is about plurality. There’s this weird tension - everyone’s being encouraged to be ‘diverse’ and get more women and minorities but those women and minorities are being told that the only way they can expect to get hired is to be more like straight white men. It’s as if they want diversity of appearance but not diversity of thought.

1 Like


Aaand we have a winner!!

You know, my husband gets this all the time at work: “We want you to come in and tell us how to adapt to this new future… so long as you don’t change anything.” Business want diversity so long as it doesn’t change anything.



I’m reading the paper and thought I’d make some notes as I go.

The Background begins by discussing the effects of single-sex education on women’s prevalence of studying STEM and the word that immediately springs to mind is ‘equivocal’. It’s all 'this study shows that it helps, but this study shows that it doesn’t. There’s certainly no clear advantage to single-sex education. One thing I thought was important to note was that one study from the UK (Robinson & Gillibrand, 2004) found that while single-sex education helped ‘higher track’ students but did nothing for those on the ‘lower track’. I think this may be a confounding variable that makes single sex private/independent schools do better. Most of the time they take students on ability (at least in part) which means that they’re removing the lower track students before they start, so of course they’re going to get better results than state schools even if there’s no benefit to single-sex education (and I write as someone who went to a girls school and went to uni to study physics).

The next subject of discussion is role models. It’s this part which prompted me to write as I went as I disagree with the definition of ‘role model’ being used. They define it as “examples of female scientists as well as interactions with actual female scientists and science teachers.” but the two contrasting papers they use to determine the utility of role models defines it as ‘female faculty’. The papers they use to show equivocal evidence for the use of role models actually, to my mind, don’t show equivocal evidence at all. The say that one study found that high school faculty didn’t impact the choice of STEM careers but university faculty did. That’s not contradicting evidence - if anything it shows that university faculty has an impact on the prevalence of women accepted onto courses, regardless of their ambition at high school. When it comes to the impact of non-faculty role models, the research seems sparse.

The Study
The aim of the study was to study “the effect of certain high school physics experiences on female students’ interest in a physical science career”.

The study looked to see if the following experiences impacted the chances of female students going on to study STEM:

  • single sex class
  • woman scientist guest speakers
  • female physics teacher
  • work of women scientists discussed
  • underrepresentation discussion

They found that only the underrepresentation discussion had any positive impact and that was slight. They did find that having a female teacher increased the likelihood of the other experiences being experienced but even when all the factors were taken together there was no effect.

The authors suggest that one of the reasons for the lack of impact for single sex education is that it is taught without ‘gender-inclusive curricular reform’ (pp6). For the gender of the teacher, they describe research that shows that the relationship between student and teacher is more important than the sex of the teacher and suggest that this is why guest speakers have little impact - the students “may not have had the opportunity to build meaningful relationships with female scientists who visited on isolated occasions” (pp6).

They also cite research that found that while female scientists felt that role models were important, they couldn’t provide “any convincing stories illus- trating the significance of role models in their lives (either in the form of actual people or in stories in books)” (pp7). My problem with the assumption being that therefore role models are meaningless is that I expect very few women (or men for that matter) can easily point to a single (or even a few) people and say ‘they’re the ones that inspired me’. It’s not about wanting to follow in a particular person’s footsteps, it’s much more subtle than that. Knowing that there’s people like you who’ve done what you want to do can give you the confidence and reassurance to carry on because what you’re trying to do is not impossible. Boys from small towns can dream of playing in the Premiership because there’s footballers who started off where they are, so however extraordinary it seems they know it’s possible because it’s been done and they watch the people who’ve done it on TV all the time.

Going back to the paper, they end with a discussion about the underrepresentation discussion (the only experience with significant impact). They suggest that the benefits of this are letting girls ‘self-realise’ (by which, from the context, I think they mean discussing what the subject means to them rather than have a teacher drone on at them) and by “helping female students realize that feelings of inadequacy or discomfort they might have stem from external norms and pressures rather than from their capabilities, interests, or values” (pp7). However, there’s no discussion of the potential problems of this sort of discussion, as highlighted by Suw.

So that’s the paper. TBH I’m not overly impressed. The dataset they used is large and their analyses seem pretty good (although I’m really confused about their Ns in table 1/figure 1. There seems to be a big difference between the Ns for each experience but I’m not really sure what the N is measuring and whether the difference has led to any weighting in the analysis. At first glance it seems that the reason that ‘underrepresentation discussion’ is able to show significance is because it’s a far bigger sample size than any of the others) but I feel that their definition of ‘role model’ is quite limited and there’s no actual explanation of what ‘under-representation discussion’ actually entailed. Given that the research was done in the US where they seem to be unable to discuss something as straightforward as sex in a sensible manner, I’m really surprised that they’ve been able to discuss the issues surrounding women in STEM in a consistent and good manner that has such a positive impact on future involvement with STEM.



Thanks for that, Fishnut! That’s a really great dissection of the paper. It does seem to be rather flawed, not least with the information that they leave out, and I’m not sure that it supports the robust conclusion that Quartz drew from it – it feels more like an attempt to back up gut feeling, rather than a thorough investigation into what really works. Without a better examination of who performs the role of role models and more detail about what the ‘under-representation discussion’ entailed, it all seems a bit muddy to me.

I’d be curious to know if they quoted any of the existing research into the impacts of role models on women’s choices? Penelope Lockwood found women role models were important, though in fairness her sample size was quite small (and I’ve not read the whole paper because I can’t access it.)

And role models in politics appear to be helpful too.

So it seems to me that the lax approach to defining role models is unlikely to be strong enough to undermine existing results.



A lot of the papers they refer to were really hard to find as a fair few don’t appear to be online. I don’t know if this is because the journals doing this sort of research aren’t online or if it’s because the research of the papers they’ve selected is so shoddy it can only get into really low impact journals without the budget to have an online presence (3 of them had a single link on google scholar which led to databases I’d never heard of before. The papers they selected were all a bit of a mishmash - UK, US, Canada, Africa - but mostly looking only at one or two schools so it’s hard to say how broadly the conclusions can be drawn and how much of a comparison between studies can be made. Again, this may be because there’s not much of this type of research going on which means they’ve got to take whatever they can get, or it could be that they’ve selected the papers that support their predetermined conclusion. Without knowing the wider literature it’s hard to say.



This is really emphasising that I need to find the money to get an academic somewhere to do a full literature review of the literature on role models. Once I’ve nailed down a couple of other projects, that one has got to go at the top of the list.



That sounds like a really good idea and like a really interesting project for someone.



Evidence that role models are important in generating innovation.

Bell et al. (2017) - Who becomes an inventor in America? The importance of exposure to innovation

[C]hildren from high-income (top 1%) families are ten times as likely to become inventors as those from below-median income families. There are similarly large gaps by race and gender. Differences in innate ability, as measured by test scores in early childhood, explain relatively little of these gaps… exposure to innovation during childhood has significant causal effects on children’s propensities to become inventors. Growing up in a neighborhood or family with a high innovation rate in a specific technology class leads to a higher probability of patenting in exactly the same technology class. These exposure effects are gender-specific: girls are more likely to become inventors in a particular technology class if they grow up in an area with more female inventors in that technology class… there are many “lost Einsteins” — individuals who would have had highly impactful inventions had they been exposed to innovation.