Thursday, 10 March 2011

The Hillary Clinton effect - how role models work for some people but not others

The benefits, or not, of reading about Ms Clinton 
Fear of prejudice can adversely affect people's performance. For example, female participants reminded of the stereotype that women are innately inferior at maths compared to men, subsequently perform sub-optimally at a maths task, especially in the company of men. This effect, known as stereotype threat, occurs at least in part because of the anxiety that one's own poor performance will be used by the ignorant to bolster their prejudicial beliefs.

An antidote to stereotype threat is to remind people of high achieving members of their in-group. For example, reminding Black Americans of President Obama's success has been shown to improve their subsequent IQ test performance. Psychologists think this 'Obama effect' occurs because the role-model's salient success takes away the burden people feel of having to represent their group.

A new study by Cheryl Taylor and colleagues has built on this literature by showing that the stereotype-busting effect of a role-model only occurs if that role-model's success is perceived as due to their own innate ability and effort. If the role-model is considered to have been lucky then their stereotype-busting power is lost. Taylor's team call this the Hillary Clinton effect.

Dozens of female undergrads rated the extent to which various successful women deserved their success, including Hillary Clinton, Paris Hilton and Oprah Winfrey. Pilot work had already established that Hillary Clinton tends to divide opinion and that was replicated here. Several months later these same female undergrads were recruited for what they thought was a separate study. Their main task was to complete a maths test. Beforehand, however, some of them were reminded of the 'women are poor at maths' stereotype. And within that stereotype-reminded group, before the maths test, half were asked to read a factual account of Hillary Clinton's life, followed by questions on it, whilst the remainder read about a successful British company (this was intended to be innocuous, just to control for the effect of completing a reading comprehension task). The key question was whether reading about Hillary Clinton would have a protective effect or not.

The classic stereotype effect was replicated. Women reminded of the sexist stereotype (and who read about a successful British company) answered 50.7 per cent of attempted items correctly compared with a success rate of 59.3 per cent achieved by women who just took the test without the stereotype reminder (there was no difference in the number of items attempted). What about the participants who read about Hillary Clinton? It depended. For the women who'd earlier said they judged Clinton's success to be deserved and due to her abilities, reading about her offered protection: they scored 62.3 per cent correct. By contrast, for the women who judged Clinton's success as down to luck and nepotism, she offered no protection: they scored just 48.9 per cent correct.

'Reading a factual biography of Hillary Clinton alleviated the performance deficits associated with mathematics stereotype threat for some women, but not for others,' the researchers said. Now more research is needed to explore this effect. For example, the perceived 'likeability', or many other characteristics of the role model, could be the key factor explaining their protective value, rather than the deservingness of their success. In the meantime, Taylor and her colleagues said the stereotype-busting effects of role-models could be enhanced and preserved by ensuring people are aware of the stable and internal causes of the role-models' successes.

ResearchBlogging.orgTaylor, C., Lord, C., McIntyre, R., and Paulson, R. (2011). The Hillary Clinton effect: When the same role model inspires or fails to inspire improved performance under stereotype threat. Group Processes and Intergroup Relations DOI: 10.1177/1368430210382680

Post written by Christian Jarrett (@psych_writer) for the BPS Research Digest.


Olin Hyde said...

Excellent research verifying that we all need to find positive role models that we perceive as people who earned their accomplishments. Put another way, this is a brilliant indictment of our culture of celebrity worship and an endorsement for us to seek out those that actually do something.

janitt dott said...

I find this iffy "study" a bit prejudiced against it's subject which appears to be The Hillary Clinton Effect. This article itself indicates to me what her popularity problem is (the CONSTANT bias she faces) To lump Hillary with Paris is and Oprah (as a group of Successful Women) is QUITE an accomplishment gap. They ARE famous for being famous. SHE is a former Senator First Lady and Secretary of State! May I also remind the author that 18 million American voters found Mrs Clinton "likeable" enough to want her to BE the first woman President of the United States! So the most curious stat to ME is that college undergrad women can READ a "factual bio" of Hillary Clinton's MANY landmark accomplishments and STILL accuse of her nepotism and luck! THAT is a bit more worthy of study than IF she can improve someone's math test!

Ben said...

Two follow up studies would be interesting:

1) Investigating the effects of self esteem and locus of control on Stereotype Effect. I highly suspect that this phenomenon will fall sharply amongst people with higher self esteem and higher internal locus of control, those who are not so heavily governed by the expectations of others. I'd also like to know if this same effect has been proven to exist in men of whichever class is dominant in the given society, if not then we could simply be seeing sample bias; lower self esteem could be the true causative variable.

2) Does using a rolemodel that's obviously skilled at the particular subject at hand combat this effect even more? For instance, in a test where men were told that women are better at psychology/psychology is a women's subject, would giving an account of Harry Harlow be equally combative of the effect as giving an account of David Beckham, or some other successful man whose skills are unrelated? Presumably the relatedness of the skills to the skill set at hand is more important than whether nepotism was involved.

Lastly, a consideration to be gained from this study is that if nepotism of rolemodels genuinely does render the stereotype effect even worse than base levels, 'Positive Discrimination' would actually be very destructive to the minorities whom it purports to help as it would place in power a series of minority professionals that have achieved their status largely through artificial nepotism, thus making the Stereotype Effect even worse in the general minority.

@ Janitt, perhaps I have misunderstood you, but 18m voters in a country of 316m is a very, very poor support base; that's 6% which means that means 94% of the country didn't want her to achieve presidency. I don't know whether that's valid or not since I'm not an American, but if it is I really wouldn't ever use that statistic again to try and prove her popularity. I also presume that the 'nepotism' students read the bio and accused her of nepotism based upon prior prejudices, not the bio itself. If the bio was deliberately engineered to suggest nepotism then that would destroy the entire study.

Jessica Chivers said...

I'm curious about whether the role model effect (i.e. the uplift in performance for women who rate Hilary Clinton as deserving of her success) holds in situations where there is no stereotype threat. I mean, could reading about a deservedly successful woman increase the reader's performance full stop? Being buoyed by reading about the role model could increase one's own sense of self-efficacy.

I'm going to share this in my "Flourishing Female" newsletter next month and discuss the power of role models so thanks for posting it Christian.

Robert Hamm said...

There is a problem with the parts of this study that are not randomized. To hold the belief that Hilary is not capable and is where she is only because of nepotism is self chosen. Perhaps these are just less intelligent people, who would be expected merely on the basis of this low intelligence to get a poor score on a math test.

Anonymous said...

I hope more tests like this come out. Im going to use this information to my advantage and read about Mea Joe Jemasen the nest time I take a test!

Post a Comment

Note: only a member of this blog may post a comment.