Sexism affects robots

The Journal of Applied Social Psychology has just pubished a study that is both bizarre and profound. It reports on two experiments that show gender stereotyping extends to robots.

(S)he’s Got the Look: Gender Stereotyping of Robots

F. Eyssel and F. Hegel

Journal of Applied Social Psychology

Previous research on gender effects in robots has largely ignored the role of facial cues. We fill this gap in the literature by experimentally investigating the effects of facial gender cues on stereotypical trait and application ascriptions to robots. As predicted, the short-haired male robot was perceived as more agentic than was the long-haired female robot, whereas the female robot was perceived as more communal than was the male counterpart. Analogously, stereotypically male tasks were perceived more suitable for the male robot, relative to the female robot, and vice versa. Taken together, our findings demonstrate that gender stereotypes, which typically bias social perceptions of humans, are even applied to robots. Implications for design-related decisions are discussed.

Sadly, the study is locked behind a paywall, which is a pity because the discussion about “implications for design-related decisions” is a sort of parallel-world look into android gender politics.

The authors discuss whether it is better to create gender free robots to fight social stereotypes or whether we should create robots that comply with society’s prejudices to make them more acceptable.

Personally, I’m all for genderqueer robots. That would really throw a spanner in the works. Or a works in the spanner.

Link to locked study (via @hysell)

BBC Future column: Wear red, win gold?

My latest column for BBC Future, a cautionary tale of scientific research, with an Olympic theme. Original here.

Studies show that wearing a particular colour increases the chances of winning a gold medal. Why this is the case serves as a timely reminder that we should always be wary of neat explanations for complex phenomena.

What does it take to be an Olympic winner? Skill? Yes. Dedication to training? Definitely. Luck? Perhaps. What colour kit you wear? Possibly.

Research conducted during the 2004 Olympic Games in Athens showed that competitors in taekwondo, boxing and wrestling who wore red clothing or body protection had a higher chance of winning. The effect wasn’t large, but when the statistics were combined across all these sports it was undeniable – wearing red seemed to give a slightly better chance of winning gold. The effect has since been shown for other sports, such as football.

The researchers had a straightforward explanation for why wearing red makes a difference. Across the animal kingdom, red colouration is associated with male dominance, signalling aggression and danger to others. The vividness of the red displayed by individuals of various species has been shown to relate to the amount of the hormone testosterone they have in the bodies, which also correlates with their physical health and eventual breeding success. The researchers claimed that humans too are subject to this “red = dominance” effect, and so, for combat sports, the athlete wearing red had a psychological advantage.

In competitive sport, small advantages like this matter. The difference between winning and losing can be milliseconds, or millimetres. So, should every country be fighting for the right for their sportspeople to wear red?

Maybe they should, but not for the reasons the study authors claimed. What happened next is a textbook case of the way in which research happens, showing us why we should always be wary of neat explanations for complex phenomena.

Close calls

Like all good science, once someone has proposed a theory, others can hold it up to scrutiny. And so it was the case with the red=dominance explanation. Another research group analysed data from a different sport at the Athens Olympics, Judo, but they found that contestants who wore either white or blue had an advantage. Instead of being an effect of evolutionary colour signals, the new claim was that the difference in performance was due entirely to the visibility of the different colours. In a combat sport the person wearing the brightest clothing will be at a disadvantage – their opponent will find it slightly easier to see where they are and anticipate their next move.

Convinced? Don’t make up your mind yet, because there’s a further twist to the tale.

This debate was resolved in the most interesting way at around the time of the Beijing Olympics in 2008. A new study suggested that the previous theories based on dominance or visibility of the competitor were wrong. The effect wasn’t anything to do with the effect of colour on the athletes, but instead to do with the effect on the referees.

I’ve written before about how we all have a tendency to look for causes that are somehow part of the essence of a person, and this seems to be another example. The statistics were correct, contestants wearing red really do win more, but we had been looking in the wrong place for an explanation. This study used digital manipulation to show experienced taekwondo referees fights that were identical, except for the colours worn by the contestants. Judging the same fights, referees awarded more points to contestants who had been photoshopped red than to contestants who had been photoshopped blue.

In any competitive sport there will be close calls, situations where the margin of victory is small, and a referee has to make a judgment to the best of their abilities in the blink of an eye. It seems that because red does have an association with victory and dominance, the judgement of these marginal situations can, occasionally, be influenced by the contestant’s clothes colour. Colour does produce a psychological effect, but it is a bias in the refs, not in the contestants.

Horse play

This story provides a classic warning for anyone trying to find psychological causes for things: the effect can just as easily be in the observers as in the thing we observe. Psychology students around the world are taught the story of Clever Hans, a horse that many believed could do arithmetic. Huge crowds would pay to see Hans, held by his trainer, being asked questions such as “what is five plus two”, and answer by stamping his foreleg seven times.

This seemed like a wondrous example of animal intelligence, until a psychologist showed that Hans was performing his trick by reading the body language of his trainer. Hans would start tapping his foot. When he got to the correct number his trainer would relax, and Hans would read this signal and stop. What looked like a miraculous ability to do maths, was really a clever – but not miraculous – ability to act according to what his trainer did.

So there the matter rests – for the moment at least. Wearing red could give you an advantage in competitive sports, but its because of the effect it has on the observers, not the observed. And, just maybe, we’ll try to be a bit more careful about calling victory as we watch contests happening in the London Olympics.

Psychosis and the fog of reality

Last May The New Yorker had a beautiful but paywalled article on on psychosis and insight. Thankfully the full text has found its way online as a pdf.

Psychosis is the psychiatric term for delusions and hallucinations, with insight being the ability to recognise that what you believe or experience is not a fair representation of reality.

The concept of insight is more easily applied to hallucinations than delusions, after all you can hallucinate patterns on the walls but realise that the patterns are not really there, but you can’t really have a belief and not believe it.

With regard to delusions, it is tested by seeing how readily people can accept that there is a chance they might be wrong. In other words, it’s an estimate of certainty with absolute certainty in a false belief being considered abnormal.

In practice, and due to the difficulties on agreeing on or verifying reality, it often comes down to whether you agree with your psychiatrist (indeed, one definition of insight, includes accepting treating as a sign of good insight), sometimes leading to situations where people with genuine psychosis completely reject any form of treatment even where it would be of clear benefit.

Rachel Aviv’s article for The New Yorker is a brilliant exploration not only of the experience of slipping into psychosis but also the politics and practicalities of insight.

By the way, Aviv has written a series of excellent articles about mental health including one called ‘Which Way Madness Lies: Can psychosis be prevented?’ for Harper’s Magazine which is also online as a pdf.

pdf of article on insight in psychosis.
pdf of article on preventing psychosis.