Are first impressions accurate?

A colleague sent me an article in the Telegraph titled “Successful male leaders have wider faces than average man” to ask for my opinion. Other venues have reported similar stories: the Daily Mail reports it as, “As plain as the nose on your face: How strangers can tell if you are powerful, intelligent or even criminal just by a quick glance at your facial features“

The first thing I always do when assessing an article about research is to look for the original source. Unlike many news articles, this one helpfully included the name and affiliation of the researcher. But after about half an hour of digging, I found that Marc Fetscherin has never conducted such a study. It seems that he is actually publicising the publication of his new book, CEO Branding, which claims to pull together research into faces and personality. I haven’t read this book, but I’m very familiar with research on social perceptions of faces, so I’m going to give a brief assessment of a few of the claims made in this article.

Can you spot a good leader?

While many studies have found that men with wider faces are more aggressive, other research has suggested these findings are more about body weight than face shape. The jury is still out on the relationship between face shape and personality characteristics, as well as the mechanisms through which such relationships could occur. Much of the research is done on “found photographs”—images of mixed martial arts fighters, hockey players, or CEOs downloaded from the internet. One intriguing possibility is that more aggressive fighters or more prestigious CEOs either pose differently (tilting your head down is a great way to increase your apparent facial width-to-height ratio) or have their photographs taken with systematically different cameras (camera settings can have huge effects on apparent face shape). On the other hand, men’s face width-to-height ratio is associated with testosterone levels, even though there aren’t any sex differences in this ratio.

Can you spot a criminal?

Many of the articles mentioned a study showing that “volunteers were able to pick out convicted criminals simply by their face shape and facial characteristics”. This study did demonstrate that people rated mugshots of convicted criminals as looking slightly more like a criminal than face photographs of non-criminals (all faces were male, White, and age 18-29). The criminals got an average rating of 4.55, while non-criminals got an average rating of 4.03 on a 7-point scale of perceived likleihood of being a criminal. But it is very hard to translate scientific measurements of difference into practical examples. So I made a computer simulation of a population with these characteristics and found that, if you randomly paired one criminal and one non-criminal (I did this 1 million times), the real criminal would look “more like a criminal” than the non-criminal only about 60% of the time, and you’d be wrong about 40% of the time. But this model assumes that people can make very fine-grained distinctions about apparent criminality; if you constrain the ratings to whole numbers on the 1-7 scale, accuracy drops to chance (50%).

Can you spot a paedophile?

Additionally, the article claims that “paedophiles [are] more likely to have minor facial flaws”. While there is research showing that paedophiles are more likley to have certian physical anomalies, it is unlikely you’d ever be able to detect a paedophile from physical appearance. In this study, the researchers measured physical anomalies in men who attended a clinic for people with illegal and/or distressing sexual behaviour. Of those men, the 24 paedophiles were more likely to have some anomalies, especially of the head, than the 52 men who were teleiophiles (attracted to adults). For example, 75% of the paedophiles had asymmetrical earlobes, while 56% of the teleiophiles did. But can you spot a paedophile by appearance alone? There just isn’t enough information in this paper to tell. It doesn’t provide any information about the rates of facial asymmetries in a typical population, just in a population of men who all have problematic sexual behaviours.

So what is the point of all this research?

I don’t want to give the impression that all of the research mentioned above is incorrect, fraudulent or trivial. It simply isn’t designed to help us figure out how to predict who will be successful CEOs or hockey stars. The face width-to-height ratio research is designed to explore the complex links among hormones, behaviour and appearance. It can also give us insight to the way very subtle correlations between appearance and behaviour can lead to widespread stereotyping. The research on the links between appearance and paedophilia was also never meant to be used to play “spot a paedophile”, but to test ideas that this type of problematic behaviour can have ties to other developmental disorders that are associated with facial asymmetries.

Dyshniku, F. et al. (2015). Minor Physical Anomalies as a Window into the Prenatal Origins of Pedophilia. Archives of Sexual Behavior. DOI: 10.1007/s10508-015-0564-7

Lefevre, C. E. et al. (2013). Telling facial metrics: facial width is associated with testosterone levels in men. Evolution and Human Behavior. DOI: 10.1016/j.evolhumbehav.2013.03.005

Valla, J. M. et al. (2011). The accuracy of inferences about criminality based on facial appearance. Evolutionary Behavioral Sciences. DOI: 10.1037/h0099274