Learning new faces - A mental ability that doesn't peak until the early thirties

Cognition researchers should beware assuming that people's mental faculties have finished maturing when they reach adulthood. So say Laura Germine and colleagues, whose new study shows that face learning ability continues to improve until people reach their early thirties.

Although vocabulary and other forms of acquired knowledge grow throughout the life course, it's generally accepted that the speed and efficiency of the cognitive faculties peaks in the early twenties before starting a steady decline. This study challenges that assumption.

A massive sample of 44,000 people, aged between ten and seventy, completed an online face learning test in which they were required to study briefly several unfamiliar faces, presented in grey scale without hair or other non-facial distinguishing features. They then had to identify those faces, shown in novel poses and varied lighting conditions, from among further unfamiliar faces.

As you might expect, performance at the task increased steadily through adolescence. But although improvement slowed once adulthood was reached, it didn't stop there. Performance in fact peaked among those participants aged 31.4 years, after which it declined slowly. The pattern of results meant that average performance by 16-year-olds matched the average performance of those aged 65.

The results suggest strongly that face learning capabilities continue to develop into the early thirties, but an alternative explanation is that the sustained changes are more generic, to do with general memory or cognitive abilities. To rule this out, a second study tested nearly 15,000 people on a face task and also a memory task involving names. As before, face learning ability peaked in the early thirties. In contrast, performance at the learning of names peaked at age 23.

A final study used children's faces, in case the earlier studies' use of more mature faces had given older participants an unfair advantage. Even with children's faces, facial learning peaked in the early thirties. However, this prolonged developmental trend wasn't found for inverted faces (performance with these peaked at age 23.5 years), thus suggesting it's specifically the ability to learn new up-right faces that continues to improve into the thirties. It remains to be seen whether this improvement reflects a kind of prolonged innate maturation process or if it's simply a consequence of more years practice at learning faces.

How big were the increases in face learning performance between the end of adolescence and the early thirties? They were modest (the effect size was d=.021) so more research is needed to find out what real life implications, if any, these lingering improvements in ability might have. Another study limitation is the use of a cross-sectional sample. Future research should study changes in ability in the same individuals over time. Notwithstanding these points, the researchers said 'our data illustrate that meaningful changes can and do occur during early and middle adulthood and suggest a need for integration of research in cognitive development and aging.'
_________________________________

Germine, L., Duchaine, B., and Nakayama, K. (2011). Where cognitive development and aging meet: Face learning ability peaks after age 30. Cognition, 118 (2), 201-210 DOI: 10.1016/j.cognition.2010.11.002

That's not a poker face, this is a poker face

What does your poker face look like? If it's the traditional, stern, emotionless expression, you may want to consider practising a new one. Erik Schlicht and colleagues report that a friendly, trustworthy face is more likely to influence your opponents, leading them to think that you've got a good hand - that you're not bluffing.

Schlicht's team had 14 relative novices play hundreds of one-shot rounds of a simplified version of Texas Hold'em poker against hundreds of different 'opponents'. Each round the participants received a two-card hand and their opponent had bet 5000 chips. They had to decide whether to 'fold' or 'call'. Folding meant they would lose 100 chips guaranteed. By calling, they would win 5000 chips if their hand was stronger then their opponent's, or lose the same amount if their hand was weaker. To boost their motivation, participants had the chance to win a small amount of money based on the outcome of one randomly chosen hand out of the 300 that they played.

Each round, before making their decision, the participants saw a picture of their opponent's face. These were morphed to appear either untrustworthy, neutral or trustworthy (see picture). Participants were told that, as in real poker, the different opponents could have different styles of play (but no mention was made of faces providing a clue to style).

Because participants played just one round against each opponent there was no opportunity to use past behaviour to make judgments about their style. This meant the only information participants had to go on was the cards in their own hand and any inferences they'd made about their current opponent's playing style based on his face. They didn't receive any feedback during play on whether they'd won a round or not.

On each round, there was an optimal decision for participants to make considering the cards in their hand and the stakes involved in holding or calling. The researchers were careful to ensure that participants' hands were of equal value across the different categories of opponent face - trustworthy, neutral, untrustworthy. Unbeknown to the participants, their opponents' hands bore no relation to their facial expression.

The key finding was that faces with neutral or untrustworthy expressions made no difference to the decisions the participants made. By contrast, if an opponent had a trustworthy face, the participants took longer to decide what to do and they made less optimal decisions. Effectively, they were behaving as if their opponent had a better hand.

'Contrary to the popular belief that the optimal face is neutral in appearance,' the researchers said, 'poker players who bluff frequently may actually benefit from appearing trustworthy, since the natural tendency seems to be inferring that a trustworthy-looking player bluffs less.' Before you try this out at your local poker den, remember the findings apply when you're up against new opposition and there's little other information to go on.
_________________________________

Schlicht EJ, Shimojo S, Camerer CF, Battaglia P, & Nakayama K (2010). Human wagering behavior depends on opponents' faces. PloS one, 5 (7) PMID: 20657772

For further info, lead author Erik Schlicht has created a webpage where he answers frequently asked questions about this research.