Are older kids and adolescents really as good as adults at recognizing emotions in faces?

Take a look at this face:


ResearchBlogging.orgDoes it look more angry or fearful? It may be rather difficult to tell: About fifty percent of adults say faces like this are angry and fifty percent say it's fearful. However, for children, the story is different. Researchers have found that small children aren't as good as adults at recognizing emotions in faces. Young children would see this picture as more fearful than angry. However, most research has suggested that kids are just about as good as adults by the time they're five years old.

But neuroscientists have consistently found that the portions of the brain responsible for processing key emotions, particularly anger and fear, continue to develop all the way through adolescence. If our brains are still changing, shouldn't we see some impact in the way kids and teens perceive emotions?

A team led by Laura Thomas felt that earlier studies of children's perception of emotion were flawed because the tests were too easy. Instead of viewing subtle emotional variations like the photo above, kids saw dramatic, obvious facial expressions -- even schematic diagrams of emotions. Could it be that what's developing as kids mature into teenagers and adults is their ability to detect subtle emotional variations?

Thomas's team showed 102 children, teens, and adults pictures of 10 different actors which had been previously rated as expressing anger, fear, or a neutral emotion. But instead of showing the most obvious emotional expressions, they used morphing software to show the viewers gradations of the emotions, like this:


For the neutral to anger morph, Picture 1 is 22.22% angry and Picture 6 is 77.77% angry. The picture at the start of this post is Picture 4 in the Fear to Anger morph, 44.44% fearful and 55.55% angry. The viewers saw the pictures in random order within blocks rating a particular emotional continuum (e.g. Neutral to Anger). Remember, there were 10 different actors portraying the emotions. After seeing each picture for 3 seconds, the viewers had to say which emotion was depicted (e.g. "Neutral" or "Angry") as quickly as possible. Here are the results for Neutral to Anger:


Adults were significantly better than both children and teenagers at detecting whether an expression was angry or neutral. Look at the ratings for face 1 and 2 for adults compared to kids and teens. These faces were 22.22% and 33.33% angry, and fewer adults rated them as angry. Similarly for faces 5 (66.66% angry) and 6 (77.77% angry), adults were more likely to say those faces were angry. When the overall pattern is analyzed statistically, the difference between adults and teens is significant. The difference between teens and children is not.

Similar patterns held for the Neutral to Fearful and Fearful to Angry morphs. The researchers say that this pattern could explain some of the difficulties parents have interacting with teenagers: the teens literally don't pick up on emotional expressions that their parents do. One interesting question: is the problem simply one of recognizing emotions in adults? Since everyone was tested on the same (adult) faces, perhaps the results would be different -- or even reversed -- if the participants viewed kids' or teens' faces.

Thomas, L., De Bellis, M., Graham, R., & LaBar, K. (2007). Development of emotional facial recognition in late childhood and adolescence Developmental Science, 10 (5), 547-558 DOI: 10.1111/j.1467-7687.2007.00614.x

More like this

The article says about 50% of adults think it's angry, whereas children all think it's fearful and concludes that adults are thus better at judging emotion. How is this evaluated? If the face is indeed fearful (as I think it is), this just means that kids are a lot better at judging emotion. But how will we know? We can obviously ask a bunch of people what they think and assume it all averages out correctly. But then if you ask mainly adults, it will be no surprise that other adults match their perceptions better.
Another way would be to ask the person making the facial expression, but first of all I don't think introspection is all that accurate and second of all, the guy is actor. The emotion is acted. What if he's just not a very good/accurate actor?

I also have severe objections to using morphing software to make a face 44.4% angry. I would say you're either angry or you're not, and then you might have gradations (a little angry vs. very angry). It seems to me that someone who is very angry has a very different favial expression than someone who is a little angry. The assumption that we can just take a neutral face and a very angry phase and morph it to get a slightly angry face seems extremely flaky to me. Putting a percentage of angryness on such a face is an even bigger assumption.

The only thing this study shows is that adults are apparently more outspoken.