Reading emotions through face masks

Tuck in to some Naked Neuroscience news with our local experts...
29 January 2021

Interview with 

Helen Keyes, ARU; Duncan Astle, Cambridge University


Brain schematic


This month, Helen Keyes looked at a paper studying the implications of adults wearing face masks on children’s abilities to interpret emotions...

Helen - This study focused on 81 children who were between the ages of 7 and 13. And that's a real key age where children start to rely heavily on using eyes to interpret emotional expression. Now, these children saw lots of images of faces and the faces had different facial expressions - sadness, anger, and fear. And the images would start off really fuzzy and they would get clearer and clearer. And the children would say which emotion was being expressed in the face. Obviously as the face became clearer, this task became easier. And the faces were either presented as a whole or presented with a face covering over them, just a surgical face mask, or they were presented as wearing sunglasses. And as was expected, children were the most accurate at identifying facial emotion when there was no face covering at all. So no face mask and no sunglasses, that's not surprising.

However, that affect was relatively small. Indeed when we look at the individual emotions, children still performed significantly greater than chance at recognising the emotion for sad faces. Even if the face is wearing a mask or wearing sunglasses, children still performed better than chance. They could still tell that the face was sad. When we look at angry faces, again wearing a face mask didn't have that much of an effect - in terms of children still performed better than chance at recognising the angry emotion, whereas sunglasses really impaired that. And then it's only really when a face was expressing fear that the introduction of any sort of face covering, so a face mask or sunglasses, really made it hard for children to interpret that emotion. They found it very difficult to tell if the face was expressing fear or surprise when you have that almost wide-eyed look, if they're wearing either a face mask or sunglasses.

So there is quite a lot of hope here, even though obviously we do better without any face coverings, the children could still largely interpret the emotions that were being conveyed.

Katie - What do you reckon to the variation in how good the kids were at picking up on the different emotions?

Helen - It's to be expected. So we know that fear and surprise are really difficult to distinguish in terms of the real main key emotions. Often anger and sadness are mixed up. So I thought it was really interesting that the children could clearly distinguish between sadness and anger.

We're talking about still images of faces here. Now in the street, if you met somebody and they were sad or angry, there's many more cues that you would rely on aside from their face. So their body language, what they're saying or shouting, that type of thing, you can clearly distinguish. But just looking at those images, it can be really tricky to distinguish between sadness and anger. And it's really a nice finding to see that the children could still make out those emotions.

Katie - How diverse were the group of kids. I'm just wondering if this kind of thing varies by individual or maybe by background or culture?

Helen - So there was a really nice spread of children. Oftentimes in this type of research, there can be a tendency to focus on one subgroup, the majority, subgroup of children, but this one had a really nice balance across different ethnicities in the children. And a fairly good gender balance between the children. And there wasn't any massively major findings there other than boys were slightly quicker at recognising the anger emotion than girls were.

Katie - The people, the faces, these were strangers, right? Just pictures on a screen.

Helen - They were all strangers. Yes. Taken from a database that is really commonly used, where we have really established that these are all sad faces and people have validated these as looking like sad faces or angry faces and so forth.

Katie - Do you suspect this might change in any way, depending on the relationship to the adult? Say, it's your parent or your teacher, you know, those really key interactions.

Helen - That's a really good point. So yes, we will definitely get better if it's a familiar face, just simply because we're much more familiar with how those faces look under different conditions. And we can assimilate that idea of someone's face in lots of different conditions into a nice template of their face. And it's really easy for us to use very few cues then with the people we're familiar with, to infer that emotion.

But in reality, the people that these children are going to see with face masks aren't predominantly going to be their parents or their teachers. Predominantly it's going to be strangers on the street.

Katie - What about kids who might find emotional communication difficult anyway? Say maybe kids who are autistic.

Helen - That's a good question because autistic children and autistic adults don't have as strong a tendency as people who aren't autistic, they don't have that strong tendency to look to the eyes and focus and linger on the eye area. So there is some evidence to suggest that people who are autistic prefer to look at featural cues, such as the mouth area to get their emotional cues. So there hasn't been a lot of research done on this in terms of face mask wearing or usage, but we would expect, yes, that to be further impaired, if your preference is not to look at the eye area, but to use the mouth for your cues for emotions. Yes, we would expect that to be a more significant impairment.

Katie - Like you said earlier, obviously emotions are quite nuanced. You know, there's facial expression, there's body language to take into account. How significant do you think this study is overall?

Helen - I think it's not massively significant in terms of whether children can ever interpret emotion in the world or their social development. I think really what it does is placate some of the worry that is really rampant at the moment that all parents are so worried about their children's social development. And it's nice to have a really small piece of comfort that says, look, well, one thing you don't need to worry about is "Oh my goodness, my children aren't getting enough input of facial emotion and they're going to be stunted for life!" So it's a nice, very small, but meaningful reassurance for parents who are already quite worried.

Duncan - So we know that the kids are better than chance in this laboratory environment, which is quite well controlled with this standard face stimuli. I was wondering whether we think that will play out in the real world, which is kind of busy and lots going on? Would that better-than-chance performance be enough for children to reliably recognise emotions out in the real world?

Helen - I think that's a really good question. And if we wanted to make this more realistic, we would, of course be introducing lots more useful social cues, as well as lots more busy-ness in the visual environment. So, yes, whereas we don't know in the real world if children would still be able to recognise these emotions through masks, they would have far much more information in terms of somebody's body language and what they were conveying in other ways, apart from their facial expressions. So it'd be very difficult to tease those two apart, which is why it's quite nice to have this in isolation, telling us just about the interpretation of facial emotion.

Duncan Astle looked at a review which explored how many teachers believed in the concept of learning styles, how likely they were to implement them in the classroom, whether this belief is reducing over time and whether intervention to counter this belief works...

Duncan - Learning styles is the concept that each person has a subtly different way of learning. So some people might prefer information to come in a visual format, some may be in an auditory format, some might prefer to learn by doing something themselves, a so-called kinesthetic learner. And the idea is that if you can match someone's learning style to the way you deliver information in the classroom, then you'll create the optimal environment for them to learn. And you will boost their learning over time.

Katie - What evidence is there to support this idea about how people learn?

Duncan - I'm afraid there isn't very much evidence. In fact, the data show that certainly kids will tell you that they have a preferred style. They will be only too quick to tell you what their preferred style is. However, the data shows that they are no better in their preferred style than in any other. So of course everyone's got a preference, but that doesn't seem to correspond at all to learning. And actually there can be some dangers of promoting the idea of learning styles, because for example, if I tell you that you're a kinesthetic learner, when I try and deliver something in class in an auditory or visual format, what's the point in you paying attention? After all this isn't in your preferred style. So the evidence that it works is null. There isn't any. And actually there's increasing realisation that there can be some negative consequences of promoting this idea.

Katie - So if there isn't any evidence that it works, then where did the idea come from?

Duncan - Well, the original idea comes from a kind of management consultancy, but the reason that it got into teaching is because - as we'll kind of come to later on - it still features in teacher training. And it's still been featured in a number of even quite recent teacher training textbooks. And that's how it's gotten into the classroom.

Katie - What did the review find then?

Duncan - Well, what they wanted to establish is whether or not teachers still believe in learning styles, and the way that they went about doing that was to do a systematic review. So they had very careful criteria for choosing different published studies. And that resulted in 33 different studies covering the views of over 15,000 teachers, surveyed between 2009 and 2020. So over a 10 year span. And of those teachers, over 89% believe in learning styles, that if you match the learning style with the delivery, you'll improve their learning. And pretty much all of those who believe in it intend to use it in their classroom practice.

Katie - What did the review find in terms of the trends? Are more or less people believing this?

Duncan - Well, it's certainly not getting any better. So what you might expect to see is if fewer and fewer people are believing in learning styles, then as time progresses - so from 2009 through to 2020 - gradually fewer teachers will say they believe it. But that's not true. There's no significant relationship with time. And actually if you split it out between trainee teachers and established teachers, actually the trainee teachers are slightly more likely to believe in learning styles, even though those are the more recent graduates of teacher training. So there isn't any evidence that it's getting better in time. It seems to suggest that basically it's pretty constant over this ten-year period.

Katie - If there is no evidence that this works, does the act of communicating that make a difference?

Duncan - It does. So there's only four studies that have tried to intervene, but they do all show that it's pretty effective. So if you deliver some kind of campaign to explain to teachers that this is nonsense, then their belief in it goes from about 78% down to 37%. So you can really change perspectives. I mean, I wonder what's happening in those 37% who presumably just don't believe the intervention, but it does show that it's pretty effective.

Katie - What does science tell us about the best ways to learn? Is a variety in learning and teaching methods useful?

Duncan - Yes, but not for the reasons that learning styles says it should be. So we know that if you deliver information from multiple different perspectives, then you get what's called deep processing. So the reason that it's good to learn information from lots of different perspectives is because you lay down more durable long-term memories associated with that, rather than a more kind of shallow trace.

Katie - By different perspectives, are you talking about reading something, versus listening to a teacher, versus doing something?

Duncan - Exactly. And the idea is because each time you re-encounter the information in a different format, you kind of reconstruct the traces from the first time. And with each reconstruction, you make the decay slope of the memory more shallow. And thus it becomes more and more durable with time.

But that's not because you're pigeonholing kids into learning in one way or another. It's just a general principle that the more variety in the way information is delivered tends to produce more durable memories.

Katie - Oh, I see. So even if I prefer learning via someone talking to me, actually doing all of those things is probably good for my learning. Is that what you're saying?

Duncan - Yeah, exactly. And you also create what's called context-independent memory. There's a really strong effect, which is that let's say you learn stuff in one format in one room. You're more likely to remember it when you have to be in that room using the same format. Whereas if you have variety in the style and locations of presentation, then you tend to get memories that are more durable across different contexts and across different formats.

Katie - What did the review make of the actual quality of the evidence they were looking at?

Duncan - Big caveats. It's really hard to get a really unbiased measure of sort of people's beliefs in something using an opportunity sample, which the vast majority of these studies are. And that's because the people who engage in that kind of thing might not be your typical teacher. And almost always these questions about learning styles and beliefs in learning styles are embedded alongside lots of other neuro-myth questions and things that might seem much less plausible. And so there could be a biasing effect in that when they come across the learning styles questions they think, "well, that seems much more plausible than the nonsense I've just read". And so all of these factors will really influence how true a measure of belief in learning styles you're actually able to get.

Katie - Okay. So overall, what do you think we can draw then from this review about the state of belief in learning styles?

Duncan - I've been talking about learning styles in schools for about 10 years. And I would say that it's still the case that most times I mention it, people believe it. So I would say their figure of sort of 80 to 90%, it's probably a little bit high for the UK, but it's not far off.

The reason behind that is because, well, there's a law called Brandolini's law, which I will not quote verbatim, but it's the idea that the energy required to refute nonsense is an order of magnitude bigger than that needed to create it. If you look for example, at the first big study showing that learning styles has a big effect, that's been cited more than 600 times.

And two years later, another study came out demonstrating the original study was nonsense and there is no effect at all. That second study has only been cited 60 times. When you get an idea out there, it gains traction very easily. And to put the genie back in the bottle takes an awful lot more energy.


Add a comment