Wearing ears may be associated with depression

[adjective][species] regularly informally collaborates with the International Anthropomorphic Research Project (IARP), a collection of psychologists and sociologists looking at the furry community. We are amateurs, they are professionals. We have the freedom to present anything we find interesting; they are constrained by the usual rules of academic engagement.

Nuka, a furry with a PhD in social psychology, is a long-time member of the IARP as well as an occasional contributor to [a][s]. In conversation about data (the best kind of conversation), he noted a surprising finding from his research:

“The IARP has found evidence that wearing ears (but not other accessories or fursuits) is particularly associated with depression and reduced self-esteem in furries.”

I used this interesting statistical tidbit during an [adjective][species] panel (Confuzzled 2015), as an example of a surprise hidden in the data: something you’d never expect, or think to look for. It provoked a few questions from the audience, which Nuka and I do our best to answer here.

First, a note about data in general. A statistical outlier implies, but doesn’t prove, a correlation. If we are working with a 95% confidence interval, then by definition a sample of random data will show a statistically significant correlation 5% of the time.

We are not making any statistical claims about a correlation between ears and depression. Nuka chooses his language carefully, as befits a professional scientist. He has not provided, and I have not asked for, a formal analysis of the data in question. Our discussion takes place at an intermediate step in the process of data analysis: something has been found but we don’t know what it means.

We ask the reader to accept our starting point at face value—there is evidence that wearing ears is particularly associated with depression and reduced self-esteem in furries—without any further statistical justification.

JM:

Hi Nuka! Thanks for contributing your thoughts.

I brought it up ears & depression in my [a][s] panel for a few reasons, mostly not related to the question at all – it was just something to engage the audience. But over the con several furs came up to me and asked about it, many with personal anecdotes, and it piqued my curiosity. I wondered if you’ve been able to shed any more light on it.

Here is my thinking. If ear-wearing is associated with depression, this means one of four things:

  1. A random quirk in the data.
  2. Furries who wear ears are more likely to suffer depression as a result of something related to the ears.
  3. There is a common root or driver for ear-wearing and depression among furries.
  4. Furries who suffer from depression are more likely to choose to wear ears.

I got the impression that furries with depression wore ears as an attempt to put on a brave, positive face to the world. I think they look in the mirror and see themselves with ears, and see that the perky ears have a positive effect on how they see themselves. I think that makes them feel less trapped in the endless grey present-tense of depression, because they can imagine the perky ear wearer being positive and doing happy things.

This fits into category (4) above, and suggests that ear-wearing is a small but positive step for sufferers of depression. And in the way that your research on fursona personality (and also the superhero/cosplay research by others) suggests – the act of dressing up can help people find a path to improved mental health.

Of course this is speculation, and tarred by the fact that all these people attended a furry convention and are probably on the happier/social end of the scale when it comes to current or former sufferers of depression.

A further interesting thought was brought up by a few people: does incidence of depression (or related results) change over time for ear-wearers? That is, is there evidence that ear-wearing is a sign of improvement? I know you’re still early in your longitudinal survey, and may not be asking the relevant questions, but I figured it was worth asking.

Nuka:

JM is absolutely correct that we’ve found a statistically significant relationship between wearing ears and reduced self-esteem in a relatively large sample of furries. JM is also correct in stating that caution is prudent when trying to interpret these sorts of results, for a few reasons, related to what’s already been mentioned above.

The observed relationship occurred without a hypothesis or theory to back it up. Theories are of incredible importance to researchers – they help us to avoid what we call “fishing expeditions”. What’s a fishing expedition? Imagine I had a dataset with HUNDREDS of variables in it. As JM correctly stated above, statistics are based on confidence intervals and probability, meaning that a percentage of our observed significant findings will be simply due to random noise in the data. As such, if I pore over hundreds or thousands of variables, eventually, I’m going to find SOMETHING that reaches a level of statistical significance. Is this a “real” finding, or just noise in the data? How do we tell?

Researchers have two tools at their disposal: theory and replication.

Theory: The difference between a genuine finding and random noise in a fishing expedition can come down to having a theory. If the researcher has a theory about how two variables interact, they create an “a priori” (before the fact) hypothesis – this means that the researcher is predicting a specific relationship between these two variables. This avoids the problem of the fishing expedition: if you find a relationship between those two particular variables, it’s FAR less likely to be random noise than if you were to trudge through hundreds of variables and stumble upon it that way. Theories are usually grounded in prior empirical work, and give researchers good reason to expect a relationship between two variables, making researchers far more confident in a finding if they “called it” beforehand.

Replication: A single study is seldom enough to close the book on a subject. A good researcher likes to be sure about what they’ve found before they consider it fact. As such, if they DO stumble upon something interesting during a fishing expedition through their data, the first thing on their mind is “will this replicate?” In other words, if I ran the study again, would I find the same result? Random noise can be surprisingly predictable. One such predictable trait is something called “regression to the mean” – an abnormally high or low result that’s due to chance will, over time, become more normal. In other words, if a finding was just a fluke the first time, you really shouldn’t see it the second time. In other words, if researchers find something during a fishing expedition and they find it AGAIN when they replicate (re-do) the study, researchers can be fairly confident it’s not just a fluke, but a genuine relationship. You’ll notice this essentially means forming an a priori hypothesis like the one mentioned above for the second round of data collection.

What does all of this statistical gobbledygook have to do with furry ears and self-esteem? Well, a lot, actually! We had no theoretical reason to predict that wearing furry ears would relate to self-esteem. Of course, we can come up with plenty of reasons to explain why these two things might be related post-hoc (after-the-fact), but it’s entirely possible that the relationship was just a fluke. Of course, now that we’ve observed this potentially-interesting relationship, the first thing on our minds is replication – can we find it again, now that we’re looking for it? This means collecting more data on the subject. If, looking at the data, we find the same relationship, then we can be reasonably confident there’s a relationship between ear-wearing and reduced self-esteem.

It’s also worth asking why these two variables might be related: why should wearing ears be related to self-esteem? If we assume that it wasn’t a statistical fluke (which remains to be seen), we see that JM has outlined three possibilities:

a) Ear-wearing causes low self-esteem
b) Low self-esteem people wear ears
c) Some other variable causes both low self-esteem and ear-wearing

Without “temporal” data—data taken at more than one point in time, it’s impossible to establish causal direction—impossible to say which variable “causes” the other to happen. That said, there are rational reasons to expect / predict all of these explanations to be true:

a) Ear-wearing causes low self-esteem: this is perhaps the weakest of the three alternatives, in part because a myriad of variables contributes to a person’s self-esteem, and it seems unlikely that any one behavior would significantly push a person’s self-esteem around. However, it may be the case that owning / wearing furry ears may be an indirect measure of how “out” a person is as a furry (how much they let other people know they’re a furry), and insofar as they’re “out” in a hostile environment (e.g., friends / family that look down upon furries), this may be bad for their self-esteem. However, for this to be true, you would have to ignore other data showing that self-disclosure, in general, is associated with better self-esteem. It’s also possible that, within the context of the furry fandom, wearing ears at conventions may make some furries feel disappointed that they can’t more fully express their fursona (e.g., having a fursuit). This, of course, makes a number of assumptions (e.g., people who wear ears really want a fursuit, and are really upset they don’t have one) that, to date, we have no data to support.

b) Low self-esteem people wear ears: there may be something to this prediction, as JM has already speculated. Past data has shown that a furry’s relationship to their fursona can help us to predict their self-esteem – namely, how “close” they feel to their fursona. Furries who feel a lot of overlap between themselves and their fursonas tend to be better off self-esteem wise, while furries who feel like there is a world of difference between themselves and their fursonas tend to have lower self-esteem. Given this, wearing furry ears may be an indicator that furries felt a lot of “distance” between themselves and their fursona (e.g., I need to wear these ears to feel more like my fursona). There is, to date, no data on this. The argument would be that furries wearing ears would be reminded of just how much difference there is between themselves and their fursona (e.g., looking in a mirror), which may contribute to negative self-esteem. Of course, the issue gets doubly messy when you realize that significant distance between oneself and their fursona may, itself, be caused by low self-esteem (my life sucks, so I’m creating a fursona who’s totally not like me to escape how much it sucks).

c) Some third variable is causing both low self-esteem and ear-wearing. This, I would argue, is perhaps the strongest candidate for explaining this potential relationship. There are numerous third variables that may explain BOTH ear-wearing and low self-esteem. For example – ears are relatively cheap to obtain (e.g., relative to, say, a fursuit, or even a tail). As such, it may be the case that wearing ears (as opposed to other pieces of furry swag) may indicate having less money at one’s disposal. Low socio-economic status may thusly predict BOTH purchasing of ears (as opposed to more expensive furry gear) AND lower self-esteem (people don’t feel good about themselves if they feel poor). Another possible third variable: perhaps ears are one of the first purchases made by newer, younger members of the furry fandom (anecdotally, a pair of ears were the first furry-themed purchase I made upon entering the fandom). It may thus be the case that being a young, new furry in the fandom may predict both buying a pair of ears AND having low self-esteem (questioning whether you belong in the fandom, feeling like a “noob” doesn’t feel very good, possibly still struggling with “coming out” to friends and family).

In summary, while I share JM’s enthusiasm for this neat, unexpected finding, I remain reserved about the findings until they’ve been replicated. At that point, once we’ve established that it’s a consistent relationship and not an anomaly, the theory-building begins: why ears and not other furry accessories? What is the causal direction? Might there be third variables? What does this tell us about the motivations underlying furry participation in the fandom? All questions for future studies!


Dr Courtney “Nuka” Plante is co-founder of the International Anthropomorphic Research Project.

About JM

JM is a horse-of-all-trades who was introduced to furry in his native Australia by the excellent group known collectively as the Perthfurs. JM now helps run [adjective][species] from London, where he is most commonly spotted holding a pint and talking nonsense.

Before posting a comment, please read our Code of Conduct

5 thoughts on “Wearing ears may be associated with depression

  1. (This is why I like this site, people Sciencing stuff!)

    As another anecdote for the pile: I’ve been wearing my rabbit ears almost non-stop for about a year. At home, at the office, everywhere. (A bit excessive maybe, but it’s a good hat!) Before that, I wore them a bunch (though not constantly) in college. The years in between when I didn’t wear them much were miserable, but there are way too many uncontrollable variables to really draw anything from that.

    It’d definitely be interesting to piece through the sources of low self-esteem. I know mine is largely derived from being transgender, which provides a whole host of reasons to lack confidence. I might be more like option B, using the ears to feel more like my (female) fursona. But that’s very rare compared to low economic status or newness, so perhaps I’m an outlier.

    There are SO MANY variables! And now I’m really curious, so good work Nuka!

  2. This rather reminds of a health-study I read about circa 1992; its long-anticipated release generated major nationwide headlines.This was at the time supposedly one of the most all-encompassing, rigid, and expensive long-term medical studies ever attempted. The goal was to take thousands of disparate lifestyle factors and cross-index them against the long-term health and mortality of the subjects, all of whom lived in New York City. The much-anticipated results…

    …revealed pretty much nothing new or unexpected, save that there was an amazingly strong statistical correlation between eating at least two slices of pepperoni pizza a week and substantially increased longevity.

    You can’t make this stuff up…

    1. Yup – it’s one of the problems with the “fishing expeditions” that I mentioned above. When you throw dozens or even hundreds of variables into an analysis, you’re going to find some that are bizarrely related just by chance. Of course, there’s nothing wrong with having a gander at your data for purely exploratory reasons IF you acknowledge that’s just what you’re doing. But it’s alarmingly common for researchers to peer through their variables, find one out of the bunch and say “ha, see, I told you so!” and ignore the fifty non-significant findings. It’s why replication is so important!

      Also, I found your anecdote about the pepperoni pizza correlation amusing. If you like that sorta of thing, you might really enjoy this site: http://tylervigen.com/spurious-correlations

Leave a Reply

Your email address will not be published. Required fields are marked *