The eyes have it.

[andrewducker]

Interesting test here which tests your ability to tell people's emotions from looking at their eyes. It's used as one of a series of tests for Aspergers and the like.
[Poll #1642444]

My score was not what I expected.

Comments

[andrewducker]

Some of them I got by trying out each of the emotions in turn and seeing where my eyes ended up. I'm not sure if that counts as cheating...

[channelpenguin.livejournal.com]

I did MUCH better than I expected. With the ones I fluffed I tended to think that none of the labels applied, that *they* had got it slightly wrong.

I'd have been concerned mind you, if I'd got full marks....

[meaningrequired.livejournal.com]

Yes, I agree, I thought the labels were wrong too :)

[princealbert.livejournal.com]

In those situations I went with my second choice, and it was always the first choice.

[cartesiandaemon.livejournal.com]

Wow. I got 31 as well. I honestly did not expect to do at all well.

It felt strange. I started off basically guessing -- except that almost all of my guesses happened to be right.

A third of the way through, I thought to try to imagine the whole face making the expressions described and see which seemed right, and that seemed to help a lot. I probably mimed making that face, but not deliberately. I started to consciously notice the position of the face and position of the eyes, which correlated with what sounded vaguely plausible for those emotions.

[cartesiandaemon.livejournal.com]

I also hypothesises that perhaps I'm bad at emotions but very good at spotting the "right" answer from sets of four, and that I should try taking it again _without_ looking at the eyes. Or that the test is actually testing something else. But I don't think it is; I imagine it is just what it says.

(I also noticed what Elmyra says)

[naath.livejournal.com]

12

Which is only slightly better than chance.

Perhaps if I'd tried the "trying out the emotions" trick I would have done better but I think that's cheating because it relies on having a short-list to pick from.

I'm not diagnosed with Aspergers or Autism.

[tamaranth.livejournal.com]

my score was also not what I expected: I don't tend to think of myself as good at reading emotion.

a couple of observations:
- I'm sure I got the same picture twice, illustrating 'fantasising' and 'thoughtful'
- I based my judgements on eye direction -- looking away, looking toward -- which makes me gloomy as eye contact is something I sometimes find difficult, and this test implies that I am giving off the wrong signals. (Not that this is a surprise, I do that a lot.)

[andrewducker]

I know they reused the same person with different expressions a few times. I didn't see the same expression twice, but I wasn't paying _that_ much attention.

[channelpenguin.livejournal.com]

I think they reused pictures too.

[marrog.livejournal.com]

They didn't reuse pictures but they did reuse the same eyes once or twice.

[channelpenguin.livejournal.com]

ah, then I got the same woman twice with very similar expressions.

[drainboy.livejournal.com]

For I AM Tim Roth.

You may think I got three wrong. No, they got three wrong.

[channelpenguin.livejournal.com]

:-) sociopaths would score very highly, I understand...

[marrog.livejournal.com]

Er...

[andrewducker]

The research I've read shows that sociopaths are frequently very adept people persons. They're just _bastards_ along with it. See, for instance, the thing I linked to the other day about people who rated highly for sociopathy being able to spot "victims" very easily.

[marrog.livejournal.com]

Lol. Oh, I know that, I wasn't questioning whether channelpenguin was right. I was 'er'-ing in light of our conversation the other day and my score.

[channelpenguin.livejournal.com]

:-) yeah, I was *relieved* to get some wrong

[andrewducker]

Aaaah! Gotcha.

I have no commentary :->

[luckylove.livejournal.com]

Oh shit. I got 35.

[channelpenguin.livejournal.com]

Also, who decides (and how) what the 'correct' emotion is? If they get people to act, they could miss the mark. If it's a consensus of many folks assesment of a picture then who is to say that that is right either? Even if they take tons of photos of people and then ask the people making the expression they may not know themselves by the time they are asked.

[andrewducker]

It doesn't really matter, to be honest. If you have a test that differentiates between people who are X and people who are Y then what you're testing on isn't that important. It could be that aspies always get something right while neurotypical people get it wrong - the test would still be predictive.

[channelpenguin.livejournal.com]

I didn't realise it was a specific test for Asperger's. It's interesting since I have been so often accused of not being able to read people right. Seems I read them just fine...just do not always react in the prescribed/expected way.

[rosamicula.livejournal.com]

I agree - I'd be interested to see which ones we got wrong and if they were the same.

[drainboy.livejournal.com]

I think a few were ambiguous. Certainly I got to a few of them, guessed what they were and that guess didn't appear on the list so I guessed something else. If the differences had been more subtle in the options then I think I would have gotten a lot more wrong.

[marrog.livejournal.com]

I only got one wrong because I second-guessed myself. But lo I am the anti-aspie, as we have established in the past.

[marrog.livejournal.com]

It's not much of a sample set yet but it's looking as though either their 'average' is a little off or you have slightly above-average friends at this sort of game.

[pigwotflies.livejournal.com]

I got 29 and was surprised by how many I got right, especially as I don't think of myself as someone who can read other people easily.

[momentsmusicaux.livejournal.com]

The test doesn't quite work for me. I got the first half dozen right and then started feeling mild panic about getting subsequent ones right too, and then making mistakes due to overthinking. Argh.

But yeah, my own problems aside, I can pretty much ace this.

[sageautumn.livejournal.com]

I did that also... totally overthinking.

I ended up with 31, but I think it should've been 33.

[stillcarl.livejournal.com]

I started out doing poorly, then about halfways through I started to get a high percentage right.

Incidentally, it resized my browser, so I wonder if that was important. And if not, then why did they do it? As all it did in my case was annoy me - and then make me resize it back again before I started. If that's not what they wanted, then a failed quiz, I reckon!

[andrewducker]

Yeah - I hate things that resize my windows.

[ladysisyphus.livejournal.com]

Ugh, there are few things that will make me more instantly pissy than a website that resizes my browser window without my consent.

[stillcarl.livejournal.com]

And I'm wondering if that's intentional. As in getting people annoyed before they start the test. Was yours resized? As I'm picking it wouldn't work with some browsers.

[ladysisyphus.livejournal.com]

Mine wasn't resized at first, but I browse in Firefox with scripts set to default to off. When I allowed the domain, though? Suddenly: tiny window.

[momentsmusicaux.livejournal.com]

Also -- wouldn't it be fun if they chucked one in of the Mona Lisa's eyes?

[ladysisyphus.livejournal.com]

31, though it should've been 33, I second-guessed myself twice.

I find it interesting how the test's images were gendered on some of the options, though.

[andrewducker]

I didn't notice any flirty men.

[marrog.livejournal.com]

Or any fantasizing or desirous men. I noticed this too.

[andrewducker]

See Elmyra's comment further down...

Edit:Aah, you already did!

[channelpenguin.livejournal.com]

Flirtatious was THE easiest one to spot.

yeah - I kept waiting for the flirty men...

[moosedevil.livejournal.com]

Wow, I got 33 correct, which is much higher than I would have expected to. Interesting test Andy, thanks for posting :)

[drainboy.livejournal.com]

33 is _exactly_ the right amount.

[davesangel.livejournal.com]

Really? Cool, that's what I got.

[elmyra.livejournal.com]

I got 29 and serious blood pressure and anger management issues.

Not a single non-white face. While there was a mix of young and old men, tehre wasn't a single wrinkle around any of the female eyes, and all the eye brows were meticulously plucked. While men were defiant and decisive and hostlie, women were fantasising (twice!) and flirtatious, and frankly mostly just bland.

This experiment set-up is deeply flawed.

(Sorry, this is not directed at you, but I just clicked through 36 images of 1950s stereotype and it's made me unexpectedly and incredibly angry, so I thought I'd share the pain. ;-)

[marrog.livejournal.com]

I had a similar irritation but I was mostly pleased with myself for my score so I didn't bother to comment. And now I feel guilty.

[andrewducker]

Yeah, I noticed that the images were rather stereotypical.

If it was on the website of the people that produced it I'd be tempted to email them.

[autodidactic.livejournal.com]

I was irritated that so many of the female emotions were based on "desire". I went out of my way not to objectify women, and my score suffered.

[kashandara.livejournal.com]

This kind of worried me too. On a couple of them I deliberately aimed for thoughful/considering type answers and was wrong on the grounds that the makers seem to think women lookingn thoughtful are thinking of what they'd like to do to them/someone...

[ashfae.livejournal.com]

This annoyed me a great deal too.

[meaningrequired.livejournal.com]

The only thing is that it gives you time to consider, and I'd like to know what the responses would be like if you were only allowed say 1 second to see it.

[marrog.livejournal.com]

The only one I looked at for more than a glance, click, glance, click, was the one I got wrong. From that I would guess that people's answers would probably be better.

[kashandara.livejournal.com]

I got a couple of wrong ones from overthinking it, going back to look at the picture zoomed in and changing my mind, so I agree that perhaps this is one of those tests where forcing folks to go on instinct might improve responses for 'normal' folks. Whether the difference would be accentuated in the autistic would be more interesting to me than the current data the test provides seems to be...

[meaningrequired.livejournal.com]

Theres something about higher order processing or something, but I can't remember exactly what it is, and it has to do with looking at faces.... unfort not my area of research.

[usmu.livejournal.com]

I scored an average 27. Funny thing was I had no trouble picking up the bad emotions. I had considerately more trouble with the interested and flirty type ones. I guess that's why I've been single all these years. :)

[woodpijn.livejournal.com]

I got 29 and was surprised how many I got right. Although several of them were kind of guesses - e.g. my reaction to the picture was "They look fed-up" and then that wasn't an option, so I compared each option to the picture and picked one of the options that wasn't a complete mismatch. (I thought a lot of them looked fed-up.)

I couldn't tell you what it is in any image that makes it look like the emotion it's supposed to look like.

I also noticed the way all the sex-related ones were women, and I'm not normally the sort of person who notices or cares about that kind of thing, so it must have been quite an extreme bias.

[cheekbones3.livejournal.com]

I got 21, and I thought I was doing quite well!

[rhythmaning.livejournal.com]

Mine neither!

I got 22. I did a lot better on "positive" behaviours than "negative". And - I think - better on reading women's eyes than men's.

[rhythmaning.livejournal.com]

Sorry, I mean positive emotions, obviously...!

Reading through the comments, I too thought that some of their descriptions were wrong.

[flickgc.livejournal.com]

Crappy flash too slow to actually complete it, but commits basic Boys Are Autiform Fail: girls are generically better at that sort of thing.

[andrewducker]

Autiform??

[flickgc.livejournal.com]

Sorry, I picked that up from (iirc) Elisabeth Moon: a disorder on the Aspergers-Autism spectrum.

[andrewducker]

Slow? Was pretty instant on my computer.

And while the majority of the people scoring at the very high end are female, there are some boys there too - my brother scores 33, and two others score 32.

[zz]

I got 31, am male, and have been diagnosed with asperger's. :>

I am amused however that the ones I got wrong are all where I assumed the person was more unhappy/hostile than they supposedly were.

[draconid.livejournal.com]

Uhm... yeah, I'm off the scale. I only got 14! Although a lot of my second 'ooops, I should have picked option X, too late' choices were right.

[heyokish.livejournal.com]

Damn. Very bad at this indeed. A sad 20. And that was all negative ones. I'm not going to think about that very much or I'll read far far too much into it.

[darkoshi]

The format of the test annoyed me. They should have made the photos larger instead of so small that you have to zoom them. And why did they make it so that the zoomed photos cover the answers, so that you have to go back and forth between zoomed and unzoomed?

I think it is problematic interpreting facial expressions, without knowing what the particular faces look like in a resting state. For example, does the face always have certain wrinkles, or is the person actively making the wrinkles? Does the person normally have squinty eyes, or are they actively squinting?

Also, I'm skeptical of how they came up with the "correct" answers. I doubt they took photos of people and then asked them what they were feeling. The answers seem to be based on some kind of stereotypes... so the test determines how familiar you are with those stereotypes.

[momentsmusicaux.livejournal.com]

Oh and also: http://www.emotioneric.com/

[in-thy-bounty.livejournal.com]

Interesting...
I always felt that I was fairly good at reading people. I seemed to struggle with apologetic though, clearly not enough people have been apologising to me.