Chess Players & Autism Quotient

So, you trust Wikipedia more than a doctor's diagnosis?

Obvioulsy you are cought by the dangerous believe that psychiatrists are compareable to real doctors.
This isn't quite the case. They are hirelings of the pharma-industry with a sole purpose: Finding something so they can prescribe some drugs.

For their diagnosis they use a book. This one: http://www.who.int/classifications/icd/en/bluebook.pdf

Look at it and you'll find that any behaviour can be considered a mental illness. Also note that there are no scientifical tests like brain-scans or anything. They simply find a match to "symptoms" within the book and voila, you are declared mentally ill.

It's nothing more but a big hoax to make money.

In this country (UK), autism is considered a medical condition and often diagnosed by a doctor.

http://www.gosh.nhs.uk/medical-conditions/search-for-medical-conditions/autism/autism-information/

So was being gay.

People are different.

Please try to consider that you are not sick just because you behave differently from the average.

My score on the test was 9 btw.

i am bona fide AUTISTIC and this TEST has proved it.

I knew it!

i am happy to be different.

I hate neurotypicals anyway.

@ Xilmi:

Fair enough. However, my point stands. I would rather take the opinion of a health professional than consult Wikipedia. You may feel differently, good luck with that.

Good suggestion! I was going to experiment with plotting the AQ scores against either rating classes or just against rating, depending on how many scores there are. It'll probably be rating classes, because that should be lossy enough that there will be no way to figure out the AQ scores of anybody who commented in the thread (without stating their AQ score) based just on their publicly available rating info at chess.com, FIDE, USCF, etc.

One complication though is that there are 5 different types of ratings it's possible to submit, and they've actually all been submitted. If there are enough data points that it makes sense to show this info for each type of rating, then I will, but I suspect that some will be much more common than others (very few FICS, for example). One way to overcome that would be to try to normalize the scores so they're comparable (e.g., FIDE = USCF-50 [or whatever], ICC=USCF+150 [...]), but I wouldn't know what numbers to choose to make them roughly comparable (obviously this is not terribly accurate, but it doesn't need to be to get a rough idea). Even if using rating classes only, it's still a problem, because should somebody whose rating is ICC=1375 be considered an E player or a D player (the standard adjustment I've heard before is 150-200, so the person might be equally likely to have a USCF rating that puts them in the D class as in the E class)?

We're at 38 data points so far, so not quite halfway to a hundred, but hopefully the thread will stay active so more people will see it and take the test and submit their scores. There have been quite a few people who have taken the test and reported their score here, without submitting it online. I was going to submit for them, but I decided against it because I have no idea whether anybody was motivated to mis-state their score (perhaps to hide their identity from me, since I know all the data, or perhaps because they wanted to score a certain way or they think it's the score they "should" have gotten). So if anybody reading did state their score in the thread but didn't submit it, please do submit.

Survey's are, by nature, easy to lie on. Psychological standards dictate that while a test or survey is a viable way to garner information and compile it as source data, they are also great ways for people to submit answers that are outside the normal expected curve. It is a sad fact really; people that essentially lie and off-balance the expected results. Because of this I doubt it matters if you submitted the scores posted on here yourself. If there is a high enough number the total should still be statistically significant on some level. That would, however, be taking a chance and could easily skew the overall due to outlier variables and false claims. But such is science.

Ok. Average the ratings. It's the only fair way I can think of. Example: I only gave one rating, my uscf. If I also gave another just average the two. If I had an ICC of 1885 and a fide of 1790 and a chess.com of 1996 and a tactics rating of 1867 on here and a fics rating of 1917 then all five would average 1891. Its a bunch of figuring out to do but ask people to submit multiple ratings if they gave them. If not just go with the one they submit.

One way to deal with extremes on both ends of the scoring range (people lying or joking around) is to automatically discard the top 10% of scores AND the bottom 10% and average the rest. Or 15% KR whatever. I believe some Olympic scoring is done like that to account for outrageous, crazy numbers screwing up the average.

15- My parents have speculated that I might have Aspbergers.

I was curious how accurate a self test would be.  May sound dumb but I took 3 times over 3 days and submitted the average.  More accurate than I thought, only changed the score I reported in the forum by 1.  I didn't submit my chess.com though because it asked for standard and I'm unrated in standard.  I gave my two OTB ratings though...

All that to say I'm not holding out on you :)  Would be interesting to see some stats.

Thanks, wafflemaster! I didn't really think anybody was intentionally holding out. I just thought some people may have overlooked the links to submit. I think the links are much more difficult to miss now though. Very interesting on the taking it 3 times and getting very similar scores. I will definitely report back with more data and analysis when there are more submissions.

MrDamonSmith:I don't think there is a problem with people lying in their submissions. The data look as I expected so far, and nobody has put in an unreasonably high rating, for example. There are not many extreme AQ numbers (only 2 over 40), and the average we're seeing so far is in the ballpark of the averages that are seen for other groups that score high (and have much higher autism rates) such as mathematicians, programmers, and engineers, which I expected to be the case. You've obviously got to take this with a huge grain of salt, and it would be silly to draw any certain conclusions from this, but I'm pleasantly surprised and grateful that nobody seems to have spammed it yet. Perhaps the fact that they would get no visible reaction or acknowledgement makes it not worthwhile for a troll to submit FIDE=2864 and AQ=50.

I got a score of 29, but I don't really know what that means. What does it mean?

Interesting how the variation seems significant in the averages versus general population.

If you've ever wondered about autism or if you have it I'd read up on it with whatever general info you can find online.

If not then personally I'd just ignore it :p

Anyway it means you self-report traits associated with autism in adults.

Lol, oh great!

I got 23 points which means I'm normal enough to not be considered autistic.

I hate to tell you...

but it definitely means you have 29 points. Not necessarily much more.

Maybe you just hate people, or are very shy, or you have a nasty girlfriend. Or you've been bullied in school. Or you like numbers in general. Or something else. Or nothing. If your life is going well, just keep going and try not to label yourself too easily. It's becoming a self sustaining thing quickly. Take most ADHD people. Label them and most of them will never discover the reason why they are hyperactive.

During my childhood some people thought I am weird.

Score 32 points means I am mostly normal, doesn't it?