[bad_alloc]

How did they ask people to locate Ukraine? Some of these points look like somebody accidentally clicked on a random spot of a map. For example, the dot below New Zealand is in the middle of the ocean (approx. -55.000000,176.000000). Some even located it within the US (Florida seems to be a popular guess). Have they found really misinformed people or is this noise in the data?

[Nacraile]

I would bet that the noise comes from people who have no idea just closing their eyes and clicking at random.

Clearly the map wasn't coloured to differentiate bodies of water, considering the clusters in the Black and Caspian seas. (But you'd think people could at least identify the major oceans...)

[AJ007]

This would correlate with 'I have no idea what I'm doing'

"In general, younger Americans tended to provide more accurate responses than their older counterparts: 27 percent of 18-24 year olds correctly identified Ukraine, compared with 14 percent of 65+ year-olds."

65+ year olds spent a big chunk of their lives in the cold war. Presumably they would know more about the geographical breakup of the Soviet Union than 18-24 year olds who didn't even exist when the Berlin Wall fell.

From a data analysis standpoint, I would say discard the outliers that look like mis-clicks. Based on the groupings you still have people that identified Alaska, Greenland, Canada, India South America, and so on. Irregardless of the colorings or map size, those should not have been chosen.

On another note, a given countries foreign policy always makes a hell of a lot more sense when placed in context on a map. I'm glad I read a lot of National Geographic as a child.

[rcthompson]

No, people who were in primary school before the breakup of the Soviet Union would probably have never seen the Ukraine on a map, since all they would see is a big red area labeled "Soviet Union". I doubt Soviet internal geography & politics were taught regularly. Younger people may not even know that the Soviet Union ever existed, but they have a much better chance of seeing Ukraine on a map in school.

[dragonwriter]

> No, people who were in primary school before the breakup of the Soviet Union would probably have never seen the Ukraine on a map, since all they would see is a big red area labeled "Soviet Union".

Many of the maps I remember seeing (in school and otherwise) as a kid in the 1980s labelled the constituent republics of the USSR.

[loqi]

Yup. No Ukraine on my globe as a kid in the 80s, just "U.S.S.R.".

[huherto]

Perhaps younger Americans took geography in school after the break out of the Soviet Union. I only know where Ukraine is because I recently looked at the map when this crisis came up.

[cafard]

It may come from growing up in a part of the country where there were a lot of Central European DPs, but I'm pretty sure that as a kid I knew where the Ukraine was.

[insuffi]

Believe it or not, but there are many real people who point to UK, when asked where Finland is.

The point is, if you know where Russia is, you likely know where Ukraine is. The clicking at random is by people who have no idea about either.

EDIT: Sorry. I meant that as a reply to your comment regarding legitimacy of the map.

[Crito]

> The point is, if you know where Russia is, you likely know where Ukraine is.

I don't know about that. There are ~15 (depending on what you are counting) countries that border Russia. And while I am fairly confident that I could label them all, Russia would certainly be the easiest to label. It's the big one; it's hard to miss.

I'm the product of American public schools for what its worth. I learned the countries back in middle school, but learning the locations of countries not in North or South America was extra credit for my class (as was learnin the state capitals, and country capitals).

[waterlesscloud]

Considering there are notable clusters in Kazakhstan and Uzbekistan, you're probably right. Those people probably located Russia, knew Ukraine is "under" Russia, so picked a couple of countries under Russia.

Incidentally, I suspect those large clusters in the central asian republics account for a large amount of the error in the median response. The lone clicks in New Zealand or whatever don't count as much as big clusters do, and those clusters are pretty near the median error of 1800 miles off.

[sentenza]

It is human not to be an expert on everything, so I'd say the people that clicked Kazachstan, Belarus, the Black Sea and the like should be forgiven.

Also, there are certainly a number (below 5%) of clicks that resulted from accident or interface understanding problems. Those could account for some of the most outlandish dots.

But then there are a lot of people that selected India. That's kind of disturbing.

[insuffi]

Fair point about the 15 countries bordering Russia. I must have not been clear. What I was trying to say is, the least you can do is point towards Russia, if you don't know where Ukraine is.

[Crito]

Ah yes, that definitely makes sense. I misunderstood you.

[dmckeon]

There seems to be a sharp drop-off line running north/south just west of Mongolia & east of Novosibirsk.

I wonder if that is from a side-scrolling artifact or drag/drop interface in the sampling (wild speculation there) or from an unconscious pruning algorithm, among people who are unaware of Ukraine's proximity to the Black Sea, roughly: "hmmh, Ukraine, medium-to-small country, former Soviet, I'll look further east, oops, there's China, better stop here.

Some more informative maps include languages and hydrocarbon pipelines:

http://news.bbcimg.co.uk/media/images/73094000/gif/_73094671...

http://www.zerohedge.com/news/2014-02-20/ukraine-situation-e...

[agapos]

"There seems to be a sharp drop-off line running north/south just west of Mongolia & east of Novosibirsk."

That line is too sharp for a 'Hey, that's China!' stop. Besides, it's unlikely that many of them could locate China correctly, due to the generic lack of geological, geographical and topographic studies in US schools.

I think they got an America (as in: continent) centred map, something like this [1], with the small difference of being cut somewhere at the eastern parts of India, instead of the eastern borders of Iran.

[1] https://upload.wikimedia.org/wikipedia/commons/8/8b/White_Wo...

[Florin_Andrei]

It's interesting how many wrong clicks are piled up in the general region north of India / east of China. In other words, where any marginally informed person would think is an area of unrest and trouble.

Kazakhstan seems to be an attractor, probably because it's so big.

"Hmmm, so Afghanistan is in that general area, and it's a troubled country, so therefore Ukraine must be thereabouts also."

[pvnick]

I imagine people think "crimean peninsula," look at Florida and go "well there's a peninsula"

[jasonwatkinspdx]

I think the real outlier dots are "I give up, so I'll just click whatever."

[azinman2]

Not surprising to me. Many Americans can't find America on a global map, including those in the military (confirmed by my shocked army friends).

[_ea1k]

Since you said "Many" (even 1% could be many by some definitions) it is hard to dispute your assertion.

Nevertheless, I think the # of Americans that can point out the US on a world map is well over 90%. I don't see any definitive data to the contrary.

[insuffi]

I think you might want to rethink that assumption. No definitive data, to be fair, but this: http://www.cnn.com/2006/EDUCATION/05/02/geog.test/

"Inside the United States, "half or fewer of young men and women 18-24 can identify the states of New York or Ohio on a map [50 percent and 43 percent, respectively]," the study said."

I don't think it's that far fetched.

[_ea1k]

The article that you referenced says that 7/10 can find China on a map. I'd be willing to bet that a far larger number can find the US on a map.

I'm pretty sure the poster that I was responding to was taken in by a fake study. The overwhelming majority of Americans can find the US on a map.

[azinman2]

37% can't according to this Gallup poll [1] from 2008. I've seen other polls with similar or worse results.

[1] http://www.huffingtonpost.com/steven-shehori/poll-37-of-amer...