6 ms·

That map is a really bad visualization. They're showing a single numeric variable with a limited scale, so all they need to express graphically is a level betwe

by Camillo 13y ago

That map is a really bad visualization. They're showing a single numeric variable with a limited scale, so all they need to express graphically is a level between "empty" and "full". They should have mapped it to a single axis that intuitively corresponds to that notion: for example, a simple scale between white and red would have been perfect. But instead, they decided to have a scale between red and blue, which goes through purple if it's an admixture of the two colors. But then they would just have had a few colors with more or less the sale lightness and saturation, but different hues, and there's really no intuitive reason why purple should be perceived as "more" or "less" than blue. So instead they made a three-point scale going from red to white and then to blue (look at the pink and light blue colors next to the center to see that this is what they did); except that now white looks like "empty", and redder and bluer colors look increasingle "full", except in opposite directions: the only situation where this makes sense is when you have relative numbers, and white is zero, red is positive and blue is negative. But then it would look like Iran has lots of depression, Brazil has no depression, and Australia has... negative depression? They could tell that something looked wrong, but what? "A-ha! Of course!" - thought the graphic designer - "a mix of red and blue is purple, not white!" So he replaced the white in the middle cell with purple, leaving the neighboring pinks and light blues unchanged. So now the scale goes "red, red but a little whiter, white but a little redder, PURPLE, white but a little bluer, blue but a little whiter, blue", and nothing makes sense. Good job!

msutherl 13y ago
I think I agree, but the existing scale does have one redeeming property which is that it makes very clear which countries have the most and which have the lest depression. Both the red and the blue stand out against the middling regions, which would not be the case if white was the bottom of the scale.
hajderr 13y ago
What do you think is the reason for this? A bunch of propaganda that we're well off here in the 'safe' part of the world? Love this comment from the article "I'm sure we've got high quality, nationally representative data out of the Congo."
- roel_v 13y ago
  Your paranoia is not only annoying, but also quite insulting to the researchers involved. Contrary to your perceived 'the Man is holding me down', not everybody is out to 'deceive' you. The reason is likely purely functional, as I explained in a comment above, and I'd even agree with the choices of colors of the authors - they show more clearly than the proposed white-red where in the world there is more depression. Furthermore, if you don't believe this particular studies' findings of depression rates, I would invite you to look into the decades of research into suicide rates, depression or just the flow of immigrants from the countries in the red in the map to those in the blue as a proxy for overall happiness and perceived life opportunities, and to draw your own conclusions.
  - hajderr 13y ago
    Sure, just check https://en.wikipedia.org/wiki/List_of_countries_by_suicide_rate https://en.wikipedia.org/wiki/List_of_countries_by_suicide_r... Where you can find a good mix of 'blue' countries such as US, Japan, India, Russia, Norway, China and Sweden. So it doesn't tell me THAT much unfortunately.
coldarchon 13y ago
>showing a single numeric variable // Actually I do not understand why they did not think of incest. In the mentioned countries the chance of a young girl getting married to her elder cousin is most of the times almost one third. They had a statistic in England that showed some immigrant groups were 11 times more likely to have a mental defect because of incest. I do not understand why science has to be political correct.
roel_v 13y ago
I agree with you that replacing white with purple is not consistent, but I don't agree with your conclusion. Depression is not something (as far as the current state of knowledge goes) that can be completely eradicated. The study was about comparing depression rates globally. In that context, taking the average (or mean, but let's assume it's distributed normally) depression rate as the 'baseline' (i.e., the middle) is not that strange. This way, the map shows not absolute depression rates, but rather the position of each country relative to the baseline value. Which is, I think, what people are most interested in - if you tell me 'country xyz has a depression rate of 2%', most people have no idea. The whole study is not about absolute depression rates, but about rates relative to other countries. I'd even go as far as defending the purple. Remember, cartography is not about choosing color according to a rigid logical scheme, it's about conveying information. Now, I think these colors convey the message of which areas in the world have the highest depression rates, and which ones the lowest, quite well. If white would have been left in as the color for countries with baseline values, those white blobs would have been distracting and would have looked as if no data was available. Instead, what we have is a map that shows in a visually uninterrupted way the rate of depression relative to the baseline, with intensity increasing where you go further away from the baseline. You proposal for using white to red would be OK if it were about showing absolute depression rates, but that's a far less interesting way of looking at the data.
- jader201 13y ago
  Your defense almost makes sense, but it still leaves out the fact that it grades from red to white, and blue to white, until it hits the middle -- at which point it suddenly turns purple. To go with your defense, it should grade from red to purple between the highest point and the mean, and blue to purple between the lowest point and the mean.
  - roel_v 13y ago
    Well actually my whole second paragraph is about the purple :) The point of going from blue to white is to to visually convey that more pale colors mean 'closer to the baseline'. I.e., pale blue is 'slightly better than baseline', pale red 'slightly worse than baseline'. If you were to blend from blue to purple, that wouldn't work any more. But if you leave in the white in the middle, it looks as if there are 'no data' spots. Hence, use variation in intensities for non-baseline, different color for baseline. Cartography is not about conceptual consistency, it's about conveying information. I challenge you to make a map that conveys the message better than the choices made in the map in the OP (apart from subjective wah wah like 'blue is democrats' as was laughably brought up elsewhere in this thread).

by Camillo 13y ago

msutherl 13y ago

I think I agree, but the existing scale does have one redeeming property which is that it makes very clear which countries have the most and which have the lest depression. Both the red and the blue stand out against the middling regions, which would not be the case if white was the bottom of the scale.

hajderr 13y ago

What do you think is the reason for this? A bunch of propaganda that we're well off here in the 'safe' part of the world? Love this comment from the article "I'm sure we've got high quality, nationally representative data out of the Congo."

roel_v 13y ago
Your paranoia is not only annoying, but also quite insulting to the researchers involved. Contrary to your perceived 'the Man is holding me down', not everybody is out to 'deceive' you. The reason is likely purely functional, as I explained in a comment above, and I'd even agree with the choices of colors of the authors - they show more clearly than the proposed white-red where in the world there is more depression. Furthermore, if you don't believe this particular studies' findings of depression rates, I would invite you to look into the decades of research into suicide rates, depression or just the flow of immigrants from the countries in the red in the map to those in the blue as a proxy for overall happiness and perceived life opportunities, and to draw your own conclusions.
- hajderr 13y ago
  Sure, just check https://en.wikipedia.org/wiki/List_of_countries_by_suicide_rate https://en.wikipedia.org/wiki/List_of_countries_by_suicide_r... Where you can find a good mix of 'blue' countries such as US, Japan, India, Russia, Norway, China and Sweden. So it doesn't tell me THAT much unfortunately.

coldarchon 13y ago

>showing a single numeric variable // Actually I do not understand why they did not think of incest. In the mentioned countries the chance of a young girl getting married to her elder cousin is most of the times almost one third. They had a statistic in England that showed some immigrant groups were 11 times more likely to have a mental defect because of incest. I do not understand why science has to be political correct.

roel_v 13y ago

I agree with you that replacing white with purple is not consistent, but I don't agree with your conclusion. Depression is not something (as far as the current state of knowledge goes) that can be completely eradicated. The study was about comparing depression rates globally. In that context, taking the average (or mean, but let's assume it's distributed normally) depression rate as the 'baseline' (i.e., the middle) is not that strange. This way, the map shows not absolute depression rates, but rather the position of each country relative to the baseline value. Which is, I think, what people are most interested in - if you tell me 'country xyz has a depression rate of 2%', most people have no idea. The whole study is not about absolute depression rates, but about rates relative to other countries. I'd even go as far as defending the purple. Remember, cartography is not about choosing color according to a rigid logical scheme, it's about conveying information. Now, I think these colors convey the message of which areas in the world have the highest depression rates, and which ones the lowest, quite well. If white would have been left in as the color for countries with baseline values, those white blobs would have been distracting and would have looked as if no data was available. Instead, what we have is a map that shows in a visually uninterrupted way the rate of depression relative to the baseline, with intensity increasing where you go further away from the baseline. You proposal for using white to red would be OK if it were about showing absolute depression rates, but that's a far less interesting way of looking at the data.

jader201 13y ago
Your defense almost makes sense, but it still leaves out the fact that it grades from red to white, and blue to white, until it hits the middle -- at which point it suddenly turns purple. To go with your defense, it should grade from red to purple between the highest point and the mean, and blue to purple between the lowest point and the mean.
- roel_v 13y ago
  Well actually my whole second paragraph is about the purple :) The point of going from blue to white is to to visually convey that more pale colors mean 'closer to the baseline'. I.e., pale blue is 'slightly better than baseline', pale red 'slightly worse than baseline'. If you were to blend from blue to purple, that wouldn't work any more. But if you leave in the white in the middle, it looks as if there are 'no data' spots. Hence, use variation in intensities for non-baseline, different color for baseline. Cartography is not about conceptual consistency, it's about conveying information. I challenge you to make a map that conveys the message better than the choices made in the map in the OP (apart from subjective wah wah like 'blue is democrats' as was laughably brought up elsewhere in this thread).