completely unrelated to the substance of this chart, but “y-axis should always start at zero” is one of those trueisms that needs to die. not all visual representation needs to be proportional, a 10 year drop in life expectancy isn’t a 14% event, it’s a 50% event, it’s war zone levels of life expectancy drop. the graph should start at somewhere around 40 years.
I very much disagree. There are extremely few cases where a y-axis shouldn’t start at zero. Anything that reasonably can be shown that way, should be, because the alternative just fucks with the monkey brain too much.
starting from zero is also fucking with your monkey brain, in this context downplaying the severity of the drop people actually experienced. all data representation is subjective. scale and origin point of charts are constantly manipulated in different scientific and statistical contexts to convey different information. obviously you have to be careful in order not to misrepresent your data, but always choosing zero can also lead to misrepresentation.
in this context downplaying the severity of the drop people actually experienced
Well… no, it isn’t downplaying it. It’s representing it perfectly. That’s quite literally the point.
When graphs aren’t zero-based, you immediately make a bad intellectual assumption as well as an ‘emotional’ reaction to the size of the graph. That immediate bad assumption poisons the way you think. We can’t eliminate that ‘emotional’ reaction, but we can (and do) standardise it and we consequently learn to gauge appropriately… only so long as people keep to that standard.
The only time you shouldn’t zero-base a graph is when the change is too small to reasonably decipher without enlarging the scale. I’ll die on that hill. If you do it for any other reason, you are just engaging in misleading emotional manipulation. There’s a place for emotional appeal, but by means of making you misunderstand reality isn’t it.
completely unrelated to the substance of this chart, but “y-axis should always start at zero” is one of those trueisms that needs to die. not all visual representation needs to be proportional, a 10 year drop in life expectancy isn’t a 14% event, it’s a 50% event, it’s war zone levels of life expectancy drop. the graph should start at somewhere around 40 years.
I very much disagree. There are extremely few cases where a y-axis shouldn’t start at zero. Anything that reasonably can be shown that way, should be, because the alternative just fucks with the monkey brain too much.
starting from zero is also fucking with your monkey brain, in this context downplaying the severity of the drop people actually experienced. all data representation is subjective. scale and origin point of charts are constantly manipulated in different scientific and statistical contexts to convey different information. obviously you have to be careful in order not to misrepresent your data, but always choosing zero can also lead to misrepresentation.
Well… no, it isn’t downplaying it. It’s representing it perfectly. That’s quite literally the point.
When graphs aren’t zero-based, you immediately make a bad intellectual assumption as well as an ‘emotional’ reaction to the size of the graph. That immediate bad assumption poisons the way you think. We can’t eliminate that ‘emotional’ reaction, but we can (and do) standardise it and we consequently learn to gauge appropriately… only so long as people keep to that standard.
The only time you shouldn’t zero-base a graph is when the change is too small to reasonably decipher without enlarging the scale. I’ll die on that hill. If you do it for any other reason, you are just engaging in misleading emotional manipulation. There’s a place for emotional appeal, but by means of making you misunderstand reality isn’t it.
I think this would work better if some comparison countries were included. Then you could drop the white space after that.