4 ms·
Statistical averages are an awful way of measuring anything beyond 100 - 1000 people and I am surprised to see statistics is still practiced in its classic form
by pdimitar 8y ago
Statistical averages are an awful way of measuring anything beyond 100 - 1000 people and I am surprised to see statistics is still practiced in its classic form to this day.
As another poster said, the average person world-wide has one testicle -- yay for valuable info? It basically tells you nothing.
As a programmer struggling in Eastern Europe at 38 year old still -- things are very far from what these rosy articles are trying to convince you of. And seeing what other people at 22-25 are going through... you don't want to know.
But people are quick to believe optimistic statistics when their livelihood isn't at stake. As the saying goes, it ain't true until it happens to you. Then it's suddenly oh so true.
- didibus 8y agoThings getting better in the average doesn't mean they're getting better for you. It does mean that they're improving though. Imagine how hard you'd have it had the average fallen on all these metrics instead of getting better. None of that means its all great and well. Which I assume is your point? Would be interesting to see a plot of this graph at the p0 and p100 (min/max) to get an idea of the baselines.
- pdimitar 8y agoI am saying that this claimed economical well-being is still largely clustered by countries or even regions. That I am struggling means nothing in the global scale, that is true -- but when you see almost everybody around you struggle then you have a very hard time believing these articles. My mother got a slightly bigger pension -- which changed it from 55 to 60 EUR a month. That's not a typo. Hooray for 10% improvement? I feel these stats would be much more credible if they gave you your actual % more direct buying power in terms of food, tech or leisure time based on your country -- not on a world-wide average which is non-informative. The way these stats articles are currently framed though, they are all but useless.
- didibus 8y agoI totally agree, some countries and regions could very well be on the decline, while the average is still going up. And that's a big problem. And if you're looking for insight into your own area, ya this is not really trying to be an article to allow you to make data driven decisions about your local politics and laws. It would have been nice of them to put all these into a tableau dashboard with many more dimensions so we could have played with the data some more and gathered better insight. That said, I disagree about these averages being useless. They're still a valid indicator of the world average in some specific metrics. And for those, they're showing a good trend for the better. If we were to monitor them, and discover a drastic change in one of them, it would be cause for concern. Similarly, seeing them continue to improve is cause for consolation.
- pdimitar 8y agoI would like even these condensed stats much more if they told us where do these average improvements come from -- and whom they benefit. I mean, it could just be 20_000 founders in SV getting rich because of the blockchain hype, and 50_000 Chinese businessmen making a lucky break, and 100_000 agriculture businesses going past $10M annual profits, etc. Add some more such data points and I still can't see how the world on average is better off -- it would still show a 5-figure number of businessmen being better off (less than 0.01% of world population), but not how an average single mother of two is. ...I guess I ended up trying to paraphrase another sibling comment of mine -- don't you think that if we see these stats without the extreme ends of the bell curve included, it would give us a better insight on how is the average folk doing?
- whatshisface 8y ago>Statistical averages are an awful way of measuring anything beyond 100 - 1000 people and I am surprised to see statistics is still practiced in its classic form to this day. That is only true in the sense that a person who won the lottery would say that "you loose money on average playing the lottery," is bad advice.
- pdimitar 8y agoYou kind of proved my point though. Statistically the winners are less than the losers and that also applies to the world economy at large. I had a friend who became millionaire at 25 (I was 28 at the time). So I guess if we get 10_000 of these people then suddenly my country's livelihood has improved by 2%? Not how this works. I'd like to see stats which don't include the extreme ends of the bell curve. I can bet my neck they will tell a very different story.
- mc32 8y ago>the average person world-wide has one testicle -- yay for valuable info? It basically tells you nothing. I disagree. If the average per person testicle count were .5 testicles, then you'd know something was amiss because you can expect it to average 1. (the average is slightly less than 1 by the way).
- pdimitar 8y agoWho decides what's small enough to be amiss, though? I've seen a lot of comment threads on articles like this one. It basically always goes on like this: mostly first-world country citizens comment, they mostly agree things are getting better, they collectively and wisely nod their heads and everything is good in the world. ...Except on the tons of places where it isn't. If it wasn't for me, my mom's increased pension (from 55 to 60 EUR) would only help her buy food for 1-2 more days -- and she'd be homeless. The global X% improvement means nothing to her and to millions of others. Let's face it, articles like these make it easy for people to entrench themselves even deeper inside their filter bubbles and echo chambers. I sometimes cynically ponder if that's not their original purpose.
- rpedela 8y agoYes, median would be better than mean. However what is most important is the overall trend. Child mortality trending downward worldwide is a great thing and says the world is getting better in terms of nutrition, disease, etc for children. Is there still child mortality? Yes. Will some parents have to go through the loss of a child? Yes. Is the world getting better overall in this regard? Absolutely.
- pdimitar 8y agoTo latch onto your example: child mortality is at an all-time low but the birth rates are quite low as well so it seems that both things cancel each other out. One has to wonder if for any positive 1% of such average global improvements there aren't many 5% localized worsenings that aren't shown in the global percentage changes. > Yes, median would be better than mean. That's my main complaint about this article.