In general, statistics are often widely misunderstood. I don't consider myself a statistician, but I do a fair amount of statistics to help make sense out of data. The number of misconceptions...
In general, statistics are often widely misunderstood. I don't consider myself a statistician, but I do a fair amount of statistics to help make sense out of data. The number of misconceptions I've heard from others at work over even the most simple of statistics (control charts, scatter plots, standard deviations) can be at times concerning, especially since I work in a field where most clinicians should have at least a cursory understanding in order to interpret findings from medical papers. I recently delivered some statistics which were a bit more involved (a series of logistic regressions) and spent literally hours talking this over with the team it was delivered to over the course of weeks. It took a lot of explaining for some concepts to sink in, and after reading this article I'm starting to see additional connections with the way the data itself is presented. I think I'm a bit partial to words which are clearly based in statistics - things like p-value and standard deviation seem to have less misconceptions than when you talk about data like odds ratios or chance. I can see how this could be extremely misleading in the court room and how both the prosecution and defense might bring on different statisticians to make their case in the most emotionally compelling way - this seems like a potential for significant injustice.
Ditto. I think my favorite statistics example was supposedly my grandfather playing Keno. He and my dad both played a pick 8. My grandfather just drew a straight line down one of the columns, to...
Ditto.
I think my favorite statistics example was supposedly my grandfather playing Keno.
He and my dad both played a pick 8. My grandfather just drew a straight line down one of the columns, to which my dad said "well that'll never win, i've never seen that happen on a keno board", to which he pointed out "its just as likely as any other combination". It really helps put into perspective just how bad we are intuitively with odds, randomness, and stats.
Worse, because people are so bad at stats, they don't understand how horribly easy they are to manipulate. The amount of shit out there that starts with "this is the narrative, find me the data to back it up" is just disgusting.
Interesting article. Reminded me of some of Vsauce2's videos about crime and stats and courtrooms which talk about the same thing: example 1 example 2 whole playlist of em Highly recommend...
Interesting article. Reminded me of some of Vsauce2's videos about crime and stats and courtrooms which talk about the same thing:
In general, statistics are often widely misunderstood. I don't consider myself a statistician, but I do a fair amount of statistics to help make sense out of data. The number of misconceptions I've heard from others at work over even the most simple of statistics (control charts, scatter plots, standard deviations) can be at times concerning, especially since I work in a field where most clinicians should have at least a cursory understanding in order to interpret findings from medical papers. I recently delivered some statistics which were a bit more involved (a series of logistic regressions) and spent literally hours talking this over with the team it was delivered to over the course of weeks. It took a lot of explaining for some concepts to sink in, and after reading this article I'm starting to see additional connections with the way the data itself is presented. I think I'm a bit partial to words which are clearly based in statistics - things like p-value and standard deviation seem to have less misconceptions than when you talk about data like odds ratios or chance. I can see how this could be extremely misleading in the court room and how both the prosecution and defense might bring on different statisticians to make their case in the most emotionally compelling way - this seems like a potential for significant injustice.
Ditto.
I think my favorite statistics example was supposedly my grandfather playing Keno.
He and my dad both played a pick 8. My grandfather just drew a straight line down one of the columns, to which my dad said "well that'll never win, i've never seen that happen on a keno board", to which he pointed out "its just as likely as any other combination". It really helps put into perspective just how bad we are intuitively with odds, randomness, and stats.
Worse, because people are so bad at stats, they don't understand how horribly easy they are to manipulate. The amount of shit out there that starts with "this is the narrative, find me the data to back it up" is just disgusting.
Interesting article. Reminded me of some of Vsauce2's videos about crime and stats and courtrooms which talk about the same thing:
example 1
example 2
whole playlist of em
Highly recommend checking them out if you enjoyed the article or find the topic interesting.
Mirror: https://archive.is/7WiJj