CailCoreshev, in a coment at Alpha Game, says:
Sometimes I think many of our problems today are due to most people having a complete lack of understanding of statistics and probabilities. Not an inability to calculate them; that does take some study. But the inability to grasp what they mean at all; to understand that “group X tends to show feature Y more than group Z” doesn’t mean X=Y or Z!=Y or X>Z or anything else except what it actually says.
He is, of course, 100% correct. Beyond them not being able to grasp simple statistical concepts even qualitatively, they eschew conversational use of statistical ideas by foisting value judgements on certain, not all, uses of them. It is a stain on one’s character to generalize, even if they just finished reading a peer reviewed study about the subject. People somehow do not grasp the difference between statistics derived from counting things that are unequivocal, and counting responses to an opinion survey. They will universally reject counted things, while they will grab every opinion survey they like and hammer us with it.
Women are worse in this regard than men. But men, we are rapidly slidding into the blissful ignorance that would say, “you can’t generalize, all people are different”. This is especially frustrating to those of us who are nerds and know how to understand statistical communication as well as to calculate and manipulate (I invite uninformed attack with this term) numbers.
I recall freshman year in college (barely). In sociology class we were taught that stereotyping is bad, while generalizing is OK. That was 1980. What happened between now and then? Doesn’t everyone take a 101 Sociology class? When did they stop eplaining the difference between stereotyping and generalizing?
It is beyond a lack of understanding. it is a visceral rejection of statistics akin to fingers in ear yelling lalalalalalalala when confronted with a fact people dislike because it makes them feel bad. Politicians and their speech writers know that people are clueless. The ignorance is a tool. Infomation such as presented in my prior post absolutely depends on not just pedestrians inability to grasp it, but on the laziness of those who perhaps could get it but are too put out by the idea of not falling in line with conventional wisdom.
Conventional wisdom IS the new statistic of preference. It started with “they say” as proof text from any claim. Today its “everyone knows”.
Everyone knows men are abusive, men cheat, men use porn, men are responsible for divorces, men abandon their kids, men run rough shod over women using scripture. Everyone knows X is unhealthy and Y is preferred, ABC parenting is superior to XYZ parenting, conservatives are closed minded biggots, women are oppressed, so forth.
Besides “everyone knows” we have solipsism and we have anecdote. Cails example “X tends to show feature Y more than group Z” is met with “my X (sister) and X (neighbor) have never shown feature Y, therefore group X absolutely does not show Y more than group Z. Some of this stuff was taught in elementary school when we learned about the unions and intersections of sets! No six sigma black belt need explain that in a bag with ten billiard balls, seven white, three black, one is more likely to pull out a white one randomly. But if that was men and women, or white people and black people, and “oulling one out of the bag” was some social pathology, you could stack divorce records or arrest records for 100 years to the moon and back, counted and sorted, and with furrowed braw and pure empathy inflection someone will tell you, nope, depends on the individual balls in the bag whether or not there is a 70% chance you’d pull out a white one.
Cails assertion that this is the source of many of our problems is, if anything, understated. It is at or near the root of most of our problems.
The population is as statistically literate as Ron Burgandy (him again) when he says “80% of the time that works every time” (or something like that).