When someone says something controversial, often the first reaction is to ask where they got their information from. What statistics are they pulling from to make the statement they have made?
Half the time, they don’t have a source at all. They say it was based on statistics or they saw it somewhere, but it’s not substantiated anywhere.
After listening to a book recently that talked about disparities and discrimination, it gave me quite a lot of new takes on how we understand statistics. Namely, that statistics can’t always be boiled down into the nice theory provers that we make them out to be. In the book, one of the prime examples of this was when the writer talks about the disparity between Hispanics in the US and Asians in the US.
A statistic that people often use to say that Hispanics don’t make as much as Asians in America is pulling the median income of both ethnicities. I don’t know the example numbers and example he gives, but I’ll give a smaller scale example.
Let’s say that Asians make $7 on average over a month’s time. Hispanics make an average of $3 over a month’s time. So clearly that shows us that Hispanics are discriminated against for getting paid as much as they should, right?
But wait, there’s at least a 3rd dimension to this…if not more. What I didn’t mention in this stat is that the median age range between these two ethnicities is quite different. Once again, not having the numbers in front of me and the book nowhere nearby, let’s say the median age of Asians is 55 and the median age of Hispanics is 27. Now, we have to not only look at the difference between ethnicities, but also ages. Are some ages paid less than other ages? Is that wrong? Does it make natural sense for a 50-year-old, on average, to get paid more than an 18-year-old?
There are maybe no clear answers, but it does raise the question about stats. Do they help answer, or only leave us with more questions?