Thursday, May 28, 2015

Be a Social Analyst

The Times has a post which is different.  We're used to taking a survey and having our results compared to the average of previous survey takers, but in this case you're asked to draw a graph.

Specifically, you're asked to plot the relationship between family income and probability of going to college.  They give a midpoint, and when you're finished compare your line to those of the previous takers and discuss the reasons for the result.  Like many people I drew an S curve, but it turns out the true graph is a straight line.  A straight line relationship is also true for some other social factors.

"Intriguingly, the relationship between parental-income rank and teen pregnancy is also quite linear, and some of the same forces are probably involved. So is the relationship between parental-income rank and a child’s future income rank.
Not every relevant relationship is linear, however. The chances that a student enrolls in the highest-quality colleges, as measured by their students’ future earnings, are a bit more complicated. These chances accelerate as incomes grow.
And enrolling is not everything. While rich children born around 1980 were nearly three times more likely to go to college than poor children, they were six times more likely to graduate, according to a study separate from the one we're showing here.


No comments: