I thought this little online calculator was a good example of taking a ton of sociological data on the backend to make a very usable product for regular joes like us who don’t want to hear all the regression statistics, but simply want to know the answer to an important question. In this case the question is: How likely are you to get divorced in the next five years?
However, it also raises—very poignantly—another problem with research on human beings: The fact that statistics don’t seem very meaningful when applied to an individual. For example, I don’t feel like the 15% chance this calculator gave me is low enough—I feel our marriage is too strong for that high of a divorce probability. Someone else in marital trouble might feel their percentage should be much, much higher. So the statistics tell good general trends for people of my age and gender, but are still not very meaningful when applied to someone personally.
But of course you already knew that about research! 🙂