NPR story (with text and audio) discusses this idea, and some related research. It also mentions, as I have, an aspect of random data (as earthquake occurrences may be): that events occurring at random times will not be evenly spaced in time, but occasionally either bunched up or further apart.
While on sabbatical in Munich, I now have the luxury of listening to NPR online, and if I listen to the archived shows (which I generally do because of the time difference), I can pick and choose the stories I want to hear. More or less this is good, but it does keep me from trying to better understand the German media, and allows me to fractionate the entire program that the NPR editors are presenting. Still, better than my first sabbatical in Germany 20 years ago, at the dawn of email but really pre-www. I had to struggle to understand German tv and newspapers, or spring for the pricey Herald Tribune once or twice a week, or go to the library to read it. Hey, come to think of it, that wasn't so bad, really.