When the researchers tried to figure out how to prevent the injuries, they learned that the only thing associated with running injuries was, in fact, running. The more people ran, the more likely it was that they would become injured.
Shoot. How could I have overlooked this foolproof method for preventing running injury?
However, Pete Larson of Runblogger has a good post pointing out that while the plural of anecdote is not data, scientific studies of running injuries have limitations, too: they're averaging over very diverse populations of runners, and not necessarily examining all the possible variables.
In terms of my own anecdotal injury experience, since reading that article on hip abductor pain that I linked to last Friday, I've been doing 2-3 sets of either lateral band walks (link to video here: http://youtu.be/ZdJ5l5Km3B8), fire hydrants (link to video here: http://youtu.be/1Cb8thnh2WU), or lateral leg raises (video here: http://youtu.be/LW_yy8KRi2s) each day. (Cranking Sir Mix-a-lot on your iPod while doing these is optional, but recommended.) It actually seems to be helping - I did a 2.5 mile run/walk on Sunday, and another today, and have experienced no hamstring pain and only slight soreness in the TFL. (Both gluteus medius muscles are sore, but in a "you worked us out" kind of way. Though I might switch to doing the exercises every other day just to make sure they get enough recovery time.)
5 days is not really enough time to build significant muscular strength, so what I'm observing is almost certainly due to either a) increased neuromuscular activation, b) the placebo effect, or c) stuff just getting better on its own. Still, it's motivation enough to keep going.