On the Apparently Inappropriate Use of Multiple Hypothesis Testing in Earthquake Prediction Studies