4 ms·

The article also mentions study of charity donations.

• Illniyar 9y ago
  Skimming the article for that study, I don't find anything wrong with it superficially, but while the effect might be statistically significant, they are tiny - I.E. from 4.2% of donors to 5.5% had K initial in Katrina, and 10% to 12% in Hurrican Mitch (though the data might be bad, since the actual data did not contain the exact information they needed). Also the effect was apparently not noticeable in Hurrican Rita. There is also no mention of the percent of people starting with the letter from the given population - since the dataset for before the hurricane was much smaller (Katrina- ~300 before, ~4000 after), it's very likely that the proportions of named individuals drifted to the average, as what happens in larger datasets. The number of hurricanes tested was also small - I.E. 3 , I'm not sure an effect can be inferred from such small numbers.