I wonder what they mean by "around" full and near moons -- if they're counting three days before and after, that covers half the lunar cycle. Add to that a small sample size (only 10 months), and you've got a pretty week for a causal connection.
I really have to wonder, too, why the sample size is so small -- every tremor is recorded and lunar cycles are perfectly predictable, so I'd think it would be pretty east to set up a computer to process 20 or 30 years' worth of data, maybe more, going back however far the seismographic measurements are considered reliable.
near = new; pretty week = pretty weak case; pretty east= pretty easy
And that with the automatic preview and everything.
Bingo!
Bulletin: 40% of sick days are taken on Fridays and Mondays!
Clear evidence of a correlation with something if you are looking for evidence of a correlation with something.