[monkeyjoe]

...and n=10, and they measured a ton of covariates and reported only the 'significant' ones. Not much actual information here.