Or at least take the 127 from the largest sample size. Openly stating "we intentionally altered our sample to fit our desired results" and still getting published is nuts.