2 ms·
Those effects were not present in the RCT. Even after reducing the p-value cutoff to 10% (rather than the standard 5%), only a few of the expected measures show
by yummyfajitas 13y ago
Those effects were not present in the RCT. Even after reducing the p-value cutoff to 10% (rather than the standard 5%), only a few of the expected measures showed a statistically significant effect (some positive, some negative), and virtually nothing persisted as far as 1st grade.
http://www.acf.hhs.gov/sites/default/files/opre/executive_summary_final.pdf http://www.acf.hhs.gov/sites/default/files/opre/executive_su...
(See also this comic if you want to point to a few blue and yellow cells in their charts. http://xkcd.com/882/ http://xkcd.com/882/ )