Harvard University

To compare, both theoretically and empirically, standard versus ideal ways of determining which kinds of statistical findings merit publication