“…Others have treated a discrete, ordinal independent variable as interval level and used standard regression-based techniques (Chandler & Pronin, 2012;Legault, Gutsell, & Inzlicht, 2011). Alternatively, investigators have modified their data to produce a dichotomous X, such as by conducting separate analyses comparing various groups of interest while discarding the remaining data (Pedersen, Denson, Goss, Vasquez, Kelly, & Miller, 2011;Ronay, Greenaway, Anicich, & Galinsky, 2012;Werle, Wansink, & Payne, 2011;Whitchurch, Wilson, & Gilbert, 2011) or collapsing multiple groups into one for comparison with another group or set of groups (Calogero & Jost, 2011;Haisley & Loewenstein, 2011;Ruva, Guenther, & Yarbrough, 2011). Another strategy used is substituting a continuous manipulation check for the multicategorical X and proceeding with a mediation analysis as if X were observed as a continuum (Forgas, 2011).…”