“…Finally, 15 dichotomous nominal variables (present vs. not present) were established for classifying sources into types, based on a list of source types extrapolated from similar studies (Albaek, Christiansen, & Togeby, 2003;Armstrong, 2004;Barr, Irlbeck, & Akers, 2012;Gasher, Hayes, Hackett, Gutstein, Ross, & Dunn, 2007;Sumpter & Braddock, 2002;Sumpter & Lukaszewski, 2001;Telg & Raulerson, 1999;Zoch & Turk, 1998). Dichotomous measurement of independent variables has been found to have little serious effect on the probability statements underlying parametric procedures of inferential statistics, that is, common inferential statistical procedures, for example, ANOVA, may be used on such data (Field, 2009).…”