“…Cross-sectional designs are traditionally less ideal for use in college impact research, given results do not represent true causal relationships and potentially introduce biased assessments of institutional quality due to a lack of control measures (Astin & Lee, 2003). However, as explained in the introduction of the conceptual model, this research used retrospective questions to capture precollege data which scholars suggest is more appropriate when measuring dependent variables with cognitive dimensions like leadership development (Howard, 1980;Howard & Dailey, 1979;Rohs, 1999Rohs, , 2002Rohs & Langone, 1997). The use of selfreported data may also be of concern as it can vary in accuracy of measurement depending on the degree to which researchers ensure both methodological and psychometric quality (Gonyea, 2005).…”