In order to determine the strength of the moderating effect of leadership behavior on therelationship between entrepreneurial orientation, social capital, and global mindsetCohen’s (1988) effect sizes were calculated. Furthermore, the strength of the moderatingeffect can be assessed by comparing the coefficient of determination value of the maineffect model with the value of the full model that incorporates both exogenous latentvariables and moderating variable (Henseler & Chin, 2010; Wilden, Gudergan, Nielsen,& Lings, 2013). Thus, the strength of the moderating effect could be expressed using thefollowing formula (Cohen, 1988; Henseler & Fassott, 2010).Moderating effect sizes () values of 0.02 can be considered as weak, effect sizes of0.15 as moderate while the effect sizes above 0.35 may be regarded as strong (Cohen,1988; Henseler & Fassott, 2010). However, according to Chin, Marcolin, and Newsted(2003) a low effect size does not necessarily means that the underlying moderating effectis insignificant. Even a small interaction effect can be meaningful under extrememoderating conditions, if the resulting beta changes are meaningful, then it is importantto take these conditions into account (Chin, Marcolin, & Newsted, 2003). Result of thestrength of the moderating effect of leadership behavior is presented in Table 4.13.Following Henseler and Fassott (2010) and Cohen (1988) rule of thumb for determiningthe strength of the moderating effect, Table 4.14 shows that the effect size for Firminternational performance was 0.43, suggesting that the moderating effect was strong(Henseler, Wilson, Götz, & Hautvast, 2007; Wilden, Gudergan, Nielsen, & Lings, 2013).18.104.22.168 Assessment of predictive relevance
The present study conjointly applied Stone-Geisser by taking a look at predictiveconnection of the analysis model victimisation blindfolding procedures (Geisser, 1974;
You've reached the end of your free preview.
Want to read all 4 pages?
- Summer '17