Results of hypothesis testing Model

I find that all f -values are all of considerable level, when the critical values proposed by Cohen (Cohen 1988] are taken as a basis. All significant paths exhibit effect sizes larger than the lowest critical value of 0,02 (see Table 17). The effects of lead usemess on domain-specific innovativeness and opinion leadership can both be regarded as weak (f (LU-DSI)=0,094 f (LU-OL)=0,078) the same applies for the effect of use experience on opinion leadership (f (UE-OL)=0,028). The effect of use experience on domain-specific innovativeness is strong (f (UE-DSI)=0,383). The analysis of the values for cross-validated redundancy shows, that all endogenous variable can be predicted by the modei ((Q (CO)=0,200) Q (DS1)=0,331 Q (OL)=0,123)). 

User embeddedness User embeddedness Domain-specific im 

Direct path not significad Direct path not significant nii rtiadii il ilaHBaM 

Fiji mediation (Direct path partly significant) Rdlinediattoo 

The significant paths in the model showed different effect sizes (see Table 19). The impact of the antecedents on cognttive empathy was weak for all three variables (f (EMB-CE]=(0,113) f (UE-CE)=0,134 f (LU-CE)=0,135). The impact of cognitive empathy on the outcome variables was moderate in two cases (customer orientation (P(CE-C0)=0,172) and domain-specific innovation (1 (CE-DSI)=0,164)), and strong in the case of opinion leadership (f (CE-OL)=0.369). The partially mediated relationship between use experience emd domain-specific innovativeness showed a moderate effects size (f (UE-DSf)=0,210). 

---