This paper presents a revised model that incorporates the performance feedback perspective from organizational learning theory into the dynamics of organizational form diffusion through replication behavior, which has been emphasized in concepts such as isomorphism in new institutional theory and exploitation of existing knowledge in evolutionary economics, and empirically tests this model using data on diversification mode selection by Korean chaebol groups. Specifically, this paper examines the effect of past diversification mode performance—both one's own and that of other organizations—on subsequent mode selection, based on the perspective that while organizations do tend to repeat and replicate existing knowledge and behavior under uncertainty, they do not simply stop there but rather make sophisticated decisions based on performance feedback from their own and other organizations' behaviors, representing more rational agents. Organizations as decision-makers who must choose diversification modes strive to make maximally rational decisions at the time of decision-making, even though the outcomes may inevitably be bounded. Therefore, this study predicts that organizations will make mode selections by considering the actual outcomes generated by their own or other organizations' behaviors—that is, the performance of their own or other organizations' past decisions—while inferring causal relationships between mode selection and performance. To verify the validity of this perspective, this study empirically analyzes data on diversification mode selection by Korean chaebol groups from 1988 to 1997, testing hypotheses that the performance of modes previously selected by the focal organization and the performance of modes previously selected by other organizations occupying higher or similar status positions will influence the focal organization's mode selection when implementing diversification. Additionally, this paper predicts that the effect of each firm's past mode performance on subsequent mode selection will be moderated by status. The empirical analysis results supported all hypotheses except for the hypothesis regarding the effect of performance feedback from the past choices of similar-status firms.