Beyond regression: an ANN approach for exploring social entrepreneurship's impact on sustainability in Korean organizations

Abstract

PurposeThis study explores the complex dynamics within Korean public organizations, specifically focusing on the interplay between social entrepreneurship, marketing innovation, and sustainability. The study aims to provide a fresh perspective by utilizing artificial neural networks (ANN) and structural equation modeling (SEM).Design/methodology/approachThe research methodology involves collecting data from a survey of 230 employees across various government-affiliated institutions in Korea. Four critical sub-dimensions of social entrepreneurship-Proactiveness, Innovativeness, Risk-taking, and Social orientation-are considered independent variables. Marketing innovation and sustainability are treated as dependent variables. The study compares the flexibility and adaptability of ANN models with SEM.FindingsProactiveness and risk-taking showed no significant influence, while innovativeness and social orientation had strong positive effects, with social orientation as the most decisive antecedent. Marketing innovation enhanced sustainability, and ANN revealed nonlinear threshold effects, where moderate proactiveness and innovativeness optimized outcomes but excessive levels reduced them.OriginalityThis study integrates Triple Bottom Line, Stakeholder, Institutional, and Schumpeter's theories to explain entrepreneurship-sustainability dynamics, highlights the decisive role of social orientation and the need for balance in entrepreneurial behaviors, and demonstrates the methodological value of ANN in uncovering hidden nonlinearities beyond SEM.