Predicting the effect of entrepreneurial stressors and resultant strain on entrepreneurial behaviour: an SEM-based machine-learning approach
本研究基于挑战-阻碍框架，利用结构方程模型和机器学习方法，分析了415名创业者的数据，发现阻碍性压力源导致创业压力，进而显著影响创新行为，且心理资本能缓解压力源的负面影响。
Purpose Considerable evidence suggests that although they overlap, entrepreneurial and employee stressors have different causal antecedents and outcomes. However, limited empirical data explain how entrepreneurial traits, work and life drive entrepreneurial stressors and create entrepreneurial strain (commonly called entrepreneurial stress). Drawing on the challenge-hindrance framework (CHF), this paper hypothesises the causal effect of hindrance stressors on entrepreneurial strain. Furthermore, the study posits that entrepreneurial stressors and the resultant strain affect entrepreneurial behaviour. Design/methodology/approach The study adopts an SEM-based machine-learning approach. Cross-lagged path models using SEM are used to analyse the data and train the machine-learning algorithm for cross-validation and generalisation. The sample consists of 415 entrepreneurs from three countries: India, Oman and United Arab Emirates. The entrepreneurs completed two self-report surveys over 12 months. Findings The results show that hindrances to personal and professional goal achievement, demand-capability gap and contradictions between aspiration and reality, primarily due to unique resource constraints, characterise entrepreneurial stressors leading to entrepreneurial strain. The study further asserts that entrepreneurial strain is a significant predictor of entrepreneurial behaviour, significantly affecting innovativeness behaviour. Finally, the finding suggests that psychological capital moderates the adverse impact of stressors on entrepreneurial strain over time. Originality/value This study contributes to the CHF by demonstrating the value of hindrance stressors in studying entrepreneurial strain and providing new insights into entrepreneurial coping. It argues that entrepreneurs cope effectively against hindrance stressors by utilising psychological capital. Furthermore, the study provides more evidence about the causal, reversed and reciprocal relationships between stressors and entrepreneurial strain through a cross-lagged analysis. This study is one of the first to evaluate the impact of entrepreneurial strain on entrepreneurial behaviour. Using a machine-learning approach is a new possibility for using machine learning for SEM and entrepreneurial strain.