2026-10-11T03:55:37Z https://api.figshare.com/v2/oai

oai:figshare.com:article/33967124 2026-09-22T17:29:10Z category_45 category_734 category_811 portal_5 item_type_3 month_year_09_2026

<p>Relevant structural analysis.</p> Song Jin (642035) Liguo Chang (5681840) NanNan Zhang (3105369) Ye Li (290115) Sociology Biological Sciences not elsewhere classified Mathematical Sciences not elsewhere classified significant mediating role remain consistently significant received extensive attention maintaining directional consistency empirical results show direct effects alone different data characteristics positive indirect effect mediating effect played exposure variables leading policy preferences play energy cooperation outcomes statistical analysis strategy cooperation policy preferences bayesian sem estimates mediation analysis based policy preferences energy cooperation mediation effect based mediation bayesian sem mediation analysis statistical performance policy formulation impact analysis example analysis outcome variables latent variables energy trade classical sem bayesian regression mediation mechanism variable uncertainty two stages standard error significance level reliable handling recent years parameter stability paper explores paper divides neighboring countries many fields making level logical chain institutional frameworks institutional framework handling latent flexible program finer mechanism <div><p>Mediation analysis is able to deeply analyze the finer mechanism of the path of exposure variables leading to outcome variables and has received extensive attention in many fields in recent years. In order to further analyze China’s energy cooperation with neighboring countries, this paper divides the two stages of policy formulation and implementation. It connects the logical chain of “cooperation policy preferences at the decision-making level, final cooperation policy preferences to energy trade and investment cooperation.” By constructing a partial Bayesian structural equation model, we verify the mediating effect played by final cooperative policy preferences. The mediation analysis based on structural equation modeling provides the possibility to analyze the complex structure among variables, including the mediation effect. And Bayesian SEM can provide a more flexible program for the construction and estimation of complex models. In addition, this paper explores the statistical analysis strategy of mediation analysis of Bayesian structural equation modeling and its statistical performance under different data characteristics through example analysis. The empirical results show that policy preferences play a significant mediating role between institutional frameworks and energy cooperation outcomes. Specifically, the Bayesian SEM estimates a positive indirect effect of 0.305, with a standard error of 0.123, a 95% confidence interval of (0.061, 0.552), and a significance level of p = 0.006. This indicates that institutional arrangements promote energy cooperation mainly by shaping policy preferences rather than through direct effects alone. Compared with classical regression-based mediation, classical SEM, Bayesian regression, and other alternative approaches, Bayesian SEM demonstrates stronger performance in handling latent-variable uncertainty, interval estimation, and parameter stability. Across the 12 estimation methods employed in this study, the estimated indirect effects range from 0.271 to 0.312 and remain consistently significant, confirming the robustness of the mediation mechanism. This indicates that while maintaining directional consistency, it offers more robust and reliable handling of latent variables and uncertainty.</p></div> 2026-09-22T17:28:52Z Dataset Dataset 10.1371/journal.pone.0358701.t007 https://figshare.com/articles/dataset/_p_Relevant_structural_analysis_p_/33967124 CC BY 4.0