Note(s): Panel A controls for potential endogeneity by including a lagged variable of trust language (LAGTRUSTTL) and examining its impact on information transparency. This approach helps in understanding the persistence or delayed effects of trust language. Panel B utilizes propensity score matching to control for selection bias and unobserved heterogeneity by comparing firms with similar characteristics but different levels of trust language usage. The regression models are adjusted accordingly to validate the robustness of the results obtained in other analyses. Standard errors in parentheses. All variables are defined in Appendix B. *p < 0.10, **p < 0.05, and ***p < 0.01