Social Influence of a Religious Hero: The Late Cardinal Stephen Kim Sou-hwan's Effect on Cornea Donation and Volunteerism

Abstract

This study examined the mediated influence of a celebrated religious hero in South Korea, Cardinal Stephen Kim, through two forms of involvement—parasocial interaction and identification—on intention toward cornea donation and volunteerism, and it investigated how the news media diffused of his death. A structural equation modeling analysis with a Web-based voluntary survey of more than 1,200 people in South Korea revealed a multistep social influence process, beginning with parasocial interaction with Cardinal Kim, leading to identification with him, which predicted intention toward cornea donation and volunteerism. Additional investigations found that news of Cardinal Kim's death diffused rapidly through media and interpersonal communication. Results of this study demonstrate that religious leaders who achieve a celebrity hero status can prompt public discussion of important issues rather quickly through extensive media coverage, enabling them to promote prosocial behavior and positively affect public health.

Acknowledgments

This study was supported by the Yeungnam University research grants in 2008. The first author thanks a research intern, Eun-Gyuhl Bae (Annenberg School for Communication, University of Southern California) for her great help on this project.

Notes

¹This study terminated data collection when the number of respondents exceeded 1,000. This number was reached on the fifth day of the survey administration. We could have a higher response rate for the sample if we left the survey open longer and sent follow-up emails to potential respondents. The response rate represents the five-day survey period.

***p < .001.

²We also tested the model with the reflected items in the confirmatory factor analysis because the model with the items theoretically drawn could yield an acceptable fit even though the items marked low factor coefficients in the current study. The test result shows that the model was barely acceptable but still exhibited a low GFI, which indicated model improvement by removing least contributing items: χ ² (492, N = 1,261) = 1,383.17; p < .001, GFI = .942, NFI = .966, CFI = .978, and RMSEA = .038. As a result, we ran the model without the reflected items and came up with an acceptable fit.