Using Hierarchical Regression to Explore the Relationships among Corporate Social Responsibility, Corporate Image and Organizational Performance

• Main Authors: Lin, Chun Han, 林淳涵
• Other Authors: Huang, Chung-Fah
• Format: Others
• Language: zh-TW
• Published: 2012
• Online Access: http://ndltd.ncl.edu.tw/handle/20576371270463613506

碩士 === 國立高雄應用科技大學 === 土木工程與防災科技研究所 === 100 === Corporate social responsibility (CSR) has gradually been integrated into the business activities of companies around the world. Under this trend, companies need to focus on what they can do for society instead of on just making profits. The primary purpose of this study was to investigate the relationships among CSR, corporate image and organizational performance. In this study, the samples are mainly aimed at A/C/E firm. A total of 600 questionnaires were sent, and 340 responses were received. After excluding the invalid ones, there were 266 valid questionnaires, indicating a 44.33% valid response rate. The questionnaire survey results were analyzed using Harman’s one-factor test, factor analysis, nested confirmatory factor analysis and hierarchical regression analysis. Based on the major results of the analysis, the following findings were reached in this study: 1. After Harman’s one-factor test it was found a single factor can only explain for 44.18% of the variances, not over 50%, so this research is credible without the problem of common method variance. 2. The results of factor analysis of the four factors of CSR, i.e. resource conservation, social participation, social commitment and pollution prevention, and the nested confirmatory factor analysis also indicate that the relationships of the four factors assumed in this study are appropriate and discriminate. 3. Through hierarchical regression analysis, we can say CSR has an indirect and positive influence on organizational performance via corporate image, which means even when the effect of corporate image is controlled for, CSR still has a strong effect on organizational performance.