Page 119 - 35-2

NTU Management Review Vol. 35 No. 2 Oct. 2025

2. Design/Methodology/Approach

To minimize common method variance, we collect data in two batches in 2023 from
209 respondents working at 64 organizations in Taiwan. These selected organizations

are representative of more than 10 industrial sectors, including manufacturing, health
care, financial insurance, higher education, and government agencies. Before formal data
collection, we conduct on-site visits and phone interviews with senior managers to confirm
that the selected organizations are actively engaged in both external and internal CSR

activities at varying levels.
In accordance with the method proposed by Muller, Judd, and Yzerbyt (2005) for
testing mediated moderation models, we use Mplus 8.3 software to test six hypotheses
in three steps. The first step is to examine the effect of the interaction between perceived

ECSR and ICSR on substantive and symbolic motivational attributions as the mediator.
The second step is to explore the relationship between substantive and symbolic
motivational attributions as the mediator and organizational identification as the dependent
variable. The third step is to examine the moderating effect of perceived ICSR on the

indirect path from ECSR perception to organizational identification, mediated by either
substantive or symbolic motivational attributions, and to explore the parallel mediated
moderation effect of perceived ICSR on the path from ECSR perception to substantive and
symbolic motivational attributions, which in turn lead to organizational identification.

3. Findings

Overall, our empirical results support all hypotheses proposed in this study.

According to our findings, employees tend to concurrently attribute both substantive and
symbolic motives to their organization’s CSR initiatives. When the interaction between
perceived ICSR and ECSR is strong, employees tend to interpret their organization’s CSR
engagement as driven by substantive motives rather than by symbolic motives. Moreover,

substantive motivational attribution significantly enhances employees’ organizational
identification, whereas symbolic motivational attribution has the opposite effect. By
applying the proposed parallel mediated moderation model, this study demonstrates that
when employees perceive a balance between ICSR and ECSR, they become more likely

111