Page 41 - 35-2

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

associated with review persuasiveness; this in turn leads to lower attitude polarization.
Attitude consistency hasno significant moderating effect. Elder and younger consumers
exhibit different assimilation biases toward AIoT health technology reviews. For younger
consumers, persuasiveness significantly mediates the relationship between attitude

consistency toward AIoT health technology and attitude polarization. The direct effect of
attitude consistency toward AIoTs on polarization is not significant; attitude consistency
is significantly and positively associated with persuasiveness, and persuasiveness is

significantly and negatively associated with polarization. Thus, young consumers have less
polarized attitudes and are more persuaded if the reviewer expresses attitudes that align
with their own. For consumers aged 36 years or older, persuasiveness does not significantly
mediate the relationship between attitude consistency and polarization. Neither the effect
of consistency on persuasiveness nor that of persuasiveness on polarization are significant

in this elder group. Interestingly, attitude consistency toward a brand differs by age.
Among younger consumers, persuasiveness remains significant regardless of whether
attitude consistency toward the brand is high or low, indicating that attitude consistency

toward AIoT health technology influences polarization through persuasiveness regardless
of attitude consistency toward the brand. For elder consumers, persuasiveness does not
significantly mediate under either high or low brand-attitude consistency.

4. Research Limitations

This study has several limitations. First, we only test two well-known medical
brands. Future work should include more obscure brands to see if attitudes toward

the brand function as a moderator. Second, our use of convenience sampling limits
generalizability. Subsequent studies should recruit a more representative, cross-cultural
sample—and ensure elder adult representation especially—to account for cultural factors
such as uncertainty avoidance and individualism. Third, we do not assess participants’
health status, perceived vulnerability, or related emotions, all of which can shape health

decisions and WOM persuasiveness; future research should include these variables.
Moreover, our focus is on the wearable heart-rate monitor, and the findings may not extend
to other AIoT applications (e.g., medical robots or drug discovery); future studies should
test their hypotheses on a diverse range of AIoT technologies. Finally, we do not explore

33