Cross‐national consumer research using structural topic modelling: Consumers' approach‐avoidance behaviours
This study introduces structural topic modelling (STM), a sophisticated unsupervised machine‐learning algorithm for text analysis, to compare Indonesian and Malaysian Muslim consumers' approach‐avoidance behaviours toward Korean beauty products using social media data. The STM results revealed...
Saved in:
Published in | International journal of consumer studies Vol. 47; no. 5; pp. 1692 - 1713 |
---|---|
Main Authors | , , |
Format | Journal Article |
Language | English |
Published |
Oxford
Blackwell Publishing Ltd
01.09.2023
|
Subjects | |
Online Access | Get full text |
Cover
Loading…
Summary: | This study introduces structural topic modelling (STM), a sophisticated unsupervised machine‐learning algorithm for text analysis, to compare Indonesian and Malaysian Muslim consumers' approach‐avoidance behaviours toward Korean beauty products using social media data. The STM results revealed 16 topics for each country, including new common themes belonging to K‐beauty culture and wannabe Korean skin. Intriguing differences were also observed between these countries. Korea‐related constructs, such as Korea's image and wannabe Korean skin, were approach factors for only Indonesians. Korean cosmetic brand‐specific topics were extracted for only Malaysians and were significantly associated with their behavioural responses. Unsuitable Korean beauty products and domestic product preferences were avoidance factors for Indonesians, but new product risks and conflicts between Muslim and Korean cultures for Malaysians. We demonstrate that STM is a helpful tool in cross‐national research for corroborating and extending the existing theoretical frameworks. The practical implications are also provided for global marketers. |
---|---|
ISSN: | 1470-6423 1470-6431 |
DOI: | 10.1111/ijcs.12923 |