Decoding Mixed Identities in Hong Kong : A Clustering Analysis of Multiple Identity Indicators

Research output: Journal PublicationsJournal Article (refereed)peer-review

1 Citation (Scopus)


Given the multiplicity and fluidity of identities, measuring mixed identities is challenging. As an epicenter of identity conflicts, Hong Kong is an ideal place to study mixed identities. However, the conventional unidimensional measurement in Hong Kong surveys and statistical correlation analysis may not accurately estimate the prevalence of different combinations of identities. Therefore, this study reassesses the patterns of identity combinations from 2016 to 2022 through K-means clustering of multiple identity indicators. The clustering analysis identifies three clusters, namely dual identity, moderate hybrid identity, and predominant Hong Kong identity. The analysis suggests that the conventional single-item indicator underestimates the proportion of mixed identities in the population. Furthermore, this study documents the rise of a predominant Hong Kong identity, particularly among young people. The regression analyses illustrate that citizens who are older, non-supporters of the pan-democratic camp, and not born in Hong Kong are more likely to have dual identity. This study contributes to the literature on measuring mixed identities by arguing that multidimensional measurement is preferable to unidimensional measurement when the main research objective is to examine the proportion of various combinations of identities.
Original languageEnglish
Pages (from-to)585-603
Number of pages19
JournalSocial Indicators Research
Issue number2
Early online date13 Dec 2023
Publication statusPublished - Jan 2024

Bibliographical note

Publisher Copyright:
© The Author(s) 2023.


  • Identity politics
  • mixed identities
  • Measurement
  • K-means clustering
  • Hong Kong


Dive into the research topics of 'Decoding Mixed Identities in Hong Kong : A Clustering Analysis of Multiple Identity Indicators'. Together they form a unique fingerprint.

Cite this