Comparing public responses to apologies: examining crisis communication strategies using network analysis and topic modeling

Sejung Park, Jin A. Choi

Research output: Contribution to journalArticlepeer-review

1 Scopus citations


This study aims to explore public perceptions and emotions toward crisis response strategies when unethical business practices occur. United Airlines’ passenger-dragging scandal is examined through the lens of the situational crisis communication theory. United Airlines employed two apology strategies, partial and full, on their Facebook page. Public responses toward the two apology strategies are analyzed through opinion mining of public comments by employing mixed text mining techniques. A total of 51,147 public comments generated by 50,103 unique users were extracted with an API-based social network analysis tool, NodeXL. First, the structural characteristics of eWOM communication in response to the two apology strategies were comparatively investigated. Next, topic modeling was employed to detect salient topics, and the intensity of positive and negative emotions to the apology strategies were compared. Additionally, semantic network analysis was used to uncover public reactions to and brand attitude toward United as a consequence of the apology strategies. The public generated a larger comment network for the full apology strategy than the partial apology. The public discussed problems with United’s apology strategies and deemed that wrong public relations strategies were used. Findings from the sentiment analysis and semantic network analysis suggest that the matched response strategy (full apology) did not change public responses and emotions. Public perception on United’s crisis management was primarily negatively associated with its hypocritical apology and image-focused reaction. This study offers insights on crisis communication strategiesfor public relations practitioners. The findings highlight that timing matters even when implementing matched response strategies in the intentional crisis condition.

Original languageEnglish
Pages (from-to)3603-3620
Number of pages18
JournalQuality and Quantity
Issue number4
StatePublished - Aug 2023


  • Crisis communication
  • Network analysis
  • Sentiment analysis
  • Situational crisis communication theory
  • Topic modeling
  • United Airlines


Dive into the research topics of 'Comparing public responses to apologies: examining crisis communication strategies using network analysis and topic modeling'. Together they form a unique fingerprint.

Cite this