Analysis of Age and Gender Differences in the Chongqing “Fat Cat”

Incident

Jihang Chen

1,*

, Yuanxi Wang

and Zonghao Zhao

Faculty of Finance, City University of Macau, Macau, 999078, China

Kunshan High School of Jiangsu Province, Suzhou, 215300, China

College of Business, Nanfang College Guangzhou, Guangzhou, 51000, China

＊

Keywords: Social Media, KOL, Opinion Leader, Reversed News Events, Social Network Analysis.

Abstract: Taking the Chongqing “Fat Cat” incident as the starting point, this study aims to analyse the behaviour pattern

of opinion leaders in news events, explore its impact on public opinion and cognition. The present study

collects data through questionnaires, and analyses the public's attention to the event, initial attitude, cognitive

change, etc. from the perspective of age and gender differences. The study found that young people (especially

18-25-year-olds) pay attention to events early, have deep emotional investment, change their attitudes, and

rely more on the comments of online opinion leaders and family members; men are more inclined to obtain

news through TikTok based on factual evidence, and have less suspicion of potential manipulation; women

are more sceptical about “fat cats”, feeling sister's speech and emotional content more sensitive, believing

that the reversal of the incident has a greater impact on social trust. This research provides a reference basis

for public opinion monitoring, crisis public relations and other fields.

1 INTRODUCTION

In the era of information explosion, KOLs play an

increasingly critical role in news dissemination,

public opinion guidance, and shaping public

cognition. Analyzing their behavioral patterns helps

comprehend the mechanisms of opinion formation,

predict trends in public discourse, and provide

references for relevant authorities in formulating

strategies for opinion guidance. However, existing

researches on KOL predominantly focuses on their

influences and characteristics, with limited empirical

studies on their behavioral patterns, particularly in the

context of news events.

This study aims to address this gap and enrich the

theoretical framework of KOL. Additionally, the

findings can be applied to fields such as public

opinion monitoring, crisis public relations, and brand

marketing, assisting institutions in better identifying,

guiding, and leveraging the influence of opinion

leaders.This study is based on the “Chongqing Fat Cat

Incident”: ”Fat cat (Mr. Liu)” committed suicide into

a river, and he accused his girlfriend “Tan Zhu”

obtaining money under false pretences before his

death; the KOL “Fat Cat’s sister” quoted this point

and carried out cyber-violence to Tan Zhu including

making up rumours and exposing Tan’s privacy.

Eventually, the police convinced most of the

discussions on Tan are fake. Meanwhile, the authors

analyze the “behavioral patterns of KOL in events,”

addressing the following key research questions:

1.What are the behavioral patterns and distinctive

characteristics of opinion leaders in news events?

2.How do the dissemination patterns of opinion

leaders and audience cognition evolve?

3.Can opinion leaders influence public opinion?

If so, through what specific behaviors?

2 LITERATURE REVIEW

The authors conducted a search for relevant academic

papers on this topic. On Google Scholar, there are

approximately 7,500 journal articles. When limiting

the searches to publications before 2021, only around

1,200 articles remained, indicating a decline in

academic attention to this subject.

The author first reviewed several research papers

on opinion leaders. The following section presents a

comprehensive analysis of the reviewed studies.

346

Chen, J., Wang, Y. and Zhao, Z.

Analysis of Age and Gender Differences in the Chongqing “Fat Cat” Incident.

DOI: 10.5220/0013991600004916

Paper published under CC license (CC BY-NC-ND 4.0)

In Proceedings of the 2nd International Conference on Public Relations and Media Communication (PRMC 2025), pages 346-350

ISBN: 978-989-758-778-8

In the study by Wu, Zhao, and Gao（Wu et al.

2019), the researchers examined the identifications

and influences of KOL on Weibo during the 2018

vaccine incident, incorporating the life cycle

characteristics of public opinion events. The event

was divided into four stages: outbreak (initial surge in

public discourse), heated discussion (sustained debate

with participation from key opinion leaders (KOLs)),

decline (controlled moderation of discourse), and

residual (stabilized attention with focus on official

investigations). The study integrated user attributes,

network features, behavioral traits, and textual

features to construct a comprehensive indicator

system, mitigating biases from single variables.

Methodologies included user feature extraction,

cluster analysis (using K-means algorithm), and time-

lag correlation analysis to assess the impact of

opinion leaders' emotional tendencies on public

sentiment. Findings revealed distinct differences in

public opinion hotspots and network structures across

stages, with media-type opinion leaders maintaining

consistent influence while individual self-media and

unverified users exhibited stage-dependent

variability. Neutral and negative sentiments from

opinion leaders preceded public sentiment shifts,

whereas positive sentiments lagged. Other

researchers also concentrate on the impact of KOL on

Weibo as well (Liu, 2020; Ma, 2022). In addition,

Previous studies also examine the impact of KOL on

public opinion (Chen & Wang, 2023; Nian & Zhang,

2005; Shen et al., 2023; Yu, 2018). Compared to other

studies, this study offered a more holistic approach to

identifying opinion leaders, particularly grassroots

influencers, and explored their role in opinion

guidance. Limitations included a single-case focus

and lack of model validation across diverse public

opinion events, suggesting future research should

expand to multi-event, multi-platform analyses for

greater generalizability.

Wang and Long （ Wang & Long, 2024)

investigated KOL on Douyin (TikTok) and their

political commentary videos, employing speech and

text sentiment analysis to empirically assess the

impact of opinion leaders' emotional tendencies on

netizen sentiment. Control variables included gender,

account type, follower count, video themes, and user

interactions. A convolutional neural network (CNN)

and panel regression model were used for analysis.

Results indicated an inverted U-shaped relationship

between opinion leaders' emotional tendencies and

netizen sentiment, with no significant effect on

sentiment polarization. Heterogeneity tests showed

male opinion leaders and hosts exerted stronger

emotional influence, while intermediate netizen

groups were more susceptible. The study innovatively

combined speech and text sentiment analysis,

providing nuanced insights for online public opinion

governance. Limitations involved imperfect

sentiment classification accuracy, limited temporal

scope, and unaddressed confounding factors,

warranting future research with expanded samples,

optimized models, and extended timelines.

Xiong and He (Xiong & He, 2013) analyzed

Weibo repost networks under the “tiered electricity

pricing” topic, proposing an improved Hyperlink-

Induced Topic Search （HITS ） algorithm (HITS-

BOWR) incorporating repost frequency and follower

count as weights to enhance opinion leader

identification. Social network analysis via University

of California at Irvine NETwork （ UCINET ）

revealed that opinion leaders occupied critical nodes

in information dissemination, with centrality scores

strongly correlated to follower counts. The study

addressed limitations of traditional PageRank

algorithms in microblog contexts but overlooked

sentiment analysis and broader topic validation.

Future research could incorporate additional

weighting metrics and expand topic coverage.

Liu and Liu (Liu & Liu,2017) studied community

network structures and opinion leader traits on Zhihu

(Quora-like platform), focusing on 1,765 users

discussing vaccines. Social network analysis (SNA)

via UCINET demonstrated sparse network topology

with rapid information diffusion, where opinion

leaders—often professionals—leveraged high-

quality contributions to sustain influence. Findings

aligned with existing literature on knowledge-sharing

platforms but were constrained by single-topic

sampling and cross-sectional study. Longitudinal

multi-topic studies were recommended.

The present study derives the following

perspectives:1. Factors Influencing Public Opinion in

Emergencies:

All five studies underscored the role of KOLs in

shaping public opinion. Wang Yijun and Long

Miaomiao highlighted their dual function in

information dissemination and opinion guidance,

with emotional tendencies directly impacting netizen

sentiment. Wu Jiang et al. further analyzed stage-

specific behavioral patterns in healthcare incidents,

while Xiong Tao and Liu Yunong emphasized

network centrality. Collectively, opinion leaders'

emotional valence, activity levels, network position,

Analysis of Age and Gender Differences in the Chongqing “Fat Cat” Incident

347

and topic relevance were identified as key

determinants.

The studies consistently demonstrated that

opinion leaders significantly sway public sentiment

during crises. Wang Yijun’s inverted U-model

suggested moderated positivity optimizes influence,

whereas Wu Jiang’s time-lag analysis confirmed

emotional leadership. Limitations included narrow

scopes and unvalidated models, calling for cross-

platform, multi-event validation.

3 METHOD

3.1 Research Subjects and Samples

The research subjects are Internet users who are over

18 years old and have known or been exposed to the

“Chongqing Fat Cat Incident”. The questionnaire is

distributed in the form of questionnaire stars through

social platforms (such as WeChat groups, QQ groups,

etc.), and the data collection time is from March to

April 2025. Using the convenient sampling method, a

total of 235 valid questionnaires were finally

obtained. The samples covered multiple age groups

and educational levels to meet the needs of follow-up

analysis.

3.2 Questionnaire Design

The questionnaire consists of five parts:

1. Basic information: including age group,

gender, education, frequency of social media use,

etc., a total of 4 items;

2. Information contact and the influence of

opinion leaders: such as whether the opinion leader's

speech changes his views, the main role of opinion

leaders, etc., a total of 6 items;

3. Factors affecting the behaviour of opinion

leaders: such as which opinions leaders' behaviours

the respondents think are the most influential (such as

publishing evidence, using emotional language,

emotional guidance, etc.), a total of 2;

4. Changes in audience attitudes and cognitions:

such as attitudes in the face of reverse news, whether

they have been misled, whether they have been

influenced by certain remarks, etc., a total of 4 items;

5. The causes and coping strategies of reversal

news: for example, the interviewees think that the

reasons for the reversal, whether it affects social trust,

how to reduce the negative impact, etc., a total of 3

items.

Most of the questions are single-choice or

multiple-choice questions. Some questions reflect

cognitive changes and behavioural changes, and are

analysed in the form of classification/ordered

variables.

3.3 Data Collection Process

The questionnaire was released through the

“Questionnaire Star” platform from March to April

2025. The questionnaire link was spread on social

media, and the respondents voluntarily participated

and filled in anonymously to ensure the authenticity

and privacy of the answers.

3.4 Data Analysis Methods

The data is analysed using questionnaire stars and

spssau, mainly including:

Descriptive statistical analysis: statistics on the

basic characteristics of samples and the frequency

distribution of core variables;

Cross-analysis: explore the relationship between

age, gender and other background variables and

cognitive attitudes;

The above analysis methods ensure that the

research conclusions are scientific, systematic and

logical.

4 RESULT

4.1 Differences in Attention to the

Incident

4.1.1 Age Characteristics

The group aged 18 and below demonstrated the

highest level of attention, with 100% showing interest

in the incident. Moreover, 36.36% followed the case

during its early stages (the death of “Fat Cat” and the

online rumor of deception), significantly higher than

other age groups.The 19–25 age group also showed

high overall attention (94.59%), yet 5.41% did not

pay attention to the incident—making it the only

young group with a noticeable proportion of

disinterested respondents.

Individuals aged 26–35 had the highest

proportion of sustained attention throughout the

incident (10.91%), with particular interest in the

reversal phase (official police statement clarifying the

fraud rumors), at 29.09%.

PRMC 2025 - International Conference on Public Relations and Media Communication

348

4.1.2 Gender Differences

Males showed slightly higher attention than females

(98.61% vs. 96.70%). However, both genders focused

most intensely on the early stage of the incident

(males 35.42% vs. females 35.16%).

4.2 Initial Attitudes Toward the

Incident

4.2.1 Age Characteristics

Younger respondents were more sympathetic toward

Fat Cat. Among those aged 18 and below, 75.76%

believed Fat Cat was a victim, which was higher than

the average of other age groups (65%).Older

respondents showed a more rational approach. In the

group aged 46 and above, 9.09% chose to “wait for

the official statement,” higher than younger groups.

4.2.2 Gender Differences

Both genders demonstrated a high level of agreement:

approximately 66% of males and 65.93% of females

believed Fat Cat was a victim, indicating minimal

gender impact.

4.3 Cognitive Shifts after the Incident

Reversal

4.3.1 Age Characteristics

Younger individuals were more likely to change their

views. In the group aged 18 and below, 51.52%

“completely changed their opinion,” significantly

higher than the 27.45% in the 36–45 age group.The

26–35 age group was most sensitive to the reversal:

60% chose “completely changed opinion,” and they

relied more on the official police statement (38.18%)

as the basis for judgment.

4.3.2 Gender Differences

Males were more accepting of the reversal: 55.56%

“completely changed opinion,” compared to 50.55%

of females. Females expressed more doubt about Fat

Cat’s sister’s statements (23.08% vs. 20.83% for

males).

4.4 Influence of Opinion Leaders

4.4.1 Age Characteristics

Younger people relied more on online influencers. In

the 19–25 age group, 59.46% believed online

influencers had the greatest influence, higher than the

46 and above group (68.18%).Among those aged 18

and below, 42.42% were most influenced by

emotional language.

4.4.2 Gender Differences

Males placed more trust in factual evidence (39.58%

vs. 32.97% for females), while females were more

sensitive to emotional guidance (41.76% vs. 40.97%

for males).

4.5 Unique Impact of Fat Cat’s Sister

4.5.1 Age Characteristics

The 36–45 age group was most affected: 66.67%

initially believed the fraud story due to Fat Cat’s

sister’s statements, a significantly higher rate than

other groups.Younger individuals were more

skeptical: 20.27% of the 19–25 age group said they

did not follow her statements.

4.5.2 Gender Differences

Females paid more attention to statements by family

members: 30.77% believed such statements affected

their perception, higher than males (20.83%).

4.6 Impact of Opinion Reversal on

Social Trust

4.6.1 Age Characteristics

The group aged 18 and below showed the lowest trust

level: 63.64% believed the reversal reduced social

trust, much higher than the 36–45 age group

(37.25%).Older respondents were more pessimistic:

59.09% of those aged 46 and above believed trust had

declined.

4.6.2 Gender Differences

Females were more sensitive: 54.95% believed social

trust declined due to the reversal, compared to

45.83% of males.

4.7 Core Findings

For the young age group, they paid attention early,

were deeply emotionally involved, and experienced

strong changes in attitude after the reversal. However,

they were also more reliant on online influencers and

family members’ statements. For males, they were

more likely to use Douyin to access news (69.44% vs.

Analysis of Age and Gender Differences in the Chongqing “Fat Cat” Incident

349

60.44% for females), were more easily convinced by

factual evidence, but showed lower levels of

suspicion toward potential manipulation (29.86% vs.

47.25% for females). For females, they were more

sensitive to statements made by Fat Cat’s sister and

emotionally charged content, and they perceived a

greater impact on social trust following the reversal.

5 CONCLUSION

Significant age differences: There are obvious

differences in the attention, attitudes, cognitive

changes and influencing factors of different age

groups in terms of the Chongqing “Fat Cat” incident.

Young people pay attention to the event early and

have deep emotional investment. After the reversal,

their attitude changes greatly, and they are easily

influenced by the voice of online opinion leaders and

families. The older group will be more rational than

the young group, and show different focusses and

attitudes in the development of events.

There are obvious gender differences: boys and

girls are different in terms of information acquisition

channels, the degree of dependence on the evidence

released, the degree of attention to the “fat cat” sister's

remarks, and the reaction dependence on social trust

affected by the reversal of events. Men get more news

from TikTok and believe in factual evidence; women

are more sensitive to emotional content and the

remarks of the “fat cat” sister, and believe that the

reversal of the incident has a greater impact on social

trust.

The research is of great significance: The results

of this research provide an important reference for the

official monitoring of public opinion, the crisis of

brand public relations, and the authenticity of online

information by netizens. Relevant institutions can

formulate more targeted strategies according to the

characteristics of different ages and gender groups to

effectively guide public opinion, respond to crises,

control the spread of public opinion, and control false

public opinion.

AUTHORS CONTRIBUTION

All the authors contributed equally and their names

were listed in alphabetical order

REFERENCES

Chen, Z. J., & Wang, N. 2023. Research on the positive

public opinion guidance role of online opinion leaders.

News Sentinel, 15: 17–19

Liu, L. Q. 2020. Discussion on the guidance of public

opinion by Weibo opinion leaders and its strategies.

New Media Research, 6(08): 23–26

Liu, Y. N., & Liu, M. R. 2017. Socialized Q&A platform

community network morphology and characteristics of

opinion leaders: Taking Zhihu.com for example. Media

Studies, 2:

Ma, Y. Y. 2022. Research on the influence of Weibo

opinion leaders on public opinion. Media Forum, 5(03):

23–26

Nian, F., & Zhang, Z. 2025. The influence of opinion

leaders on public opinion spread and control strategies

in online social networks. IEEE Transactions on

Computational Social Systems

Shen, B., Wen, Q., & Zhou, Y. 2023. The influence of key

opinion leader’s attitudes toward netizens on public

opinion. Lecture Notes in Education Psychology and

Public Media, 3: 36–43

Wang, Y. J., & Long, M. M. 2024. Be in the mood: The

influence of opinion leaders on netizens’ emotions in

current affairs and political events. Journal of

University of Electronic Science and Technology of

China (Social Science Edition), 26(5): 95–112

Wu, J., Zhao, Y. H., & Gao, J. H. 2019. Research on Weibo

opinion leaders identification and analysis in medical

public opinion incidents. Social Science Journal, 3(4):

53–62

Xiong, T., & He, Y. 2013. The identification and analysis

of micro-blogging opinion leaders in the network of

retweet relationships. Information Studies, 6: 55–62

Yu, Z. Y. 2018. Research on key opinion leaders' public

sentiment guidance in the era of self-media (Master's

thesis, Jilin University)

PRMC 2025 - International Conference on Public Relations and Media Communication

350