Bits and Biases: Exploring Perceptions in Human-like AI Interactions

Using the Stereotype Content Model

Fernando Jorge F. Macieira

, Diego Costa Pinto

, Tiago Oliveira

and Mitsuru Yanaze

NOVA Information Management School (NOVA IMS), Universidade NOVA de Lisboa,

Campus de Campolide, Lisboa, Portugal

Escola de Comunicações e Artes, Universidade de São Paulo, Av. Prof. Lúcio Martins Rodrigues, 443, São Paulo, Brazil

Keywords: SCM, CASA, AI, Chatbot, Anthropomorphism.

Abstract: In an AI-infused world, user trust in responses generated by autonomous systems is of critical importance.

Building upon the work of Ahn, Kim, and Sung (2022), this study examines the impact of stereotypes

attributed to chatbots on user trust using the Stereotype Content Model (SCM), which relies on dimensions

like warmth and competence for universal cross-culture social judgment. This research investigates how age-

related stereotypes influence user perceptions of anthropomorphic AI, specifically chatbots, and their

perceived warmth and competence. We conducted two experiments: Study 1 used AI-generated illustrations

to present "young" and "old" chatbot personas, while Study 2 used realistic photos. Participants watched pre-

recorded interactions with the chatbot "Dave" and evaluated its warmth and competence on a 9-point Likert

scale. Data were collected through Prolific, ensuring a diverse sample. Study 1 found no significant

differences in perceptions of warmth and competence between the young and old chatbot personas. However,

Study 2 revealed that the younger persona was perceived as warmer than the older one, indicating that the

realism of the chatbot's appearance affects stereotype activation. These results underscore the importance of

aligning chatbot personas with user expectations to enhance trust and satisfaction.

1 INTRODUCTION

Artificial Intelligence (AI) language models are the

most recent “hype” in technology. New models and

generative AI are “born” every day, and AI usage has

become almost ubiquitous. However, its acceptance

may depend on how trustful (Choung, David, and

Ross, 2023) and friendly it is considered to be (Tay,

Jung, and Park, 2014) and its “human likeness” (Kim,

Kang, and Bae, 2022).

To facilitate this interaction and improve AI’s

trustfulness, research suggests that anthropomorphism

(i.e. applying human characteristics to inanimate

objects) can make human-AI agent interactions more

familiar—nonetheless, guided by the same norms

governing interpersonal relationships. (Aggarwal and

McGill, 2012; Ahn, Kim, and Sung, 2022; Sreejesh

and Anusree, 2017)

Hence, like human interactions, these relationships

are also influenced by how we perceive and evaluate

others’ “personality” traits (Ahn, Kim, and Sung,

2022). Since people interact with anthropomorphic

robots and computers as if they were real humans

(Nass et al., 1995; Nass, Steuer, and Tauber, 1994),

various social traits, gender, and personality are

important for interpersonal relationships, affecting

relationships and evoking social stereotypes (Tay,

Jung, and Park, 2014).

Though most studies regarding AI and stereotypes

are focused on biases that appear in the agent

responses, a few scholars have studied how these

stereotypes may affect the trust in AI agent’s

responses. The study of Ahn, Kim and Sung (2022) is

an example; it explores the effects of gender

stereotypes on evaluating AI recommendations, using

the Stereotype Content Model (SCM) of Fiske,

Cuddy and Glick (2007) which relies on dimensions

like warmth and competence for universal cross-

culture social judgment (Fiske, 2017; Fiske, Cuddy,

and Glick, 2007). Others like Liu et al. (2022) have

used the same scale applied to brands (Kervyn, Fiske,

and Malone, 2022) and different anthropomorphic

representations (El Hedhli et al., 2023).

Despite their interesting findings, the authors

acknowledge the need for more research using other

types of stereotypes, such as age, religion, social

Macieira, F. J. F., Pinto, D. C., Oliveira, T. and Yanaze, M.

Bits and Biases: Exploring Perceptions in Human-like AI Interactions Using the Stereotype Content Model.

DOI: 10.5220/0013192700003956

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

In Proceedings of the 7th International Conference on Finance, Economics, Management and IT Business (FEMIB 2025), pages 161-166

ISBN: 978-989-758-748-1; ISSN: 2184-5891

161

status, etc., and cross-cultural versions (Ahn, Kim,

and Sung, 2022). Building on this, we conducted

experiments focusing on age as the stereotype to be

elicited..

This paper is structured in 6 sections: Section 2

offers a brief theoretical background on AI and

human interaction and a deeper one on the concept of

the “Computers Are Social Actors (CASA)”

paradigm, as well as the Stereotype Content Model—

the scale used to measure the expected results. In

Section 3 and its subsections, we explain the

experiments used and discuss the findings. Section 4

comprises conclusions and a brief general discussion;

sections 5 and 6, address implications followed by

limitations and future research suggestions.

2 THEORETICAL FRAMEWORK

AI technology is reshaping the service industry,

creating new service experiences with substantial

implications for both customers and managers (Doorn

et al., 2017) and to make these human-machine

interactions more familiar and easier, companies and

developers use anthropomorphism (applying human

characteristics to inanimate objects) (Aggarwal and

McGill, 2012; Ahn, Kim, and Sung, 2022; Sreejesh

and Anusree, 2017).

Individuals tend to attribute human-like qualities

to entities that exhibit distinct human traits, like

smiling expressions (Epley, Waytz, and Cacioppo,

2007). Prior research has highlighted the

effectiveness of specific cues in conferring human-

like attributes on inanimate objects and social agents.

Notably, characteristics like human-like physical

forms, perceived animacy, and interactivity induce

anthropomorphism in objects (El Hedhli et al., 2023).

Researchers on human-machine interaction like

Nass, Steuer and Tauber (1994) and Nass et. al (1995)

have discovered that people interact with

anthromorphic robotic/machine entities as if they

were human, resembling regular human-human

interactions; therefore, proposing a paradigm called

CASA—Computers Are Social Actors.

The CASA paradigm avers that people will infer

computers’ “personalities” through cues received

during interactions and will respond to these

personalities as if they were human interactions (Nass

et al., 1995). Accordingly, the CASA paradigm is

often used to understand how people perceive AI in

the context of human-computer interaction (Hong,

Choi, and Williams, 2020). More recently, De

Kervenoael et al. (2024) reinforced the CASA

paradigm when interacting with service robots in

retail environments, meaning that the concept still

applies to human-computer interactions.

CASA studies consider human-human interactions

and expect similar results when replicating them in a

human-computer interaction (Edwards et al., 2019).

Hence, the present study focuses on experimenting

with human age-related stereotypes and technology

knowledge in a chatbot interaction, as it would occur

in a common human-human interaction.

Considering these similarities with human-human

interactions, we propose to use the Stereotype

Content Model (Fiske et al., 2002; Fiske, Cuddy, and

Glick, 2007; Fiske, 2017; Kervyn, Fiske, and Malone,

2022) to assess how people evaluate machine entities

regarding prejudices and social groups.

The Stereotype Content Model (SCM) posits that

the stereotypes are encapsulated by two fundamental

dimensions: warmth and competence (Fiske et al.,

2002; Fiske, Cuddy, and Glick, 2007; Cuddy et al.,

2009).

When individuals interact with someone for the

first time, they instinctively assess whether the person

exhibits benevolent and cooperative intentions,

reflecting qualities associated with warmth. These

qualities encompass traits like honesty, friendliness,

and sincerity. Concurrently, they evaluate the person's

capacity to translate these intentions into action,

marking competence—characterized by attributes like

knowledge, creativity, and efficiency (Cuddy, Glick,

and Beninger ,2011; El Hedhli et al., 2023).

Thus, warmth signifies how individuals perceive

others’ intentions. When someone is perceived as

well-intentioned, they are generally considered

trustworthy; in contrast, competence pertains to an

individual's ability to actualize their intentions.

Despite some concerns expressed by Friehs et al.

(2022), the SCM has been used as a validated scale in

various stereotype studies, regarding humans, and

brands (Kervyn, Fiske, and Malone, 2022; Fournier

and Alvarez, 2012; Liu et al., 2022), artificial

intelligence (Kim, Kang, and Bae, 2022; Ahn, Kim,

and Sung, 2022), robots (Tay, Jung, and Park, 2014)

and even virtual influencers (El Hedhli et al., 2023).

Fiske et al. (2002) explicitly report in SCM

studies about the age stereotype, especially related to

older people being considered as having lower

perceived competence and higher warmth. Thus, the

main contribution brought by this article is integrate

SCM and CASA, using experimental design studies

to assess how people evaluate their dimensions of

warmth and competence when interacting with

different chatbot personas (younger vs. older).

We hypothesize that chatbot’s persona (young vs.

old) has a more positive impact on warmth when

FEMIB 2025 - 7th International Conference on Finance, Economics, Management and IT Business

162

older and a more positive impact on competence

perception when younger. Accordingly, younger

chatbot personas will enhance the perception of

competence while simultaneously diminishing

perceptions of warmth. On the other hand, at least

regarding technology, older AI personas are

evaluated as being warmer but presenting less

competence than the younger ones.)

3 OVERVIEW OF STUDIES

In our research, we designed two experiments (Study

1 and Study 2) to assess the influence of the age of AI

personas on their evaluation, grounded in the SCM

framework (Fiske, Cuddy, and Glick, 2007). In these

studies, we intend to systematically vary AI personas’

age cues to assess their respective impacts on the

perceived dimensions of warmth and competence.

The study was designed by requesting participants

to evaluate a pre-recorded interaction between a

chatbot named “Dave” and a user seeking

recommendations for purchasing a new wireless

mouse. We chose the wireless mouse product based

on previous studies where the product was found to

be relevant but utilitarian, lacking significant

emotional involvement or hedonic features (Ahn,

Kim, and Sung, 2022). To provide hints about the

chatbot's personality, we used AI-generated images.

All chatbots presented in the pre-recorded

interactions were referred to simply as "Dave." Other

chatbot characteristics, such as the text used,

remained identical for all interactions.

In both studies, participants were randomly

directed to one of the interactions (young vs. old) and

responded to six questions related to the SCM using

a 9-point Likert scale, attention checks, and

manipulation checks. Additionally, some questions

related to participant profiles, such as age and gender,

were included.

3.1 Study 1

In this study, we aimed to identify the impacts on the

"warmth" and "competence" dimensions resulting

from variations in the age of the persona adopted by

the chatbot.

Using the Prolific platform, we collected 203

observations, one of which did not pass the attention

check (“What was the name of the chatbot on the

interaction you just saw?”). This was excluded from

the study, leaving 202 valid observations, which were

divided into 100 in the "Young Dave" condition

(watching the interaction with the young chatbot) and

102 in the "Old Dave" condition (watching the

interaction with the old chatbot). Figure 1 shows the

pictures used to represent the chatbot’s persona in the

first experiment.

Besides seeking confirmation of the chatbot's

name (Dave), we created a manipulation check, in

which the participant had to evaluate the chatbot's

age. Participants rated Old Dave's age as 59 on

average (Std. Deviation = 19,8) and Young Dave's

age as 28 on average (Std. Deviation = 13,5).

Therefore, we considered all 202 observations as

valid. Female participants comprised 60% of the

sample and male participants 40% (one participant

did not disclose gender), all of them declaring English

to be their main language. The average age of the

participants was 37,1 years—37,7 for female

participants and 36,3 for males.

Figure 1: Illustrations used as stimuli to impersonate

Old/Young Dave chatbot in study 1.

Preliminary analysis using SPSS indicated that

the questionnaire related to the SCM demonstrated

satisfactory validity (COMPETENCE Cronbach’s

Alpha = 0.914 / WARMTH Cronbach’s Alpha =

0,910), confirming the scale's applicability.

3.2 Results

We conducted a MANOVA to identify significant

differences in the means of the "warmth" and

"competence" dimensions among participants

assigned to each of the conditions of AI_PERSONA

(Old Dave and Young Dave). The multivariate

analysis did not reveal any significant differences in

the perceptions of these two dimensions.

Contrary to previous work on SCM and CASA, it

seems—at least relating to illustration-like chatbot

personas—that age stereotypes are not elicited.

Therefore, our main hypothesis, that the “older” Dave

chatbot would elicit the elderly people stereotype

(being warmer and less competent than “young”

Dave) was not supported. Due to the non-significant

differences in the previous study, we redesigned the

experiment using more realistic photos, instead of

illustrations – as the use of illustration-like stimuli

could interfere with the perceived anthropomorphism

level.

Bits and Biases: Exploring Perceptions in Human-like AI Interactions Using the Stereotype Content Model

163

Table 1: Descriptive Statistics and MANOVA results for

study 1 – (Young Dave vs Old Dave).

3.3 Study 2

Due to our results and considerations on Study 1, we

replicated the same experiment, with the same pre-

recorded interaction, but with more realistic

anthropomorphism (Figure 2) instead of illustration-

like pictures. Like the previous experiment, the

following images were AI generated.

Figure 2: More realistic pictures used as stimuli to

impersonate Real Old/Young Dave chatbot in Study 2.

Using Prolific again, we collected 206 completed

questionnaires. Four respondents did not pass the

attention check (“What was the name of the chatbot

on the interaction you just saw?”) and were excluded,

leaving 202 valid observations: 100 in the "Young

Dave" condition (watching the interaction with the

young chatbot) and 102 in the "Old Dave" condition

(watching the interaction with the old chatbot).

Female participants constituted 60,8% and male

participants (two participants identified as non-

binary) 39,1%, all of whom declared that English was

their primary language. The average age of the

participants was 38,2 — 38,0 for female participants

and 38,9 for males.

Besides seeking confirmation of the chatbot's

name (Dave) as an attention check, we also conducted

a manipulation check, in which the participant had to

evaluate the chatbot's age (N=200). Participants rated

Old Dave's age as 44,8 on average (N=100; Std.

Deviation = 16,0) and Young Dave's age as 27,2 on

average (N=100; Std. Deviation = 11,3).

3.4 Results

Again, after verifying SCM scale validity

(COMPETENCE Cronbach’s Alpha = 0.883 /

WARMTH Cronbach’s Alpha = 0,904), we

conducted a MANOVA to identify significant

differences in the means of the "warmth" and

"competence" dimensions among respondents

assigned to each of the conditions of chatbot’s

persona (Old Dave vs. Young Dave).

Table 2: Descriptive Statistics and MANOVA results for

study 2 – young/old with more realistic images.

In accordance with our initial analysis and

understanding, more realistic representations used as

stimuli were in fact able to create differences in

respondents’ perceptions. However, while we

expected the older persona to be rated warmer, the

REAL YOUNG DAVE was rated warmer than the

REAL OLD DAVE instead (OLD DAVE = 5,6340;

YOUNG DAVE = 6,1961).

Since the average estimated age of REAL OLD

DAVE was 45, we postulate that the perceived age

was not significant enough to characterize the picture

as “elderly”. Therefore, the associated prejudices did

not arise.

4 CONCLUSIONS

Our research explores the impact of age-related

variations in a chatbot's persona on the two

fundamental dimensions of "warmth" and

"competence." In two separate studies, we presented

participants with interactions featuring a chatbot

named "Dave" portrayed as young and older

individuals.

However, a preliminary analysis in the first study

did not reveal any significant differences in the

perceptions of these dimensions between the two age

groups. This outcome suggests that the abstract and

less anthropomorphic nature of illustrations might

have hindered the activation of age-related

stereotypes. The result goes against previous SCM

and CASA research, implying that at least when

experimenting with illustration-like AI generated

images as chatbot’s personas age stereotypes are not

elicited. Consequently, considering these results and

theoretical base, we conducted a follow-up

experiment replacing the illustrations with more

realistic photos to better understand whether the

persona's realism would influence the participants'

judgments. In contrast, in the second experiment, the

more realistic stimuli had effect on perceptions –

highlighting how more lifelike depictions may

prompt stronger stereotype-related judgments.

Dependent Variable df F Sig. Chatbot Persona Means Std. Deviation N

Old Dave 6,4854 1,71992 103

Young Dave 6,5657 1,66834 99

Old Dave 6,0065 1,88272 103

Young Dave 5,9394 1,88283 99

0,737

0,8

COMPETENCE

WARMTH

0,113

0,064

Dependent Variable df F Sig. Chatbot Persona Means Std. Deviation N

Old Dave 6,2680 1,82938 102

Young Dave 6,6800 1,51455 100

Old Dave 5,6340 2,06789 102

Young Dave 6,2000 1,45412 100

COMPETENCE 1 3,034 0,083

WARMTH 1 5,046 0,026

FEMIB 2025 - 7th International Conference on Finance, Economics, Management and IT Business

164

However, contrary to the hypothesized outcome

that the older persona would exhibit higher warmth,

the younger persona was rated warmer, likely because

the perceived age of the older persona was

insufficient to evoke elderly-related biases. These

findings highlight the importance of ensuring that

stimuli used in stereotype studies effectively convey

the intended traits, since prior research suggests that

stereotype activation depends on clear and

pronounced cues (Fiske, 2017).

5 MANAGERIAL IMPLICATIONS

On a theoretical level, our study contributes to the

ongoing discourse surrounding human-AI interactions

by shedding light on the complex interplay between

anthropomorphism, stereotypes, and user perceptions.

By exploring the impact of age stereotypes on

evaluations of chatbot personas, we extend existing

research on the "computers are social actors"

paradigm, demonstrating its relevance in

contemporary AI applications. Moreover, our

findings raise questions about the role of stereotypes

in shaping user behaviors in and evaluations of

human-AI interactions.

Partially reinforcing the principles outlined in the

"computers are social actors" paradigm (Nass et al.,

1995), our study offers insights and questions for

managers considering the deployment of or already

having chatbots deployed for customer service –

where chatbots are seen as a promising technology for

service providers (Nicolescu & Tudorache, 2022). It

is necessary to seriously research and study which

"persona" or personality to attribute to chatbots

because they directly influence the perceptions

consumers derive from their interactions (Lian &

Lian, 2023). Choosing the appropriate persona for

chatbots becomes relevant as it directly influences

consumers' perceptions of warmth and competence,

which in turn affect their trust (Choung et al., 2023)

and satisfaction levels (Hsu & Lin, 2023). Thus, we

emphasize the importance of matching chatbot

personas with user expectations to optimize trust and

engagement, a critical insight for digital customer

service managers seeking to enhance user satisfaction

and the effectiveness of their channels.

6 LIMITATIONS

Despite its limitations, our study contributes to the

dynamics of human-AI interactions, particularly

regarding the implications of anthropomorphism and

stereotypes on user perceptions. While the use of pre-

recorded interactions in this study offers a controlled

environment for assessing respondents’ perceptions,

it lacks the complexity of real-world interactions.

Future research could benefit from incorporating live

interactions between participants and AI systems to

capture the dynamic nature of human-AI interactions

more accurately. By doing so, researchers could

observe how perceptions of warmth and competence

evolve over the course of a conversation, providing

insights into the sustained impact of

anthropomorphism and stereotypes in longer

interactions.

Moreover, the study's focus on a single utilitarian

product (wireless mouse) limits the generalizability

of its findings to other contexts and how different

personas may have different evaluations according to

the subject of interaction. To address this limitation,

future research could explore interactions with a

broader range of products, including those with

varying levels of emotional involvement or hedonic

features.

Furthermore, while the study examines age

stereotypes, there is a notable absence of exploration

into other types of stereotypes that may influence

perceptions of chatbot personas. Future research

could expand upon this by investigating how

stereotypes related to gender, ethnicity, socio-

economic status, or technological expertise impact

users' willingness to accept suggestions and engage

with AI systems. By examining a wider array of

stereotypes, researchers can gain a more

comprehensive understanding of the factors that

shape users' perceptions and behaviors in human-AI

interactions.

REFERENCES

Aggarwal, Pankaj, and Ann L. McGill. 2012. “When

Brands Seem Human, Do Humans Act Like Brands?

Automatic Behavioral Priming Effects of Brand

Anthropomorphism.” Journal of Consumer Research

39 (2): 307–23.

Ahn, Jungyong, Jungwon Kim, and Yongjun Sung. 2022.

“The Effect of Gender Stereotypes on Artificial

Intelligence Recommendations.” Journal of Business

Research 141: 50–59.

Choung, Hyesun, Prabu David, and Arun Ross. 2023.

“Trust in AI and Its Role in the Acceptance of AI

Technologies.” International Journal of Human–

Computer Interaction 39 (9): 1727–39.

Cuddy, Amy J. C., Susan T. Fiske, Virginia S. Y. Kwan,

Peter Glick, Stéphanie Demoulin, Jacques-Philippe

Bits and Biases: Exploring Perceptions in Human-like AI Interactions Using the Stereotype Content Model

165

Leyens, Michael Harris Bond, et al. 2009. “Stereotype

Content Model across Cultures: Towards Universal

Similarities and Some Differences.” British Journal of

Social Psychology 48 (1): 1–33.

Cuddy, Amy J.C., Peter Glick, and Anna Beninger. 2011.

“The Dynamics of Warmth and Competence

Judgments, and Their Outcomes in Organizations.”

Research in Organizational Behavior 31 (January): 73–

98.

De Kervenoael, Ronan, Alexandre Schwob, Rajibul Hasan,

and Evangelia Psylla. 2024. “SIoT Robots and

Consumer Experiences in Retail: Unpacking Repeat

Purchase Intention Drivers Leveraging Computers Are

Social Actors (CASA) Paradigm.” Journal of Retailing

and Consumer Services 76 (January): 103589.

Doorn, Jenny van, Martin Mende, Stephanie M. Noble,

John Hulland, Amy L. Ostrom, Dhruv Grewal, and J.

Andrew Petersen. 2017. “Domo Arigato Mr. Roboto:

Emergence of Automated Social Presence in Organiza-

tional Frontlines and Customers’ Service Experiences.”

Journal of Service Research 20 (1): 43–58.

Edwards, Chad, Autumn Edwards, Brett Stoll, Xialing Lin,

and Noelle Massey. 2019. “Evaluations of an Artificial

Intelligence Instructor’s Voice: Social Identity Theory

in Human-Robot Interactions.” Computers in Human

Behavior 90 (January): 357–62.

El Hedhli, Kamel, Haithem Zourrig, Amr Al Khateeb, and

Ibrahim Alnawas. 2023. “Stereotyping Human-like

Virtual Influencers in Retailing: Does Warmth Prevail

over Competence?” Journal of Retailing and Consumer

Services 75 (November): 103459.

Epley, Nicholas, Adam Waytz, and John T. Cacioppo.

2007. “On Seeing Human: A Three-Factor Theory of

Anthropomorphism.” Psychological Review 114 (4):

864–86.

Fiske, Susan T. 2017. “Prejudices in Cultural Contexts:

Shared Stereotypes (Gender, Age) Versus Variable

Stereotypes (Race, Ethnicity, Religion).” Perspectives

on Psychological Science 12 (5): 791–99.

Fiske, Susan T., Amy J. C. Cuddy, Peter Glick, and Jun Xu.

2002. “A Model of (Often Mixed) Stereotype Content:

Competence and Warmth Respectively Follow from

Perceived Status and Competition.” Journal of

Personality and Social Psychology 82 (6): 878–902.

Fiske, Susan T., Amy J.C. Cuddy, and Peter Glick. 2007.

“Universal Dimensions of Social Cognition: Warmth

and Competence.” Trends in Cognitive Sciences 11 (2):

77–83.

Fournier, Susan, and Claudio Alvarez. 2012. “Brands as

Relationship Partners: Warmth, Competence, and in-

Between.” Journal of Consumer Psychology 22 (2):

177–85.

Friehs, Maria-Therese, Patrick F. Kotzur, Johanna

Böttcher, Ann-Kristin C. Zöller, Tabea Lüttmer, Ulrich

Wagner, Frank Asbrock, and Maarten H. W. Van Zalk.

2022. “Examining the Structural Validity of Stereotype

Content Scales – A Preregistered Re-Analysis of

Published Data and Discussion of Possible Future

Directions.” International Review of Social Psychology

35 (1): 1.

Hong, Joo-Wha, Sukyoung Choi, and Dmitri Williams.

2020. “Sexist AI: An Experiment Integrating CASA

and ELM.” International Journal of Human–Computer

Interaction 36 (20): 1928–41.

Hsu, Chin-Lung, and Judy Chuan-Chuan Lin. 2023.

“Understanding the User Satisfaction and Loyalty of

Customer Service Chatbots.” Journal of Retailing and

Consumer Services 71 (March): 103211.

Kervyn, Nicolas, Susan T. Fiske, and Chris Malone. 2022.

“Social Perception of Brands: Warmth and Competence

Define Images of Both Brands and Social Groups.”

Consumer Psychology Review 5 (1): 51–68.

Kim, Juran, Seungmook Kang, and Joonheui Bae. 2022.

“Human Likeness and Attachment Effect on the

Perceived Interactivity of AI Speakers.” Journal of

Business Research 144 (May): 797–804.

Lian, Lee Kim, and Song Bee Lian. 2023. “Examining

Anthropomorphism of Chatbots and Its Effect on User

Satisfaction and User Loyalty in the Service Industry.”

Electronic Journal of Business and Management 8 (1):

1–14.

Liu, Fu, Haiying Wei, Zhenzhong Zhu, and Haipeng

(Allan) Chen. 2022. “Warmth or Competence: Brand

Anthropomorphism, Social Exclusion, and

Advertisement Effectiveness.” Journal of Retailing and

Consumer Services 67 (July): 103025.

Nass, Clifford, Youngme Moon, B. J. Fogg, Byron Reeves,

and D. Christopher Dryer. 1995. “Can Computer

Personalities Be Human Personalities?” International

Journal of Human-Computer Studies 43 (2): 223–39.

Nass, Clifford, Jonathan Steuer, and Ellen R. Tauber. 1994.

“Computers Are Social Actors.” In Proceedings of the

SIGCHI Conference on Human Factors in Computing

Systems Celebrating Interdependence - CHI ’94, 72–

78. Boston, Massachusetts, United States: ACM Press.

Nicolescu, Luminița, and Monica Teodora Tudorache.

2022. “Human-Computer Interaction in Customer

Service: The Experience with AI Chatbots—A

Systematic Literature Review.” Electronics 11 (10):

1579.

Sreejesh, S., and M.R. Anusree. 2017. “Effects of

Cognition Demand, Mode of Interactivity and Brand

Anthropomorphism on Gamers’ Brand Attention and

Memory in Advergames.” Computers in Human

Behavior 70 (May): 575–88.

Tay, Benedict, Younbo Jung, and Taezoon Park. 2014.

“When Stereotypes Meet Robots: The Double-Edge

Sword of Robot Gender and Personality in Human–

Robot Interaction.” Computers in Human Behavior 38

(September): 75–84.

FEMIB 2025 - 7th International Conference on Finance, Economics, Management and IT Business

166