Table 1 Literature review table Study...

Table 1

Literature review table

Study	Theory	Methodology	Findings
Luo et al. (2019)	Emotion management theory	Experimental design	Appraisals and post-recovery emotions sequentially mediated the relationship between emotion regulation and consumer word-of-mouth
Chin et al. (2020)	Not specified	Experimental design	Empathy emerged as the most effective strategy to mitigate aggressive behavior
Sheehan et al. (2020)	Not specified	Experimental design	Unresolved chatbot errors decreased adoption and perceived humanness
Castillo et al. (2021)	Not specified	Interview	Customer resource loss determined customer coping strategies in AI-based service failure and customers passed the blame to technology for service failure
Choi et al. (2021)	Not specified	Experimental design	Warmth robots increased customer dissatisfaction during failures. Humanoids could effectively recover trust with sincere apologies or explanations
Seeger and Heinzl (2021)	Theory of anthropomorphism	Experimental design	Interaction with a human-like chatbot compared to a machine-like chatbot considerably decreased customer trust
Crolic et al. (2022)	Functionalist theory of emotion	Text analysis, experimental design	Too much chatbot humanization undermined service evaluation
Filieri et al. (2022)	Not specified	Text analysis	Customers’ emotions in chatbot interaction included positive emotions such as joy, surprise, interest and excitement. Robots malfunction decreased satisfaction
Huang and Dootson (2022)	Theory of stress and coping	Experimental design	High customer participation increased emotion-focused coping (i.e. frustration, aggression) when the availability of a human assistant was disclosed early on
Pantano and Scarpi (2022)	Multiple intelligences, social interaction theory	Survey after interaction with AI	Visual spatial intelligence affected positive emotions, no effect on negative emotions. Social intelligence affected positive and negative emotions. Verbal intelligence did not affect any type of emotions. Processing speed only affected negative emotions
Brendel et al. (2023)	Frustration–aggression theory	Experimental design	Perceived humanness directly increased the frustration with the chatbot when it produced errors. Perceived humanness increased service satisfaction which in turn reduced frustration. Perceived humanness influences the nature of aggression when users become frustrated
Herhausen et al. (2023)	Theory of arousal	Experimental design, text analysis	High- versus low-arousal emotions reduced gratitude. Active listening and empathy in the firm response de-escalated high arousal emotions and increased gratitude. For low-arousal emotions, there were diminishing effects for active listening while the effect of empathy varied across studies
Liu et al. (2023)	Implicit personality theory	Experiment	Using humorous emojis by chatbots increased consumers' reuse intention through consumers' perceived intelligence
Zhang et al. (2023)	Lay belief and emotional competence	Experimental design	Chatbot apology led to lower customer satisfaction than symbolic recovery from human employees due to chatbots lack of emotional competence
Chen et al. (2025)	Benign violation theory, relief theory	Experiment	Chatbot humor and informal language increased customer perceived failure. Chatbot failures were misunderstanding, lack of competence, personalization, and assurance
Liang et al. (2024)	–	Field and lab experiment	Chatbot gender mattered dealing with angry customers. Male chatbots were suitable for using apology, while female chatbots were suitable when using appreciation strategy
Ozuem et al. (2024)	Frustration–aggression theory	Qualitative design, interview	Customers' frustration and aggression affected both customer loyalty and technology adoption
Tang et al. (2024)	Social affordance	Experiment	Customer anger decreased customer satisfaction and chatbot empathy mitigated such effect
Zhang et al. (2024)	Stress-and-coping theory	Semi-structured interviews	Chatbot failure capabilities were misunderstanding, failing to solve problems, requesting sensitive information, faking humanization. Customers emotions were anger, frustration, betrayal and defeat
This study	Social response theory	Text analysis	This study identified customer incivility behaviors and customer emotions in chatbot interactions, identified chatbot capability failure and recovery-oriented capability and revealed a paradoxical effect of chatbot empathy in chatbot failure contexts

Study	Theory	Methodology	Findings
Luo et al. (2019)	Emotion management theory	Experimental design	Appraisals and post-recovery emotions sequentially mediated the relationship between emotion regulation and consumer word-of-mouth
Chin et al. (2020)	Not specified	Experimental design	Empathy emerged as the most effective strategy to mitigate aggressive behavior
Sheehan et al. (2020)	Not specified	Experimental design	Unresolved chatbot errors decreased adoption and perceived humanness
Castillo et al. (2021)	Not specified	Interview	Customer resource loss determined customer coping strategies in AI-based service failure and customers passed the blame to technology for service failure
Choi et al. (2021)	Not specified	Experimental design	Warmth robots increased customer dissatisfaction during failures. Humanoids could effectively recover trust with sincere apologies or explanations
Seeger and Heinzl (2021)	Theory of anthropomorphism	Experimental design	Interaction with a human-like chatbot compared to a machine-like chatbot considerably decreased customer trust
Crolic et al. (2022)	Functionalist theory of emotion	Text analysis, experimental design	Too much chatbot humanization undermined service evaluation
Filieri et al. (2022)	Not specified	Text analysis	Customers’ emotions in chatbot interaction included positive emotions such as joy, surprise, interest and excitement. Robots malfunction decreased satisfaction
Huang and Dootson (2022)	Theory of stress and coping	Experimental design	High customer participation increased emotion-focused coping (i.e. frustration, aggression) when the availability of a human assistant was disclosed early on
Pantano and Scarpi (2022)	Multiple intelligences, social interaction theory	Survey after interaction with AI	Visual spatial intelligence affected positive emotions, no effect on negative emotions. Social intelligence affected positive and negative emotions. Verbal intelligence did not affect any type of emotions. Processing speed only affected negative emotions
Brendel et al. (2023)	Frustration–aggression theory	Experimental design	Perceived humanness directly increased the frustration with the chatbot when it produced errors. Perceived humanness increased service satisfaction which in turn reduced frustration. Perceived humanness influences the nature of aggression when users become frustrated
Herhausen et al. (2023)	Theory of arousal	Experimental design, text analysis	High- versus low-arousal emotions reduced gratitude. Active listening and empathy in the firm response de-escalated high arousal emotions and increased gratitude. For low-arousal emotions, there were diminishing effects for active listening while the effect of empathy varied across studies
Liu et al. (2023)	Implicit personality theory	Experiment	Using humorous emojis by chatbots increased consumers' reuse intention through consumers' perceived intelligence
Zhang et al. (2023)	Lay belief and emotional competence	Experimental design	Chatbot apology led to lower customer satisfaction than symbolic recovery from human employees due to chatbots lack of emotional competence
Chen et al. (2025)	Benign violation theory, relief theory	Experiment	Chatbot humor and informal language increased customer perceived failure. Chatbot failures were misunderstanding, lack of competence, personalization, and assurance
Liang et al. (2024)	–	Field and lab experiment	Chatbot gender mattered dealing with angry customers. Male chatbots were suitable for using apology, while female chatbots were suitable when using appreciation strategy
Ozuem et al. (2024)	Frustration–aggression theory	Qualitative design, interview	Customers' frustration and aggression affected both customer loyalty and technology adoption
Tang et al. (2024)	Social affordance	Experiment	Customer anger decreased customer satisfaction and chatbot empathy mitigated such effect
Zhang et al. (2024)	Stress-and-coping theory	Semi-structured interviews	Chatbot failure capabilities were misunderstanding, failing to solve problems, requesting sensitive information, faking humanization. Customers emotions were anger, frustration, betrayal and defeat
This study	Social response theory	Text analysis	This study identified customer incivility behaviors and customer emotions in chatbot interactions, identified chatbot capability failure and recovery-oriented capability and revealed a paradoxical effect of chatbot empathy in chatbot failure contexts

Source(s): Authors’ own work

[ViewLarge]

Sharing Unavailable