Loneliness Among Adolescents Who are Users of Chatbot Companionship and Non-Users

 

Loneliness Among Adolescents Who are Users of Chatbot Companionship and Non-Users.

Anmol Purbia¹, Elizabeth Jasmine², Jayashree S³

¹Student, Masters in Counseling Psychology, Indian Institute of Psychology and Research (IIPR), Bangalore, India.

²Professor of Psychology, Indian Institute of Psychology and Research (IIPR), Bangalore, India.

³Assistant Professor, Department of Psychology, Indian Institute of Psychology and Research (IIPR), Bangalore, India.

 

Abstract

This research explores the intricate relationship between chatbot companionship and adolescent loneliness among students in Secondary Education and Higher Secondary Education. As loneliness remains a widespread concern among adolescents, exacerbated by digital interactions that often fall short of fulfilling emotional needs, this study examines how chatbot companionship, a rising AI-driven technology, affects loneliness and explores potential gender differences. Using a quantitative research design, 153 Indian adolescents aged 14-18 participated through a non-probability sampling method. Participants completed demographic questionnaires and the UCLA Loneliness Scale. Data analysis focused on variables including chatbot companionship (usage vs. non-usage), gender (male vs. female), and loneliness. The findings reveal that loneliness is a significant issue among Indian adolescents, with many experiencing moderate to high levels of emotional isolation. Gender differences in loneliness levels were not statistically significant, although females reported slightly higher mean loneliness scores. No significant interaction effect between chatbot companionship and gender on loneliness was found. Notably, experiencing personal loss emerged as a significant predictor of increased loneliness. Additionally, while longer device usage was associated with lower loneliness scores, the daily time spent specifically on chatbot interactions did not significantly influence loneliness. This study contributes to the existing body of knowledge by exploring the specific impact of chatbot companionship on adolescent loneliness in the Indian context, with a focus on gender differences.

Keywords: Loneliness, AI Chatbots, Indian Adolescents, Chatbot Companionship.

 

Introduction

One of the most defining characteristics of being human is our basic need to be with others (Baumeister & Leary, 1995; Chang et al., 2020). If such an essential need is not met, it is common for individuals to experience feelings of loneliness (Chang et al., 2020). Loneliness among adolescents has become a pressing concern, exacerbated by increasing digital interactions that often fail to fulfill emotional needs. The advent of AI-driven technologies, such as chatbots, presents a new dimension in addressing loneliness. This study investigates the relationship between chatbot companionship and loneliness among Indian adolescents, with a particular focus on potential gender differences. AI companions have emerged as sophisticated computer programs designed to simulate human conversation, offering companionship and support. Their roots can be traced back to early attempts at natural language processing, with ELIZA, created in the 1960s, serving as a pioneering example. These early chatbots laid the foundation for subsequent advancements in NLP and artificial intelligence. Chatbots, also known as conversational agents or dialogue systems, are computer programs designed to simulate human conversation through text or voice interactions (Adamopoulou & Moussiades, 2020). These artificial intelligence (AI) powered systems are capable of understanding natural language, processing user inputs, and generating appropriate responses (Følstad & Brandtzaeg, 2017). The increasing use of chatbots for social interaction and support highlights the need to understand their efficacy in mitigating loneliness. While some studies suggest that chatbots can offer valuable emotional support (Rodríguez-Martínez et al., 2024), others report limited benefits. This research aims to fill gaps in the literature by exploring the specific effects of chatbot companionship on loneliness and how these effects may vary by gender in the Indian context.

 

Methods

The study aimed to explore the intricate relationship between chatbot companionship and adolescent loneliness, with a particular focus on potential gender differences. The participants were 153 Indian adolescents, aged 14-18, from Secondary and Higher Secondary Education institutions, who were selected using a non-probability sampling method. Data collection involved a socio-demographic questionnaire, which included questions about personal loss, device usage, chatbot use and purposes, and the UCLA Loneliness Scale to measure loneliness levels. The study employed a quantitative research design, utilizing a 2x2 factorial design that examined two independent variables: chatbot companionship (users versus non-users) and gender (male versus female). Loneliness was the primary dependent variable under investigation. The analysis was conducted using a two-way ANOVA to assess the main effects of chatbot companionship and gender, as well as the interaction effects between these two variables on loneliness levels. Post-hoc tests were performed to explore pairwise comparisons, and an independent sample t-test was conducted to compare loneliness levels between different groups. Additional analyses were performed to explore the potential impact of personal loss and the duration of overall device usage on loneliness. All the collected data were entered into and analyzed using SPSS software, ensuring statistical accuracy and clarity.

 

Results and Analysis

The analysis of this study revealed interesting insights into adolescent loneliness and the potential influence of chatbot companionship. Through a two-way ANOVA, it was found that there were no statistically significant main effects of chatbot usage or gender on loneliness. However, the mean loneliness scores for chatbot users were slightly lower than those of non-users, indicating a possible trend worth further investigation. Gender differences, while not statistically significant, did show that females reported higher average loneliness scores than males. The interaction between chatbot usage and gender did not produce significant results either, suggesting that the relationship between these variables may not be straightforward. An additional analysis focusing on personal loss showed that adolescents who had recently experienced a loss reported significantly higher loneliness levels. This finding highlights the importance of contextual factors in understanding loneliness and suggests that emotional distress from personal experiences like grief may outweigh any benefits of chatbot companionship in alleviating loneliness. Furthermore, the duration of device usage was found to be inversely related to loneliness scores, with longer overall device use correlating with lower loneliness. However, the time spent specifically on chatbots did not appear to have a unique effect on loneliness, reinforcing the idea that general digital engagement, rather than chatbot interaction specifically, may contribute to emotional well-being.

 

Table 1.1

Descriptive Statistics for Loneliness Scores

			Statistic	Std. Error
Loneliness Score	Mean		48.18	.699
	Std. Deviation		8.641

 

The descriptive statistics for the loneliness scores indicate that the sample population had a mean score of 48.18, suggesting a moderate level of loneliness. The standard deviation of 8.641 indicates a moderate degree of variability in the scores, suggesting that while most participants reported moderate levels of loneliness, there was a range of experiences within the sample.

 

Table 1.2

Descriptive Statistics for Loneliness Scores

Depended Variable:		Loneliness Score
Use of Chatbot	Gender	Mean	Std. Deviation	N
User of Chatbot	Male	46.62	8.821	29
	Female	46.71	5.797	24
	Total	46.66	7.534	53
Non-User of Chatbot	Male	43.87	8.657	23
	Female	43.00	8.794	10
	Total	43.61	8.569	33
Total	Male	45.40	8.772	52
	Female	45.62	6.889	34
	Total	45.49	8.038	86

 

The descriptive statistics reveal that there is no significant difference in loneliness scores between males and females. The mean values suggest a slight tendency for chatbot users to report higher levels of loneliness compared to non-users. The mean loneliness score for the entire sample is 45.49, with a standard deviation of 8.038.

 

Table 1.3

ANOVA Analysis of Loneliness Scores by Use of Chatbot and Gender

	df	Mean Square	F	Sig.
Use of Chatbot	1	189.979	2.941	.090
Gender Use of Chatbot * Gender	1 1	2.784 4.172	.043 .065	.836 .800

The ANOVA analysis and revealed that neither chatbot usage nor gender, nor their interaction, significantly predict loneliness scores in the sample. The R Squared value of .036 indicates that a negligible amount of the variance in loneliness scores can be explained by the combined effects of Use of Chatbot, Gender, and their interaction. This suggests that the model is a poor fit for the data. Further research is needed to identify these factors and understand their relationship with loneliness.

 

Table 1.4

Descriptive Statistics for Loneliness Scores by Experiencing Personal Loss

Experiencing Personal Loss	Mean	Std. Deviation	N
Yes (Grieving)	51.64	8.194	67
No (Not Grieving)	45.49	8.038	86
Total	48.18	8.641	153

 

The adolescents who experienced personal loss reported significantly higher levels of loneliness (M=51.64) compared to those who did not (M=45.49). While the standard deviation for both groups was similar, indicating comparable variability in loneliness scores, the significant difference in mean scores suggests a strong association between personal loss and higher levels of loneliness among adolescents.

 

Table 1.5

ANOVA Results for the Effect of Experiencing Personal Loss on Loneliness Scores

	df	Mean Square	F	Sig.
Experiencing Personal Loss	1	1425.984	21.700	<.001
Total	153

 

The ANOVA analysis revealed a significant difference in loneliness scores between individuals who experienced personal loss and those who did not. The F-statistic of 21.700 and a p-value of less than 0.001 indicate that the observed differences are unlikely to be due to chance. This suggests that personal loss is a significant predictor of loneliness, revealing that individuals who have experienced personal loss are more likely to report higher levels of loneliness.

 

Conclusion

The present study provides valuable insights into the complex relationship between chatbot companionship, gender, and loneliness among Indian adolescents. While initial hypotheses posited significant effects of chatbot usage and gender on loneliness, the findings did not support these assumptions, revealing no substantial differences in loneliness levels based on these factors. However, the study underscores the profound impact of personal loss on adolescent loneliness, indicating that those grieving are significantly more prone to Loneliness. Additionally, the duration of device usage emerged as a noteworthy factor, with longer usage associated with lower loneliness scores, although the time spent specifically on chatbot interactions did not significantly influence loneliness. This study contributes to the existing body of knowledge by exploring the specific impact of chatbot companionship on adolescent loneliness in the Indian context, with a focus on gender differences.

 

References

Adamopoulou, E., & Moussiades, L. (2020). Chatbots: History, technology, and applications. Machine Learning with Applications, 2, 100006. https://doi.org/10.1016/j.mlwa.2020.100006

Baumeister, R. F., & Leary, M. R. (1995). The need to belong: Desire for interpersonal attachments as a fundamental human motivation. Psychological Bulletin, 117(3), 497-529.

Chang, E. C., Muyan, M., & Hirsch, J. K. (2020). Loneliness, positive life events, and psychological maladjustment: When good things happen, even lonely people feel better! Personality and Individual Differences, 160, 109936. https://doi.org/10.1016/j.paid.2020.109936

Følstad, A., & Brandtzaeg, P. B. (2017). Chatbots and the new world of HCI. Interactions, 24(4), 38-42. https://doi.org/10.1145/3085558

Rodríguez-Martínez, M. C., Vivas, A. B., Reneses, B., & Rodríguez-Sánchez, J. M. (2024). Effectiveness and acceptability of chatbots for mental health support: A systematic review and meta-analysis. Journal of Medical Internet Research, 26(1), e45634. https://doi.org/10.2196/45634