Background

The global use of social media and its regulation among students remains a significant public-health challenge. Social media use has been associated with increased productivity among undergraduates, coinciding with an expanding user base 1,2. Social-media platforms—websites or applications that students use to interact with peers 3, share ideas 4, and enhance social visibility 5—served multiple educational functions during the coronavirus disease 2019 (COVID-19) pandemic 6, when social gatherings were restricted. Most students possess social-media accounts; for example, Facebook reports 2.9 billion monthly active users 7. Students find these platforms helpful for maintaining long-distance friendships and family relationships 8. Moreover, they can provide access to scholarships 9 and educational events 10,11. Systematically evaluating the advantages of Facebook, YouTube, WhatsApp, Instagram, Snapchat, Twitter, Google+, Pinterest, and online forums is, however, time-consuming. Many students remain online for prolonged periods 12,13, thereby dedicating less time to study and potentially impairing preparation for quizzes, examinations, and overall academic performance 14. Empirical studies indicate that social-media exposure may modulate memory processes 15,16. Working memory is a critical cognitive resource for academic activities 17, enabling students to solve problems and comprehend instructional materials 18,19. Nevertheless, the mechanisms through which social media influences working memory are not yet fully elucidated. Some investigations have reported beneficial effects 4,2021 that may facilitate educational and scientific activities and ultimately enhance academic performance 14. Despite these findings, only a minority of students employ social media primarily for educational purposes 22, and most studies have detected no significant association between overall social-media use and grade-point average (GPA) 5,23,24. Furthermore, few studies have examined social-media engagement in relation to students’ working memory. Although good academic performance does not necessarily imply optimal memory, academic failure is more frequent among students with memory impairments 25.

The present study therefore aimed to investigate the impact of social-media use on working memory and academic performance among undergraduates from three universities—two in Tbilisi and one in Batumi. We analysed the associations between social-media exposure, working-memory indices, and GPA. The resulting evidence is expected to inform students about the evolving influence of social media on their academic and professional development.

Methodology

Results

Students’ demographic distribution

Among the participants, 58.7 % were female (Fig. 1.0); the mean (± SD) age was 21.94 ± 2.8 years, and the majority were international students (Fig. 2.0) from all three universities (BSU, UG, and EEU; Fig. 3.0). Figure 4 demonstrates that the prevalence of social media usage disorder was higher in the categories of pre-occupation (41.6 %), persistence (60.9 %), displacement (44.3 %), and escape (67.6 %). Supplementary Table S1 shows that 16.2 % of students reported needing improvement in academic performance. Students with elevated scores (> 25) in the storage (4.4 %), attention (6.2 %), and executive (4.4 %) domains also reported corresponding complaints (Table S).

According to Table 1, 80.3 % of students with good academic performance were at risk of neglecting activities such as hobbies, sports, and class assignments owing to social media use (OR = 0.63; χ² = 5.133; p = 0.023; 95 % CI [0.425–0.942]). In the same table, 90 % of these students were also at risk of engaging in arguments with friends due to social media usage (OR = 1.13; χ² = 5.368; p = 0.021; 95 % CI [1.099–3.470]).

Table 2 indicates that 94 % of students with good working memory were at risk of withdrawal complaints (OR = 0.34; χ² = 6.865; p = 0.009; 95 % CI [0.154–0.793]). Additionally, 93.4 % reported conflicts with parents, siblings, or partners attributed to social media use.

Academic Performance Interval	Yes	No	chi-square	P-value	OR	95% Confidence
Preoccupation						Lower	Upper
Good Performance	85.7%	82.5%	1.324a	0.250	1.271	0.844	1.913
Needs Improvement	14.3%	17.5%
Tolerance
Good Performance	86.9%	83.2%	1.032a	0.310	1.341	0.76	2.365
Needs Improvement	13.1%	16.8%
Withdrawal
Good Performance	84.8%	83.4%	.227a	0.633	1.113	0.718	1.725
Needs Improvement	15.2%	16.6%
Persistence
Good Performance	81.8%	86.9%	3.242a	0.072	0.680	0.446	1.037
Needs Improvement	18.2%	13.1%
Displacement
Good Performance	80.3%	86.6%	5.133a	0.023	0.633	0.425	0.942
Needs Improvement	19.7%	13.4%
Problems
Good Performance	90%	82.2%	5.368a	0.021	1.953	1.099	3.470
Needs Improvement	10%	17.8%
Deception
Good Performance	85%	83.3%	.323a	0.570	1.138	0.728	1.779
Needs Improvement	15%	16.7%
Escape
Good Performance	84.4%	82.5%	.442a	0.506	1.152	0.759	1.747
Needs Improvement	15.6%	17.5%
Conflict
Good Performance	86.9%	83.2%	1.032a	0.310	1.341	0.76	2.365
Needs Improvement	13.1%	16.8%

	Working memory
Storage Domain	Yes	No	chi-square	P-value	OR	95% Confidence Interval
Preoccupation						Lower	Upper
Good	95.3%	97.6%	2.879a	0.090	0.496	0.217	1.132
Needs Improvement	4.7%	2.4%
Tolerance
Good	94.7%	97.5%	3.644a	0.056	0.458	0.202	1.040
Needs Improvement	5.3%	3.9%
Withdrawal
Good	94%	97.8%	6.865a	0.009	0.349	0.154	0.793
Needs Improvement	6%	2.2%
Persistence
Good	96.1%	97.5%	1.020a	0.312	0.633	0.259	1.547
Needs Improvement	3.9%	2.5%
Displacement
Good	95.3%	97.8%	3.324a	0.068	0.466	0.201	1.078
Needs Improvement	4.7%	2.2%
Problems
Good	94.7%	97.2%	2.378a	0.123	0.511	0.214	1.217
Needs Improvement	5.3%	2.8%
Deception
Good	95.7%	97.1%	0.946a	0.331	0.660	0.284	1.533
Needs Improvement	4.3%	2.9%
Escape
Good	96.3%	97.4%	0.622a	0.430	0.687	0.269	1.754
Needs Improvement	3.7%	2.6%
Conflict
Good	93.4%	97.3%	4.776a	0.029	0.390	0.163	0.934
Needs Improvement	6.6%	2.7%

Discussion

The present study quantified the prevalence of social-media addiction and investigated its associations with academic performance and working memory among undergraduate students. Patterns of social-media use varied according to participants’ age, gender and geographical origin 34. Consistent with these findings, Meng et al. 35 reported a global upward trend in social-media addiction, which was also evident in our predominantly international sample. Social media exerts a profound influence on individuals and society, shaping perspectives, interpersonal behaviours and interactions across genders 36. It provides a platform for emotional expression and positive reinforcement, but also propagates unrealistic beauty standards and fosters feelings of inadequacy and dominance 3,5,37. Such exposures may elicit social comparison, diminished self-esteem, jealousy and anxiety 38, factors that potentially mediate the development of social-media addiction observed in this cohort.

‘Escape’ behaviour was particularly prevalent among female students. Societal expectations and gender-specific pressures may render female students more vulnerable to stress and anxiety. Escape behaviours include mindless scrolling, online gaming, passive video consumption and seeking validation through ‘likes’ and comments. Accordingly, female students may employ social media as a coping mechanism to obtain psychological relief 39. Male students, however, reported greater ‘conflict’ with family members regarding their social-media use, whereas female students tended to express emotions to avert discord and maintain relationships 40. Prevailing masculinity norms may discourage males from activities perceived as frivolous, thereby reducing the prevalence of escape behaviours in this group 39,41,42. Algorithmic reinforcement, intentionally optimised to prolong user engagement, may exacerbate interpersonal conflict by decreasing receptivity to reasoned discourse and entrenching pre-existing positions. The absence of face-to-face cues on social media can promote misinterpretation and aggressive exchange, further intensifying conflict 43.

Shimoga et al. 44 similarly noted that students who limited social-media time were more likely to engage in productive physical activity, a trend mirrored in our data. In contrast, students reporting no physical activity frequently exhibited symptoms of ‘preoccupation’, ‘tolerance’ and ‘displacement’. Consistent with previous work 45,46, heightened ‘preoccupation’ was associated with extended time spent online. Adverse online interactions may impair concentration, disrupt time management and precipitate procrastination. Platform design, peer influence, escapism, instant gratification and fear of missing out collectively drive excessive use among students 46. A paucity of offline support may lead students to rely on social media as their primary avenue for social interaction. Stress, low self-esteem, inadequate coping strategies, peer influence, anonymity, escapism and attention-seeking behaviour may underlie displacement and problematic use 47.

Our study did not identify a significant association between social media addiction and grade point average (GPA), although students with higher academic performance still reported ‘displacement’ and ‘problems’. Previous research has reported that social media use can negatively affect academic performance by reducing the time available for learning and classroom activities, particularly among students requiring academic improvement 48. The high prevalence of students who reported neglecting other activities to engage with social media may, however, reflect that they also employ these platforms for educational purposes 9. Nevertheless, prolonged daily exposure to social media has been associated with deterioration in academic performance 49. The widespread personalization and modification of social-networking sites among students therefore remains a major concern with respect to their study habits 50.

With respect to cognitive outcomes, our findings indicate that social media addiction did not exert a broad impact on working memory; only ‘withdrawal’ and ‘conflict’ complaints were reported in connection with this domain. These data therefore suggest no direct influence of social media addiction on working memory, although indirect effects mediated by the neglect of essential activities cannot be excluded. A comparable investigation likewise reported that working memory was unlikely to be affected by social media use in healthy participants 35. However, the same study demonstrated that excessive exposure among individuals with psychiatric disorders such as depression can impair working memory—an aspect beyond the scope of the present work 35. Another study observed that memory performance was negatively influenced by social media use, potentially owing to stress-related neurobiological mechanisms 51. Although a large proportion of our participants exhibited adequate working-memory capacity, many still reported ‘withdrawal’ and ‘conflict’, underscoring the need for targeted interventions to preserve and enhance this cognitive function.

Conclusion

Our study identified physical activity and complaints related to physical displacement and other difficulties as significant predictors of academic performance. Students should be mindful that excessive engagement with social media may substantially impair their studies. Accordingly, access to counselling services is recommended for students reporting such complaints, as early intervention may promote optimal working memory and prevent psychosocial sequelae that were beyond the scope of the present investigation.

This cross-sectional design limits causal inference between social media use disorder, academic performance, and working memory because longitudinal trajectories could not be evaluated. Our statistical approach may also have introduced error, thereby affecting the reliability of the findings. Although the sample encompassed a substantial cohort of students from three universities in Viet Nam’s two most populous cities, generalisation to the entire student population remains restricted; in addition, most participants were international students at various stages of training.

The instruments used to assess social media use, academic performance, and working memory exhibit inherent validity and reliability constraints, which could influence result accuracy. The study systematically applied all nine DSM-5 criteria for behavioural addiction; however, these criteria constitute a subjective self-report scale and did not allow differentiation among specific platforms (e.g., Facebook, Instagram, YouTube), a notable limitation. Moreover, unmeasured confounders may have affected the observed associations. Because data were self-reported, response bias is possible. Finally, we were unable to perform formal clinical diagnoses of addiction owing to the large sample size; nevertheless, the findings underscore the need to raise awareness of social media addiction among university students.

Abbreviations

APS: Academic Performance Scale; BSU: Batumi State University; CI: Confidence Interval; COVID-19: Coronavirus Disease 2019; DSM-5: Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition; EEU: East European University; GPA: Grade-Point Average; OR: Odds Ratio; SD: Standard Deviation; SMU: Social Media Usage (Scale); SPSS: Statistical Package for the Social Sciences; UG: University of Georgia; UGREC: University of Georgia Research Ethics Committee; WMQ: Working Memory Questionnaire.

Author Contributions

M. E. N: Original draft, conceptualization, Formal analysis, Data curation, Writing—review & editing. M. E. M. and C.P: Methodology, Investigation, Formal analysis, Writing—review & editing. C. R and V. N: Methodology, Data curation, Writing—review and editing. Z. N: Original draft, Conceptualization, Methodology, Formal analysis, Data curation, Writing—original draft, Writing—review and editing.

All authors have read and agreed to the published version of the manuscript.

Funding

This study received no external funding.

Institutional Review Board Statement

The study was conducted in accordance with the Declaration of Helsinki and approved by the Bioethics council of the University of Georgia (UGREC-31-23).

Informed consent

Comprehensive instruction was provided in the survey including a consent form in which the participant can discontinue the study at any time.

Data Availability Statement

The data that support our findings are available from the first author upon reasonable request.

Acknowledgments

The authors wish to thank all participants for the time dedicated to completing this study.

Conflicts of interest

The authors have no conflict of interest.