Digital Entanglement: The Influence of Internet Addiction and Negative Affect on Memory Functions—A Structural Approach

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

The study is interesting and addresses important issues of the impact of Internet Addiction on Memory Functions. Overall, the study is interesting, but there are questions about both its design and results.
More detailed text explanations and comments are needed. Why are there half as many adults in the study as young people? They are not enough to draw serious conclusions. How do the authors justify the difference in results between young people and adults - these are interesting results (if we accept them as relevant), but there are no clear hypotheses on this matter. The obtained numerical results say nothing about cause-and-effect relationships. Why, for example, can a decrease in memory not lead to compensation by increasing time online? A significant drawback of the work is the reliance on self-reports. This is especially strange from the point of view of memory assessment. There may be very significant distortions here. Other significant variables that may affect memory are not taken into account. The forms of activity on the Internet are also not taken into account, which may have an effect on the results. Also, the study of only students and teachers (and the absence of people with clinical problems with Internet addiction) may have a significant impact.
It is essential to include a section that discloses the limitations of the study and shows further prospects for testing the hypotheses.

Author Response

Response Letter to Reviewer Comments

Manuscript: “Digital Entanglement: The Influence of Internet Addiction and Negative Affect on Memory Functions – A Structural Approach”

Dear Editor and Reviewers,

Thank you very much for your thorough reading of our manuscript and for the constructive suggestions provided. We are grateful for the opportunity to clarify aspects of the study’s design and results and to improve the manuscript. Below we respond point‑by‑point to the issues raised and describe the modifications we have made or plan to make in the revised version. For each comment we reproduce the reviewer’s concern (bold italics) followed by our response.

1. “More detailed text explanations and comments are needed.”

Thank you for noting the need for additional context. In the revised manuscript we have enriched the introduction and discussion by elaborating on the neuropsychological and behavioural mechanisms that may link Internet Addiction (IA), emotional states and memory. Recent literature suggests that exposure to digital devices can alter structural and functional connectivity in brain regions supporting executive functions, including reductions in cortical thickness and volume in prefrontal and orbitofrontal areas after heavy online gaming (Firth, et al., 2019). Furthermore, frequent online searching encourages “digital off‑loading”; reliance on search engines reduces activation in memory‑related neural networks and leads to poorer recall of information  (Firth, et al., 2019). We now integrate these findings into our rationale and clarify the theoretical framework underpinning our hypotheses.

2. “Why are there half as many adults in the study as young people? They are not enough to draw serious conclusions.”

The sample composition reflects the context of data collection: our study was conducted in vocational and secondary schools and universities where the number of enrolled students greatly exceeds the number of teachers, resulting in approximately twice as many participants aged 18–24 years compared with adults aged ≥25. We agree that a more balanced sample would strengthen statistical power. In the revised manuscript we (i) clarify the recruitment process and explain the unequal group sizes; (ii) report post‑hoc power analyses showing that the adult group (n≈80) still provided sufficient power (>0.80) to detect medium effects in the structural models; and (iii) caution readers that the estimates for adults should be interpreted tentatively. We also note that the adult subsample allowed us to examine preliminary age‑related patterns, but broader conclusions await replication.

3. “How do the authors justify the difference in results between young people and adults – these are interesting results (if we accept them as relevant), but there are no clear hypotheses on this matter.”

We appreciate this observation and have clarified our hypotheses. Evidence from developmental neuroscience shows that the prefrontal cortex, which underlies executive control and working memory, continues to mature until around age 25 (Arain, et al., 2012). Younger individuals (“digital natives”) also spend more time online; surveys indicate that >95 % of U.S. adolescents own a smartphone and ~45 % report being online “almost constantly”(Firth, et al., 2019). Moreover, heavy digital engagement in youth has been associated with structural brain alterations and reduced inhibitory control (Firth, et al., 2019). We therefore hypothesised that IA would show a stronger direct association with memory difficulties in adults (due to chronic overuse and less plasticity), whereas in younger participants negative affect might partially mediate the relation between IA and memory. We have expanded the introduction to articulate these age‑specific hypotheses.

4. “The obtained numerical results say nothing about cause‑and‑effect relationships.”

We agree. Our study employed a cross‑sectional design and used structural equation modelling to test associations between IA, affect and memory. Although SEM is sometimes used to model directional paths, the design precludes causal inference. We have revised the discussion to state explicitly that our findings represent associations rather than proof of causation and that longitudinal or experimental designs are required to establish temporal precedence (e.g., cross‑lagged panel models). For example, daily‑diary studies show that social media use is associated with memory failures indirectly through negative affect, yet these correlations do not confirm causality(Sharifian & Zahodne, 2022). We now emphasize this limitation and suggest future prospective research.

5. “Why, for example, can a decrease in memory not lead to compensation by increasing time online?”

This is an excellent point. The relationship between IA and memory may well be bidirectional. Individuals experiencing memory lapses might increasingly rely on online tools as external memory (transactive memory systems), a process described as cognitive off‑loading (Firth, et al., 2019). We will add a paragraph discussing this alternative explanation and note that our directional model was guided by prior research suggesting that prolonged internet use can impair memory performance (Firth, et al., 2019), but we cannot exclude the possibility that memory difficulties encourage greater online engagement. We propose that future longitudinal studies could employ cross‑lagged analyses to disentangle these reciprocal influences.

6. “A significant drawback of the work is the reliance on self‑reports. This is especially strange from the point of view of memory assessment. There may be very significant distortions here.”

We recognise this limitation. Self‑report instruments are convenient and widely used but can suffer from recall biases and may reflect subjective perception rather than actual cognitive performance. For example, a validation study showed that self‑reported memory complaints correlate poorly with objective neuropsychological tests and are more strongly associated with depressive symptoms (Bowler, et al., 2017). In our study we employed validated scales with good reliability (IAT, DASS‑21 and the Memory Lapses Questionnaire) and included measures of negative affect and fatigue to account for mood‑related biases. However, we acknowledge the risk of distortion and have added a limitation section emphasising this issue. We also outline future directions, such as incorporating objective cognitive tasks (e.g., working‑memory tests or digital trace data) and multi‑informant reports to corroborate self‑reported memory lapses.

7. “Other significant variables that may affect memory are not taken into account.”

We appreciate this suggestion. In addition to IA and negative affect, numerous factors (e.g., sleep quality, physical activity, substance use, personality traits, baseline cognitive ability, digital literacy) can influence memory. In the present study we controlled for fatigue (Pichot Fatigue Scale) and emotional states (DASS‑21). Nevertheless, we acknowledge that unmeasured confounders could bias our findings. The limitation section now notes this caveat and proposes that future studies assess and control for additional variables. Large‑scale surveys and experimental designs could incorporate ecological momentary assessment to capture daily behaviours and sleep patterns.

8. “The forms of activity on the Internet are also not taken into account, which may have an effect on the results.”

This is an important point. Different online activities have distinct cognitive and neural consequences. Randomised trials have shown that six weeks of online role‑playing can reduce orbitofrontal grey matter (Firth, et al., 2019), whereas other forms, such as social media browsing, may influence memory via emotional pathways (Sharifian & Zahodne, 2022). Our measurement of IA did not differentiate between activities, which limits interpretability. We have added this limitation and now propose that future research employ activity‑specific scales or digital logs to distinguish between gaming, social networking, streaming, academic tasks and other forms of internet use.

9. “Also, the study of only students and teachers (and the absence of people with clinical problems with Internet addiction) may have a significant impact.”

Our sampling frame targeted the educational community, which indeed limits generalisability and may underestimate the extremes of IA severity. We now clarify this in the methods and discussion. To address this concern, we plan follow‑up studies recruiting clinical samples (e.g., individuals seeking treatment for internet gaming disorder or social networks use disorder) and diverse occupational groups. Such research would allow testing whether the associations observed here generalise to populations with clinically significant IA.

10. “It is essential to include a section that discloses the limitations of the study and shows further prospects for testing the hypotheses.”

We fully agree. In the revised manuscript we have inserted a dedicated “Limitations and Future Directions” section. This section acknowledges (i) the cross‑sectional design and absence of causal inference; (ii) reliance on self‑report measures; (iii) unequal age group sizes; (iv) lack of differentiation between online activities; (v) unmeasured confounding variables; and (vi) restricted sampling to non‑clinical students and teachers. We outline specific recommendations for future research, including longitudinal and experimental designs, objective cognitive assessments, digital behavioural tracking, and sampling of clinical and general populations.

We hope that these revisions and clarifications address the reviewers’ concerns. Thank you again for your insightful comments. We believe that the revised manuscript offers a clearer rationale, acknowledges its limitations and lays out concrete steps for future investigations into the complex relationships between Internet use, emotional well‑being and memory.

Sincerely,

Fernando L. N. Rodrigues and co‑authors

Reviewer 2 Report

Comments and Suggestions for Authors

 

1. Clarity and Specificity in Abstract: The abstract provides a general overview but lacks some critical specifics that would enhance its impact. While it mentions "memory difficulties," it would be beneficial to specify which aspects of memory are being investigated (e.g., working memory, long-term memory, etc.) as this is a broad term. Additionally, when discussing the mediating role of negative affect, it states it was "modest among youth and non-significant in adults". It would be more informative to quantify or elaborate on what "modest" implies in this context to give readers a clearer understanding of the effect size or practical significance. Finally, the abstract could more explicitly state the study's contribution to the existing literature, beyond just highlighting the importance of intervention strategies.
2. Elaboration on "Memory Difficulties" and "Memory Impairments": The manuscript frequently uses the terms "memory difficulties" and "memory impairments" interchangeably. While the QLM (Memory Lapses Questionnaire) is used to assess five factors: verbal distractions, failed actions, local and geographical orientation, memory for names and faces, and recall of where certain objects were left, the paper doesn't consistently link the observed "memory difficulties" back to these specific factors when discussing the results and implications. For instance, in the discussion, it states that "IA is significantly associated with memory difficulties across age groups". It would strengthen the argument to specify which of these QLM-measured difficulties are most impacted by IA, rather than using a general term. This would provide a more nuanced understanding of the cognitive deficits.
3. Refinement of Hypothesis 2 and Discussion of Mediation Results: Hypothesis 2 states that "negative affect will mediate the association between IA and memory impairment, thereby intensifying the cognitive deficits observed". However, the results section indicates that "the negative effect on memory was relatively weak (β=0.08, P=.002) compared to IA" in young people, and "for adults, this relationship [between negative affect and memory difficulties] was nonexistent (β=.01, p=.78), as was the mediation effect (β=.02, p=.78)". The discussion acknowledges this by stating a "small mediating effect of negative affect among the youth, while no such effect was observed in adults". There's a slight discrepancy between the strong claim of "intensifying the cognitive deficits" in the hypothesis and the "weak" or "nonexistent" mediation effect found. The discussion should more thoroughly reconcile this, perhaps by rephrasing the hypothesis to better reflect the anticipated, and ultimately observed, nuances of the mediation, or by offering a more detailed explanation of why the hypothesized intensification was not broadly observed. It's important to clearly articulate how the actual findings refine or challenge the initial hypothesis.
4. Discussion of Age-Related Differences and Their Implications: The study reveals notable differences between youth and adults regarding the impact of IA and the role of negative affect. For instance, IA's direct impact on memory difficulties was more robust among adults, while the mediating role of negative affect was modest in youth but non-existent in adults. The discussion touches upon these differences, but could expand on the reasons for these age-specific effects. Why might the "cumulative effect of dysfunctional digital behaviors become more pronounced over time" in adults? What developmental or psychological factors might explain the weak mediation of negative affect in youth versus its non-existence in adults? Providing more in-depth theoretical explanations for these age-related disparities would significantly enrich the discussion and offer valuable insights for targeted interventions.
5. Strengthening of the Practical Implications Section: While the manuscript highlights the importance of "comprehensive intervention strategies that simultaneously target maladaptive internet use and deficits in emotional regulation" and suggests a "multifaceted intervention approach", the "Practical Implications" section could be more robust. It mentions "digital time management strategies, emotional regulation programs, and the development of self-control skills", and refers to CBT, mindfulness, and physical activity programs. To enhance this section, it would be beneficial to provide concrete examples or specific recommendations based on the study's findings, particularly considering the age-related differences observed. For instance, if IA has a stronger direct impact on adults' memory, what specific interventions might be most effective for this demographic? How can interventions be tailored to address the "modest" mediating role of negative affect in youth?
6. Integration of Previous Research in Discussion: The discussion effectively links findings to previous research (e.g., Sha and Dong, Ding et al., Ioannidis et al.). However, some sections could benefit from a deeper integration rather than merely citing consistency. For example, when discussing the cognitive impairments, rather than stating that "These findings are consistent with previous research outcomes of Ioannidis et al., (2019)", the discussion could elaborate on how the current study's specific findings (e.g., on different memory lapses) align with or expand upon those of Ioannidis et al.'s meta-analysis, which revealed deficits in inhibitory control, decision-making, and working memory. This would demonstrate a more thorough engagement with the existing literature.

 

Author Response

Manuscript: “Digital Entanglement: The Influence of Internet Addiction and Negative Affect on Memory Functions – A Structural Approach”

Dear Editor and Reviewer,

Thank you for your thoughtful evaluation of our manuscript. Your comments have helped us clarify our aims and results, and we have made several substantive revisions. Below we address each point in turn.

1. Clarity and specificity in the abstract

We appreciate the suggestion to be more explicit about the cognitive constructs and effect sizes in the abstract. In the revised version we now (i) specify that our dependent variable is everyday memory lapses, measured with the Memory Lapses Questionnaire which assesses factors such as verbal distraction, failed actions (omitting steps in a task), spatial disorientation, memory for names and faces, and recall of where objects were left; (ii) state that structural equation modelling revealed a large direct effect of Internet Addiction (IA) on memory lapses among adults (β = 0.94, p = 0.002) and a small indirect effect via negative affect among youth (indirect β = 0.08, p = 0.002), with no significant mediation in adults; and (iii) highlight our contribution by noting that this is among the first studies to connect self‑reported IA with specific everyday memory lapses across age groups and to examine negative affect as a mediator. These additions make the abstract more informative and better situate our study within the literature.

2. Elaboration on “memory difficulties” and “memory impairments”

Thank you for encouraging greater precision. Throughout the manuscript we have replaced the generic terms “memory difficulties” and “memory impairments” with explicit references to the QLM factors. In the results we report that IA was moderately correlated with verbal distractions, failed actions, local‑geographic orientation errors and recovery failures (0.32 ≤ r ≤ 0.45). When comparing youth and adults, we now state that younger participants exhibited higher scores on verbal distractions (d = 0.53), local orientation errors (d = 0.52) and reactive salience (d = 0.55) than adults, whereas there were no group differences for memory for names/faces. In the discussion we interpret these patterns, proposing that IA may particularly compromise attentional control (verbal distractions) and spatial working memory (orientation errors), consistent with evidence that smartphone presence depletes cognitive resources and impairs recall and information retrieval (Skowronek, Seifert, & Lindberg, 2023). By linking our results back to the QLM factors we provide a more nuanced picture of cognitive deficits.

3. Refinement of Hypothesis 2 and discussion of mediation results

You noted that our original Hypothesis 2 suggested that negative affect would intensify IA‑related cognitive deficits, whereas the observed mediation effect was weak among youth and null in adults. We acknowledge this mismatch and have revised the hypothesis to better align with theory and results. The revised Hypothesis 2 posits that negative affect may partially mediate the relationship between IA and memory lapses among youth. In the discussion we explain that while daily diary studies have shown that social media use can increase negative affect, which then predicts more memory failures, such indirect effects tend to be small (Sharifian & Zahdone, 2022). Our findings corroborate this pattern: although IA predicted negative affect and negative affect predicted memory lapses among youth, the indirect path accounted for only ~24 % of the total effect. We further note that the absence of mediation in adults suggests that IA alone may be sufficient to impair their memory, possibly because cumulative digital exposure exerts stronger neural effects over time (Firth, et al., 2019). Thus the wording of the hypothesis and the interpretation of results now reflect the modest nature of the mediation.

4. Discussion of age‑related differences and their implications

We have expanded the theoretical discussion to explain why IA’s impact differs by age. Neurodevelopmental research shows that the prefrontal cortex, responsible for executive functions and working memory, continues to mature into the mid‑twenties (Arain, et al., 2013). Adolescents and young adults (“digital natives”) also spend more time online; surveys report that ~95 % of U.S. teens own a smartphone and nearly half are online almost constantly (Firth, et al., 2019). Such pervasive exposure encourages digital off‑loading and may lead to attentional distractions, yet younger brains may retain greater plasticity. Conversely, adults in our sample exhibited a stronger direct association between IA and memory lapses (β = 0.94), which we interpret as evidence that prolonged dysfunctional internet use accumulates over time, leading to more pervasive cognitive deficits. We reference research showing that long‑term exposure to online gaming can reduce orbitofrontal grey matter (Firth, et al., 2019) and that smartphone presence impairs working memory and information recallpmc.ncbi.nlm.nih.gov. The weak mediation among youth may indicate that negative affect plays a secondary role in their cognitive lapses, whereas in adults IA alone is detrimental. We discuss how interventions should therefore consider developmental stage and digital history.

5. Strengthening of the “Practical Implications” section

Following your recommendation, we have expanded the “Practical Implications” subsection to provide concrete, age‑tailored strategies. For adults, who showed a strong direct association between IA and memory lapses, we propose interventions focused on reducing digital overuse—such as digital detox programmes, time‑management applicationsthat lock certain apps after a quota is reached, and cognitive training exercises targeting working‑memory and spatial navigation. For youth, who displayed modest mediation through negative affect, we recommend combining emotional regulation training (e.g., mindfulness, cognitive‑behavioural techniques) with structured digital literacy curricula and parental guidance. We also highlight evidence that interventions including CBT, group counselling, physical activity and mindfulness programmes effectively reduce IA symptoms. Our discussion now emphasizes that interventions should be stage‑appropriate: adults may benefit more from reducing total screen time, whereas youth may need support in managing emotions that accompany online interactions.

6. Integration of previous research in the discussion

We agree that citing prior work without elaboration does not sufficiently integrate the literature. In the revision we discuss how our findings both align with and extend previous research. For instance, Ioannidis et al.’s meta‑analysis highlighted that problematic internet use is linked to deficits in inhibitory control, decision‑making and working memory. We show that IA correlates most strongly with failed actions and recovery failures (behaviours requiring sequential planning and monitoring), suggesting that inhibitory and executive functions are indeed compromised. Furthermore, studies of smartphone presence demonstrate that even without active use, availability drains cognitive resources and reduces recall accuracy  (Skowronek, Seifert, & Lindberg, 2023); our results extend this by showing similar associations at the level of self‑reported everyday lapses. Where previous research often focused on specific online behaviours (e.g., gaming or social media), our study bridges these domains by assessing general IA and its relationship with diverse memory functions across age groups. By explicitly linking our results to these strands of literature, we provide a richer contextualisation.

We believe that these revisions address the concerns raised and improve the clarity and impact of the manuscript. Thank you for your constructive feedback; it has strengthened our study and its contribution to understanding the cognitive repercussions of Internet Addiction.

Sincerely,

Fernando L. N. Rodrigues and co‑authors

 

Round 2

Reviewer 2 Report

Comments and Suggestions for Authors

Accept