Academic review of a master's thesis examining social media's impact on health, assessing argument structure, evidence quality, and academic rigor.
This is a real analysis generated by AnalyzeMyDocument. Try your own for free.
This analysis examines Sarah M. Richardson's master's thesis investigating the relationship between social media usage and mental health outcomes in adolescents. The thesis employs a mixed-methods design combining quantitative survey data (N=342) with qualitative interviews (N=24) to explore associations between social media use and anxiety/depression symptoms among 13-18 year-olds. While the study demonstrates significant positive correlations between heavy social media use and mental health symptoms, the analysis reveals both methodological strengths and critical limitations that affect the validity and generalizability of findings.
Strengths:
Weaknesses:
The theoretical grounding in Social Comparison Theory and the Displacement Hypothesis is appropriate but underdeveloped. The thesis fails to adequately operationalize these theories in the research design. Social Comparison Theory predicts specific mechanisms (upward comparison, self-esteem mediation) that are neither measured nor tested. The Displacement Hypothesis requires measurement of displaced activities, which the study lacks.
The literature review demonstrates good awareness of the field's contradictory findings, appropriately presenting both negative effects (Twenge et al., 2018; Haidt & Twenge, 2023) and null/positive findings (Orben & Przybylski, 2019; Valkenberg et al., 2022). However, the review exhibits several limitations:
The identification of methodological concerns in existing literature is thorough and accurate. The thesis correctly identifies cross-sectional designs, self-report measures, crude screen time proxies, and publication bias as major limitations. Ironically, the study then replicates many of these same limitations, particularly the reliance on total screen time as a "crude proxy."
Appropriate Elements:
Concerns:
The reported effect sizes (R² = .18 for anxiety, .14 for depression) are moderate and practically meaningful. However, the thesis overstates the strength of these associations. With social media explaining 9-14% of variance in mental health outcomes after controlling for demographics, substantial variance remains unexplained.
The qualitative component demonstrates several strengths:
The three identified themes (social comparison, FOMO, cyberbullying) are well-supported and theoretically coherent. However, the analysis lacks depth in several areas:
Causal Language vs. Cross-sectional Data:
The thesis repeatedly implies causal relationships ("social media affects mental health") despite using cross-sectional data. This represents a fundamental misalignment between methodology and conclusions.
Social Media Measurement:
Despite critiquing "crude screen time proxies" in the literature review, the study relies on a single self-report item measuring daily hours. This approach fails to distinguish between:
Sample Limitations:
The suburban Massachusetts sample (67% White, middle-class) severely limits generalizability. Social media effects may vary significantly across socioeconomic, cultural, and geographic contexts.
Selection Bias:
Participants were recruited during "school advisory periods," potentially excluding students with attendance issues who might have different social media usage patterns and mental health profiles.
Measurement Timing:
No information provided about when surveys were administered relative to significant events (exam periods, holidays, social conflicts) that could influence both social media use and mental health.
The study's practical implications are generally reasonable but lack specificity:
For Parents: The recommendation to "monitor and discuss" is vague and unsupported by intervention research.
For Educators: Digital literacy programs are appropriate but need specific curriculum components based on the identified themes.
For Policymakers: Age verification and algorithmic transparency recommendations exceed the scope of the study's findings.
The suggestion to screen for social media-related distress in clinical settings is valuable but requires development of validated screening tools, which the study does not provide.
The study demonstrates appropriate ethical safeguards:
The study largely confirms existing correlational findings without advancing causal understanding. The methodological limitations prevent meaningful advancement of scientific knowledge in this area.
1. Longitudinal Design: Essential for establishing temporal precedence and causal relationships
2. Objective Usage Data: Integration of screen time tracking apps with self-report measures
3. Platform-Specific Analysis: Separate examination of different social media platforms
4. Ecological Momentary Assessment: Real-time sampling of mood and social media use
1. Multidimensional Social Media Assessment: Measure active vs. passive use, content types, and social features
2. Mediation Analysis: Test proposed mechanisms (social comparison, sleep displacement)
3. Positive Outcomes: Include measures of social connection, creativity, and learning
4. Individual Differences: Examine personality, family environment, and peer relationships as moderators
Future research requires samples that include:
This thesis represents competent graduate-level work that makes a modest contribution to the field. However, significant methodological limitations prevent it from substantially advancing scientific understanding of social media's impact on adolescent mental health. The study would benefit from major revisions to align methodology with research questions and to address the fundamental limitation of inferring causation from correlational data.
Richardson's thesis addresses an important contemporary issue using appropriate mixed-methods methodology. While the findings are consistent with emerging concerns about social media's impact on adolescent mental health, the cross-sectional design and measurement limitations significantly constrain the study's scientific contribution. The work demonstrates competent execution within these constraints but fails to advance the field beyond existing correlational findings. Future research incorporating longitudinal designs, objective usage measures, and diverse samples will be necessary to establish causal relationships and inform evidence-based interventions.