Mental health and depression as mediators between social media use screen time and academic integrity among tertiary students in Ghana.

Increased social media use and screen time were significantly associated with poorer mental health outcomes among tertiary students in Ghana.

• The model explained 31% of variance in mental health outcomes (R² = 0.31, p < .001)
• Late-night screen usage was the strongest negative predictor of mental well-being (β = -0.29, p < .001)
• Emotional investment in social media was equally strong as a negative predictor of mental well-being (β = -0.30, p < .001)
• Sample consisted of 970 students from public Colleges of Education across Ghana selected through stratified random sampling
• Data were analyzed using Pearson correlations, multiple regression, and mediation/moderation models (Hayes PROCESS v4.0)

Depression significantly mediated the relationship between social media use and academic performance, representing partial mediation.

• Indirect effect through depression: β = -0.14, 95% CI [-0.21, -0.07], p < .001
• The mediation model accounted for 27% of the variance in depression and 23% in academic scores
• The direct path from social media use to academic performance remained significant after accounting for depression (β = -0.17, p < .001)
• The total effect of social media use on academic performance was stronger (β = -0.31, p < .001), indicating partial rather than full mediation

Depression significantly predicted increased assessment dishonesty among tertiary students.

• Depression predicted assessment dishonesty (β = 0.29, p < .001)
• The relationship between depression and assessment dishonesty was amplified among students with high non-academic screen time exposure
• Correlation analysis showed significant positive associations between depression and academic integrity violations (r = .25 to 0.33, p < .01)

Screen time was significantly negatively associated with assessment scores.

• Correlation coefficients between screen time and assessment scores ranged from r = -0.28 to -0.32, p < .01
• The study used a cross-sectional correlational design
• Associations were detected across multiple assessment score measures as indicated by the range of correlation coefficients

Late-night screen usage and emotional investment in social media emerged as the strongest predictors of poor mental well-being.

• Late-night screen usage had a standardized regression coefficient of β = -0.29, p < .001
• Emotional investment in social media had a standardized regression coefficient of β = -0.30, p < .001
• Both predictors were identified within a model explaining 31% of variance in mental health outcomes (R² = 0.31, p < .001)
• These two variables were described as 'the strongest negative predictors of mental well-being'