Department of Psychology, The University of Hong Kong, Pokfulam, Hong Kong.
Department of Psychology, University of Oslo, Oslo, Norway.
J Med Internet Res. 2022 Jan 10;24(1):e27000. doi: 10.2196/27000.
BACKGROUND: As social media is a major channel of interpersonal communication in the digital age, social media addiction has emerged as a novel mental health issue that has raised considerable concerns among researchers, health professionals, policy makers, mass media, and the general public. OBJECTIVE: The aim of this study is to examine the prevalence of social media addiction derived from 4 major classification schemes (strict monothetic, strict polythetic, monothetic, and polythetic), with latent profiles embedded in the empirical data adopted as the benchmark for comparison. The extent of matching between the classification of each scheme and the actual data pattern was evaluated using sensitivity and specificity analyses. The associations between social media addiction and 2 comorbid mental health conditions-depression and anxiety-were investigated. METHODS: A cross-sectional web-based survey was conducted, and the replicability of findings was assessed in 2 independent samples comprising 573 adults from the United Kingdom (261/573, 45.6% men; mean age 43.62 years, SD 12.24 years) and 474 adults from the United States (224/474, 47.4% men; mean age 44.67 years, SD 12.99 years). The demographic characteristics of both samples were similar to those of their respective populations. RESULTS: The prevalence estimates of social media addiction varied across the classification schemes, ranging from 1% to 15% for the UK sample and 0% to 11% for the US sample. The latent profile analysis identified 3 latent groups for both samples: low-risk, at-risk, and high-risk. The sensitivity, specificity, and negative predictive values were high (83%-100%) for all classification schemes, except for the relatively lower sensitivity (73%-74%) for the polythetic scheme. However, the polythetic scheme had high positive predictive values (88%-94%), whereas such values were low (2%-43%) for the other 3 classification schemes. The group membership yielded by the polythetic scheme was largely consistent (95%-96%) with that of the benchmark. CONCLUSIONS: Among the classification schemes, the polythetic scheme is more well-balanced across all 4 indices.
背景：随着社交媒体成为数字时代人际交流的主要渠道，社交媒体成瘾已成为一个新的心理健康问题，引起了研究人员、卫生专业人员、政策制定者、大众媒体和公众的极大关注。
目的：本研究旨在从 4 种主要分类方案（严格单分型、严格多分型、单分型和多分型）中检验社交媒体成瘾的流行程度，并以实证数据中嵌入的潜在特征为基准进行比较。采用敏感性和特异性分析评估每种方案的分类与实际数据模式的匹配程度。还研究了社交媒体成瘾与 2 种共病心理健康状况（抑郁和焦虑）之间的关联。
方法：采用横断面网络调查，在 2 个独立样本（英国 573 名成年人，其中 261/573 名男性，平均年龄 43.62 岁，标准差 12.24 岁；美国 474 名成年人，其中 224/474 名男性，平均年龄 44.67 岁，标准差 12.99 岁）中评估了研究结果的可重复性。两个样本的人口统计学特征与各自人群相似。
结果：不同分类方案的社交媒体成瘾患病率不同，英国样本为 1%至 15%，美国样本为 0%至 11%。潜在特征分析为两个样本确定了 3 个潜在群体：低风险、有风险和高风险。除了多分型方案的敏感性相对较低（73%-74%）外，所有分类方案的敏感性、特异性和阴性预测值均较高（83%-100%）。然而，多分型方案的阳性预测值较高（88%-94%），而其他 3 种分类方案的阳性预测值较低（2%-43%）。多分型方案得出的群体归属在很大程度上与基准一致（95%-96%）。
结论：在分类方案中，多分型方案在所有 4 项指标上都更加平衡。
Medicina (Kaunas). 2020-5-6
J Med Internet Res. 2024-3-6
Behav Sci (Basel). 2025-5-13
Front Psychiatry. 2024-9-16
Nat Sci Sleep. 2024-5-27
Front Psychiatry. 2021-3-19
JAMA Netw Open. 2020-9-1
Zotero 插件
