OBJECTIVES

In the last several decades, screen-based media devices have become an integral part of the lives of children and adolescents for education, communication, and recreation. Yet, there’s a lack of a comprehensive list of areas covered by the literature on screen media use. This study provides a comprehensive summary of topics on screen media use among children and adolescents using machine learning algorithms.

METHODS

We obtained literature relevant to screen media use among children and adolescents from PubMed, Scopus, CINAHL, PsycINFO, and EconLit. We used natural language processing and unsupervised machine learning algorithms, specifically topic modeling, to derive clusters of research topics. The records were classified into structured abstracts and unstructured abstracts. We extracted the objective sections from structured abstracts and used the full texts from unstructured abstracts. The texts were then tokenized and lemmatized before fitting two separate topic models using Latent Dirichlet Allocation (LDA).

RESULTS

The search yielded a total of 37,354 retrieved records, of which 14,314 were structured abstracts and 23,040 were unstructured abstracts. We identified 25 topics from structured abstracts and unstructured abstracts separately. Based on the number of articles in each topic, the highly researched topics are the use of social media platforms for health information and communication, screen media use and health behaviors (e.g., physical activity, smoking, sleep, eating, etc.), and computer-assisted intervention for disease diagnosis and treatment. The least explored topics are dietary management, healthcare-related screen media use for injury rehabilitation, and web information and body image concerns.

CONCLUSIONS

Our study confirms the feasibility of using topic modeling to derive a comprehensive summary of research on screen media use among children and adolescents. While extensive research has been conducted on screen media use, our findings suggest some unexplored areas, such as racial/ethnic disparities and the economic impact of screen media use.