Get started
Work

An exploratory latent profile analysis of digital addiction symptoms, physical education predisposition, and psychological well-being in middle school students

Year: 2026
Type: article
Abstract: This study aimed to identify latent profiles of middle school students based on digital addiction symptoms, physical education (PE) predisposition, and psychological well-being using a person-centered approach. A total of 736 middle school students (50.7% female; grades 5–8) from state middle schools in Antalya and Osmaniye provinces were assessed using the Digital Addiction Scale for Children (DASC), the Physical Education Predisposition Scale (PEPS), and the Stirling Children’s Well-being Scale (SCWBS). Data were analyzed via Latent Profile Analysis (LPA) in R, with between-profile differences examined through ANOVA, post-hoc tests, and multinomial logistic regression. Under the retained EEE covariance specification, LPA yielded three profiles (Entropy = 0.81): Moderate Symptom Severity ( n = 597, 81.1%), characterized by moderate digital addiction symptom severity, favorable PE predisposition, and positive well-being; Lower Well-Being ( n = 39, 5.3%), with a numerically higher symptom severity mean (M = 45.0 against 41.2) and the lowest well-being in the sample (M = 22.4); and High Symptom Severity ( n = 100, 13.6%), with the highest symptom severity in the sample (M = 72.5) and a mean well-being score numerically similar to that of the first profile (M = 45.4 against 46.3). DASC showed the largest overall effect size (η² = 0.544) and SCWBS the second largest (η² = 0.317); because these indicators generated the profiles, such values describe the retained solution and do not validate it. After accounting for classification uncertainty, female gender remained associated with membership in the High Symptom Severity profile (OR = 1.68, 95% CI [1.04, 2.73]), whereas the grade-level association with the Lower Well-Being profile did not remain significant. Sensitivity analyses indicated that the retained three-profile partition depended on covariance specification and class number; rescaling the indicators left the EEE solution essentially unchanged (ARI = 0.945), but removing three multivariate outliers reduced the smallest profile below 5% of the sample. The retained three-profile solution should be interpreted as one plausible, specification-dependent representation of heterogeneity in this two-province sample, not as evidence of fixed student types, and the smallest profile was its least stable part. Whether the observed configurations have implications for psychological-risk assessment requires replication using independent outcomes and more representative samples.
DOI: 10.1186/s12889-026-30009-4
Language: English
Open Access status: gold
Estimated APC paid: $3,290
OpenAlex ID: W7223180995
FWCI:
Cites:
Cited by:
Related to: