JMIR Publications

ABSTRACT

Background:

Digital hygiene refers to health protective behaviors that help adults regulate screen use, reduce cyber risk, and engage with online content more safely and reflectively. Brief, psychometrically robust measures are needed for population research and intervention evaluation.

Objective:

We aimed to develop and validate the Digital Hygiene Behavior Scale (DHBS) for adult internet users.

Methods:

We developed the DHBS in five phases from April 2022 to December 2024: conceptualization, item generation, expert content validation, pilot testing, and nationwide validation. We refined an 89 item pool through a 2 round Delphi panel with 10 experts and a pilot study (n = 436), which yielded 39 items. Participants from two independent nationwide Polish samples recruited via quota sampling completed computer assisted web interviewing (Sample 1 n = 1067; Sample 2 n = 1000). Using these data, we evaluated psychometric properties with exploratory factor analysis based on a polychoric correlation matrix (robust unweighted least squares extraction, oblique rotation) and confirmatory factor analysis using WLSMV. We also assessed convergent validity (Spearman rho), known groups validity (one-way ANOVA with Games Howell tests and omega squared), internal consistency (Cronbach alpha, McDonald omega), and temporal stability over a two-week interval in a subsample (n = 50) using intraclass correlation coefficients.

Results:

We reduced the scale to 17-items across four correlated domains: boundary setting and control (5 items), security behaviors (5 items), content and information management (3 items), and other health promoting behaviors (4 items). Confirmatory factor analysis indicated good fit: chi-square = 298.17, p < 0.001; RMSEA = 0.057 (90% CI 0.049 to 0.065); CFI = 0.975; TLI = 0.970; WRMR = 0.92. Standardized loadings ranged from 0.57 to 0.84 (all p < 0.001). The total score showed high internal consistency (alpha = 0.85; omega = 0.86), and subscales showed acceptable to good consistency (alpha 0.71 to 0.84; omega 0.73 to 0.84). Correlations supported convergent validity, with positive associations with self-rated health (rho = 0.11, p < 0.001) and attention to health related behaviors (rho = 0.29, p < 0.001), and negative associations with psychosomatic complaints (rho minus 0.17 to minus 0.23, all p < 0.001). Known groups analyses showed differences by intention to improve digital hygiene, self-reported screen overuse, and weekday and weekend smartphone time (all p < 0.001). The total score demonstrated excellent test retest reliability (ICC = 0.89, 95% CI 0.81 to 0.94, p < 0.001). We derived norms by transforming raw scores to a 0 to 100 metric and generating sex and age stratified sten scores.

Conclusions:

The DHBS is a brief, reliable, and valid instrument for assessing the frequency of digital hygiene behaviors in adults. Its four-domain structure supports population monitoring, education, and evaluation of interventions promoting healthier and safer digital practices.