JMIR Publications

ABSTRACT

Background:

Digital technologies are reshaping healthcare, making digital health literacy (DHL) a critical competency for navigating online health information. Although widely used as a unidimensional measure of DHL, the literature increasingly supports a multidimensional framing of the eHealth Literacy Scale (eHEALS). Studies propose alternative factor structures, but many analyses have not accounted for the ordinal nature of eHEALS response data.

Objective:

This study aimed to identify and validate an alternate multidimensional structure of eHEALS accounting for its ordinal response scale.

Methods:

Data were drawn from the 2022 GetCheckedOnline community survey of consenting English-speaking British Columbia residents aged ≥16 years who reported sexual activity in the past 12 months. Participants were recruited through geo-targeted digital advertisements, community outreach, and in-person recruitment at public events, and community locations. DHL was measured using eHEALS, with responses collected on a five-point Likert scale. Descriptive statistics summarized eHEALS responses using means, medians, and interquartile ranges. Exploratory and confirmatory factor analyses (EFA and CFA) were used to assess the scale’s structure using polychoric correlations and standard model fit indices. Reliability and validity were evaluated using polychoric ordinal alpha, average variance extracted, and composite reliability, with missing data addressed via multiple imputation.

Results:

Overall, 1,657 participants met inclusion criteria (mean age = 33.0, SD = 11.77); 47.3% identified as women, and 30.4% as racialized minorities. Most participants (80.5%) reported easy internet access, and 62.1% had post-secondary education. Across eHEALS items, median scores were 4.0 (interquartile ranges were between 1.0-2.0), with excellent internal consistency (polychoric ordinal α = 0.92). Exploratory factor analysis supported a three-factor solution explaining 65.7% of the variance, demonstrated through CFA (χ² = 71.680, df = 17, p < 0.001, RMSEA = 0.059, SRMR = 0.026, CFI = 0.969, TLI = 0.948). The final model included Information Navigation (loadings: 0.765–0.917), Resource Appraisal (0.825–0.892), and Confidence in Use (0.803 for both items), with composite reliability (0.784–0.900), and average variance extracted (0.503–0.738) supporting construct validity.

Conclusions:

This study confirms a multidimensional structure of eHEALS, identifying Information Navigation, Resource Appraisal, and Confidence in Use as key dimensions of DHL. The model enhances measurement precision and informs targeted interventions. Future research should examine how these factors shape access to digital health services, and scale development must evolve to reflect emerging technologies such as AI and social media algorithms.