Development and Validation of an eHealth Literacy Scale: Measuring a Broad Spectrum of eHealth Skills in Web 3.0 Context (Preprint)

BACKGROUND eHealth literacy (EHL) refers to a mixture of capabilities that enable individuals to deal with health information via e-approach, and apply it to solve health problems. With the digitization of health care and the wide availability of health applications, a more diverse range of eHealth skills is required to properly use such health facilities. Existing EHL measurements focus mainly on the health skill of information obtaining (Web 1.0), yet skills of online interactivities (Web 2.0) and self-data managing and applying (Web 3.0) have not been well measured. OBJECTIVE This study aimed to (1) develop a measurement of EHL comprising a comprehensive spectrum of Web 1.0, 2.0, and 3.0 skills (eHLS-Web3.0) and (2) evaluate its validity and reliability along with the measurement invariance among college students. METHODS In Study 1, 421 Chinese college students (mean age = 20.5 ± 1.4 years; 51.8% female) and 8 health experts (mean age =38.3 ± 5.9 years; 87.5% female) were involved to develop the eHLS-Web3.0. The scale development included three steps: item pool generation, content validation, and exploratory factor analysis (EFA). In Study 2, 741 college students (mean age = 21.3 ± 1.4 years; 52.2% female) were recruited from four Chinese cities to validate the new-developed eHLS-Web3.0. The construct validity, convergent validity, concurrent validity, internal consistency reliability, test-retest reliability, and measurement invariance across gender, major, and region were examined by a series of statistical analyses, including confirmatory factor analysis (CFA), and multi-group CFA analysis with SPSS 25.0 and Mplus 7. RESULTS Based on the item pool of 374 statements collected during the conceptual developments, 24 items (4~10 items per subscale) were generated and adjusted after cognitive testing and content validity examination. Through the EFA, we finally developed a three-factorial eHLS-Web3.0, including Acquisition (8 items), Verification (6 items), and Application (10 items). In Study 2, CFAs supported the construct validity of the 24-item three-dimensional eHLS-Web3.0 (χ2 = 903.076, χ2/df =3.701, CFI = .924, TLI = .914, RMSEA = .060, SRMR = .051). The average variance extracted (AVE) of .58 and high correlation between eHLS-Web3.0 subscales and eHEALS (r = .725- .880, P < .001) supported good convergent validity and concurrent validity of the eHLS-Web3.0. Results also supported satisfactory internal consistency reliability (α = .976, ρ = .934 - .956) and test-retest reliability (r = .858, P < .001) of the scale. Multi-group CFAs demonstrated that the 24-item eHLS-Web3.0 to be invariant at all configural, metric, strong, and structural levels across gender (female/male), major (sport-related/medical/general), and region (Yinchuan/Kunming/Xiamen/Beijing). CONCLUSIONS The 24-item three-dimensional eHLS-Web3.0 is developed and verified to be a reliable and valid measurement of EHL in Web 3.0 context among Chinese college students.