# Development of Japanese and Indonesian Versions of the electronic-Health Literacy Scale

## TL;DR

Researchers developed and validated Japanese and Indonesian versions of a scale to measure electronic health literacy among nursing students.

## Contribution

The study confirms the validity and reliability of culturally adapted e-Health Literacy Scale versions in Japan and Indonesia.

## Key findings

- The e-HLS showed no item skewness or ceiling/floor effects in either country.
- Fourth-year nursing students scored higher in health literacy than earlier-year students.
- Cross-cultural validity supported configural and metric invariance but not scalar invariance.

## Abstract

Reportedly, electronic-Health (eHealth) literacy is lower in younger people than in those in their 40s and 50s, despite the Internet being an essential information source for the youth. We developed Japanese and Indonesian versions of the e-Health Literacy Scale (e-HLS).

Conceptual equivalence was verified using the forward-backward translation method. First- to fourth-year nursing students at two universities in Japan and Indonesia were surveyed using a web-based questionnaire, including a 12-item e-HLS. Data analysis involved item analysis, factor analysis, and validation of convergent and discriminant validity, reliability, discriminative validity, and cross-cultural validity. The study was approved by the Indonesian and Japanese ethics boards.

Ninety-nine Japanese and 407 Indonesian participants responded to the survey (response rate: Japan 23.9%, Indonesia 89.6％). Neither country showed item distribution skewness, and no ceiling or floor effects were detected. Confirmatory factor analysis supported the same three-factor structure as the original e-HLS. Discriminative validity indicated that fourth-year students had higher overall, functional, and interactive health literacy scores than first- to third-year students. There were no significant differences between the two groups in terms of critical health literacy. Cronbach’s alpha coefficients were 0.86 overall and 0.74-0.83 for each of the three domains. Regarding cross-cultural validity, configural invariance and full metric invariance models showed a good fit, but scalar invariance was not supported.

From the results, the content validity, construct validity (structural, cross-cultural, convergent, and discriminant validity), and reliability (internal consistency) of the Japanese and Indonesian versions of the e-HLS were confirmed.