Factor structure of self-reported health literacy scales: a large-scale cross-sectional study among Japanese-speaking adults

Authors:
Publisher website
Google Scholar
Abstract
The importance of health literacy has become widely recognized in the fields of psychology and public health over past decades, during which more than 100 health-literacy scales have been developed. These scales typically have unique targets (or subscales), leading to an enormous number of factors proposed across scales. However, conceptual overlaps and redundancy are left unresolved. This study therefore explored the factor structure of 219 items extracted from 11 existing scales assessing subjective health literacy, each of which was administered to Japanese-speaking online panels (N = 16 097; 7722 women; mean age = 54.9 years). Exploratory factor analysis suggested seven factors, three of which represented general health literacy and the other four pertained to social health literacy, mental health literacy, digital literacy, and numeracy. The three factors of general health literacy cover slightly different constructs (e.g. ability vs. motivation to find health information) and may reflect differences in the rating scales (easy to difficult vs. agree to disagree). Confirmatory factor analysis on independent testing data supported this seven-factor structure with selected items. Each factor showed, overall, small correlations with mental and physical health and healthy lifestyles, which leaves the discriminant validity open for future research. These results suggest that subjective health literacy scales can be reduced to a smaller, more parsimonious number of factors. Some of the administered scales were combined as a single latent factor, whereas scales targeting specific aspects (social, mental, digital, and numeracy) were recognized as independent factors. Recommendations for assessment and battery designs are discussed.
Funding Information
  • Council for Science, Technology and Innovation (JPJ012248)

This publication has 66 references indexed in Scilit: