Development and validation of assessments of adolescent health literacy: a Rasch measurement model approach

Background Health literacy (HL) is implicated in improved health decision-making and health promotion, and reduced racial, ethnic, and socioeconomic health disparities. Three major areas of HL include functional, interactive, and critical HL. HL skills develop throughout the lifespan as individuals' psychosocial and cognitive capacities develop and as they accumulate experiences with navigating health systems. Though adolescence is marked by increased involvement in health decision-making, most HL studies and measures of HL have focused on adults. Both the adult and adolescent HL literature are also limited by the paucity of validated test-based measures for assessing HL. The existing test-based validated HL measures for adolescents were originally designed for adults. However, adolescents are at an earlier phase of developing their HL skills (e.g., fewer experiences navigating the health system) compared to adults and measures originally designed for adults may assume prior knowledge that adolescents may lack therein underestimating adolescents' HL. This study developed and validated test-based assessments of adolescents' functional, interactive, and critical HL. Methods Items were generated in an iterative process: focus groups with adolescents informed item content, cognitive interviews with adolescents and expert consultation established content and face validity of the initial items, and items were revised or removed where indicated. High school students (n = 355) completed a measurement battery including the revised HL items. The items were evaluated and validated using Rasch measurement models. Results The final 6-item functional, 10-item interactive, and 7-item critical HL assessments and their composite (23 items) fit their respective Rasch models. Item-level invariance was established for gender (male vs. female), age (12-15-year-olds vs. 16-18-year-olds), and ethnicity in all assessments. The assessments had good convergent validity with an established measure of functional HL and scores on the assessments were positively related to reading instructions before taking medicine and questioning the truthfulness of health information found online. Conclusions These assessments are the first test-based measures of adolescents' interactive and critical HL, the first test-based measure of functional HL designed for adolescents, and the first composite test-based assessment of all three major areas of HL. These assessments should be used to inform strategies for improving adolescents' HL, decision-making, and behaviors.