Format

Send to

Choose Destination
BMC Health Serv Res. 2018 Jun 28;18(1):506. doi: 10.1186/s12913-018-3275-7.

Establishing the HLS-Q12 short version of the European Health Literacy Survey Questionnaire: latent trait analyses applying Rasch modelling and confirmatory factor analysis.

Author information

1
Department of Public Health and Department of Nursing, Faculty of Social and Health Sciences, Inland Norway University of Applied Sciences, PO Box 400, N-2418, Elverum, Norway. hanne.finbraten@inn.no.
2
Department of Health Sciences, Faculty of Health, Science and Technology, Nursing science, Karlstad University, SE-65188, Karlstad, Sweden. hanne.finbraten@inn.no.
3
Department of Health Sciences, Faculty of Health, Science and Technology, Nursing science, Karlstad University, SE-65188, Karlstad, Sweden.
4
Department of Nursing, Faculty of Social and Health Sciences, Inland Norway University of Applied Sciences, PO Box 400, N-2418, Elverum, Norway.
5
Department of Nursing and Health Promotion, Faculty of Health Sciences, OsloMet - Oslo Metropolitan University, PO Box 4, St Olavs plass, N-0130, Oslo, Norway.
6
Norwegian Centre for Science Education, Faculty of Mathematics and Natural Sciences, University of Oslo, PO Box 1106, Blindern, N-0317, Oslo, Norway.

Abstract

BACKGROUND:

The European Health Literacy Survey Questionnaire (HLS-EU-Q47) is widely used in assessing health literacy (HL). There has been some controversy whether the comprehensive HLS-EU-Q47 data, reflecting a conceptual model of four cognitive domains across three health domains (i.e. 12 subscales), fit unidimensional Rasch models. Still, the HLS-EU-Q47 raw score is commonly interpreted as a sufficient statistic. Combining Rasch modelling and confirmatory factor analysis, we reduced the 47 item scale to a parsimonious 12 item scale that meets the assumptions and requirements of objective measurement while offering a clinically feasible HL screening tool. This paper aims at (1) evaluating the psychometric properties of the HLS-EU-Q47 and associated short versions in a large Norwegian sample, and (2) establishing a short version (HLS-Q12) with sufficient psychometric properties.

METHODS:

Using computer-assisted telephone interviews during November 2014, data were collected from 900 randomly sampled individuals aged 16 and over. The data were analysed using the partial credit parameterization of the unidimensional polytomous Rasch model (PRM) and the 'between-item' multidimensional PRM, and by using one-factorial and multi-factorial confirmatory factor analysis (CFA) with categorical variables.

RESULTS:

Using likelihood-ratio tests to compare data-model fit for nested models, we found that the observed HLS-EU-Q47 data were more likely under a 12-dimensional Rasch model than under a three- or a one-dimensional Rasch model. Several of the 12 theoretically defined subscales suffered from low reliability owing to few items. Excluding poorly discriminating items, items displaying differential item functioning and redundant items violating the assumption of local independency, a parsimonious 12-item HLS-Q12 scale is suggested. The HLS-Q12 displayed acceptable fit to the unidimensional Rasch model and achieved acceptable goodness-of-fit indexes using CFA.

CONCLUSIONS:

Unlike the HLS-EU-Q47 data, the parsimonious 12-item version (HLS-Q12) meets the assumptions and the requirements of objective measurement while offering clinically feasible screening without applying advanced psychometric methods on site. To avoid invalid measures of HL using the HLS-EU-Q47, we suggest using the HLS-Q12. Valid measures are particularly important in studies aiming to explain the variance in the latent trait HL, and explore the relation between HL and health outcomes with the purpose of informing policy makers.

KEYWORDS:

Confirmatory factor analysis of categorical data; HLS-EU-Q47; HLS-Q12; Health literacy; Rasch modelling; Short version; Validation

PMID:
29954382
PMCID:
PMC6022487
DOI:
10.1186/s12913-018-3275-7
[Indexed for MEDLINE]
Free PMC Article

Supplemental Content

Full text links

Icon for BioMed Central Icon for PubMed Central
Loading ...
Support Center