Validity and reliability of the Turkish health literacy scale – Short form (THLS-SF) in the older people: A practical tool for assessment of health literacy

Abstract

Although various long and complex scales related to health literacy exist — including those that assess individuals’ ability to exhibit appropriate behaviours to improve their health — there is no known short health literacy scale applicable to the general population, especially older adults. The aim of this study was to develop a short scale form that could measure health literacy levels in a practical and valid manner. During the development of the short form, methodological issues such as a decrease in validity and reliability, loss of content representation, and inconsistency across different groups were encountered. These issues were overcome through expert opinions, comprehensive psychometric analyses (EFA, CFA), the use of statistical criteria for item selection, and test applications across different demographic groups. The study is a methodological study conducted on 1094 cognitively and mentally healthy community-dwelling elderly individuals aged 65 and over. Data were collected through face-to-face interviews conducted by the researchers in participants’ homes. The questionnaire included sociodemographic characteristics, Turkish Health Literacy Scale-32, World Health Organisation Quality of Life Questionnaire - Elderly Module and Barthel Activities of Daily Living Index. In the study, reliability and validity analyses of the 8-item short form of the THLS, which was formed with one item from each dimension, were conducted. Psychometric properties were analysed by applying confirmatory and explanatory approaches (item analysis, Rasch analysis, internal consistency, discriminant validity, known groups validity). 71.3% of the participants in the research group are in the 65–74 age group, 52.1% are women, 77.9% are primary school graduates, and 63.8% have an excellent health perception. The short form has a unidimensional structure, with floor and ceiling effects of 2.1% and 11.6%, respectively. Cronbach’s alpha value of the form is 0.928, and the explained variance of the unidimensional structure is 66.8%. Confirmatory factor analysis summary fit values for construct validity are RMSEA=0.007, CFI=1.00, and χ2/sd=1.04. The item difficulty and discrimination level of the created eight-item form are sufficient, Rasch analysis item difficulty and distribution properties are compatible, reliability is 0.93, and PSI is 3.59. Scale items do not show DIF in gender and age distributions, do not violate local independence, and have a unidimensional structure. The scale significantly correlates with THLS-32 and Barthel and WHOQOL-OLD (p<0.05). The THLS-SF score significantly differed across gender, age, education level, income level, and health status (p<0.05). In conclusion, the eight-item TSOY-SF created from the THLS-32 is a valid and reliable measurement tool for assessing health literacy when applied to older adults. The short form obtained saves time and reliably measures the basic dimensions of health literacy.