6533b823fe1ef96bd127e21f

RESEARCH PRODUCT

The SF-8 Spanish Version for Health-Related Quality of Life Assessment: Psychometric Study with IRT and CFA Models.

Irene FernándezLaura GalianaJosé M. Tomás

subject

MaleLinguistics and LanguagePsychometricsmedia_common.quotation_subjectLogistic regressionLanguage and LinguisticsStructural equation modeling03 medical and health sciences0302 clinical medicinePerceptionItem response theoryPersonalityHumans030212 general & internal medicinePatient Reported Outcome MeasuresGeneral Psychologymedia_commonAgedAged 80 and over030505 public healthModels StatisticalSpanish versionMiddle AgedMental healthTest (assessment)SpainQuality of LifeFemale0305 other medical sciencePsychologyClinical psychology

description

AbstractThe aim of current research is to analyze the psychometric properties of the Spanish version of the SF–8, overcoming previous shortcomings. A double line of analyses was used: competitivestructural equations modelsto establish factorial validity, andItem Response theoryto analyze item psychometric characteristics and information. 593 people aged 60 years or older, attending long life learning programs at the University were surveyed. Their age ranged from 60 to 92 years old. 67.6% were women. The survey included scales on personality dimensions, attitudes, perceptions, and behaviors related to aging. Competitive confirmatory models pointed out two-factors (physical and mental health) as the best representation of the data: χ2(13) = 72.37 (p< .01); CFI = .99; TLI = .98; RMSEA = .08 (.06, .10). Item 5 was removed because of unreliability and cross-loading. Graded response models showed appropriate fit for two-parameter logistic model both the physical and the mental dimensions. Item Information Curves and Test Information Functions pointed out that the SF–8 was more informative for low levels of health. The Spanish SF–8 has adequate psychometric properties, being better represented by two dimensions, once Item 5 is removed. Gathering evidence on patient-reported outcome measures is of crucial importance, as this type of measurement instruments are increasingly used in clinical arena.

10.1017/sjp.2018.4https://pubmed.ncbi.nlm.nih.gov/29562950