作者: Agnes Hamon , Mounir Mesbah
DOI: 10.1007/978-1-4757-3625-0_13
关键词:
摘要: Quality of life studies must be concerned with the accuracy, or reliability (as it is usually called in psychometrics) measurement. A coefficient gives an evaluation a questionnaire’s ability to yield interpretable statements about construct being measured. There are two classes models for analyzing unidimensional scales: classical and Item Response Theory (IRT) models. Classical based on linear decomposition score, estimated using Cronbach’s alpha coefficient. IRT focus relationship between probability correct answer latent variable. The Rasch model special which having good measurement properties. In context model, general, no similar Cronbach Alpha clearly defined. Some authors have studied Fisher information parameter, provides measure accuracy estimator. this chapter, we present how models, propose model. Simulated data derived from communication subscale Sickness Impact Profile used illustrate methods.