作者: Paul K. Crane , Kaavya Narasimhalu , Laura E. Gibbons , Dan M. Mungas , Sebastien Haneuse
DOI: 10.1016/J.JCLINEPI.2007.11.011
关键词: Cognitive Abilities Screening Instrument 、 Cognitive skill 、 Cognition 、 Cognitive test 、 Dementia 、 Psychometrics 、 Psychology 、 Item response theory 、 Statistics 、 Level of measurement
摘要: Abstract Objective To cocalibrate the Mini-Mental State Examination, Modified State, Cognitive Abilities Screening Instrument, and Community Instrument for Dementia using item response theory (IRT) to compare screening cut points used identify cases of dementia from different studies, measurement properties tests, explore implications these on longitudinal studies cognitive functioning over time. Study Design Setting We cross-sectional data three large ( n > 1000) community-based in elderly. IRT scales performed simulations studies. Results varied quite widely across The four tests have curvilinear scaling levels precision, with more error at higher functioning. In simulations, scores always better than standard scoring, whereas a strategy account varying precision had mixed results. Conclusion Cocalibration allows direct comparison any tests. Standard scoring appears be poor choice analysis testing data. More research is needed into precision.