Friday Harbor Laboratory University of Washington August 22-26, 2005 Advanced Psychometric Workshop on Methods for Cross-Cultural Cognitive Assessment Friday Harbor Laboratory University of Washington August 22-26, 2005 Friday Harbor Psychometrics Workshop 2005
Background Neuropsychological tests are widely used to detect cognitive changes associated with brain disease and injury Test results are used in diagnosing brain disease and injury Interpretation of test results can have important clinical implications Much of the research in this area has been done with Caucasians Friday Harbor Psychometrics Workshop 2005
Background Cognitive impairment associated with diseases of aging is a major clinical problem Alzheimer’s disease, cerebrovascular brain injury, other degenerative dementia, etc. Neuropsychological tests are widely used in clinical evaluation Much of the research in this area has been done with Caucasians Friday Harbor Psychometrics Workshop 2005
Background Well established finding that ethnic minorities have lower average performance on neuropsychological tests Cutpoints based upon Caucasian samples result in high false positive rates in minority populations Raises important question about measurement bias How do we use neuropsychological test results to measure cognitive changes of dementing illnesses in ethnically diverse groups Friday Harbor Psychometrics Workshop 2005
Psychometric Theory - Measurement of Ability Domain Friday Harbor Psychometrics Workshop 2005
Determinants of Cognitive Test Performance Friday Harbor Psychometrics Workshop 2005
Determinants of Cognitive Test Performance Friday Harbor Psychometrics Workshop 2005
Determinants of Cognitive Test Performance Friday Harbor Psychometrics Workshop 2005
Determinants of Cognitive Test Performance Friday Harbor Psychometrics Workshop 2005
Measurement Bias Differential validity across groups Validity is purpose specific Therefore, bias is purpose specific In context of diseases of aging, primary validation is ability to detect and monitor disease Friday Harbor Psychometrics Workshop 2005
Sources of Measurement Non-Invariance (Bias) Across Ethnic Groups Item Level Differential Item Function (DIF) Items have different relationship to ability in different groups Scale Level Differential relationship of ability to disease in different groups Friday Harbor Psychometrics Workshop 2005
Determinants of Cognitive Test Performance Scale Level Bias Item Level Bias Friday Harbor Psychometrics Workshop 2005
Differential Item Function (DIF)-Free Test Friday Harbor Psychometrics Workshop 2005
Test with Differential Item Function (DIF) Friday Harbor Psychometrics Workshop 2005
Determinants of Cognitive Function Friday Harbor Psychometrics Workshop 2005
Determinants of Cognitive Function Friday Harbor Psychometrics Workshop 2005
Determinants of Cognitive Function Friday Harbor Psychometrics Workshop 2005
Determinants of Cognitive Function Friday Harbor Psychometrics Workshop 2005
Purpose of Workshop Develop psychometric methods and skills for cognitive assessment in culturally diverse populations Presentations of conceptual and methodological issues and approaches Share experience and skills Apply methods to existing dataset to address important measurement issues Friday Harbor Psychometrics Workshop 2005
Studies/Manuscripts Identification of DIF and assessment of impact on scale level scores External validation of neuropsychological test results in a multiethnic sample Construct validation of reading tests as measures of premorbid ability Dimensional structure of tests of executive function in multiethnic sample Friday Harbor Psychometrics Workshop 2005
Dataset Spanish and English Neuropsychological Assessment Scales (SENAS) Item and scale level data Demographic and linguistic variables Clinical diagnosis Quantitative MRI Friday Harbor Psychometrics Workshop 2005
Schedule Monday Evening Introduction and Overview Overview of SENAS project Friday Harbor Psychometrics Workshop 2005
Schedule Tuesday AM Neuropsychology, neuroimaging, and dementia Cultural and demographic influences on test scores Tuesday Afternoon Free Tuesday Evening Logistic regression approaches to DIF IRT approaches to DIF Friday Harbor Psychometrics Workshop 2005
Schedule Wednesday AM Covariance structure/ Mplus applications to measurement Description of SENAS dataset Overview of manuscripts and workgroups Wednesday PM Workgroups Wednesday Evening Report of Workgroups Friday Harbor Psychometrics Workshop 2005
Thursday Schedule Morning Open didactic/discussion slot Workgroups Afternoon Evening Report of workgroups Friday Harbor Psychometrics Workshop 2005
Friday Schedule Morning Open didactic/discussion slot Workgroups Afternoon Evening Report of workgroups Friday Harbor Psychometrics Workshop 2005