|
Monday, September 20
|
||
|---|---|---|
|
Evening
|
Welcome to Friday Harbor; Introduction to conference | Gerald van Belle, Paul Crane, Bill Lee |
| Broad overview to AD epi questions for non-AD-epidemiologists: missing data, longitudinal analyses, risk factor analyses, screening, two stage sampling | Gerald van Belle | |
| Broad overview of IRT for non-test theorists: θ, I(θ), rational test construction, DIF |
Paul Crane Slides |
|
|
Tuesday, September 21 |
||
|
Morning |
Issues in longitudinal data analysis and the advantages of mixed effects model regression formulations | Richard Kryscio, Brenda Kurland |
| Basic introduction to M+ and PARSCALE | Rich Jones, Paul Crane | |
| Broad overview of the CHS dataset: design, data elements available, work done to date; Discussion: hypotheses to test: risks for cognitive decline over time in the CHS dataset | Alice Arnold, Annette Fitzpatrick | |
|
Afternoon |
Work session 1: PARSCALE of the 3MSE | Paul Crane |
| Work session 2: set-up of mixed-effects models (other than cognitive scores); testing of assumptions | Richard Kryscio | |
|
Evening |
Report of PARSCALE group | Paul Crane |
| Report of mixed-effects models group | Richard Kryscio | |
|
Wednesday, September 22 |
||
|
Morning |
Issues in scale co-calibration / score equating | Rich Jones, Paul Crane, Richard Swartz |
| Discussion: how should we incorporate SE(theta) into mixed effects regression models? | Gerald van Belle | |
| Excel spreadsheet analysis of IRT models | Dan Mungas | |
|
Afternoon |
Work session 1: M+ analysis | Rich Jones |
| Work session 2: incorporation of 3MS theta and SE(theta) into mixed effect regression models | Richard Kryscio | |
| Work session 3: MMSE and 3MSE co-calibration / score equating | Paul Crane | |
| Item 6 question | Jim Scanlan | |
|
Evening |
Report of M+ analysis group | Rich Jones |
| Report of mixed effects regression group | Richard Kryscio, Marta Mendiondo | |
| Report of co-calibration group | Paul Crane | |
| Item 6 question group | Jim Scanlan | |
|
Thursday, September 23 |
||
|
Morning |
Testing IRT assumptions: unidimensionality using MPLUS; local independences | Rich Jones, Paul Crane |
| DIF detection using ordinal logistic regression and integration into hypothesis testing; DIF detection using MIMIC formulation | Paul Crane, Rich Jones | |
| Missing data and imputation: NORM from Penn State | Richard Kryscio, Marta Mendiondo | |
|
Afternoon |
Free | |
|
Evening |
Report of MPLUS group | Rich Jones |
| Report of DIFdetect group | Paul Crane | |
| Report of mixed effects models group | Richard Kryscio | |
| Change with time | Xiaowei Song | |
|
Friday, September 24 |
||
|
Morning |
Comparison of standard sum scores to DIF-adjusted IRT scores; design of validation analysis | Dan Mungas |
| Rash model 2PL comparisions | Deborah Schnipke | |
| Issues of diagnosis; optimal screening test properties; use of ROC curves to determine area in which to maximize information | Paul Crane | |
| CAMCOG | Rosemary Abbott | |
| Discussion: final mixed effects models incorporating all of the things we have considered: item-level DIF incorporation, missing data, 3MSE and MMSE, and how to write up | Richard Kryscio | |
| Discussion: Manuscript preparation and dissemination plan | Paul Crane, Rich Jones, Gerald van Belle, Dan Mungas | |
|
Afternoon |
Work session 1: Validation study of DIF-adjusted IRT scores | |
| Work session 2: 3MSE screening test optimization | ||
| Work session 3: Mixed effects model group, incorporating DIF-adjusted IRT scores | ||
|
Evening |
Risk factor group | Jim Scanlan Slides |
| M+ working group; APoE effect; growth modeling by subgroups as predictive model; DIF analysis | Rich Jones, Dagmar Amtmann, Rosemary Abbott Jones Slides Amtmann Slides |
|
| Report of mixed effects model group: final conclusions about risk factors | Richard Kyrscio |
|
| ROC Group | Paul Crane |
|
|
Saturday, September 25 |
||
|
Morning |
Wrap-up session: lessons learned from conference, future plans | All |
Advanced Psychometric Methods for Cognitive Aging Research (ΨMCA)
Psychometric and statistical methods for cognitive aging science