karolina-m/catcdi_validation_eng_pl

AI Summary: Investigate why a Computerized Adaptive Test (CAT) algorithm, specifically the CAT-CDI implemented using the mirt package in R, is selecting items with low discrimination parameters. This involves analyzing the provided plot showing item difficulty and discrimination, understanding the CAT algorithm's item selection process, and potentially debugging the R code or consulting with the mirt package author.

AI Summary: Investigate whether the observed weak correlation (r = .37) between item frequencies in English and Polish CAT-CDI administrations is due to differences in test length. This involves recalculating item frequency per item administered instead of per administration, and assessing the impact of this change on the correlation and item ranking.

AI Summary: The task involves resolving a discrepancy in the interpretation and usage of IRT parameters 'b' (difficulty) and 'd' (easiness) within the context of a manuscript on the validation of the CAT-CDI. The goal is to determine which parameter ('b' or 'd') is more appropriate and consistent with the literature for describing item difficulty in the manuscript, ensuring clarity and accuracy in the presentation of results.