Enhancing the utility of CLSA cognitive measures: Handling non-ignorable missing data for more accurate inferences

Project Summary

Understanding how cognitive function changes as people age is a key objective of aging research. The Canadian Longitudinal Study on Aging collects a wide range of cognitive data to help researchers explore these changes. However, many participants do not complete all parts of the assessments, often due to the difficulties related to their cognitive abilities. This missing information can lead to biased results if not handled properly. This project will use advanced statistical methods to better account for missing cognitive data. By modelling potential causes of the missing responses, we can estimate each person’s cognitive ability even if they didn’t complete every assessment item. These models allow us to estimate individuals’ cognitive abilities even when some test items are incomplete. The resulting scores are more accurate, account for measurement error, adjust for inconsistencies across groups, and make full use of the available data.