Analysis of Unbalanced Simultaneously Clustered and Longitudinal Data Using Quasi-Least Squares in SAS

are measured on people who live within defined geographic regions. The measurements are then clustered within regions at each time point, but also comprise longitudinal measurements on each person across time. These data might be expected to have two potential sources of correlation, due to the similarity of measurements within and between the people in a region across time. While the former example considered multiple outcomes (e.g. systolic and diastolic blood pressure) measured repeatedly on each subject, this example considers multiple people assessed repeatedly in each region (e.g. in the same census tract). We will primarily consider subjects with multiple serial outcomes when we describe our software, but one should keep in mind that our results can also be applied to serial measurements of one outcome on different people within geographic regions.