Recently there has been considerable interest and developments in the analysis of longitudinal data. One such area is latent growth curve modeling. However, it could be very difficult to analyze count data using such models, especially when there is a preponderance of counts that are zero (zero inflation). This seminar will use data on criminal behavior (where most of the counts of criminal acts are 0) in a longitudinal context to understand and predict criminal behavior over time. Examples will be provided using Mplus version 3.
- View the powerpoint slides (as a .pdf file)
- View the Mplus input, output, and graph files (as a .zip file)