Discussion Paper no. 01-17


This paper uses longitudinal data to examine negative child outcomes among young children in Canada. In following a cohort of children as they move through their early elementary school years, a slight decline in hyperactivity is documented, as is a slight increase in internalizing difficulties. After introducing a longitudinal dimension to our analysis, both family structure and income poverty are found to have a weaker effect on child outcomes than was initially anticipated. The current analysis also demonstrates the utility of LISREL’s “weighted least squares” estimation procedure in the development of structural equation models while working with ordinal and/or censored variables.