Missing data is one of the most frequent problems to be addressed in longitudinal data analysis. Missing data in a longitudinal study is often due to attrition, unit non response or item non response. However, omissions can also be a consequence of the design of the study: in a planned missing design part of the missingness is due to an option made by the researcher to avoid the burden on the respondent and, hence, increase the quality of the data that are available (C.K. Enders, 2010). The statistical analysis of longitudinal data can be done using latent growth curve models (LGCM), which allow to capture information about interindividual differences in intraindividual change over time. The patterns of change are summarized in relatively few parameters: the means and variances of the random effects, as well as the covariance between intercept and slope (Bollen & Curran, 2006). This talk presents the main results and conclusions from a Monte Carlo simulation study conducted to investigate the effect of non-response due to a planned missing design on parameter estimates, standard errors and fit measures. LGCMs with unconditional linear growth (and three or four time points) are considered. Sample sizes of 100, 250 and 500 observations are used. The impacts of different patterns and percentages of missingness are discussed.

--