Developmentalists are often interested in understanding change processes, and growth models are the most common analytic tool for examining such processes. Among the growth models, nonlinear growth curves are especially valuable to developmentalists because the defining characteristics of the growth curves such as initial levels, rates of change during growth spurt, and asymptotic levels can be estimated. A variety of growth models were introduced and used for model data fit in this study. Statistical model fit was examined using statistics such as chi-square statistic, comparative fit index, Tucker-Lewis index, root mean square error of approximation, Bayesian information criterion , and Akaike’s information criterion.
Comments & Questions.
1. Actually, there is no model can fit to data according to the criteria of the fit indices. But the authors say that because the residual variances were very small, and the model accounted for at least 95% of the variability at each occasion…etc. So the Preece-Bainesb was accepted for the best model. But there were no such data for other models were represented.
2. In this study, it seems that nonlinear growth curves are especially valuable for description of change processes. It is probable to use the item response theory modes in these kinds of data, but I don’t know how to implement it.