Fig. 4From: Longitudinal change in restricted and repetitive behaviors from 8-36 monthsBest-fitting model estimates for Repetitive Motor scores plotted over raw data for (a) MNLFA-derived factor score estimates and (b) raw mean scores. The model outcome for the raw mean score is converted to probability, as a logistic linking function was used. Y-axes for both plots are scaled with limits of mean outcome +/− 2.5 SDs. c Impact of sex on intercept and slope. Parents of females reported fewer repetitive motor behaviors at 18 months, and their rate of decline was more rapid relative to malesBack to article page