Single subject experimental designs for behavior modification analysis and monitorage, are considered among the most significant contributions from contemporary behavioral psychology to the applied analysis of behavior. Behavioral analysts have for many years shown a preference for these designs mainly because they offer: a) more accuracy in evaluating the individual performances as opposed to inferring them from group averages; b) greater flexibility in the experimental design as a function of induced behavioral modifications; c) the experimental analysis of rare events, for which it is too difficult to find groups of subjects; d) more evidence for main effects, due to greater control over the causes of behavioral variability. More recently statistical procedures based on time series analysis provide these designs with both (a) internal and (b) external validity, permitting: (a) control of the cause-effect relation between independent (i.e. educational procedures) and dependent variables (behavior), and (b) generalization of the results for different subjects and for many settings. We present here an instance of the single subject experimental design, for time series, planned to modify the verbal behavior of a 7 year old child with syndrome of Down and autistic aspects in the behavior. In our presentation, the emphasis is laid on the statistical procedure we used to verify the effectiveness of the educational treatment programmed to modify the child’s behavior.