Date of Graduation


Document Type


Degree Type



College of Education and Human Services


Learning Sciences and Human Development

Committee Chair

Dan Hursh.


Over the past decade, systematic applications of Applied Behavior Analysis (ABA) have proven an effective means of improving the prognosis for many young children with autism. But in every study indicating that a number of participants have experienced good outcomes from these interventions, there remain a sizeable number of participants who have not had the same success. This study was undertaken to examine the differences between the parental rankings for each of eight characteristics across three points in time for children with autism to determine who succeeds with this kind of treatment and who does not. These parental reports were also examined to determine which rankings, if any, of these characteristics at the time of diagnosis might predict a positive outcome. Participants were recruited through four different Internet listserves dedicated to the use of ABA interventions with young children with autism. Eighty-five respondents, all parents of young children with autism, answered a 50 question survey that included, among other questions, a ranking of their child's ability in speech, social attachment, toy play, sensory problems, peer play, tantruming, self-stimulation, and toilet training, across three points in time: at diagnosis, when treatment ended, and when the survey was completed. The results of this study showed, like many before it, that most young children with autism will improve in their functioning, at least in these characteristics, after at least one year of discrete trial training, a method of ABA. In this study, the children not only improved overall based on the comparison of characteristics ranks across time, but they also showed increases in the sums of all eight of the characteristic variables. These results also indicated that, of the eight variables, the ranking for speech increased the most for most children, followed by toilet training and toy play. For the larger group, those that had not yet completed discrete trial training, no one variable, or groups of variables within the eight characteristics, were found to significantly contribute to the outcome (recovery score) more than the others. For the group that had completed treatment (n = 29), one variable, toy play, was negatively correlated with a positive outcome from treatment.