Repetitive and obsessive behaviors are a serious symptom of many clinical disorders. While these behaviors can be debilitating and greatly reduce quality of life, their temporal structures and developmental paths are still largely unknown. Following up on research done by Tanimura, Yang, Ottens, and Lewis (2009), we have conducted a study of a mouse model of repetitive behavior in order to better classify development paths and behavior groupings.

Trajectory models were used to describe behavioral evolution and classify developmental pathways. In the course of our research, we reaffirmed the existence of three quantitatively distinct trajectory groups found in the previous paper within our larger and more complete data set. The larger data set also allowed us to implement point process models, clustering, and semiparametric models to identify patterns within single observational periods and across developmental stages. Finally, hidden Markov models were used to further describe mouse behavior and aid in clustering. These findings help give a better understanding of the temporal structures of repetitive behavior in our mouse models and to better classify individual mice in early stages of development, which will allow for more accurate physiological study.