Recently, there has been a call for more advanced analytic techniques in violence against women research, particularly in community interventions that use longitudinal designs. The current study re-evaluates experimental evaluation data from a sexual violence bystander intervention program. Using an exploratory latent growth curve approach, we were able to model the longitudinal growth trajectories of individual participants over the course of the entire study. Although the results largely confirm the original evaluation findings, the latent growth curve approach better fits the demands of "messy" data (e.g., missing data, varying number of time points per participant, and unequal time spacing within and between participants) that are frequently obtained during a community-based intervention. The benefits of modern statistical techniques to practitioners and researchers in the field of sexual violence prevention, and violence against women more generally, are further discussed.