We tested if violent versus non-violent and competitive versus cooperative versus solo videogames affect teenagers' levels of aggression. This study had a 2 x 3 between-subjects design whereby participants were randomly assigned to one of the two-game violence and one of the three interaction type conditions. Aggression levels were measured by having teenagers interact with a confederate (i.e., someone that is secretly acting part of the study) and having third-party observers rate the aggression levels of the teenagers (on a scale from 1 = not at all aggressive to 100 = very aggressive). Below is the data that was collected: