The distinctive lesson from this project--the BrainPower Program to reduce peer directed, reactive aggression in elementary-school students--concerned students' perceptions of observers who watched and recorded their behaviors on the playground. The purpose of the observations was to record any differences in the aggressive behaviors of children who were participating in the BrainPower Program and children who were not participating. The observers reported insignificant differences in the behaviors of the two groups of children in their playground activities across four school sites. These observer findings contrasted with the significant changes in participants compared to controls indicated in other evaluation methods, such as teacher ratings and student self-reports. This article suggests that students' awareness of additional adults observing them on the playground caused all of the students, whether controls or program participants, to curtail their aggressive behaviors with peers. This leads to the question about whether simply adding more adults to children's environments of interaction with other children or reducing the adult-to-child ratio in school settings can be an effective intervention. This possibility poses a hypothesis for a future evaluation. In terms of using data-collection methods that produce valid findings regarding an intervention, this article suggests that the use of observers who are visible to those being evaluated is not a sound evaluation method. If possible, current audiovisual technologies should be used to produce a more valid record of behaviors in various settings. 223 references