Observer behavior as a potential source of reactivity: Describing and quantifying observer effects in a large-scale observational study of police

One theme in the field research literature is that observer effects (or observer behavior) can act as a source of bias in observational data. This article uses both descriptive and quantitative data on reactivity from a large scale observational study of police to (1) describe and categorize instances of observer effects (in the form of observers helping officers), (2) specify hypothesized effects between rapport and observer effects, and (3) assess the impact of observer effects on patrol officer arrest and use of force behavior. Descriptive accounts indicate that observer behavior at times significantly altered the course of events. In addition, bivariate analyses suggest that observer effects are likely officer initiated rather than observer initiated. Finally, both bivariate and multivariate analyses revealed that observer effects are significant predictors of use of force but not arrest behavior net of a variety of controls for acclimation between officer and observer over the course of fieldwork, officer work orientation, suspect characteristics, and situational factors. The implications of these findings for future research are discussed.