Our understanding of social information processing has been limited by the fact that observed behaviors conflate information processing with resulting responses. Here, we develop a mathematical model that isolates how individuals assess their environment separately from individual decision-making. We use this model to examine how social information affects environmental assessments and consider the influence of personal and social sampling efforts, social sampling methods, and types of environmental variation. We show that social information use often improves but sometimes impairs environmental assessments and that the magnitude of the effect increases with environmental variation. Furthermore, while there may be different ways to collect a given amount of information from social partners, our analyses suggest these differences have relatively little impact on the accuracy of environmental assessments.