Different social contexts have been used when measuring distributional preferences. This could be problematic as contextual variance may inadvertently muddle the measurement process. We use a within-subjects design and measure distributional preferences in resource allocation tasks with role certainty, role uncertainty, decomposed games, and matrix games. Results show that, at the aggregate level, role uncertainty and decomposed games lead to higher degrees of prosociality when compared to role certainty. At the individual level, we observe considerable differences in behavior across the social contexts, indicating that the majority of people are sensitive to these different social settings but respond in different ways.