Young children show remarkably sophisticated abilities to evaluate others. Yet their abilities to engage in proportional moral evaluation undergoes protracted development. Namely, young children evaluate someone who shares "absolutely" more as being "nicer" than someone who shares "proportionally" more (e.g., sharing 3-out-of-6 is nicer than sharing 2-out-of-3, because 3 > 2, even though 3/6 < 2/3), whereas adults think the opposite. We investigate the hypothesis that this prior work underestimates children's proportional social reasoning by relying on discrete and spatially separated quantities (e.g., individual stickers), which can hinder proportional reasoning even outside social contexts. In three experiments we examine whether 4- and 5-year-old children's social evaluations are impacted by the discreteness and spatial separation of the resource and compare their behavior to adults (18 to 63 years; across all samples: 38% girls/women, 62% boys/men; no other demographic data was collected). We find that children are sensitive to these features: when the resource was divided into discrete units (Experiment 1) or spatially separated (Experiment 2) children were more likely to use absolute amount, as opposed to proportion, relative to when the resources were not divided and remained spatially connected. However, adults were highly sensitive to proportion regardless of the display's perceptual features (Experiment 3), and children's use of proportion remained below adult-levels. These results suggest that perceptual features influence children's use of absolute versus proportional information in their social evaluations, which has theoretical and methodological implications for understanding children's conceptions of fairness.