Are group variability judgments made in an on-line or memory-based fas
hion? In a first experiment addressing this question, subjects made ju
dgments of a group's variability, a judgment intended to be on-line (l
iking), or a judgment intended to be memory-based (religiousness). Var
iability judgments were made more slowly than on-line judgments and at
the same speed as memory-based judgments. Independently of this effec
t, in-group variability was judged more rapidly than out-group variabi
lity. In addition to replicating these results, a second experiment us
ing a minimal group paradigm demonstrated that the amount of similarit
y information recalled predicted both the latency and extremity of var
iability judgments. We discuss the implications of our conclusion that
variability judgments are predominantly memory-based for models of gr
oup variability judgments.