Two experiments examined preschool children's ability to draw inferenc
es about numerosity from correspondences between sets. In Experiment 1
, 3- and 4-year-old children made numerical inferences about a hidden
set from their own counts of a corresponding visible set and also from
numerical information about that set stated by the experimenter. Expe
riment 2 contrasted a count condition with a move condition, in which
children's attention was not explicitly drawn to the numerosity of the
visible set. Again, children were able to make numerical inferences a
s early as 3 years of age. However, differences between the 2 conditio
ns implicate production deficiencies in young children's use of counti
ng as a problem-solving strategy when they are not explicitly told to
count.