This study was designed to test whether calling to mind an initial belief a
nd presenting information that challenges that belief affects the extent to
which preschoolers will modify it. The belief that was challenged in a con
trolled demonstration concerns the effect of the size of an object on its s
inking speed (holding weight constant). In addition, children's belief abou
t the effect of weight on sinking speed (holding size constant) was examine
d, a belief that was confirmed in a demonstration. The final belief about s
ize for those who received nothing but empirical demonstrations was less li
kely to be compatible with the demonstration than the final belief of those
in two other conditions. Children in the other conditions were given the o
pportunity in the context of interviews to form expectations about how size
and weight separately relate to sinking speed, in addition to receiving th
e demonstrations. An interview either directly preceded the demonstration f
or the variable concerned (coordinated sequence) or did not (uncoordinated
sequence). The tendency for the final belief about size to be compatible wi
th the demonstration was related more strongly to age in the condition with
an uncoordinated sequence than in either of the other conditions. Some chi
ldren among those whose final belief about the effect of size on sinking sp
eed was compatible with the demonstration also refined their belief about t
he effect of weight, suggesting that these two beliefs may cohere as a syst
em. These findings show that a relatively short experimental procedure can
be an effective means of bringing about some refinement of a young child's
beliefs. (C) 2001 Elsevier Science Inc. All rights reserved.