Models are one example of external spatial representations which provide in
formation about referent spaces of the wider world. The ability to understa
nd and use such representations is an important one, since they can provide
previously unknown information about the real world, as well as affecting
how we think about and visualise it. Previous research has provided differi
ng indications as to the age at which such understanding emerges in childre
n. These differences may be the result of the different methods and tasks w
hich have been used to assess children's understanding of spatial represent
ations, since these tasks and methods may not necessarily be equivalent in
terms of the cognitive processes involved. The two experiments reported her
e provide a starting point for a systematic investigation into two tasks an
d two methods used. Performance on a hide-and-seek type task is compared to
performance on a positioning task, and the ability of children to make inf
erences about a referent space on the basis of a representation is compared
to their ability to make the converse inferences. Results suggest that you
ng children may find it easier to infer from a referent space to a represen
tation, but that this may depend upon the particular task they are asked to
carry out. Possible reasons for these differences are discussed, as are th
e implications for assessing children's actual competence in this domain, a
s opposed to the cognitive load of the experimental design.