Children's memories according to fuzzy-trace theory: an endorsement of the theory's purpose and some suggestions to improve its application.

This reflection on the work of Brainerd and Reyna (1998, this issue) focuses on the goals of mathematical modeling of memory. Suggestions are made that might improve the experimental design used to examine the model in children, and points are raised regarding the boundary conditions of the model. It is concluded that the model currently cannot handle situations in which partial verbatim information is used along with partial gist information or in which information has an inhibitory effect. Both encouragements and precautions for the use of modeling are offered.