G. Andrews et Gs. Halford, Children's ability to make transitive inferences: The importance of premise integration and structural complexity, COGN DEV, 13(4), 1998, pp. 479-513
Four experiments investigated 4- to 6-year-olds' transitive inferences. In
Experiments 1-3, there was a nonmapping condition in which inferences were
made either about stacked blocks, or about sticks ordered left to right. In
the mapping condition, inferences were made by mapping either from blocks
to sticks, or the reverse. In Experiments 2-4, relational complexity was ma
nipulated by requiring either 1 or 2 premise relations to be processed in a
single decision. Mapping was harder than nonmapping, but relational comple
xity was the main source of variance, with 2 relations being harder in both
mapping and nonmapping conditions. The percentage of participants integrat
ing 2 relations in a single decision was estimated at 20% at age 4, 53% at
age 5 and 57% at age 6, suggesting gradual development of transitive infere
nce ability. Results suggest that relational complexity has a strong effect
on transitive inference in 4- to 6-year-old children.