Reasoning from memory: a lifespan inquiry into the necessity of remembering when reasoning about class inclusion.

In the current experiment, we evaluated the relationship between reasoning and remembering across the lifespan (10-, 13-, 15-, 23-, and 67-year-olds) using color and number class-inclusion tasks. Subjects made judgments when memory load or information load was either high or low. Reading latency, choice latency, and choice accuracy data were collected and analyzed using formal models that partitioned memory and reasoning as well as type of reasoning strategy. The results showed that: (1) study skills, as indexed by reading latency, were sophisticated across the lifespan; (2) the difference between color and number choice latencies increased with age; (3) reasonably consistent effects were obtained as a function of increasing memory load, with the quality of both reasoning and remembering being adversely affected; (4) different effects were obtained at each age as a function of increasing information load; and (5) the major developmental trend across the lifespan was a tradeoff between class-inclusion and subclass-subclass forms of reasoning. A number of extant theoretical approaches were discussed, none of which were able to account for variations in class-inclusion performance across the lifespan. What was clear, however, was that memory-reasoning dependencies must be incorporated into any comprehensive theory of cognitive development across the lifespan.