Effects of giftedness and achievement on the training and transfer of a strategy for solving analogies

This research explored group differences according to giftedness and achievement in the acquisition and generalization of a strategy for solving analogies. A distinction was made between proximal and distal transfer, with the latter expected to differentiate between gifted and nongifted cognition. Underachievement in gifted children was expected to reflect either strategy deficits, or the absence of performance differences in a theoretically important cognitive skill (generalization) between the so-called "underachieving" gifted and other bright but nongifted children. 162 seventh and eighth graders were selected according to intelligence and achievement scores, academic program, and teacher opinion, and assigned to one of four groups: high achieving gifted, underachieving gifted, high achieving nongifted, and average achieving nongifted. Each child was seen individually for two sessions, and solved a total of five sets of ten multiple-choice analogies. The first session included two baseline trials (one verbal and one figural set), followed by training in the use of a strategy. The second session included a proximal transfer trial (same analogy type as used at training), and a distal transfer trial (analogies from the never-trained domain). All analogies were solved orally, and strategy use was determined from audio-recordings. The results showed that the high achieving gifted children were more spontaneously, frequently, and successfully strategic than the other three groups, as well as most accurate following the decision not to use a strategy. They were also the only group to show performance increases at distal transfer. In terms of gifted underachievement, there was evidence to support both hypotheses. The underachieving gifted children showed qualitative deficits in strategic functioning as compared to their high achieving gifted counterparts, and also tended to "look" like the high achieving nongifted group in their patterns of performance. These results were discussed in terms of the likelihood of subgroups of underachieving gifted children, and their implications for education and the identification of giftedness.