This study, which assessed hypothesis-testing abilities using a discrimination-learning paradigm, found that 16 language-impaired primary-level children solved fewer problems than 16 controls equated on cognitive level, but the 2 groups used similar hypothesis types to solve the problems. Type of verbal feedback (explicit versus nonexplicit) did…

The question of how children represent and use comparative or partially ordered information is examined. Two experiments tested a conjecture that a common representation, a linear order, underlies the processing of all comparatives. (Author/MS)