Nonsense syllables in associative recognition tasks: Implications for global memory models

Recent research into global memory models has focused on gaining empirical support for the predictions these models make. Clark and his colleagues produced no conditions in which a test item and a distracter shared a common component within a number of experiments using an associative recognition task and have reasoned that recall might play a role in the task. This experiment used nonsense syllables to reduce further the role of recall in the associative recognition task and produce the predicted advantage from overlap. The manipulations produced equal performance between test conditions (overlapping vs nonoverlapping) and superior performance on a long list versus a short list, i.e., a negative list-length effect. The implications of these findings for various global memory models are discussed.