Iwan S J, Siegel G M
J Speech Hear Res. 1982 Jun;25(2):224-9. doi: 10.1044/jshr.2502.224.
Preschool children were paired in 12 speaker-listener dyads in which the speaker described common, familiar items and the listener attempted to guess their identity. Objects were presented until the listener had successfully guessed five objects and had missed five objects. Successfully guessed objects were conspicuously placed in an E-Z box whereas those missed were conspicuously placed in a HARD box. These 10 objects were then presented to the same team for identification once again, with each object being removed from the E-Z or HARD box with appropriate comments about its difficulty offered by the experimenter. Postfeedback, the speakers used longer and more informative descriptions for items that had originally been failed, and shorter, less informative descriptions for those that had been successfully guessed on the first identification trials. The changes in the form of the messages for the success items were somewhat greater than for the failure items, suggesting that children might not have mastered specific strategies for composing maximally useful messages, even when the children were disposed to respond to listener communicative needs.
学龄前儿童被分成12组说话者-倾听者对子，其中说话者描述常见的、熟悉的物品，倾听者试图猜出它们是什么。不断展示物品，直到倾听者成功猜出5个物品且猜错5个物品。成功猜出的物品被明显地放在一个“简易”盒子里，而猜错的则被明显地放在一个“困难”盒子里。然后，这10个物品再次展示给同一组，让他们再次辨认，实验者会从“简易”或“困难”盒子中取出每个物品，并对其难度给出适当的评价。反馈后，说话者对最初猜错的物品使用更长、信息更丰富的描述，而对第一次辨认试验中成功猜出的物品使用更短、信息更少的描述。成功物品的信息形式变化比失败物品的变化更大，这表明即使孩子们倾向于回应倾听者的交流需求，他们可能也没有掌握构建最有用信息的具体策略。
