The Effect of Initial Accuracy on the Learning and Retention of Novel Words

During word learning, incorrect guesses about meaning are inevitable, particularly in difficult learning situations. Prior research has shown that optimally difficult tasks improve word learning and retention, but found no item-level effects of first-guess inaccuracy on learning, raising the question of how difficulty and guessing accuracy interact in this process. Effects of first-guess accuracy on retention are unknown. To unpack the effects of task difficulty and first-guess accuracy on word learning and retention, we conducted two cross-situational word-learning (CSWL) studies. In Study 1, 49 English-speaking adults completed a difficult CSWL task with four items presented on each trial. Their word learning was tested immediately following the learning phase and then again 48 h later. Data were analyzed with respect to the accuracy of the first guess of a word's meaning. Participants in this study learned and retained words best when their first guess had been correct. Study 2 simplified learning, presenting three items per trial. Fifty-two English-speaking participants showed high levels of learning and retention, with no significant differences based on first-guess accuracy. Comparing data across these studies revealed effects of both task difficulty and first-guess accuracy on learning, as well as an interaction of these factors, revealing a greater cost to inaccurate first guesses on the more difficult task. Retention of word meanings, however, was associated only with task difficulty. These results have implications for theories of error-driven versus difficulty-driven learning.