Tauber Sarah K Uma, Dunlosky John, Rawson Katherine A
1 Department of Psychology, Texas Christian University, Fort Worth, TX, USA.
2 Department of Psychology, Kent State University, Kent, OH, USA.
Exp Psychol. 2015;62(4):254-63. doi: 10.1027/1618-3169/a000296. Epub 2015 Jan 1.
The positive effect of delayed retrieval practice on subsequent test performance is robust; by contrast, making delayed judgments of learning (JOLs) encourages covert retrieval but has a minor influence on final test performance. In three experiments, we experimentally established and explored this memory-metamemory paradox. After initial study of paired associates (e.g., husky - ram), participants either were explicitly tested (husky - ?) or made a JOL. In Experiment 1, we adopted the standard JOL method, using a short retention interval, whereas in Experiments 2 and 3, we used a common testing-effect method involving a longer retention interval. Delayed JOLs did not boost test performance, but explicit delayed tests boosted memory after a longer retention interval. As important, participants spent less time to make JOLs than to retrieve responses. These data indicate that differences in the dynamics of retrieval for practice tests versus delayed JOLs are responsible for the paradox.
延迟检索练习对后续测试表现的积极影响是稳固的；相比之下，进行延迟学习判断（JOLs）会促进隐性检索，但对最终测试表现的影响较小。在三项实验中，我们通过实验确立并探究了这一记忆元记忆悖论。在对配对联想词（如哈士奇-公羊）进行初始学习后，参与者要么接受明确测试（哈士奇-？），要么进行学习判断。在实验1中，我们采用标准的学习判断方法，使用较短的保持间隔，而在实验2和实验3中，我们使用一种涉及较长保持间隔的常见测试效应方法。延迟的学习判断并未提高测试表现，但明确的延迟测试在较长的保持间隔后增强了记忆。同样重要的是，参与者进行学习判断所花的时间比检索答案的时间少。这些数据表明，练习测试与延迟学习判断在检索动态方面的差异是造成这一悖论的原因。
