The Development Of Judgments Of Learning

Judgments of learning (JOLs) are assessments that people make about how well they learned particular information. The mechanism study focuses on the magnitude and accuracy of JOLs. The magnitude refers to the prediction about the future test performance on recently. The accuracy refers to the consistency between JOL magnitude and test performance. The developmental study focus on the age and individual differences which in the magnitude and accuracy of JOLs. In the present studies, 4～6^th graders of primary school participated in the experiment, and the concrete noun pairs were selected as experimental materials. Five experiments were conducted to explore the magnitude, accuracy and bias of JOLs with the age.This paper involved two studies. The first study included Experiment 1 and 2. The second study consisted of Experiment 3,4 and 5.The first study was conducted to explore the developmental characteristics of magnitude and accuracy of item-by-item JOLs and aggregate JOLs under the immediate judgment and delayed judgment conditions. 110 students participated in Experiment 1, they were chosen from the 4～6^th grades of primary school with normal vision and intelligence. The same subjects participated in Experiment 2.In Experiment 1, we constructed 30 word pairs which included 17 unrelated and 13 associated pairs. The materials in Experiment 2 also included 17 unrelated and 13 associated pairs. The pair was composed of two Chinese words. Those Chinese words were all concrete nouns which were familiar to the students. The materials in Experiment 1 and 2 are homogeneous. The dependent variable was the number of the pairs in Experiment 1, and the same as Experiment 2. Magnitude of JOL was the number of the pairs that the student predicted how many he could recall in the future test. The recall score was the number of the pairs that how many he recalled in the future test really. The accuracy refers to the difference between magnitude and recall score.The second study investigated the foresight bias of the JOL accuracy. Experiment 3 explored the foresight bias of immediate JOL. 125 fourth-, fifth-, and sixth graders with normal vision and intelligence participated in the experiment. The materials included 20 associated pairs composed of two equal sets of 10 pairs which were matched in terms of the strength of the forward and backward associations. In addition to the 20 asymmetrical pairs, 10 unrelated pairs with zero associative strength were included in the list as the filler materials . Experiment 4 explored the foresight bias of delayed JOL. The participants were 117 students. The materials were the same as Experiment 3. Experiment 5 explored the mechanism of foresight bias. 36 fifth-graders participated in this experiment. The materials were the same as Experiment 3. The dependent variable was the number of the pairs in Experiment 3, and the same as Experiment 4 and 5.The results showed that (1) the trends of the JOLs magnitudes of the students were consistent. The magnitudes of immediate item-by-item JOLs were higher than under the delayed item-by-item JOLs, and The magnitudes of immediate aggregate JOL and delayed aggregate JOLs were lower. (2) There was age difference in the accuracy of immediate item-by-item JOLs. The JOL accuracy of fourth-and fifth graders was both overestimated, and the JOL accuracy of sixth-graders was good. No significant age difference was found in the accuracy of delayed item-by-item JOL, in which the accuracy of all the graders was good. (3) There was age difference in the accuracy of immediate aggregate JOLs. The JOL accuracy of fourth- and fifth graders was good, and the JOL accuracy of sixth-graders was underestimated. No significant age difference was found in the accuracy of delayed aggregate JOL, in which the students were all underestimated. (4)There was age difference in the foresight bias. When students making immediate JOL, there were foresight biases in the JOLs of fifth-and sixth-graders, but not in the JOLs of fourth-graders. When students making delayed JOL, there was foresight biases in the JOLs of fourth-graders, but not in the JOLs of fifth-and sixth-graders. (5) When the free recall test was required, there were no foresight biases in the JOLs of fifth-graders.In summary, the higher grade students in the elementary school use different cues such as encoding fluency and retrieval fluency that result in the differences of their JOL accuracy. Dual-factors hypothesis can explain the characteristics and the trend of the JOLs with the age. Foresight bias induces students to overestimate their memory. Delayed JOL can reduce foresight bias.