An Experimental Study On Judgment Of Learning And Its Accuracy

Judgment of Learning refers to the prospective judgment of the recall performance after studying items. It is one of the most important forms of metacognitive judgment. Researchers continuously focus on the mechanism of judgment of learning and find some factors influencing judgment of learning, including the subjective judgment of the overall memory effects, the memory effects in different conditions, such as study time and encoding strategy, the relative degree of difficulty, expectant demands on test, such as the speed of memory and the accuracy of memory, and the experienced memory effects of specific task, and so on.In order to explain the mechanism of the judgment of learning, researchers presented several hypotheses, including mainly retrieval hypothesis, monitoring-retrieval hypothesis, monitoring-dual-memories hypothesis, accessibility hypothesis, competition hypothesis, cue utilization model, and fluency hypothesis.Retrieval hypothesis stresses the function of mnemonic trace in forecasting future recall performance. The mnemonic trace will deepen and enlarge with the increasing of study time and extent. The easier one reads the mnemonic trace formed during learning, the higher one makes judgment of learning. Fluency hypothesis emphasizes the speed of processing. The faster one processes information and the earlier the process ends, the higher subjective probability one retains and recalls an item. The finding of UWP effects indicates that with the prolonging of study time and the increasing of study trials, learners'confidence of recall decreases. The degree of process influences judgment of learning in two exclusive ways, the conclusion is inconsistent with the anticipation made by cue utilization model that judgment of learning is sensitive to the intrinsic cues but is insensitive to extrinsic cues. This thesis examines the influence that study time imposes on judgment of learning and recall performance in the perspective of the relationship between metacognitive monitoring and control.The experiments are made on the computers. After participants studied two kinds of association paired-words with different degree of difficulty, they were asked to forecast the recall performance of the paired-words in the future on the 5-points scales. And then make the participants recall the target words with the help of cue words. There are two experiments in this thesis. Experiment 1 attempt to validate that judgment of learning is driven by data in the condition in which control affects monitoring. 20 participants were randomly assigned to a condition of a 2(difficulty of items)×2(type of judgment) mixed design. Between-subjects variable was type of judgment and degree of items was within-subject variable. The dependent variables were participants'study time, grade of judgment and recall performance. Computers recorded the dependent variables automatically. Experiment 2 investigated the assumption that judgment of learning is driven by goals of learners in the condition in which monitoring affects control. 30 participants were randomly assigned to a condition of 2(difficulty of items)×2(level of incentives) within-subjects design. Factor 1 was the same with that in Experiment 1. Level of incentives was divided into two levels, the high level of incentive and low level of incentive. The dependent variables were the same with that in Experiment 2. The following conclusions were drawn from the above two experiments.1. Learners'control over their study can affect the forecast of their recall performance. In this condition it's the property of items which drives the judgment of learning. The characteristic of this kind of learning is that control affects monitoring.2. The perception of importance of learning materials can influence judgment of learning directly. Grades of judgment of learning increase with the increasing of processing extent. The factor which drives judgment of learning is some motive factors such as the goal of learning. The characteristic of that kind of learning is that monitoring affects control.3. Judgment of learning is sensitive not only to intrinsic cues but extrinsic as well, e.g. study time, learning intention, and test demands. The sensitivity to extrinsic cues is achieved through intrinsic cues.4. When the property of material and importance of learning goal are perceived by learners simultaneously, data-driven and goal-driven processing coexists. The goal-driven operation is more obvious than data-driven operation. Goal-driven operation has the priority over data-driven operation.