An Empirical Study About The Underconfidence-With-Practice Effect In The Judgment Of Learning

As an important metamemory monitoring judgement, judgment of learning is the prediction about the future performance of learned materials. The underconfidence-with-practice effect ,which describes an underconfidence phenomenon that people underestimate their recall performance during a test on their repeated study, is an intriguing and counter-intuitive phenomenon in the field of judgment of learning. There were many researches on the underconfidence-with-practice effect at abroad, most of the experimental evidence supports that the underconfidence-with-practice effect in the judgment of learning is robust and pervasive, whereas some other evidence demonstrated some constraints on the occurrence of the effect. With respect to the functional mechanism of this underconfidence effect, the researchers have proposed different theoretical models to investigate this effect in depth, but has yet to form a comprehensive system to explain the effect of the theoretical model. The purpose of this study is to reveal the mechanism of the underconfidence-with-practice effect, to further explore the theoretical model which can explain the effect. At the same time researching the properties of materials how to effect the underconfidence-with-practice effect, providing in-depth study of reference for the effect.This study attempted to verify the previous theoretical model, expansion and proposed a comprehensive system theoretical model to explain the underconfidence-with-practice effect. There are two experiments in this study. In experiment one, we discussed the role of memory for past test in the underconfidence-with-practice effect and analysised the role of memory for past test in the second judgement in the two types of judgement. The results show that in the immediate judgement all the items appear the underconfidence-with-practice effect. Memory for past test can better predict the second judgement of learning. Unrecalled items on Test 1 that were subsequently remembered on test 2 contributed disproportionately to the UWP effect. In the delayed judgement, all the items don't appear the UWP effect, only in the first judgement of learning, which show that over-confidence. The second recalled results can be better forecast the second judgement of learning. The second judgement of learning for items that were forgotten on Test 1 but remembered on test 2 isn't different with items that were remembered on both trails. The Results verify the MPT hypothesis, but there are not the UWP effect in both the related and unrelated pairs. Based on the study of experiment 1, The experiment 2 further research the properties of materials how to effect the UWP effect in the immediate judgement and the availability of the MPT hypothesis to explain the UWP effect in the different attributes of experimental material. The results show that in the immediate judgement the UWP effect in both the related and unrelated pairs produced a separation, only in the unrelated pairs. Further analysis found that in two types of pairs, the previous test results have the largest contribution to the second judgement of learning. Due to the impact of material properties, in the related pairs, overestimation of items that were unrecalled on both trails in the second judgement of learning can be offset underestimation of items that were unrecalled on Test 1 but remembered on test 2, so not appear the UWP effect. But in the unrelated pairs, overestimation can't be offset, still show the phenomenon of under-confidence. And to investigate the effect of a mechanism, at the same time pointed out the possible future research directions.