Studies On L2 Discourse Testing Effect Of Chinese English Learners

In recent years,a large number of researches suggested that testing is an effective method to promote learning performance and that the testing effect has become a hot topic in the field of cognition and education.Studies have proved that the testing effect tends to appear in a delayed final test rather than an immediate final test.At the same time,the testing effect is prone to show up in simple materials with low element interactivity.However,whether the testing effect exists in complex materials with high element interactivity is still controversial.On the other hand,there were a few researches to investigate the influence of individual differences on the testing effect,such as,language proficiency and working memory capacity but without a unified conclusion.L2 discourse is common and complex in teaching and learning process,but testing effect in L2 discourse is not well explored.Therefore,this study explores L2 discourse testing of Chinese English learners and how language proficiency(LP)and working memory capacity(WMC)affect the outcomes based on cognitive load theory.The research questions are as follows:1.Is there an immediate/delayed testing effect in L2 discourse comprehension?2.How does LP affect the immediate/delayed testing effect in L2 discourse comprehension?3.How does WMC affect the immediate/delayed testing effect in L2 discourse comprehension?In this study,134 students of non-English major in a university participated in the main experiment,and eight students with similar language proficiency took part in the pilot experiment.The experimental material was a L2 complex discourse with higher element interactivity.The participants were randomly divided into four groups:immediate restudy(IR)group,immediate testing(IT)group,delayed restudy(DR)group and delayed testing(DT)group.One week before the main experiment,the LP and WMC of the participants were measured to ensure that there was no significant difference among the four groups.In phase one of the main experiment,all participants studied the material for the first time;in phase two,the participants in IR and DR groups studied the material for the second time while the participants in IT and DT groups recalled the material freely for a test;in phase three,participants in IR and IT groups carried out the final test immediately after phase two,whereas participants in DR and DT groups took part in the final test one week later.In all procedures of the experiment,the four groups were given the same amount of time.After the whole experiment process,there were two participants selected randomly in each group for the post-interview.The major findings were shown as follows:1.Performance in the restudying and testing groups had no significant difference in immediate/delayed L2 discourse comprehension.This was mainly due to the high element interactivity of the complex material and participants with low LP which resulted in an unsatisfactory understanding of the text in the initial study.Therefore,the initial testing in the second phase couldn’t promote the understanding of the L2 discourse.To conclude,there was no immediate/delayed testing effect in the comprehension of factual and inferential questions.2.In IR,IT and DT groups,performance in factual questions(FQ)was significantly positively correlated with LP.This was mainly because intrinsic cognitive load decreased while the germane cognitive load increased due to the higher LP of participants.They were likely to recall the content of the text in the final testing immediately after restudy/testing,and the higher the LP,the stronger the participants’ability to remember the information.In the DT group,participants were asked to recall the discourse freely depending on the memory of the content,which promoted the retrieval in the delayed final testing.On the contrary,participants in the DR group made a relatively superficial understanding of the passage,which could not guarantee the long-term retention of the memory.On the other hand,there was no correlation between the LP and the comprehension of inferential questions(IQ).This was mainly because the knowledge transfer ability was a kind of high-order cognitive ability,but the participant’s overall LP is low,neither restudying nor testing the passage could help them analyze the discourse in depth.3.In the IT group,WMC was significantly positively correlated with the performance in the comprehension of FQ and IQ.This is mainly because participants with high WMC in the IT group could understand the text very well at the initial study,reducing the extraneous cognitive load during the free-recall test.Therefore,the information could be retrieved and stored in memory to a greater extent and participants could apply related information in the final testing.However,this kind of memory retention couldn’t last a long time because the information processing is not profound.In contrast,the lack of resources was not so serious in restudy groups that WMC didn’t play a role.In this study,it is found that the testing did not significantly promote the memory retention and knowledge transfer of L2 discourse in both immediate and delayed final tests indicating that the testing effect was affected by the material complexity.At the same time,whether the testing effect appears was also influenced by learners’LP and WMC.Therefore,teachers should select appropriate teaching methods according to the material complexity,learners’LP and WMC in order to optimize the L2 teaching effect.