Effects Of L2 Working Memory And Language Level Factors On Anaphora Resolution Of Senior High School Students

In the literature,there is no unified conclusion about the factors affecting anaphora resolution.Some scholars think that anaphora resolution is related to antecedents(caramazza et al.,1977;Nicol et al.,1989;Zhao Dongmei,Liu Zhiya,2006;Li Rong,2016;Wu Mingjun,Yang Yulan,Wu Xiaoming,2018;Wang Wenbin,Tao Yan,2019),some scholars think that anaphora resolution is related to language level(Pretorius,2005;Kim et al.2015;Yeon hee Choi,Herim Ahn and Jinyong Lee,2018),and the other scholars believe that anaphora resolution is related to working memory(Mohammad,2019;Wang Hui;2015).Therefore,the author takes 50 students in a class of Grade one of a key senior high school in Yan'an City as the research subjects.Through the language proficiency test,online working memory and online anaphora resolution test,the data obtained are analyzed by One-Way ANOVA,two–factor variance analysis,t-test and multiple regression analysis of SPSS 21.In this way,the author attempts to explore the following problems: 1)Does the language level affect working memory? 2)Can working memory and language proficiency predict anaphora resolution in senior high school students? 3)Does L2 working memory and language level affect anaphora resolution of senior high school students?The results are as follows:(1)Taking the total score of working memory as the dependent variable and the high,medium and low language level of the subjects as the independent variable,the analysis of one-way ANOVA was carried out.The difference between the groups was not significant with P = 0.293 > 0.05,indicating that working memory could not distinguish the language level significantly.(2)With language level and working memory as independent variables and anaphora processing and anaphora understanding as dependent variables,multiple regression analysis was conducted respectively.Working memory can predict anaphora processing with Beta =-0.298 and P = 0.03 < 0.05 and anaphora understanding with Beta = 0.956 and P = 0.00 <0.05;and working memory processing with Beta = 0.554 and P = 0.00 and storage with Beta =0.634 and P = 0.00 can predict senior high school students' anaphora understanding;In anaphora processing,the working memory processing Beta =-0.093 and P = 0.52 can notpredict anaphora processing,while the working memory storage process Beta =-0.290 and P = 0.04 can predict anaphora processing.(3)There was no significant difference between the high and low language level subjects in the online anaphora resolution test.The p value of independent sample t-test was 0.295 and P = 0.384 respectively,which were greater than 0.05.(4)Using anaphora processing and anaphora resolution as dependent variables,working memory and language level as independent variables,two-way ANOVA were conducted.In anaphora processing,the main effect of working memory in anaphora processing was significant with F(1,49)= 5.133 and P = 0.039 < 0.05,the main effect of language level was not significant with F(1,49)= 3.033 and P = 0.111 > 0.05,the interaction effect of the two was not significant with F(2,49)= 131.170 and P = 0.067 > 0.05;In anaphora understanding,the main effect of working memory was significant with F(1,49)= 16.833 and P = 0.003 <0.05,the main effect of language level was not significant with F(1,49)= 1.231 and P =0.444 > 0.05,the interaction between the two was not significant with F(2,49)= 0.705 and P= 0.721 > 0.05.Research shows:(1)Working memory can not distinguish language level significantly,which is inconsistent with the research conclusions of Xu Luna,Chang Xin and Wang Pei(2017),but consistent with the research conclusions of Hae et al.(2020).(2)Working memory can predict anaphora processing and understanding,in which both processing and storage of working memory can predict anaphora understanding,while storage of working memory can predict anaphora processing,and processing of working memory can not predict anaphora processing,which is inconsistent with the conclusion of Mohammad(2019).(3)There was no significant difference between high and low language level subjects in the online anaphora resolution test,and the influence of language level on anaphora resolution was not significant,which was consistent with Lozano(2017)and Mohammad(2019),but was not consistent with Pretorius(2005),kimet al.(2015)and Yeon et al.(2018).And working memory has a significant effect on anaphora resolution,which is consistent with the research conclusions of Wang Hui(2015)and Mohammad(2019).(4)The above findings can be well explained by the two-stage serial model((Frazier,1995)and multicomponent model(Baddeley,2000).Therefore,we believe that high school teaching should strive to achieve the following things:(1)Try to understand the differences of students' working memory and teach students in accordance with their aptitude.(2)In order to improve students' understanding and mastery of anaphora,targeted training should be carried out for anaphora resolution.