| Working memory (WM), an important cognitive construct, has been investigated extensively in the field of cognitive psychology and second language acquisition (SLA). Generally speaking, learners with high WM always perform better than learners with low WM. It is hypothesized that the constraint of WM on language acquisition can be mediated by different types of feedback. However, only a few studies have attempted to investigate the relationship between working memory capacity and corrective feedback. Three factors are included in the present study: working memory capacity, corrective feedback and time. The intention of the study is to explore the relative efficacy of working memory capacity and corrective feedback on the acquisition of the past tense forms. The research questions are listed as follows:1. Is there a statistically significant difference between two different working memory capacity groups (high working memory capacity versus low working memory capacity) in regards to students' repeated measures of the three test-scores in the acquisition of past tense forms (receptive knowledge and productive knowledge)?2. Are there statistically significant differences among the corrective feedback types (recast versus metalinguistic correction versus control) in regards to students' repeated measures of the three test-scores in the acquisition of past tense forms (receptive knowledge and productive knowledge)?3. Is there an interaction of the working memory capacity (high working memory capacity versus low working memory capacity) and the corrective feedback types (recast versus metalinguistic correction versus control), in regards to students' repeated measures of the three test-scores in the acquisition of past tense forms (receptive knowledge and productive knowledge)?4. Are there significant interactions among students' repeated measures of the three test-scores in the acquisition of past tense forms (receptive knowledge and productive knowledge), comparing the working memory capacity (high working memory capacity versus low working memory capacity) and the corrective feedback types (recast versus metalinguistic correction versus control)?The study adopts a 2 (working memory capacity) x 3 (corrective feedback types) x 3 (time) factorial design with two between group factors (working memory capacity and feedback type) and one within group factor (time). Both receptive and productive knowledge of past tense are considered as dependent variable, with the measurements of untimed Grammatical Judgment Test (GJT) and Oral Imitation Test (OIT).120 Chinese students as EFL learners from 4 intact classes from Nanjing Tech University participated in the present study. They were first been divided into two working memory capacity groups (high working memory group and low working memory group) according to the scores of Operation Span Task (OSPAN), and then they were randomly assigned into recast, metalinguistic correction and control (no feedback) groups. All the subjects took part in a picture description task, their errors on past tense were treated with different corrective feedback types which were provided by two native speakers according to their group assignment.The repeated measure mixed MANOVA and further mixed ANOVA were employed and the major findings were shown as the followings:1. There is a statistically significant difference between two working memory capacity groups in regards to students'repeated measures of the three test-scores in terms of the acquisition of past tense forms (for receptive knowledge:F=30.05, p =.000<.05, ?2=.209; for productive knowledge:F= 12.55, p=.001<.05, ?2 =.099). Subjects with high working memory capacity always perform better than those who with low working memory capacity.2. There are statistically significant differences among corrective feedback in regards to students' repeated measures of the three test-scores in the acquisition of past tense forms in terms of receptive knowledge (F= 4.744, p=.000<.05, ?2 =.077). The results of post hoc test have shown that, in the immediate posttest, metalinguistic correction is more effective than recast, and in the delayed posttest, recast is more effective than metalinguisitc correction. However, the differences are not statistically significant in terms of productive knowledge (F= 1.99, p=.141>.05, ?2?.034).3. Working memory capacity and corrective feedback has a significant interaction only in immediate posttest for both of the two tests. In terms of receptive knowledge (F= 4.614, p=.012<.05, ?2=.075), for subjects with high working memory capaicty, recast is more effective than metalinguistic correction, for subjects with low working memory capacity, no significant interaction is found; and in terms of productive knowledge (F= 4.135,p=.018<.05, ?2=.052), for subjects with high working memory capaicty, metalinguistic correction is more effective than recast, for subjects with low working memory capacity, no significant interaction is found.4. Working memory capacity and corrective feedback have a significant interaction among students' repeated measures of the three test-scores in the acquisition of past tense forms in terms of receptive knowledge (F=.058, p=.009 <.05, ?2=.058), that is to say, with the influence of time, significant interaction is found in immediate posttest instead of delayed posttest; however, no interaction is found in terms of productive knowledge (p=.133>.05), although significant interaction is found in immediate posttest, but the effect of time is not strong enough to make the interaction significant among all the three independent factors.The present study has significant theoretical and pedagogical implications. On one hand, teachers should choose the feedback type according to the working memory capacity of EFL learners; on the other, enough attention should be paid to help students reviewing and practicing certain linguistic structure, since time is one of the factors that will affect the effectiveness of grammar acquisition. |
PDF Full Text Request