Assessing Chinese Students’ Writing Performance in an American University: The Relationship between Selected Written Errors, Teacher’s Feedback, and Learners’ Interlanguage Experiences

The study examined Chinese students’ writing performance through the lens of corrective feedback and learners’ interlanguage experiences. It concludes that coding on paper may work only on learners who pay much attention to teachers’ feedback. It is always the work of both students and teachers to improve the accuracy in English writing.


Introduction
The effectiveness of error correction on ESL writing has been a topic of controversy in the past decades.
Preceding 1996 studies on teachers' feedback assumed that corrective feedback might benefit English language learners in improving the accuracy of writing. Although some studies like Robb et al. (1986) claims that teachers' corrective feedback has no significant effectiveness on 134 Japanese students' writing, no face-challenges are made as a review article published on the major language learning journal. Truscott (1996) publishes the review article "The case against grammar correction in L2 writing classes" to allege that grammar correction in L2 writing should be abolished because it is harmful and ineffective, the idea of which triggers a long-term debate on the effectiveness of error correction on L2 writing. In fact, Ferris (1999Ferris ( , 2004 and Truscott (1996)  experimental studies to make up this deficiency by focusing on different groups of English language learners and by emphasizing the functions of different feedback types in L2 writing. However, limited studies investigates students' feedback to teachers' correction and students' explanations on the reasons why they have these errors and how their interlanguage backgrounds interact with the reasons. My study attending to these needs covers both the effectiveness of a writing teacher's corrective feedback to English language learners' use of article and students' thoughts on their performance.

The Effectiveness of Corrective Feedback on Grammar
After Truscott (1996Truscott ( , 1999 repeatedly claims the futileness and harms of corrective feedback, many studies burst to examine the effectiveness of corrective feedback on L2 writing; meanwhile, the opposite sides sustains their arguments about the non-effectiveness of corrective feedback as well. These features of studies are tabulated as follow:  Bitchener's (2008) contributes to the argument that groups with direct feedback outperformed the group without feedback.
In the meantime, some studies have claimed that corrective feedback is not effective. Kepner (1991) examines 60 intermediate-level Spanish learners and comes out that direct error correction does not affect the effectiveness of students' output in writing with the limitations that no pre-test measurement is provided and that researcher does not control over the length of journal entry. In addition, Polio et al. (1998) maintain that differences of post-test scores for experimental and control groups was not notably, but the different instruments in post-test eliminate the influence on the findings. Otherwise, Truscott and Hsu (2008) tests the differences between the writing results of group with underlined error and group with no marks with the result of no meaningful differences.
While the existence of many studies demonstrates the non-effectiveness of corrective feedback, they also have significant limitations which may change the results of the experiments. Therefore, my assent to these studies is given with a hook at the end.

Studies Analyzing Different Types of Corrective Feedback on L2 Writing
Besides the discussion on the effectiveness of corrective feedback on L2 writing, a number of studies have expended to the field of examining the effective of different types and strategies of written corrective feedback. Bitchener et al. (2005) define both "direct or explicit feedback" and "indirect strategies" in their study "The Effect of different types of corrective feedback on ESL student writing" as follow: direct feedback refers to teachers' detection of an error and gives corrections beside the wrong form, whereas indirect feedback is demonstrated by the situations when teachers indicates that an error without providing a correction, therefore needing the student to "diagnose and correct" it.
Based on my findings, the feedback can be subsumed under two broad categories: Correcting wrong forms and structures; Writing "meta-linguistic explanations" (Bitchener, 2008, p. 105).

Indirect feedback:
Coding; Recording numbers of errors in front of a line; Circling or underscoring the errors The categories of feedback are expanded to another dimension of teacher-student interaction in addition to the dimension of direct-indirect feedback. Considering the teacher-student interaction, when the situation [teacher +, student +, essay +] appears, feedback strategies include teacher-student individual conference. When it is [student +, essay +], peer review and individual error logs can be utilized to give feedback. In the situation [teacher +, student-, essay +], direct corrective feedback contains crossing unnecessary part of speech, the insertion of part of speech, correction of wrong forms and sentence structures, and "meta-linguistic explanations" (Bitchener, 2008, p. 105 A student conducted by Chandler (2003) demonstrates both advantages and disadvantages of indirect and direct feedback. According to Chandler (2003), the pre-eminence of correction "may be due to the fact that students are able to correct significantly more of their errors on their revisions after this method than after teacher responses either describing the type or nothing the location of errors made, or both" (p. 291). Ferris (2004) also argues in "The 'Grammar Correction' debate in L2 writing: Where are we, and where do we go from here? (and what do we do in the meantime…?)" that explicit feedback like correction prevents adult learners from fossilization and pause to make progress in improving language accuracy. On the other hand, the participant in Chandler's (2003) study whose claim is used as a conclusion on the majority of participants credits the indirect corrective feedback "underline with description" by stating that this feedback helps her/him a lot. It seems that direct correctness may save students' time on correcting and may straight display learners' errors, but considering the further influence on learners' language learning, indirect corrective feedback weighs more.
The flip side of some types of indirect feedback also eliminates their benefits. Chandler (2003) claims that "marginal description of type of error had the most negative effect on accuracy of subsequent writing of any of the feedback methods used in either study, perhaps because of student attitudes toward it" (p. 292) Because revising errors relying on marginal description is cognitively demanding, students may lack of patience to figure out their mistakes by themselves. It also takes more time to check with teachers for the correct answers.

Studies on "Coding", one of indirect corrective feedback
Some studies (Table 3) have discussed the indirect feedback "coding" on students' essays. Neither Ferris and Roberts (2001) nor Robb et al. (1986) spotted many distinctions between coded and uncoded options. Robb et al. (1986) claims that the more direct methods of corrective feedback raters use, the less students pay attention to these errors. Although Robb et al.'s (1986) study results in the non-effectiveness of corrective feedback, their analyses on the improvement of accuracy still contribute to the discussions on direct and indirect feedback. Similarly, in Ferris and Roberts's (2001) study, there are no significant differences between the "coding" and "uncoding" groups.  Selinker (1972) brought up the definition of "interlanguage" to refer to the stage that a learner's language is moving to the target language. In terms of the special situation of interlanguage, influence from L1 and L2 functions in a learner's interlanguage system which can be recognized as the third language besides L1 and L2. Since interlanguage is governed the learner's internal reorganization of both L1 and L2 grammar, the learner's performance is a significant approach to test the mechanism of interlanguage. In addition, Corder (1974)

Participants
There are two participants named L and D, all of who are college students from an American university.

Settings
The goals of E120 English Composition emphasize the process of writing in which students practice and improve their organization, development, and style in essay writing. The class met for 50 min three times a week over 15 weeks. During these 37 hours and a half of class time, knowledge about essay construction, organization, and MLA citation as well as grammar and punctuation are taught and practiced, both to improve international students' understanding of and application to writing rules in English.

a. Participants' essays
Writing samples are collected from participants writing assignments from E120 English Composition where students are required to write 5 papers based the whole semester. Because of my limited time of research, the first four essays are collected from three participants to examine their errors. The instructor of English Composition handed out a list of correction symbols. After students turn in their essays, the teacher codes on their essays by applying the correction symbols. Students are required to revise their essays based on the teacher's corrective feedback before the submission of final copies.
Students are welcomed to ask any question and make an appointment with the teacher to talk about their essays before final submission. The teacher scores students' essays on a scale of 1 to 50, and students receive their essays with attached score sheets: Topic: _________________ Participants are required to rank each item on the Likert scale: from 1 "like very much" to 5 "dislike very much". The Part III is designed to obtain participants' individual language learning strategies in both oral and written. Many of the questions depict supposed situations, and participants are asked to relate their resolution to these vignettes.
The questionnaire is sent separately to participants via email, and they are asked each question orally via mobile phone with tape-recording the conversation. The researcher transcribed all the radio tapes during data collection. The questionnaire is modified during the oral question-answer sessions if necessary. It helps to refine questions and targets the subjects the study examines.

d. Error Explanation Interview
The interview is used to explore participants' explanation of their errors in writing. The interview is to probe the participants' understanding of the reasons that they make mistakes. It helps participants to reflect on their errors and internalize the rules. for time for English, she told me that she was exposed to English except that 4 hours in Chinese. She is taking 5 courses this semester, and one is Introduction of Literature, which is given a devil course by L.L. She spent 5-8 hours on reading in week days, and she told me she sometime did not understand what her literature professor was talking about.

Results and Discussion
She was diligent to learn English because she had a vocabulary pamphlet with her. She always excerpted some interesting sentences from English magazines or novels she read. She told me her thoughts on English grammar which is interesting, "I think English grammar is easy to learn but hard to use. When I write a composition, I find that grammar rules are not isolated from each other. English grammar is a mechanism which works in cooperation with each other. It turns out to be that the grammar rules I learnt from class work out only 60% on their original definitions". Her best harvest from the E120 was the knowledge about English punctuation. She had never learnt English punctuation in English classes in China, and she felt that the system of punctuation assisted her in understanding the system of English grammar.
Our topics covered her favorite language learning activities which were translation between Chinese and English, Role Play, small group work. In fact, she ranked all of these activities 1 on the Likert scale.
She also praised the coding system which was applied in E120 due to her first time to know about the She had bunches of language learning strategies shared when I asked hypotheses in Part III. From her answers, it was no doubt that she was the positive and active language learner who had a low affective filter. She would like to ask any native speakers any question to check if she understood his or her meanings. She said that the only reason that she gave up to probe to the bottom of the business was that she did not want to disappoint her American friends. She did not fear for laughing when she used inappropriate vocabulary to express her meanings.

b. Errors in essays
In the error analysis on essays of L, there are total 11 errors in the use of article spotted. The error rate is 8.2% total. The error rates in each essay are 6.9%, 3.6%, 5.9%, and 5.3%, from which there are no significant changes in the use of article. In terms of the data, the participant has problems in the use of "the", and she appears to substitute definite article for indefinite article. For example, in essay 2, she wrote "Sometimes, the person who was supposed to be my friend was not the good friend indeed" (not a god friend indeed); however, in essay 4, the similar error is made by her again: "I chose the English major in undergraduate degree" (an English major). It seems for her that the coding system does not work well for her correction on the error in the use of article. Because her misuses of article maintained a little regular pattern, it might be better to examine her explanation of her errors.
c. Students' Explanation of Errors L attributed the misuse of "the" to her understanding of the rules of article. She thought that definite article was used in front of the noun which emphasized the unique and also was used in the situation which was mentioned before. In the sentence "I chose the English major in undergraduate degree", she interpreted her understanding, "I thought there was only one English major at Saint Mary's University, so everyone should know it. When I mentioned English major, it is the one in our school". She looked not to be interested in the errors which were made by careless, and she would not want to discuss these errors. She understood the grammar rules of articles, but she did not know how to use them in essays because there was not article in Chinese. This find corresponds to those reported in Master (1997)'s argument on English article system. Master (1997) argues that English language learners from the language system without article are hard to acquire English article system. From L's explanation, it is improved under this circumstance.

Participant: D
a. Interlanguage Background The average time D exposed to his first language is about 4 hours during which he read Chinese news, facebooked, and chatted with Chinese students. The average time D exposed to English is about 6 In order to communicate fluently with native speakers, D spent at least 2 hours on learning English every day. The learning composed of completing the homework for ELB program (reading and writing assignments) and self-studying (enlarge vocabulary and watching newspaper). For most times, D found his vocabulary was enough for him to fluently communicate with classmates and teachers exception some situations. One typical situation was when the topic was too deep that involved too many terminologies. D recorded all the new words from notebooks and often took a look to refresh his memory. He would like to try out all the new words, but sometimes it depended on the nature of the vocabulary was whether a terminology that seldom appeared or a common words that needed to be used very often. D tried to use the new words he learned in his writing, and he thought he used them in a right way because he checked the context of when to use those new words and recreate a similar context for applying those words.
When we talked the learning strategies, D said he did not on purpose to find all the possible ways to learn a language, but just memorized whatever new to him. He always kept curious on new things, and he would like to probe the world. D did not attach too much importance on his mistakes in English compositions because he knew that he was a foreigner. It was for sure to make mistakes for a foreigner.
He never tracked his mistakes, nor did his professor. When he did not understand the corrections from professors, he would turn to his professor for help, and next time, he tried to practice the corrected forms.
b. Errors in essays D had total 15 errors in the use of article, and the total error rate was 6.25% which was a relative low percentage compared with L.L. His error rates in the use of article fluctuated from essay 1 to 4. In essay 1 and 3, the error rates were as high as 7.5% on average. On the contrary, in essay 2 and 4, the rates fell to 4.85%. The unsteady performance deserved to discuss in his explanation of errors. Otherwise, his misuse of definite article "the" took a great part in his errors. 12 out of 15 errors were related to the use of "the". The most significant evidence appeared in the data of the participant was that the change of using "the" from first two essays to the last two essays. As for the errors in the first two essays, "the" was deleted by rater, whereas in the last two essays, "the" was added in his sentences. It may indicate that the coding system may work well in D's case. It displayed a few evidence showing that coding system worked in help with D.X.'s correctness on uses of article; however, it barely aroused D's awareness to his errors in use of "the". It might suggest that the teacher should notice the changes of students and keep them in mind for the future revision to their feedback-giving types.
c. Students' Explanation of Errors D ascribed his errors to the confused understanding of the rules. He thought he might misunderstand these rules. When he chose wrong article to use, he thought it was his intuition to choose these words.
He told me that, "when I am not sure whether or not I should add 'the' in front of a noun, I usual add one". He argued that Chinese influenced him a lot, and he always translated sentences from Chinese to English. He also acclaimed the provision of correctness because he believed that answers from teachers were always right, and he should memorize these answers for future use.

Conclusion
From evidence the researcher collected, there are two conclusions summarized. From D's case, coding may work to some students, but it is not enough to have only coding on students' paper. It echoed to Ferris (2004), "teachers must prepare themselves to effectively treat students' written errors, […] once teachers themselves are prepared, the effective treatment of students' written error must include a variety of carefully integrated components" (p. 59). D and the like had high attention on their errors, but they lack of teachers' diversified feedback. For example, one-on-one conference on discussing the revision of essays may increase students' opportunity to solve their grammar problems. Therefore, teachers should do more preparations on students' performance.
Based on L's case, it concludes that students' attentions to their errors are also important. Chandler (2003) points out "that if students did not revise their writing based on feedback about errors, having teachers mark errors was equivalent to giving no error feedback" (p. 280). If students do not response to a teacher's corrective feedback, they cannot improve the accuracy in English writing. It is always the work of both students and teachers.