Can Indirect, Delayed Error Correction Improve Students’ Willingness to Speak in the Target Language?

At a high school where I have worked as a visiting lecturer, students of English as a Foreign Language (EFL) tend to use their mother tongue (L1) instead of the target language (L2) in communicative activities. Many factors are to blame for this issue, such as the seating arrangement, the language curriculum, the infl uences from the interlocutors, and the teacher correction methods (Pham 2005). This study aims to propose a strategy to correct student errors during communicative activities. Data analysis shows that the strategy of error correction that a teacher uses will have a substantial impact on student use of L2 in oral activities. Drawing from such fi ndings, I suggest some practical teaching strategies to maximize L2 and minimize L1 in speaking activities.


Introduction
To strengthen students' ability to use English communicatively, the Communicative Language Teaching (CLT) approach is widely used in most of the English classes at a high school in the south of Vietnam. As a result, these students have more chances to be involved in oral activities than they would being taught through other teaching methods. However, one problem that typically occurs during speaking activities at this school is the students' failure to practice the target language. Instead, they end up chatting in their mother tongue. Several reasons for this reversion include the seating arrangement, the language curriculum, the infl uences from their interlocutors, and the teacher error-correction methods (Pham 2005). This study seeks to explore the eff ectiveness of an innovation in correcting student errors. The following research question guides this study: Can indirect, delayed error correction improve students' willingness to speak in the target language?

Defi nition of key terms
Errors refer to inaccurate forms found in student speech or writing and are considered inevitable in the process of learning. Errors are an integral part of the learning process and it is impossible to learn a language without fi rst committing errors (Dulay et al. 1982). There are two kinds of errors: performance errors are caused by a student's inattention and fatigue, while competence errors are the result of a student's lack of knowledge of language rules (Dulay et al. 1982). In this study, error is used to refer to any deviation from language areas (i.e. grammar, pronunciation and so forth) made in a student's oral activities.
Feedback is defi ned as one of the micro-environmental factors aff ecting the quality and rate of second language acquisition. Feedback refers to the reader's or listener's response to the learner's speech/writing (Dulay et al. 1982) or performance (Song et al. 2017). It is assumed that feedback can facilitate the learner's language learning and help reduce the number of errors as a result (Carroll et al. 1992).
Error correction is the recognition of the existence of errors (Lee 1997) and its ultimate goal is "to equip students with a range of strategies to help them become more independent editors and better writers" (Truscott 1996, 171). Error correction can be categorized into indirect explicit correction and direct explicit correction. Indirect explicit error correction refers to the provision of feedback on errors -that is, the teacher provides prompts about the location and/or the nature of errors by means of codes. Direct explicit error correction refers to the overt correction of errors -that is, the teacher provides the correction of students' errors (Lee 1997).

The nature of errors in Second Language Acquisition (SLA)
Many SLA researchers have pointed to the nature of errors in the language learning process. According to Doff (1988, 187), "language is a system of rules that learners have to acquire, and (…) 'trying out' language and making errors are a natural and unavoidable part of this process". Nunan (2001, 88) also believes that "errors were seen not as evidence of pathology on the part of the learners, but as a normal and healthy part of the learning process" and "errors made by learners were systematic, rather than random". Moreover, Littlewood (1998, 22) claims that "learners' errors need not be seen as signs of failure; on the contrary, they are the clearest evidence for the learners' developing systems".

Reasons for student errors
Contrastive analysis, namely the comparison between mother tongue (L1) and target language (L2) has pointed out the reasons for making errors. Nunan (2001, 89) states that "where L1 and L2 rules are in confl ict, errors are likely to occur which are the result of interference between two languages". Second, hypotheses formation should also be taken into consideration. Ellis (1999, 182) considered "simplifi cation" as one of the strategies used to form hypotheses. Inside "simplifi cation", there are two other strategies: "transfer", which involves the use of learners' L1 as a basis for forming hypotheses about the L2, and "overgeneralization", which involves extending existing L2 knowledge to a new interlanguage form. Larsen-Freeman and Long (1991, 247) maintain that "certain errors are caused by the learners' failure to observe the boundaries of rules, which are classifi ed as overgeneralization. Other errors are attributed to simplifi cation". Ellis (1999, 68) also states that "learning does not only involve the formation of correct hypotheses; it also involves interim hypotheses which are systematically amended until the correct hypotheses are arrived at". Brown (2000, 94-97), however, believes that "transfer is a general term describing the carryover of previous performance or knowledge to subsequent learning and overgeneralization is the incorrect application -negative transfer -of previously learned second language material to a second language context".

Some eff ects of correcting student errors
Some studies (Chandler 2003;Diab 2005;Guenette 2007) have pointed out the importance of feedback for student language development. Even though errors are unavoidable in the language learning process, teachers cannot leave students free to make errors. Ellis (1997, 15) suggests some reasons to explain why teachers should give more attention to errors. First, "it is useful for teachers to know what errors learners make". Second, "it is possible that making errors may actually help learners to learn when they self-correct the errors they made" (1997,15). Doff (1988, 187) maintains that "we need to correct some errors, to help students to learn the correct forms of the language. But it does not mean that we have to correct students all the time". However, teachers must exercise caution when correcting student errors. Ellis (1999, 65) observes that "the provision of negative feedback does not appear to lead to more accurate performance, at least not immediately". It can be inferred that if teachers in a language classroom do not know how to appropriately correct student errors, they may risk making students reluctant to use L2 in oral activities.

Students' feedback preference
In general, students want and expect teacher feedback on their errors (Ferris 1995) and almost every student said they wanted errors to be corrected by their teacher (Ferris and Roberts 2001). Enginarlar (1993) surveyed the attitudes of 47 college freshmen on two modes of feedback employed by two instructors. One was coded feedback and the other was coded content feedback. The survey showed that an overwhelming majority (98%) reported their support of feedback. Students considered eff ective feedback to comprise attention to grammatical errors, guidance on writing skills, and comments on content (Enginarlar 1993). Kasanga (2001, 7) demonstrated that, "feedback provided by the teacher is still highly valued by the student-writers because of his/her traditional role as 'evaluator'". Basing on studies, one could argue that most students want feedback from their teachers. However, because of their tendency to be shy and reluctant to request feedback, there is little evidence among the Vietnamese population to support this assertion. Therefore, the purpose of this research is to fi nd out whether Vietnamese high school students desire feedback and whether such feedback will boost their confi dence in using the L2 in oral activities.

The teaching context
The teaching context plays an important role in the success of innovation in the tertiary, secondary, primary (TESEP) environment. Holliday (1994) stated that TESEP institutions have a diff erent purpose socially from British, Australian and North American (BANA) institutions and therefore innovations in TESEP institutions using methodology from BANA countries will not work unless local classroom cultures are taken into account. As a result, some general features of the Vietnamese teaching context and classroom culture should be taken into consideration before innovation is conducted.

Teachers
At the high school where I have worked as a visiting lecturer, my colleagues and I are under greater pressure than those who teach in language centers or universities because we have no choice in what we teach in the classroom. We must follow what has been inscribed by the Ministry of Education and are not allowed to deviate. Our teaching ability is assessed by the students' results. If the results are excellent, we are ranked as an excellent teacher, and if the results are not satisfactory, we are ranked as a teacher in need of improvement. Furthermore, reading and grammar -not listening and speaking -are tested in the end-of-semester exam so our teaching staff focuses on helping students with grammar and reading exercises, leaving second-language speaking as an afterthought. As a result, some of my colleagues are reluctant to innovate. Generally speaking, due to the pressure, we try to teach quickly (which means immediate and direct correction of errors) but eff ectively. We do not have much freedom in teaching and innovation will almost invariably encounter many barriers.

Students
The students in my school, however, may be more accepting of innovation. Born in the era of technology, they typically fi nd it easy to adapt to changes. And like most students at other Vietnamese schools, my students prefer accuracy in their every action (Minh 2000); immediate and direct error correction can instill in them a sense of failure. Therefore, innovation in the methods that teachers apply to correct errors may receive their support.

Innovation and rationale
In my high school, one common problem in oral activity is the students' use of their mother tongue. It has been found that one of the factors that causes this problem is the teachers' reaction towards students' errors. Most of us correct errors immediately and we think that this is a good way to help students. However, when working in groups, students have the feeling of losing face if they are corrected by the teacher with their partners around. This will inhibit students from practicing the target language because of their fear of making errors and being corrected.
To support this point, I decided to conduct a small survey of 50 students in my school. Each student was invited to answer a short pre-innovation questionnaire about error correction methods and submit it to their English teacher in charge. The study participants were invited to complete a questionnaire that aims to collect information on their age, gender, frequency of making errors when speaking English, teacher's reaction toward errors, and preferences for how errors should be corrected.
The results of this survey show that 100% of the students in this class make errors when they speak English. However, when these students were asked about the way errors are corrected in the classroom, 90% said that immediate error correction done openly and in front of their peers makes them feel as if they have lost face. The other 10% of the students thought that immediate and direct correction is easier to remember and saves time. Student preferences for specifi c error-correction methods, on the other hand, vary and can be divided into three groups: Group A: Teachers should write errors on the board, then give hints for students to guess the correct form. Group B: Teachers should correct orally by giving corrections immediately. Group C: Teachers should show the names of students who make errors and then provide correction for the whole class.
The number of students in each group is shown in Figure 1. As can be seen from this fi gure, most students would prefer teachers to innovate and fi nd new ways to correct errors. Instead of correcting immediately and directly, teachers may note the errors that students make when they speak and delay the work of correcting until the end of the task, namely to the post-speaking stage. Additionally, as a classroom exercise it might be better if teachers anonymously correct errors made by individuals. By using this indirect and delayed method, all students can learn from the errors of others without individual students experiencing the stigma of being named. To sum up, innovation in teacher error correction is conducted to encourage students to enhance their use of the target language in oral activities.

1. School
The innovation was conducted in a local high school in the south of Vietnam. Like most other state-owned schools, student tuition is less than that of private schools and the class sizes can be quite high, typically around 50 students. The administrative structure in this school is like that of most other Vietnamese high schools, with a principal and two vice-principals who check all of the general aspects of teaching and learning to make sure everything is carried out properly. There are 12 diff erent departments under the control of the principal. The head of each department oversees all academic aff airs in his area, such as teaching schedules, teaching methodologies and fi nal assessments. However, the high school does not design the course book and the curriculum, both of which come from the Ministry of Education. The teachers have to follow what has been inscribed. English is taught as a compulsory subject and receives no special attention from the management board. An English fi nal exam, which focuses on grammar and reading and ignores listening and speaking, is given at the end of the semester.

Students
As this is a state-owned school, the average number of students in each class (approximately 50) is greater than that in private language centers; all classes comprise diff erent English-ability levels. While Vietnam is undoubtedly a homogeneous society, any given classroom consists of students from diverse familial backgrounds. Some students come from families that stress the importance of education while others come from families that pay little attention to their children's school performance. Some students are keen on learning English, as they believe that it is an important part of their lives. Others, however, think that learning English is less important than working part-time to help support their families. They have little interest in practicing skills at home or school.

Innovation and its features
6.1. The type of social change Markee (1997) posits that immanent change takes place when both recognition for the need of change and proposal of solution to a problem belong to the same social system. This is true for this innovation because the problem of using the mother tongue in oral activities is found in the researcher's classrooms. Moreover, the innovation takes place because of the researcher's insistence and not because of the management board's willingness. As a matter of fact, the researcher is the owner of the innovation and has the role of internal change agent. White (1988) holds the view that a problem-solving approach, whose aim is to make use of research in modifying and improving curriculum practice, is also at the basis of action research, thus having a direct relationship to innovation and reform. In fact, this innovation belongs to the problem-solving model because it takes the form of action research and is designed by the teacher with full ownership and responsibility. The model of social-interaction should also be taken into consideration. According to Markee (1997), the most important insight that this model off ers change agents is the claim that diff usion is nothing less than a form of communication. Should this innovation become successful, it will be spread to other classrooms in the same school by way of social communication and informal report and of course, will be adopted by more users who face the same problem. Through the social network, this innovation will be made popular to all the people concerned.

Roles of the stakeholders
Both White (1988) and Markee (1997) acknowledge that while instructors play key roles in language teaching innovations, other individuals -called the stakeholders -are also involved. There are three roles for the teacher who conducts the innovation. First, he is the adopter who makes the decision to change the way errors are corrected. Second, he holds the role of implementer. Third, he is the change agent with his own innovation management throughout the innovation process. The students are the clients who will receive the innovation; certain students are early adopters who help their peers with the adaptation to changes. The teacher receives help from his colleagues with observation and student involvement during the innovation. This innovation is fully controlled by the teacher-in-charge; therefore, it receives no resisters from the school's management board or students.

The participants
The innovation took place in a class of 50 students who were around 17 years old. The main English course book (Hoang et al. 2006), which is copyrighted by the Ministry of Education, covers four skills and all language areas for pre-intermediate learners. There are three periods for English per week, each lasts 45 minutes. It is the normal class with students of mixed English levels. The students' English results from the last semester are shown Figure 2. illustrates that the students in this class are in mixed levels, ranking from high distinction to pass. This class was chosen for innovation for two reasons. First, the researcher was also the teacher in this class, so it was convenient to carry out. Second, despite being a pre-intermediate level class, the students' English abilities varied. It is hoped that this can ensure generalized results so the success of this innovation can be applied to other classes with the same problem.

The phases of innovation
The innovation took place in six weeks. Table 1 represents the timeline for the innovation.

Table 1. Innovation timeline
Week What was carried out -collect the questionnaire and analyze data

Evaluation instruments
To collect the most reliable data, two main evaluation instruments were employed:

a. Observation
Observation is conducted with the help of three colleagues, as the researcher has to be in the role of implementer during the oral activities. The observers who are instructed how to use the observation form in the second week enter the class at the beginning of the lesson and sit at the back of the class during the pre-speaking stage. When students practice speaking, the observers walk around the class and get closer to the groups (but not too close as students may be uncomfortable with strangers) to listen to the language used in the activities and determine whether errors are made. The teacher, however, still works with the students to help with ideas or vocabulary and takes note of all the errors made. The observation sessions start in the third week rather than the fi rst for two reasons. First, students take some time to get familiar with the new error correction methods. Second, the initial result of the innovation requires at least two weeks to achieve. Observation is conducted in four sessions in four weeks, as the time for this innovation is limited. The data from the observation are then collected and ready for analysis.

b. Questionnaire
The use of a questionnaire is for assessing the overall eff ectiveness of this innovation and to check the reliability of observation. Therefore, students receive a post-innovation questionnaire only in the fi fth week and they have one week to complete it. The questionnaire is designed in the third week with closed and open questions to solicit opinions from students for their evaluation of this innovation. In this questionnaire, students refl ect their opinion on the new error correction method (positive or negative), their confi dence for speaking English (little or much), and whether they make the same error that was previously corrected.
Moreover, students are assured that the information they give will be confi dential and for research purposes only. Questionnaires are given to all the 50 students in the class. All questionnaires are returned to the researcher on time thanks to student cooperation.

Results and discussion
Most of the data from the observation and questionnaires are collected at the end of the sixth week and are displayed in the following table: Table 2 illustrates how each session produces diff erent results. In the four sessions, the number of groups and students per group remains static. This is to ensure that each student has the same chance to speak in each session. The percentage of English and Vietnamese heard in oral activities is calculated by taking the average from the three observers' results. The observation results shows that the percentage of Vietnamese heard through the four sessions decreases steadily while English increases, but still quite slowly. These results inspire hope for the success of the innovation. To this point, student expectations are to receive immediate error correction and the idea of self-correction through indirect and delayed teacher correction is foreign. As a consequence, the teacher not only has to give hints to his students to self-correct but also provide the proper forms of some common errors in speech when the students fail to do so on their own. It should be pointed out that the students' speed of self-correction is quite slow in the four sessions, most likely due to their lack of familiarity with the indirect correction method. However, the students' attitudes in oral activities and their low rate of making already-corrected errors are the rewards for the teacher's eff ort.

Observation
The students tended to be more engaged and interested in speaking English in later sessions with no fear of being discouraged. In summation, it can be seen from the table that the factors that decide on the success of the innovation, such as the percentage of English heard and the students' desire to participate in oral activities, show an upward trend. On the contrary, those factors that lower the success of this innovation, such as Vietnamese heard and the rate of making already-corrected errors, are in an opposite direction.  Table 3 illustrates many of the positive takeaways from the innovation conducted in this class. Eighty-four percent of the students are in favor of changes in error correction and have no feelings of discouragement with the new method. Students' overall confi dence in speaking English rises and the vast majority of them do not want to revert to the old method. Some other students, however, are still unfamiliar with this new way of correcting errors and have problems with speaking English in oral tasks; their switching to Vietnamese is an unavoidable outcome. These students expect knowledge to be provided by their teacher. The number of such students accounts for a minority of the class. It is hoped that with time and the assistance of early adopters who have received the benefi ts of the new method this problem can be overcome.

Evaluation
Based on the data collected through the observation and the questionnaire, the innovation can claim some success. Factors pointing to this conclusion include an increase in the percentage of English heard, the students' desire to participate in oral activities, and the students' preference toward the changes and interest in oral activities.

Strength
This innovation has been partially successful in a small-scale teaching environment. It produces positive results since there is a need for change from insiders and the innovation is conducted with a form of action research. Due to the variation in ability among the students in the class, it can be asserted that the innovation will work with students of diff erent levels and therefore guarantee the generalizability of the innovation's result. The innovation addresses a common problem presently found in Vietnamese language classrooms and can be applied in places that share the same problem.

Weakness
Because of the short length of time (six weeks) and a relatively small sample size (50 students), there is a possibility that the results achieved in this class may not be replicated at other schools or universities. As a result, this innovation should be considered carefully before it is applied in a larger-scale environment. Yet another unsolved problem found in the innovation is the limited ability of the students to self-correct, even after hints from the teacher. This may be explained by the students' learning habits but can be overcome with peer assistance or early adopters. This may suggest another topic to be researched.

Recommendation
Based on gains from the innovation, some implications for error correction should be pointed out. First, language teachers should provide adequate correction in the classroom, especially as it applies to speech. Not all errors should be corrected. Martínez (1999) thinks that excessive feedback on errors can have a negative eff ect on motivation. Teachers should correct serious errors and give suggestions for other errors, allowing students to self-correct. This applies to correcting students' pronunciation in speaking lessons. If all pronunciation were to be corrected, it is safe to assume that most students would be dissuaded from speaking English in public for fear of being corrected. Second is the need for students to self-correct their errors. The purpose of self-correction is, according to Karra (2006), to help students know how to identify an error in order to avoid it in the future. Instead of off ering immediate error corrections, teachers can ask students to think of possible correct forms by giving them suggestions. Also, teachers can get students to self-correct pronunciation by letting them listen to authentic materials by native speakers from a recording or online dictionary.
Finally but importantly, language teachers should act as a good model of language to minimize the chances of students' making errors. All SLA researchers know that input is converted into intake; learners make use of this material for dual purposes, namely, comprehension and acquisition. Therefore, if students are given an incorrect form of language, they will process and acquire that form, which may result in incorrect language use. As a result, language teachers must give students correct forms of language while teaching to reduce the burden of errors they have to correct.

Conclusion
It is expected that the innovation can be a tool for language teachers to overcome the problem of student use of L1 in classrooms by means of delayed and indirect error correction. The innovation, however, still has an unsolved problem. That is the speed of students' self-correction. A future study can investigate this topic in another teaching context.