The comparative effects of intensive versus extensive recasts through grammar-focused tasks on Iranian intermediate EFL learners’ grammatical accuracy in writing

محورهای موضوعی : Curriculum Design and Development

Somayeh Hosseini ¹ ( Department of English Language, Shahr-e-Qods Branch, Islamic Azad University, Tehran, Iran )
Mahdieh Shafipoor ² ( Department of English Language, Shahr-e-Qods Branch, Islamic Azad University, Tehran, Iran )
Morteza Teimourtash ³ ( Department of English Language, Science ad Research Branch, Islamic Azad University, Tehran, Iran )

تاریخ دریافت : 1402/07/28 تاریخ پذیرش : 1402/12/06 تاریخ انتشار : 1402/12/20

کلید واژه: : Grammar-Focused Tasks, Extensive Recast, Intensive Recast, Recast, EFL Writing,

چکیده مقاله :

Grammatical accuracy has been controversial in L2 writing, and the utility of different types of corrective feedback has been enigmatic. Therefore, the present study attempted to explicate the impact of intensive versus extensive recasts through grammar-focused tasks on Iranian EFL learners’ grammatical accuracy in writing. To do so, 60 Iranian male and female EFL learners were assigned into two experimental groups (i.e., intensive and extensive groups) and one control group. After taking the Oxford Placement Test (OPT), the experimental groups received intended corrective feedback, while the control group did not receive any feedback. Then, the data was analyzed deploying both descriptive and inferential statistics. The results revealed the outperformance of experimental groups compared to the control group in the writing post-test. Also, it was concluded that extensive recast can improve EFL learners’ writing performance much better than intensive recast. The findings have some fruitful pedagogical implications for both teachers and materials developers.

چکیده انگلیسی:

Grammatical accuracy has been controversial in L2 writing, and the utility of different types of corrective feedback has been enigmatic. Therefore, the present study attempted to explicate the impact of intensive versus extensive recasts through grammar-focused tasks on Iranian EFL learners’ grammatical accuracy in writing. To do so, 60 Iranian male and female EFL learners were assigned into two experimental groups (i.e., intensive and extensive groups) and one control group. After taking the Oxford Placement Test (OPT), the experimental groups received intended corrective feedback, while the control group did not receive any feedback. Then, the data was analyzed deploying both descriptive and inferential statistics. The results revealed the outperformance of experimental groups compared to the control group in the writing post-test. Also, it was concluded that extensive recast can improve EFL learners’ writing performance much better than intensive recast. The findings have some fruitful pedagogical implications for both teachers and materials developers.

منابع و مأخذ:

Arege, J. B. (2010). Exploring some effects of different types of error correction feedback on ESL student
Arnaudet, M. L., & Barrett, M. E. (1990). Paragraph development: A guide for students of English. Prentice Hall Regents.
Ashwell, T. (2000). Patterns of teacher response to student writing in a multiple-draft composition classroom: Is content feedback followed by form feedback the best method? Journal of Second Language Writing. 9(3), 227-258.
Batstone, R. (1994). Grammar. Oxford University Press.
Bitchener, J., Young, S., & Cameron, D. (2005). The effect of different types of corrective feedback on ESL student writing. Journal of Second Language Writing, 14(3), 191-205. https://doi.org/10.1016/j.jslw.2005.08.001
Carroll, S. E. & Swain, M. (1993). Explicit and implicit negative feedback: An empirical study of learning. Studies in Second Language Acquisition, 15(3), 357-386. https://doi.org/10.1017/s0272263100012158
Chandler, J. (2003). The efficacy of various kinds of error feedback for improvement in the accuracy and fluency of L2 student writing. Journal of Second Language Writing, 12(3), 267–296. https://doi.org/10.1016/s1060-3743(03)00038-9
Chastain, K. (1988). Developing second language skills: Theory and practice. Harcourt Brace Jovanovich, Inc.
Chun, A. E., Day, R. R., Chenoweth, A., & Luppescu, S. (1982). Errors, interaction, and correction: A study of native-nonnative conversations. TESOL Quarterly, 16(4), 537-547. https://doi.org/10.2307/3586471
Ellis, N. C. (2008). The dynamics of second language emergence: Cycles of language use, language change, and language acquisition. The Modern Language Journal, 92(2), 232-249. https://doi.org/10.1111/j.1540-4781.2008.00716.x
Ellis, R. (2001). Introduction: Investigating form‐focused instruction. Cambridge University Press.
Ellis, R., Loewen, S., & Erlam, R. (2006). Implicit and explicit corrective feedback and the acquisition of L2 grammar. Studies in Second Language Acquisition, 28(3), 339-368. https://doi.org/10.21832/9781847691767-015
Fahim, M., & Montazeri, M. (2013). The impact of metalinguistic corrective feedback on EFL learners’ levels of lexical resource and grammatical range and accuracy in their oral proficiency. International Research Journal of Applied and Basic Sciences, 4(7), 1776-1782.
Fathman, A. K., & Whalley, E. (1990). Teacher response to student writing: Focus on form versus content. In B. Kroll (Ed.), Second language writing: Research insights for the classroom (pp. 178-190). Cambridge University Press.
Ghoorchaei, B., Tavakoli, M., & Ansari, D. N. (2010). The impact of portfolio assessment on Iranian EFL students’ essay writing: A process-oriented approach. GEMA Online Journal of Language Studies, 10(3), 122-142. https://doi.org/10.47191/ijsshr/v4-i11-20
Greenslade, T., & Félix-Brasdefer, J. C. (2006). Error correction and learner perceptions in L2 Spanish writing. In C. A. Klee & T. L. Face (Eds.), 7th conference on the acquisition of Spanish and Portuguese as first and second languages (pp. 185-194). Cascadilla Proceedings Project.
Harmer, J. (2003). How to teach writing. Longman-Pearson Publications.
Hyland, K. (1998). Genre-based pedagogies: A social response to process. Journal of Second Language Writing, 12(1), 17-29. https://doi.org/10.5040/9781350037939.0034
Hyland, K. (2003). Second language writing. Cambridge University Press.
Jacobs, H. L., Zinkgraf, S. A., Wormuth, D. R., & Hartfield, V. F. (1981). Testing ESL composition: A practical approach. Newbury House.
Kamiya, N. (2015). The effectiveness of intensive and extensive recasts on L2 acquisition for implicit and explicit knowledge. Linguistics and Education, 29(1), 59-72. https://doi.org/10.1016/j.linged.2014.11.002
Kepner, C. G. (1991). An experiment in the relationship of types of written feedback to the development of second-language writing skills. The Modern Language Journal, 75(3), 305-313. https://doi.org/10.1111/j.1540-4781.1991.tb05359.x
Kim, J. H. (2004). Issues of corrective feedback in second language acquisition. Working Papers in TESOL & Applied Linguistics, 4(2), p.1-24. https://doi.org/10.1017/9781108589789.003
Kim. H., Mathes, G. (2001). Explicit and Implicit corrective Feedback. The Korea TESOL Journal, 4(1), 57-72.
Liu, Y. (2008). The effects of error feedback in second language writing. Journal of Second Language Acquisition and Teaching, 15(1), 65-79.
Long, M., Inagaki, S., & Ortega, L. (1998). The role of implicit negative feedback in SLA: Models and recasts in Japanese and Spanish. Modern Language Journal, 82(3), 357–371. https://doi.org/10.1111/j.1540-4781.1998.tb01213.x
Lyster, R. (2004). Differential effects of prompts and recasts in form-focused instruction. Studies in Second Language Acquisition, 26(3), 399-432.
Lyster, R., & Ranta, L. (1997). Corrective feedback and learner uptake: Negotiation of form in communicative classrooms. Studies in Second Language Acquisition, 22(1), 37-66. https://doi.org/10.1017/s0272263197001034
Nunan, D. (2003). Practical English language teaching. McGraw-Hill Publications.
Rezaei, S., Mozaffari, F., & Hatef, A. (2011). Corrective feedback in SLA: Classroom practice and future directions. International Journal of English Linguistics, 1(1), 21. https://doi.org/10.5539/ijel.v1n1p21
Richards, J. C., & Schmidt, R. W. (2002). Longman dictionary of language teaching and applied linguistics. Routledge.
Sarandi, H. (2015). Reexamining elicited imitation as a measure of implicit grammatical knowledge and beyond…? Language Testing, 32(4), 485-501. https://doi.org/10.1177/0265532214564504
Selinker, L. (1992). Rediscovering interlanguage. Longman Publication.
Semke, H. (1984). The effects of the red pen. Foreign Language Annals, 17(2), 195-202. https://doi.org/10.1111/j.1944-9720.1984.tb01727.x
Sheen, Y., Wright, D., & Moldawa, A. (2009). Differential effects of focused and unfocused written correction on the accurate use of grammatical forms by adult ESL learners. System, 37(4), 235-255. https://doi.org/10.1016/j.system.2009.09.002
Sheppard, K. (1992). Two feedback types: Do they make a difference? RELC Journal, 23(1), 103- 110.
Sukur, M., & Demerican, F. (2020). The effects of intensive and extensive recasts on learning the third person singular. Research on Youth and Language, 14(2), 58-75. https://doi.org/10.1111/modl.12387
Suzuki, W., Nassaji, H., & Sato, K. (2019). The effects of feedback explicitness and type of target structure on accuracy in revision and new pieces of writing. System, 81(1), 135-145. https://doi.org/10.1016/j.system.2018.12.017
Tavakoli, H. (2012). A dictionary of research methodology and statistics in applied linguistics. Rahnama Press.
Tedick, D. J., & de Gortari, B. (1998). Research on error correction and implications for classroom teaching. ACIE Newsletter, 1(3), 1-6.
Truscott, J. (1996). The case against grammar correction in L2 writing classes. Language Learning, 46(3), 327-369. https://doi.org/10.1111/j.1467-1770.1996.tb01238.x
Van-Ha, X., & Murray, J. (2023). Corrective feedback: Beliefs and practices of Vietnamese primary EFL teachers. Language Teaching Research, 27(1), 211-240. https://doi.org/10.1177/1362168820931897
Wang, Y. H., & Liao, H. C. (2008). The application of learning portfolio assessment for students in the technological and vocational education system. Asian EFL Journal, 10(2), 132-154.
Williams, C., & Polio, J. (2009). Teaching and testing writing. In M. H. Long & C. J. Doughty (Eds.), The handbook of language teaching (pp. 134-167). Wiley Online Library.
writing [Unpublished doctoral dissertation]. University of Chicago.
Wu, C. H. (2007). Spoken grammaticality and EFL teacher candidates: measuring the effects of an explicit grammar teaching method on the oral grammatical performance [Unpublished doctoral dissertation]. The Ohio State University.
Zemach, D. E., & Rumisek, L. A. (2005). Academic Writing: From Paragraph to Essay. Macmillan.
Zheng, Y., & YU, S. (2018). Student engagement with teacher written corrective feedback in EFL writing: A case study of Chinese lower-proficiency students. Assessing Writing, 37(1), 13-24. https://doi.org/10.1016/j.asw.

متن کامل:



Curriculum Research		Volume 4, Issue 4 March 10, 2024
The comparative effects of intensive versus extensive recasts through grammar-focused tasks on Iranian intermediate EFL learners’ grammatical accuracy in writing
Article info	Abstract
Article Type:	Grammatical accuracy has been controversial in L2 writing, and the utility of different types of corrective feedback has been enigmatic. Therefore, the present study attempted to explicate the impact of intensive versus extensive recasts through grammar-focused tasks on Iranian EFL learners’ grammatical accuracy in writing. To do so, 60 Iranian male and female EFL learners were assigned into two experimental groups (i.e., intensive and extensive groups) and one control group. After taking the Oxford Placement Test (OPT), the experimental groups received intended corrective feedback, while the control group did not receive any feedback. Then, the data were analyzed deploying both descriptive and inferential statistics. The results revealed the outperformance of experimental groups compared to the control group in the writing post-test. Also, it was concluded that extensive recast can improve EFL learners’ writing performance much better than intensive recast. The findings have some fruitful pedagogical implications for both teachers and materials developers.
Original Research
Authors:
Somayeh Hosseini1, Mahdieh Shafipoor2, Morteza Teimourtash3
Article History:
Received: 2023/10/20 Accepted: 2024/02/25 Published: 2024/03/10
	Keywords: Grammar-Focused Tasks, Extensive Recast, Intensive Recast, Recast, EFL Writing

 

---

[1]  . Department of English Language, Shahr-e-Qods Branch, Islamic Azad University, Tehran, Iran.

[2]  . Department of English Language, Shahr-e-Qods Branch, Islamic Azad University, Tehran, Iran (Corresponding Author). [e-mail] 

[3]  . Department of English Language, Shahr-e-Qods Branch, Islamic Azad University, Tehran, Iran.

1. Introduction

Writing is a fundamental communication skill (Chastain, 1988). It is a ubiquitous skill and a significant tool for learning and conveying messages (Harmer, 2003). Writing does not occur in a vacuum, and it is always embedded in a web of relationships among the components of writing (Williams & Polio, 2009).

It is generally agreed that the ability to write efficiently in a second language (L2) has become vital for many language learners worldwide (Ghoorchaei et al., 2010). The importance lies in the fact that EFL learners of English often need more linguistic means to convey their thoughts in written English (Nunan, 2003). 

        In second language writing (L2), the place of grammatical accuracy has been debated over the recent years. Before the advent of communicative language teaching, grammar was central to teaching methods (Ellis, 2008). Batstone (1994) affirmed the crucial role of grammar in understanding language since it offers a framework for learners to come up with their experience of learning a foreign language. In addition, grammar plays a vital role in language processing because it assists learners in getting involved in the surrounding world and building the order and structure of information (Ellis, 2008). 

From a host of techniques suggested for bolstering L2 writing, feedback has embarked on a good body of scholarly research (e.g., Ashwell, 2000; Liu, 2008) in second language acquisition. Feedback is commonly used in SLA research and second language pedagogy. Feedback deals with any information a teacher offers in response to a learner’s production (written or spoken) and is mostly used to address inaccuracy rather than accurate production (Tavakoli, 2012). In other words, feedback can be defined as “comments or other information that learners receive concerning their success on learning tasks or tests, either from the teacher or other persons” (Richards & Schmidt, 2002, p.199). As Lyster and Ranta (1997) put it, teachers use different kinds of corrective feedback when correcting learners’ errors, from which metalinguistic feedback includes any “comments, information, or questions related to the well-formedness of the students’ utterance, without explicitly providing the correct form” (p. 46).

        Intensive and extensive recasts have yet to be given due attention from different kinds of corrective feedback in L2 writing. Intensive recast is provided on a single, pre-selected linguistic structure, whereas extensive recast is provided through various linguistic forms (Ellis, 2001). From a theoretical point of view, intensive recasts can be more effective compared to extensive ones since language learners usually care more about the type of feedback that concerns a single error type instead of directing at a broad type of error (Suzuki et al., 2019). However, despite the fact that some research applying intensive recasts have revealed positive results, some evidence exists that extensive feedback is reasonably practical.

The results of the studies focusing on the role of corrective feedback in L2 learning are controversial, and an ample number of studies have been conducted in the field, which has a long history. One of the major dilemmas is that the majority of findings on corrective feedback have been conflicting (Hyland & Hyland, 2006). 

        Recently, a good deal of scholarly research has been performed on teacher feedback. Numerous studies have attempted to shed light on various types of corrective feedback (Ashwell, 2000; Bitchener et al., 2005; Liu, 2008). Furthermore, according to Ellis (2008, p. 106), “there is no corrective feedback recipe” to apply a specific method to all learners. Similarly, Tedic and Gortari (1998) recommend that teachers provide their learners with a bundle of feedback techniques since different techniques might attract different learners. This may be justified by considering the learners’ characteristics (e.g., age, needs, objectives, and proficiency level) which determine the right kind of feedback.  

        In addition, communicative and task-based language teaching has recently been given colossal attention. Within these approaches, there is a tendency toward eliminating correction on learners’ linguistic production. The proponents of these methods believe that correction puts too much pressure and hinders learners’ production (Fahim & Montazeri, 2013). Many language scholars have challenged this claim over the last decade (Hyland & Hyland, 2006). Given this, the ultimate role of corrective feedback has been the bone of contention among many scholars and further research in this realm seems inevitable. 

        Another critical issue is that the role of different types of recast through grammar-focused tasks has been given a leap service in L2 writing. Lyster and Ranta (1997, p. 46) defined recast as “the teacher's reformulation of all or part of a student's utterance, minus the error”. Extensive recast is provided on a wide range of linguistic forms (Ellis, 2001). According to Brown (2014), extensive recasts occur when the feedback is not limited to a single target structure and learners receive feedback on many structures that occur incidentally during the instruction. However, intensive recast is provided on a single, pre-selected linguistic structure (Nassaji, 2017).Moreover, some scholars believe extensive recast is highly time-consuming and ineffective since students are only sometimes ready to participate (Hyland & Hyland, 2006).

        Second/foreign language teachers and researchers who have worked on corrective feedback have long observed that corrective feedback is an essential strategy for developing writing skills, and providing it by language teachers assists learners in learning the appropriate linguistic forms. Consequently, finding out the best techniques for corrective feedback to enhance student’s writing performance has always been a primary concern for teachers. 

        It can be claimed that the “most highly valued and desired classroom activities” is the teacher feedback which is considered as the most appropriate error feedback (Kim & Mathes, 2001, p. 56). In addition, Selinker (1992) confirms that errors are considered an indispensable language-learning process and must be performed to assist learners in producing the L2 more accurately. So, students tend not only to receive feedback from their teachers but also a preference toward certain types of teacher feedback (Kim & Mathes, 2001). The goal of providing feedback is simple: To help the learners to notice the problem in their production and correct it after following the feedback. It should be mentioned that scholarly articles in this field reflect the impact of different kinds of feedback on learners’ accuracy in L2 writing. 

The study of different types of teacher feedback has been the focus of many scholarly studies for many years. This importance can be attributed to the fact that many scholars have attempted to deal with the link between teacher feedback and L2 learning (Rezaei et al., 2011). In the midst of all these scholarly attempts, the study of influential types of teacher feedback has received more attention. Early studies cast doubt on its ultimate utility and efficacy (e.g., Kim, 2004), but recent studies point out its fruitful results in the classroom context (Caroll & Swain, 1993; Long et al., 1998).

        As was mentioned before, in reviewing the literature, the findings of studies regarding the role of error correction, which is one type of teacher feedback, could be more varied and precise. Some studies, such as Truscott (1996) and Kepner (1991), found no significant impact on L2 writing, while some studies pointed to the difference it can make in students’ L2 writing (Chandler, 2003). Here, another look at teacher feedback will be taken, and the discussion on different types of teacher feedback will be expanded. 

        Chun et al. (1982) carried out a study on corrective feedback and claimed that teacher feedback rarely occurred in the language classroom, and in most cases, it was employed carelessly and was not noticed by the learners. In the same line, Sheppard (1992) applied two different kinds of feedback to an essay writing class. The learners consisted of two groups: Group A received coded error feedback, focusing on the error’s type and location. Group B received feedback on the content of their writing as well as the clarification requests, which was a written feedback offered in the margin of their papers. The results depicted no noticeable difference in the revised papers of the two groups’ writing. 

        Similarly, Semke (1984) carried out a study about narrative writing among 141 German students for ten weeks and claimed no significant difference in the case of providing feedback in classes. Truscott (1996) referred to this study and claimed that if error correction were fruitful, it would have significantly improved students’ narrative writing accuracy. Nevertheless, the gentle point neglected by Truscott and rightfully mentioned by Arege (2010) is that both studies did not have any control groups, and the amassed results can be attributed to this negligence.

        Bitchener et al. (2005), in a research on corrective feedback, dealt with three groups differentiating on the hours learners attended the class. The full-time class was provided with direct written corrective feedback accompanying teacher’s explicit correction. The 10-hour class received only direct written corrective feedback. In contrast, the four-hour class was provided with no corrective feedback but had feedback based on the quality of their writing. The results revealed that the group that was provided with direct feedback outperformed significantly in terms of accuracy. This study overcame the discrepancies of the previous studies by entailing a control group and having treatments with appropriate lengths. 

All the studies above suggested that error feedback, no matter which type was provided, had no or at least lesser effect on improving students’ L2 writing accuracy. The following studies, however, showed that teacher feedback can yield more fruitful results.

        Fathman and Whally (1990) conducted a study in which 72 ESL students at the Intermediate level took part in a writing class and were asked to write an essay using a picture sequence. The allocated time was 30 minutes. The participants in four groups were provided with four different types of treatment: (1) They received no feedback, (2) They received feedback on content, (3) The feedback was on grammar and content, and (4) The feedback included only grammar. Learners in group 3 had the privilege of receiving feedback through underlined errors and written comments. The findings indicated that all students improved in grammatical accuracy, either students who had received grammar feedback only or those who had received grammar and content feedback. Unfortunately, errors were not classified and were marked comprehensively. It must be clear whether improvements were made on any specific errors.

        Furthermore, Ashwell (2000) conducted a long-term study that lasted nearly one year. This study considered four types of feedback: (1) form-focused feedback after content feedback, (2) content feedback after form-focused feedback, (3) form and content feedback simultaneously, and (4) no feedback. All the learners were asked to write a draft twice (D1 and D2) before coming up with the final version (D3). The form-focused feedback was provided with underlining or circling lexical or grammatical errors. Content feedback included organization, paragraphing, relevance and cohesion. The findings depicted that all three aforementioned groups improved significantly in their writing accuracy, but the group that received simultaneous form-content feedback outperformed the other groups.

        Also, Hong (2004) examined the effect of coded versus un-coded feedback. The first group was provided with feedback in which errors were underlined, while the second group received both coded and underlined feedback for their errors. Syntactic, lexical and mechanical errors were focused on. The improvement of accuracy in the revised drafts of both groups was manifested, but the coded feedback group outperformed the other group. 

        In addition, Liu (2008) investigated the impact of error feedback on L2 writing. The study was an attempt to examine learners’ abilities to self-correct their writings considering two feedback conditions: (1) direct correction with the help of the teacher through corrective feedback and (2) indirect correction through just designating the error by the teacher without correcting it. Consequently, the results depicted that both kinds of feedback assisted learners in revising their drafts. Despite the fact that direct feedback decreased the rate of learners’ errors, it did not increase learners’ accuracy in another paper. Therefore, it was claimed that offering corrective feedback was not sufficient to improve learners’ writing accuracy.

        Arege (2010) targeted the role of different kinds of correction feedback on language learners’ writing. Fluency, successful correction and accuracy development were considered. The findings showed no improvement in fluency. However, for accuracy development, a controversial result was reached. Also, learners had a positive attitude toward receiving error correction feedback.

        Sarandi (2015) explored the third person –S in grammatical accuracy and claimed that the group that provided the corrective recast performed much better than the group that received no feedback.

Also, a study by Kamiya (2015) showed that extensive and intensive recasts were highly significant in developing learners’ explicit knowledge.

Finally, in a study, Nassaji (2017) found that extensive recast had a more critical role in improving writing compared to intensive recast. Thus, to shed more light on this, the current study aimed to investigate the role of recasts in both classroom settings to explore the effect of intensive vs. extensive recasts on EFL learners’ grammatical accuracy in writing. Thus, the following research questions were posed:

1. Does using extensive recast through grammar-focused task have any statistically significant effect on Iranian intermediate EFL learners’ grammatical accuracy in writing?

2. Does using intensive recast through grammar-focused task have any statistically significant effect on Iranian intermediate EFL learners’ grammatical accuracy in writing?

3. There was not any statistically significant difference between extensive versus intensive recast on Iranian intermediate EFL learners’ grammatical accuracy in writing?

 

2. Methodology

In order to explore the comparative effects of intensive versus extensive recasts through grammar-focused tasks on Iranian intermediate EFL learners’ grammatical accuracy in writing, this study deployed a pre-test post-test control group design.

 

The schematic representation of the design is depicted as follows:

EG (intensive) T1 X1 T2

EG (Extensive)                                               T1  X2 T2

CG T1 --- T2

X stands for treatment; EG represents the experimental group, and T1 and T2 stand for pre and post-tests.

        Regarding participants, 100 Iranian EFL learners were selected in this study based on their willingness to participate in all the study phases. They were first pre-tested through the Oxford Placement Test (OPT). OPT aimed to guarantee the students’ homogeneity and neutralize the effect of other factors that could change the study’s outcome. Moreover, pre-testing was performed to ensure all subjects had the same language proficiency level. The population in this study contained both male and female students. In terms of age, the subjects ranged from 19 to 24. The due learner participants were categorized as two experimental groups (i.e., intensive versus extensive) and a control group. They were intermediate language learners studying English at the Mehr Institute.

        APA ethical guidelines were all taken into consideration to select subjects. As a result, both confidentiality and informed consent were included. All the participants took part in the study voluntarily, and their identity and performance were kept confidential.

        In terms of materials applied to this study, the OPT, one of the most famous and standard tests, was utilized to specify English language learners’ level of L2 proficiency. The sample of the OPT employed in this study included two sections: Section A: 40 items, and Section B: 20 items. Section A contained 25 MC pictorial items, 15 questions in MC cloze text format, and 20 grammatical MC items. Section B contained ten MC cloze text items and ten MC vocabulary items. The whole test lasted 50 minutes.

        Furthermore, the following books were used for the instruction purposes in the present study: (1) Paragraph Development, by Arnaudet and Barrett (1990); (2) Academic Writing from Paragraph to Essay, by Zemach and Rumisek (2005).

        In order to gain the purpose of the research, the following procedure was carried out: I) sampling, II) pre-test, III) treatment, and finally IV) post-test. Before commencing the instruction, 100 male and female EFL learners (i.e., as the study population) who matched the selection criteria were recruited. The selected participants were intermediate language learners based on the evaluation made at the Mehr Institute. However, to guarantee the homogeneity of the participants, the OPT was given to a population of 100 learners. After scoring the test, the data was analyzed, and 60 learners (i.e., as the study sample) whose scores were one standard deviation above and one below the mean were chosen and assigned into two experimental groups (i.e., extensive and intensive) and a control group equally (n=20).

        Following the pre-test, the learner participants in experimental and control groups received the intended treatments and placebo, respectively. Before the instruction, the teacher also provided the participants with the necessary information about what they were required to do. For the experimental groups, in every session, the teacher chose a particular type of genre (i.e. example, cause-and-effect analysis, classification, comparison vs. contrast and argumentative essays), discussed the genre by modelling and elaborating the building blocks of the genre, and the communicative purpose which it served. Next, in the joint construction stage, the students were asked to reconstruct, revise and paraphrase the text in their own words. Then, having prior understanding and knowledge of the structure of the genre, the students were asked to write a paragraph. The teacher offered either intensive feedback (i.e., on a single, pre-selected linguistic structure) or extensive feedback (i.e., on a wide range of linguistic forms). Finally, the participants in experimental groups were provided with error grammar-focused tasks in which they were required to revise the erroneous sentences. Furthermore, they could consult their peers as well as their teacher. These tasks were deployed to enhance students’ grammatical accuracy. The whole treatment for both experimental groups lasted for ten sessions.

        Moreover, placebo was provided for the control group. Having delineated the structure of the genre, the learner participants were asked to write a paragraph. Next, their writings were partially commented on in their papers. Finally, the writing post-test was given to them to deal with the effect of the two types of treatment.

        In addition, for assessing the participants’ writings, Jacobs et al. (1981) suggested that researchers should obtain at least two writing samples from each participant in order to have a reliable representative of students’ performance. To do so, two tasks were explicitly designed to fulfil the purpose of the study. The topics were assigned in a way that elicited genuine engagement by placing writers in authentic situations. Before implementing the treatment, a writing pre-test was performed to obtain the initial differences among the test takers. The pre-test was an in-class writing task in which the participants were supposed to write a paragraph within an hour. To reach this, the learners were provided with three topics, and they were asked to choose one of the topics and to write a paragraph with at least 150 words. To select the writing topics, some factors were taken into consideration, such as moderate difficulty, not depending on learners’ background knowledge, and including everyday issues. The topics were chosen based on participants’ lives and society to stimulate participants to write enthusiastically. Afterwards, two qualified raters scored the writing samples in accordance with the writing scoring rubric following Wang and Liao (2008). This rubric includes a comprehensive framework with an emphasis on the main factors in the scoring process involving organization, focus, support convention and vocabulary. The Pearson correlation coefficient was further estimated to ensure a suitable inter-rater reliability level. The reliability was 0.78, which seemed to be an acceptable measure of inter-rater reliability. After the treatment, a writing post-test was administered to check the effect of the two types of treatment on the groups. Like the pre-test, the writing post-test was a one-hour, in-class task paragraph writing. Similarly, the learner participants were supposed to write a paragraph with a minimum of 150 words, choosing among the provided topics. To select the writing topics, the same aforementioned guidelines were taken into account.

        To control the subjectivity of the data collection, several forethoughts were considered. First, two raters rated the writing papers to ensure the reliability of the scoring. The raters followed a taxonomy of the grammatical errors, through which they could recognize and measure the number of errors each participant committed in the writing. Secondly, the raters were trained to learn about the grammatical errors in the learners’ writing by using the coding list of errors.

 

3. Results and Discussion

The descriptive statistics of the learner participants’ scores in both pre and post-test, including experimental and control groups, are depicted in Table 1.

Table1. The Descriptive Statistics for the Participants’ Pre-Test and Post-Test Scores in the Groups

Descriptive Statistics
	N		Minimum		Maximum		Mean		Std. Deviation			Skewness			Kurtosis
	Statistic		Statistic		Statistic		Statistic		Statistic			Statistic		Std. Error	Statistic	Std. Error
Groups		60		1.00		3.00		2.0000		.82339	.000		.309		-1.526	.608
Pre-Groups		60		15.00		21.00		17.5833		1.30243	.257		.309		-.163	.608
Post-Groups		60		17.00		25.00		20.9783		2.03439	-.027		.309		-.532	.608
Pre-extensive		20		15.75		21.00		17.7500		1.59151	.522		.512		-.826	.992
Pre-intensive		20		16.00		19.00		17.6375		.85638	-.201		.512		-.092	.992
Pre-Control		20		15.00		20.00		17.3625		1.38477	-.078		.512		-.650	.992
Post-extensive		20		19.00		25.00		22.6500		1.66307	-.521		.512		-.416	.992
Post-intensive		20		18.00		24.00		20.8125		1.40693	-.024		.512		.308	.992
Post-Control		20		17.00		22.00		19.4725		1.65811	-.064		.512		-1.204	.992
Valid N (listwise)		20

 

        Table 1 reveals the descriptive statistics of the learner participants’ scores in both pre and post-test, including experimental and control groups. Following Table 1, the mean scores of the extensive, intensive, and control groups in the pre-test were 17.75, 17.63, and 17.36 which raised to 22.65, 20.81, and19.47 in the post-test in post-test, respectively.

In addition, Skewness and Kurtosis values for both the pre-test and post-test were within the ranges of +/- 2, which proved the normality of the data. This normality was also traced in the normal distribution curves and box plots as well.

        To check the normality assumptions for running ANCOVA, the normality of distribution of test scores, homogeneity of regression slopes, linearity of slope of regression lines, and homogeneity of error variances were calculated. To do so, first, the normality of distribution of test scores was checked through the Kolmogorov-Smirnov and Shapiro-Wilk’s tests. Table 2 depicts the results of these tests.

 

Table 2. The Kolmogorov-Smirnov Test for Normality of the Distribution of the Test Scores

Tests of Normality
	Kolmogorov-Smirnova			Shapiro-Wilk
	Statistic	df	Sig.	Statistic	df	Sig.
Pre-Groups	.181	20	.084	.918	20	.089
Post-Groups	.192	20	.053	.933	20	.176
Pre-extensive	.181	20	.084	.918	20	.089
Pre-intensive	.186	20	.068	.914	20	.076
Pre-Control	.177	20	.099	.948	20	.332
Post-extensive	.192	20	.053	.933	20	.176
Post-intensive	.153	20	.200*	.955	20	.443
Post-Control	.129	20	.200*	.934	20	.187
*. This is a lower bound of the true significance.
a. Lilliefors Significance Correction

        As Table 2 shows, the Kolmogrov-Smirnov and the Shapiro-Wilk values were not significant. Then, this was resulted in observing the normality assumption of the test (P>.05). Next, the homogeneity of regression slopes was dealt with by measuring the interaction between groups and the covariate (pre-test).

 

        Table 3. Test of between Subjects Effect

Tests of Between-Subjects Effects
Dependent Variable: Post Groups
Source	Type III Sum of Squares	df	Mean Square	F	Sig.
Corrected Model	176.333a	5	35.267	28.066	.000
Intercept	5.421	1	5.421	4.314	.043
Groups	2.744	2	1.372	1.092	.343
Pre.Groups	63.988	1	63.988	50.924	.000
Groups * Pre.Groups	1.270	2	.635	.506	.606
Error	67.853	54	1.257
Total	26649.615	60
Corrected Total	244.187	59
a. R Squared = .722 (Adjusted R Squared = .696)

 

        As Table 3 depicts, the obtained value [F (2, 54) = .506, Sig =.606)] was larger than P< .05, which revealed that the interaction between the covariate and independent variable was not statistically significant and the assumption of the homogeneity of the slope of regression lines was held. Second, the linearity of the slope of regression lines was checked through a scatterplot. As Figure 9 presents, there was a linear relationship between the pre and the post-test scores, which indicated that the linearity assumption of regression lines was also met. 

 

Figure 1. Linear relationship among regression lines

 

        After checking the normality assumptions, the Levene’s statistic was deployed to learn about the homogeneity of error variances. Levene’s statistic deals with the assumption that the error variance is equal for all the groups.

 

Table 4. Levene's Test of Equality of Error Variancesa

Dependent Variable: Post Groups
F	df1	df2	Sig.
.216	2	57	.807
Tests the null hypothesis that the error variance of the dependent variable is equal across groups.
a. Design: Intercept + Pre Groups + Groups

 

        As displayed in Table 4 above, the results of Levene’s test were not statistically significant for the post-test (F (2, 57) = .216, Sig = .807). Considering the obtained results, it was revealed that there was not a statistically significant difference among the groups’ variances. Having checked the normality assumptions, an analysis of covariance (ANCOVA) was run to deal with the research hypotheses and remove the effect of the pre-test on learner participants’ performance in the post-test. Table 5 reflects the results.

 

 

 

Table 5. Tests of Between-Subjects Effects for the Analysis of Covariance (ANCOVA)

Dependent Variable: Post.Groups
Source	Type III Sum of Squares	df	Mean Square	F	Sig.	Partial Eta Squared
Corrected Model	175.063a	3	58.354	47.275	.000	.717
Intercept	10.476	1	10.476	8.487	.005	.132
Pre.Gropus	73.273	1	73.273	59.361	.000	.515
Groups	81.059	2	40.530	32.835	.000	.540
Error	69.124	56	1.234
Total	13705.215	60
Corrected Total	244.187	59
a. R Squared = .717 (Adjusted R Squared = .702)

 

        

 

 

 

 

 

 

 

 

 

        In Table 5, the Group row indicates the significant effect of the treatment on the dependent variable. Having the pre-test scores adjusted, the results revealed a significant effect of the groups (F (2, 56) = 32.83, P= 0.00, partial η² = .540)]. As the P-value was smaller than 0.05, it could be claimed that there were statistically significant differences between the post-test and the pre-test’s mean scores. Next, the marginal means were estimated. Although the F-value of 32.83 depicted a statistically significant difference between the mean scores of all the groups on the post-test, the post-hoc comparison tests were utilized to compare the groups two by two to check the validity of the null hypotheses posed in this study. Tables 6 and 7 reflect the results of the analyses.  

Table 6. Group Estimates

Dependent Variable: Post Groups
Groups	Mean	Std. Error	95% Confidence Interval
			Lower Bound	Upper Bound
Extensive	22.506a	.249	22.007	23.005
Intensive	20.766a	.249	20.268	21.264
Control	19.663a	.250	19.163	20.163
a. Covariates appearing in the model are evaluated at the following values: Pre Gropus = 17.5833.

 

        Table 6 indicates that the estimated marginal mean score for the extensive recast group (22.50) was higher than those obtained by the intensive recast group (20.76) and the control group (19.66).

 

Table 7. Pairwise Comparisons for the Groups' Performance

Pairwise Comparisons
Dependent Variable: Post Groups
(I) Groups	(J) Groups	Mean Difference (I-J)	Std. Error	Sig.b	95% Confidence Interval for Differenceb
					Lower Bound	Upper Bound
Extensive	Intensive	1.740*	.352	.000	.873	2.608
	Control	2.843*	.354	.000	1.970	3.717
Intensive	Extensive	-1.740*	.352	.000	-2.608	-.873
	Control	1.103*	.353	.008	.232	1.973
Control	Extensive	-2.843*	.354	.000	-3.717	-1.970
	Intensive	-1.103*	.353	.008	-1.973	-.232

Based on estimated marginal means

*. The mean difference is significant at the .05 level.

b. Adjustment for multiple comparisons: Bonferroni.

 

        According to the results revealed in Tables 6 and 7, the following can be concluded:

There was a statistically significant difference between the results obtained from the extensive group (M=22.50) and the control group (M= 19.66), (MD = 2.84, P < .05). Therefore, the first null hypothesis, “Using extensive recast through grammar-focused task does not have any statistically significant effect on Iranian intermediate EFL learners’ grammatical accuracy in writing” was rejected. This implied that the extensive recast group performed better than the control group on the post-test.

Moreover, there was a statistically significant difference between the mean scores of the intensive recast group (M = 20.76) and the control group (M = 19.66) (MD = 1.10, P < .05). Accordingly, the second null hypothesis as “Using intensive recast through a grammar-focused task does not have any statistically significant effect on Iranian intermediate EFL learners’ grammatical accuracy in writing” was also rejected. This implied that the intensive recast group also had a better performance compared to the control group on the post-test.

        Additionally, there was a statistically significant difference between the mean scores of the extensive recast group (M = 22.50) and the intensive recast group (M = 20.76), (MD = 1.74, P > .05). Therefore, the third null-hypothesis that “There was not any statistically significant difference between intensive versus extensive recast on Iranian intermediate EFL learners’ grammatical accuracy in writing” was also rejected implying that extensive recast could affect EFL learners’ writing performance more than intensive recast. 

Concerning the first research hypothesis (i.e., using intensive recast through grammar-focused tasks does not have any statistically significant effect on Iranian intermediate EFL learners’ grammatical accuracy in writing), the present study’s findings revealed that the intensive recast group outperformed significantly compared to the control group. 

        Moreover, regarding the second research hypothesis (Using extensive recast through grammar-focused tasks does not have any statistically significant effect on Iranian intermediate EFL learners’ grammatical accuracy in writing), it was revealed that the extensive recast group performed significantly better than the control group in writing post-test.

        Consequently, this study’s findings align with some previous studies (e.g., Sukur & Demircan, 2020). There are, however, some discrepancies. For example, Ellis et al. (2008) found no significant difference between intensive and extensive corrective feedback, claiming that both feedback types were much better than no correction in the narrative writing post-test. Sheen et al. (2009) also confirmed that the focused corrective feedback group significantly outperformed the control group in the narrative writing post-test, whereas the unfocused corrective feedback group did not show any outperformance. Moreover, Sarandi (2015) focused on the third person and claimed that the group receiving corrective recast performed much better than the no-feedback group on the post-test oral production. In addition, considering Kamiya’s (2015) study, it was revealed that both intensive and extensive recasts had significant roles in improving explicit knowledge. On the other hand, Nassaji (2017), in a study, concluded that extensive recast was more beneficial for the accurate use of articles and grammaticality judgment tasks compared to intensive recast.

 

4. Conclusion

Grammatical accuracy has been a controversial issue in L2 writing. A number of questions have been asked by researchers studying L2 writing over the last decades. Also, the role of corrective feedback on the acquisition of grammatical accuracy has always been a controversial issue in the last couple of decades. Thus, the present study explored the impact of intensive versus extensive recasts through grammar-focused tasks on Iranian EFL learners’ grammatical accuracy in writing.

        The findings revealed that using both extensive and intensive recasts through grammar focused task had statistically significant effect on Iranian intermediate EFL learners’ grammatical accuracy in writing. However, extensive recast has more highlighted role in improving EFL learners’ writing performance compared to intensive recast.

        Consequently, in the classroom settings, the English teachers can provide their students with intensive recasts through grammar-focused tasks since intensive recasts may help the learners notice the target structure more easily and focus on how and where that structure should be used. This can be beneficial especially for the learners who do not have much background knowledge about the target structure.

        In addition, material developers are suggested paying attention to the recurring types of grammatical errors across different types of paragraphs. They can also devote supplementary parts and activities in the books for introducing the grammar-focused tasks to help learners produce the target grammatical structures or revise the common errors in students’ performance through extensive recasts.

Furthermore, teacher educators can make teachers aware of the common types of grammatical errors in learners’ L2 writing or inaccuracies of students’ performance across different types of paragraphs.

Moreover, the application of intensive vs. extensive recasts through grammar-focused tasks can be investigated in other skills. For example, the impact of different kinds of corrective feedback can be investigated in speaking. To add, future research can be applied to compare the effects of extensive vs. intensive recasts at different levels in terms of their readiness to obtain the target structure.

 

References

Arege, J. B. (2010). Exploring some effects of different types of error correction feedback on ESL student

Arnaudet, M. L., & Barrett, M. E. (1990). Paragraph development: A guide for students of English. Prentice Hall Regents.

Ashwell, T. (2000). Patterns of teacher response to student writing in a multiple-draft composition classroom: Is content feedback followed by form feedback the best method? Journal of Second Language Writing. 9(3), 227-258.

Batstone, R. (1994). Grammar. Oxford University Press.

Bitchener, J., Young, S., & Cameron, D. (2005). The effect of different types of corrective feedback on ESL student writing. Journal of Second Language Writing, 14(3), 191-205. https://doi.org/10.1016/j.jslw.2005.08.001 

Carroll, S. E. & Swain, M. (1993). Explicit and implicit negative feedback: An empirical study of learning. Studies in Second Language Acquisition, 15(3), 357-386. https://doi.org/10.1017/s0272263100012158 

Chandler, J. (2003). The efficacy of various kinds of error feedback for improvement in the accuracy and fluency of L2 student writing. Journal of Second Language Writing, 12(3), 267–296. https://doi.org/10.1016/s1060-3743(03)00038-9  

Chastain, K. (1988). Developing second language skills: Theory and practice. Harcourt Brace Jovanovich, Inc.

Chun, A. E., Day, R. R., Chenoweth, A., & Luppescu, S. (1982). Errors, interaction, and correction: A study of native-nonnative conversations. TESOL Quarterly, 16(4), 537-547. https://doi.org/10.2307/3586471 

Ellis, N. C. (2008). The dynamics of second language emergence: Cycles of language use, language change, and language acquisition. The Modern Language Journal, 92(2), 232-249. https://doi.org/10.1111/j.1540-4781.2008.00716.x 

Ellis, R. (2001). Introduction: Investigating form‐focused instruction. Cambridge University Press.

Ellis, R., Loewen, S., & Erlam, R. (2006). Implicit and explicit corrective feedback and the acquisition of L2 grammar. Studies in Second Language Acquisition, 28(3), 339-368. https://doi.org/10.21832/9781847691767-015 

Fahim, M., & Montazeri, M. (2013). The impact of metalinguistic corrective feedback on EFL learners’ levels of lexical resource and grammatical range and accuracy in their oral proficiency. International Research Journal of Applied and Basic Sciences, 4(7), 1776-1782.

Fathman, A. K., & Whalley, E. (1990). Teacher response to student writing: Focus on form versus content. In B. Kroll (Ed.), Second language writing: Research insights for the classroom (pp. 178-190). Cambridge University Press.

Ghoorchaei, B., Tavakoli, M., & Ansari, D. N. (2010). The impact of portfolio assessment on Iranian EFL students’ essay writing: A process-oriented approach. GEMA Online Journal of Language Studies, 10(3), 122-142. https://doi.org/10.47191/ijsshr/v4-i11-20  

Greenslade, T., & Félix-Brasdefer, J. C. (2006). Error correction and learner perceptions in L2 Spanish writing. In C. A. Klee & T. L. Face (Eds.), 7th conference on the acquisition of Spanish and Portuguese as first and second languages (pp. 185-194). Cascadilla Proceedings Project.

Harmer, J. (2003). How to teach writing. Longman-Pearson Publications.

Hyland, K. (1998). Genre-based pedagogies: A social response to process. Journal of Second Language Writing, 12(1), 17-29. https://doi.org/10.5040/9781350037939.0034 

Hyland, K. (2003). Second language writing. Cambridge University Press.

Jacobs, H. L., Zinkgraf, S. A., Wormuth, D. R., & Hartfield, V. F. (1981). Testing ESL composition: A practical approach. Newbury House.

Kamiya, N. (2015). The effectiveness of intensive and extensive recasts on L2 acquisition for implicit and explicit knowledge. Linguistics and Education, 29(1), 59-72. https://doi.org/10.1016/j.linged.2014.11.002 

Kepner, C. G. (1991). An experiment in the relationship of types of written feedback to the development of second-language writing skills. The Modern Language Journal, 75(3), 305-313. https://doi.org/10.1111/j.1540-4781.1991.tb05359.x 

Kim, J. H. (2004). Issues of corrective feedback in second language acquisition. Working Papers in TESOL & Applied Linguistics, 4(2), p.1-24. https://doi.org/10.1017/9781108589789.003 

Kim. H., Mathes, G. (2001). Explicit and Implicit corrective Feedback. The Korea TESOL Journal, 4(1), 57-72.

Liu, Y. (2008). The effects of error feedback in second language writing. Journal of Second Language Acquisition and Teaching, 15(1), 65-79.

Long, M., Inagaki, S., & Ortega, L. (1998). The role of implicit negative feedback in SLA: Models and recasts in Japanese and Spanish. Modern Language Journal, 82(3), 357–371. https://doi.org/10.1111/j.1540-4781.1998.tb01213.x 

Lyster, R. (2004). Differential effects of prompts and recasts in form-focused instruction. Studies in Second Language Acquisition, 26(3), 399-432.

Lyster, R., & Ranta, L. (1997). Corrective feedback and learner uptake: Negotiation of form in communicative classrooms. Studies in Second Language Acquisition, 22(1), 37-66. https://doi.org/10.1017/s0272263197001034 

Nunan, D. (2003). Practical English language teaching. McGraw-Hill Publications.

Rezaei, S., Mozaffari, F., & Hatef, A. (2011). Corrective feedback in SLA: Classroom practice and future directions. International Journal of English Linguistics, 1(1), 21. https://doi.org/10.5539/ijel.v1n1p21 

Richards, J. C., & Schmidt, R. W. (2002). Longman dictionary of language teaching and applied linguistics. Routledge.

Sarandi, H. (2015). Reexamining elicited imitation as a measure of implicit grammatical knowledge and beyond…? Language Testing, 32(4), 485-501. https://doi.org/10.1177/0265532214564504 

Selinker, L. (1992). Rediscovering interlanguage. Longman Publication.

Semke, H. (1984). The effects of the red pen. Foreign Language Annals, 17(2), 195-202. https://doi.org/10.1111/j.1944-9720.1984.tb01727.x 

Sheen, Y., Wright, D., & Moldawa, A. (2009). Differential effects of focused and unfocused written correction on the accurate use of grammatical forms by adult ESL learners. System, 37(4), 235-255. https://doi.org/10.1016/j.system.2009.09.002 

Sheppard, K. (1992). Two feedback types: Do they make a difference? RELC Journal, 23(1), 103- 110.

Sukur, M., & Demerican, F. (2020). The effects of intensive and extensive recasts on learning the third person singular. Research on Youth and Language, 14(2), 58-75. https://doi.org/10.1111/modl.12387

Suzuki, W., Nassaji, H., & Sato, K. (2019). The effects of feedback explicitness and type of target structure on accuracy in revision and new pieces of writing. System, 81(1), 135-145. https://doi.org/10.1016/j.system.2018.12.017 

Tavakoli, H. (2012). A dictionary of research methodology and statistics in applied linguistics. Rahnama Press.

Tedick, D. J., & de Gortari, B. (1998). Research on error correction and implications for classroom teaching. ACIE Newsletter, 1(3), 1-6.

Truscott, J. (1996). The case against grammar correction in L2 writing classes. Language Learning, 46(3), 327-369. https://doi.org/10.1111/j.1467-1770.1996.tb01238.x

Van-Ha, X., & Murray, J. (2023). Corrective feedback: Beliefs and practices of Vietnamese primary EFL teachers. Language Teaching Research, 27(1), 211-240. https://doi.org/10.1177/1362168820931897 

Wang, Y. H., & Liao, H. C. (2008). The application of learning portfolio assessment for students in the technological and vocational education system. Asian EFL Journal, 10(2), 132-154.

Williams, C., & Polio, J. (2009). Teaching and testing writing. In M. H. Long & C. J. Doughty (Eds.), The handbook of language teaching (pp. 134-167). Wiley Online Library.

writing [Unpublished doctoral dissertation]. University of Chicago.

Wu, C. H. (2007). Spoken grammaticality and EFL teacher candidates: measuring the effects of an explicit grammar teaching method on the oral grammatical performance [Unpublished doctoral dissertation]. The Ohio State University.

Zemach, D. E., & Rumisek, L. A. (2005). Academic Writing: From Paragraph to Essay. Macmillan.

Zheng, Y., & YU, S. (2018). Student engagement with teacher written corrective feedback in EFL writing: A case study of Chinese lower-proficiency students. Assessing Writing, 37(1), 13-24. https://doi.org/10.1016/j.asw.

---