• No results found

The Effectiveness of Different Types of Corrective Feedback for Grammar Acquisition: a Swedish Perspective

N/A
N/A
Protected

Academic year: 2021

Share "The Effectiveness of Different Types of Corrective Feedback for Grammar Acquisition: a Swedish Perspective"

Copied!
24
0
0

Loading.... (view fulltext now)

Full text

(1)

CULTURE-LANGUAGES-MEDIA

Independent Project with Specialization in English

Studies and Education

15 Credits, First Cycle

The Effectiveness of Different Types of

Corrective Feedback for Grammar

Acquisition: a Swedish Perspective

Effekten av olika typer av korrigerande feedback för grammatikinlärning: ett

svenskt perspektiv

Nils Annertz

Theodor Sj

ö

lund

Master of Arts in Upper Secondary Education, 300 credits or in Secondary Education, 240 credits

English Studies and Education 2021-01-16

Examiner: Maria Graziano Supervisor: Shaun Nolan

(2)

Individual contribution

We hereby certify that both authors of this paper have put in an equal amount of work. This goes for the research and all the parts of the paper.

Authenticated by: Nils Annertz Theodor Sjölund

Abstract

Both grammar and corrective feedback (CF) are important for second language acquisition, though they are not mentioned explicitly in the Swedish upper secondary school curriculum. Moreover, it is not clear which type of CF is most effective in grammar acquisition. This paper aims to consolidate, compare and contrast the findings of several articles examining the effect of CF on grammar acquisition. Two databases were used to find articles applicable to answering the research question. After excluding those that did not meet the criteria, nine articles were used in the end. Although the findings show that CF is effective for grammar acquisition in general, they do not show any type of CF to be superior to another. The factors identified as affecting the results of the studies are target structure, method, language proficiency and L1. Through the studies analyzed, it is not possible to establish the long-term effects of the various types of CF. Nevertheless, the findings suggest that oral meta-linguistic explanation (ME) is beneficial for teachers due to its time-efficiency. In

conclusion, CF, being shown to facilitate grammar acquisition for L2 learners in general, has positive effects on English learning in the Swedish classroom. However, more research is needed in order to establish its effectiveness long-term and on a more detailed level.

(3)

Table of Contents

1. Introduction 3 2. Aim 6 3. Method 6 Table 2 8 4. Results 9

4.2 Comparison and Contrast Based on Feedback 13

4.2.1 Direct Corrective Feedback 13

4.2.2 Meta-linguistic Explanation 13

4.2.3 Recasts 14

4.2.4 Indirect Corrective Feedback 14

4.2.5 Findings not Directly Related to Type of Feedback 15

4.2.6 Result Conclusion 16

4.3 Discussion 16

5. Conclusion 19

(4)

1. Introduction

The teaching of English grammar is important for second language acquisition (Ellis, 2006). Therefore, one can argue that grammar should be a part of English language education in Swedish schools, as is the case in the Swedish curriculum for lower secondary schools (Skolverket, 2011a). However, grammar is not explicitly mentioned in the curriculum for upper secondary levels, yet one can interpret from phrases in the curriculum that grammar could be included in one’s teaching. For example: “[Students should learn] how words and phrases in oral and written communications create structure and context by clarifying introduction, causal connection, time aspects, and conclusions.” (Skolverket, 2011b, p. 3). As a result of this, it is not clear if or how grammar should be taught. However, Ellis (2006) stated, “In short, there is now convincing indirect and direct evidence to support the teaching of grammar.” (p. 86), indicating that there is indeed a place for grammar in second language education.

Additionally, feedback is an essential factor in order to improve students’ learning outcomes and language accuracy (Hattie & Timperley, 2007). This claim is further justified by Dabaghi (2006) who asserted, "Error correction has been widely recognized as vital in second language acquisition and teaching" (p. 10). There are many different forms of feedback, but according to Hattie and Timperley, corrective feedback (CF) is the most common one. CF has been defined by Ellis, Loewen and Erlam (2006) as responses to students’ utterances or phrases that contain errors. However, CF in itself can be of different natures, for example, implicit/explicit or direct/indirect. Explicit CF makes it clear that an error has been made, while implicit CF does not point out the error or make the learner overtly aware that the error has been made. Feedback that provides the actual correct answer is sometimes referred to as direct CF, while indicating only that there is an error is a more indirect approach (Babanoğlu & Ağçam, 2015). Furthermore, Lyster and Ranta (1997) came up with six categories of CF, of which four are relevant to this paper. These categories are presented in Table 1 below with definitions adapted from Lyster & Ranta and Shao (2015); however, it should be noted that the different terms describing the type of feedback intertwine. For example, direct CF is very similar to explicit correction.

(5)

Table 1

Type of CF Definition

Explicit correction (similar to direct CF) The teacher informs the student that something went wrong, and provides the correct answer directly.

Recast If a phrase uttered contains an error, the

teacher restates it minus the actual error. Meta-linguistic feedback (often referred to

as meta-linguistic explanation)

The teacher informs the student that there is an error without providing the correct answer. In some cases the teacher gives a comment regarding the nature of the error, for example verb tense or subject-verb agreement.

Repetition (also a type of direct CF) The teacher repeats the phrase in its entirety, but calls attention to the error by changing the intonation.

CF can be studied through different learning theory-perspectives derived from cognitivism. Among the more researched theories, the interaction hypothesis and the noticing hypothesis are the most prominent ones (Chen, Lin & Jiang, 2016). The interaction hypothesis focuses on language acquisition through communication in the form of input, output, and feedback, and it is through communicative pressure that languages are learnt. From this perspective, CF through

communication is a way for learners to become aware of what is acceptable in the target language, and what is not. Furthermore, learners realize through the communicative feedback that there is a gap between their output and what the target language allows, which subsequently gets filled, resulting in an intake (Chen et al., 2016, Lightbown & Spada, 2012). The noticing hypothesis by Richard Smith proposes that in order to learn something, it must first have been noticed. When it comes to the relationship between CF and the noticing hypothesis, the theory is that the feedback helps learners to notice differences between the target-language and their output, resulting in

(6)

Another learning theory that can be connected to second language (L2) learning is the skill acquisition theory (SAT). According to this theory, the acquisition of a skill has three stages of development. It starts with theoretical knowledge of the skill, referred to as declarative knowledge. The application of this knowledge through repetition and practice marks the second stage, called the proceduralization. Lastly, after sufficient use of the procedural knowledge of the skill, it becomes automated. This means that the learner can make use of the skill without cognitive processing of the declarative knowledge. This theory is often applied to different skills in sports, but as stated above, it can also be applied to L2, or in this case L2 grammar acquisition (DeKeyser 2015, Wagner & Wulf 2016).

Furthermore, knowledge of language learning can be divided into two types: explicit and implicit knowledge. Additionally, there is a discussion on how grammar is acquired for L2 learners. Explicit knowledge can be described as a meta-linguistic and an aware knowledge that is used consciously, whereas implicit knowledge is used or applied to a context without the learner being aware of it (Ellis, 2005). However, it can be disputed whether the knowledge has been acquired or learnt. Nassaji (2017a) presented older arguments from Stephen Krashen’s comprehensible input

hypothesis that there is a distinction between acquisition and learning. Acquisition is described as the unconscious learning process of a language, while learning is more explicit and focused on providing rules of the language. However, more recent discussions aim to erase the differences between learning and acquisition, thus implying that both terms are identical (Wilson-Gegeo & Nielsen, 2003). In this paper, acquisition and learning will be treated synonymously, as it has been in several of the studies selected for this analysis.

Regarding studies and research on language acquisition and the teaching of L2, two terms are generally used. Firstly, English as a second language (ESL) is used when English is not the first spoken language, but it is one of the official languages spoken of the country. Secondly, English as a foreign language (EFL) is used when English is neither the first spoken language nor an official language of the country (Nayar, 1997). Therefore, the term which should be used in Sweden is EFL; although, one can argue that the term ESL can be used because the English language proficiency in Sweden is consistently ranked as very high by the Education First organization (EF, 2020).

(7)

There is a considerable number of studies investigating the role of feedback in ESL and EFL. However, it is problematic that there is little or no consensus in the research regarding which type of feedback, be it for example written or oral, is most effective in facilitating grammar acquisition (Shao, 2015). Consolidating the relevant research available and investigating this matter further could prove to be most significant for ESL and EFL teachers. The reason being that it could optimize grammar education for both teachers and students in terms of achieving the learning outcomes presented in the curriculum. Additionally, feedback should be used on a daily basis by teachers, and this research could be beneficial in other aspects of language teaching as well.

2. Aim

The aim of this research paper is to investigate if or how CF can improve English learners’ grammatical competences. The feedback investigated includes both written and oral feedback. To investigate this, the present paper will comprise an analysis and comparison of the relevant research available. Therefore, the following research question is proposed:

“Do different kinds of corrective feedback by teachers affect grammar acquisition for learners of English, and if so, how and to what extent?”

3. Method

When researching articles for this paper two databases were chosen and used simultaneously: Education Research Complete (ERC) and Education Resource Information Center (ERIC). The following search terms were used: corrective feedback, oral feedback, written feedback, L2, ESL, EFL, grammar and grammar acquisition. The type of feedback was not limited to either written or oral, for two reasons: firstly, both apply to teachers, and secondly, finding any differences in the results on written versus oral feedback can prove to be an interesting find. Hence, the search terms included both types. Moreover, since this paper is about grammar acquisition for learners of English in Sweden, it was deemed relevant to use search terms directed towards learners of a second or

(8)

foreign language, thus, the use of both ESL and EFL, as well as L2. For this very same reason, the research question being focused on syntax, grammar and grammar acquisition was included.

The research question targets the relationship between CF and grammar acquisition for learners of English. Therefore, the empirical studies examined for this paper includes various kinds of CF and how it affects various kinds of grammar for L2 learners. To make sure that the research was up to date, only scientific empirical studies conducted between 2010 and 2020 were used. This would ensure that the studies were made in the context of curricula being up to date or recent at the time of writing. Additionally, in order to reinforce trustworthiness for this paper, studies which were neither empirical, nor published in an academic journal or peer-reviewed were excluded. For the same reason, studies conducted on less than 30 participants were excluded. The research question of this paper being focused on CF prompted the exclusion of any studies not conducted on that type of feedback. Furthermore, since both authors of this paper are trainee teachers for students of ages 13 and above, and students of this age in Sweden are usually at an intermediate level or higher (EF, 2020), studies with participants at a beginner level and younger children were ruled out.

In order to establish the research relevant to this paper, the search terms - “corrective feedback”, “oral feedback”, “written feedback”, “L2”, “ESL”, “EFL”, “grammar”, and “grammar acquisition” - were used in different combinations. When using every search term in the databases 382 articles were found. The number of articles was narrowed down, by changing the combinations of the search terms. For example, using “oral feedback”, “ESL or EFL or L2”, and “grammar”, 19 articles were found, and 4 of these 19 articles were chosen due to the relevancy of their title and abstract. Additionally, when using “written feedback, “ESL or EFL or L2”, and “grammar” as search terms, 70 articles were found, and by using the same method described above, 7 of these articles were used. “Corrective feedback”, “grammar” and “EFL or ESL or L2” resulted in 162 articles, of which 10 were deemed relevant, mainly through readings of title and abstract. Several of the 10 articles found were the same as had already been selected in the previous search; nevertheless, it resulted in 2 new articles, making the number of articles selected 12.

After scrutinizing the 12 articles, 2 were excluded because of the aforementioned exclusion criteria, resulting in 10 empirical studies. One of the studies’ results were not statistically significant and was therefore omitted from this paper, making the total number of studies selected for this analysis 9.

(9)

These studies are listed in table 2 below, as well as their main area of CF and target structure, number of participants, age, and language proficiency level (if applicable).

Table 2

Author Type of CF Area of grammar focused on

Participants Language proficiency level

Benson, S., & DeKeyser, R. (2019)

Written Direct Feedback, and written meta-linguistic explanation

Verb tenses: simple past and present perfect

n=165. College students Low intermediate - advanced Bitchener, J., & Knoch, U. (2010) Written meta-linguistic explanation, indirect written feedback, and oral meta-linguistic explanation Articles n=63. Age 18-20 Advanced Ji Hyun Kim. (2016)

Oral Meta-linguistic and Recast

Articles n=63.

University students

Intermediate

Li, S., Zhu, Y., & Ellis, R. (2016)

Different timing of feedback. Immediate and delayed recast.

Past passive construction n=120. Ages 13-15.

Studied English for an average of 6.1 years Mansourizade

h, K., & Izwan Abdullah, K. (2014)

First direct written feedback, then oral and written meta-linguistic explanation

Subject-verb agreement n=43. Ages 19-20 Participated in an academic writing course Nassaji, H. (2011) Negotiated- limited negotiated and non-negotiated oral CF

Articles and prepositions n=31. Ages 18-48

Intermediate

Nassaji, H. (2017)

Oral feedback: intensive and extensive recasts

Articles n=48. Ages 18-37. Mean age: 23 Intermediate Shintani, N., Ellis, R., & Suzuki, W. (2014)

Written direct corrective feedback & written meta-linguistic explanation

Indefinite article and the hypothetical conditional n=171. Ages 18-21 Pre intermediate Wagner, J. P., & Wulf, D. J. (2016)

Indirect WCF Subject-verb agreement,

verb tense, singular/plural, and article usage. n=33. High-school students age 14-18 Intermediate-advanced

(10)

4. Results

All of the articles reviewed in this paper include studies on various types of CF. Some of them investigated several different types of CF, and others only one. The main focus of this section is the results of those studies and how the different types of feedback compare to each other in order to facilitate grammar acquisition. This section consists of three parts, a short summary for each of the nine studies which are organized according to Table 2, a comparison and contrasting of the studies, and finally a discussion of the findings. Furthermore, for information about the different feedback types see Table 1.

4.1 Summaries of Studies

Benson and DeKeyser (2019) investigated the effect of written direct feedback and meta-linguistic explanation (ME) for ESL learners on the simple past and the present perfect. Furthermore, they explored how learners’ attributes of the language, referred to as language-analytic ability (LAA), correlates with feedback. Their study was conducted on 151 participants in a state college in the USA and the participants’ origin varied greatly. The students were divided into groups depending on their LAA and origin, then randomly divided into three groups: two treatment groups for each type of feedback, and one control group. The task was to write essays on four occasions. The feedback was given on the two first essays, and directly after the second essay’s feedback the participants wrote again, and unannounced on an occasion four weeks later. The authors found both feedback types showed a short-termed improvement; only direct feedback showed a long-term improvement on the simple past. The researchers concluded the feedback being beneficial for enhancing grammar accuracy, yet more research is needed, and the research should include LAA on various components to see the correlation between aptitude and feedback.

Bitchener and Knoch (2010) investigated the effect of oral and written ME, and indirect written CF on the English article system over a 10-week period of time. The participants were 63 advanced L2 learners taking an introductory course at a university in the USA, originating from East and South Asia. The participants were separated into three treatment groups and one control group, and all of them took the same test: writing a story based on pictures. This was conducted three times: one pretest, one post-test, and one post-test 10 weeks after without any interference. The researchers

(11)

found that all treatment groups advanced directly with the feedback. However, the indirect feedback group did not maintain the knowledge for the post-test 10 weeks after. The authors concluded that CF can, with only one treatment, improve advanced learners' writing accuracy; although, they stressed that more research is needed.

Kim (2016) conducted a study to see the effects of recast and oral ME on EFL learners’ explicit and implicit accuracy on the English article system. The researcher used 63 participants who all were intermediate English learners at a university in Korea. The participants were divided into two treatment groups and one control group. One treatment group received recast and the other meta-linguistic. The students were tested on producing the correct articles orally, and there was a pretest, a post-test, and a delayed post-test two weeks after. The study found both recast and ME provided the learners with explicit knowledge, both short-termed and longitudinal. However, only ME provided longitudinal implicit knowledge. The author concluded that feedback provides EFL learners with grammatical accuracy, although Kim argued more research is needed within the field, suggesting more longitudinal studies, and studies conducted on learners with little to no knowledge of the target structure.

Ellis, Li, and Zhu (2016) investigated if the timing of CF, in this case immediate and delayed recast, has any effect on grammar acquisition on EFL learners. Additionally, they explored how feedback affected learners of little to no previous knowledge of the target structure, and if it facilitates either explicit or implicit knowledge. They included 120 participants of the ages 13-15 at a middle-class school in China, and their target structure was the past passive. The participants were divided into four groups, one for each type of feedback, one for the task instructions, and one control group. Three tests on oral dictogloss tasks were conducted, a pretest, a post-test, and a delayed post-test. The study showed explicit knowledge was obtained for both immediate and delayed feedback, the former being slightly more effective. However, the study showed almost no improvement on implicit knowledge for the learners. The authors concluded immediate feedback could have a

positive effect on learners with almost no prior knowledge of the target structure, and more research is needed.

(12)

taking an advanced English communicative class in Malaysia. The researchers divided the participants into three groups: one oral, one written, and one oral and interaction group. The participants took a short essay pretest which they received direct written feedback on. Thereafter, they received their respective meta-linguistic treatment and wrote a post-test. The researchers found that both types of feedback greatly enhanced the learning, but oral showed greater accuracy gain than written. They reasoned this being a pedagogical advantage since it was conducted as a mini-lesson, thus making it time-efficient. The authors concluded that ME could improve language acquisition for second language learners, especially oral feedback. Additionally, they argued more research is needed on the topic, particularly more longitudinal research.

Nassaji (2011) researched the effect of oral feedback on how well students were able to correct their errors. He also compared oral negotiated feedback with limited negotiation and no negotiation, as well as the effectiveness of negotiated feedback on prepositions and articles. Two Canadian ESL classes of 15 and 16 students respectively participated, all at an intermediate language proficiency level. During three weeks, the participants wrote journals and received feedback on them. Each week focused on one type of feedback, either non-negotiated, limited negotiated, or negotiated. An immediate and a delayed post-test was conducted every week. The results showed that negotiated feedback was most effective, followed by limited negotiated and non-negotiated feedback for both post-tests. Concerning target structures, the same result was shown for articles on both post-tests, but there was no statistically significant difference for prepositions. The author concluded that negotiated feedback seems to be more effective than non-negotiated feedback. However, the long-term effects are not clear and more research on the subject is needed.

The effect of extensive vs intensive recasts on the acquisition of English articles has been researched by Nassaji (2017b). The effect of recasts in general, and if different outcome measures affected the results were also researched. Forty-eight intermediate ESL learners were divided into three groups, one receiving extensive recasts, i.e. feedback on all types of errors, one receiving intensive recasts, i.e. only feedback on one target structure, and a control group. All groups received their respective treatments on an oral and a written task. Three outcome measures, an oral, a written, and a grammar test, were used at pretest, immediate post-test, and delayed post-test. The results showed that the groups receiving recasts outperformed the control group for all outcome measures. However, for the written test, the results were only statistically significant for the delayed post-test. The extensive

(13)

recast group performed better than the control group for the other two outcome measures, but the intensive recast group did not. The author concluded that while recasts are effective when it comes to the acquisition of English articles, more longitudinal studies are needed to establish the long-term effect.

Shintani, Ellis and Suzuki (2014), investigated if direct CF and ME respectively improve accuracy for indefinite articles and hypothetical conditionals compared to no feedback, and how the two types of feedback compared. The researchers had 171 participants of ages 18-21 divided into five groups, two groups receiving either direct CF or ME, two receiving direct CF or ME plus revision, and one control group. Over the course of five weeks, three written dictogloss tasks were completed and the experimental groups received their respective treatments. The results showed that for hypothetical conditionals, all experimental groups outperformed the control group short-term; however, only the group receiving direct CF plus revision increased their accuracy long-term. There was no statistically significant difference for any of the groups regarding indefinite articles, because, the authors

speculated, the participants focused on the more complex target structure, the hypothetical conditional. They added that this was also the reason for CF not having a long-term effect. The authors concluded that the findings showed both types of feedback were effective short-term, but only direct CF with revision were effective long-term.

In a study by Wagner and Wulf (2016), they investigated the effect of indirect written CF on overall errors, as well as grammatical, mechanical, and word usage. They also researched if different types of grammar errors are equally affected by indirect written CF, and if the proficiency level of the

participants affects the effectiveness of indirect written CF. The learners, 17 intermediate and 16 advanced, were divided into four groups and produced written paragraphs every three weeks for 12 weeks. However, only two of the groups received indirect written CF on the paragraphs. The results showed that indirect written CF was effective for overall errors, as well as grammatical, mechanical and word usage. However, the improvement regarding mechanical errors was not statistically significant. Indirect written CF was also shown to be effective for some grammatical errors (subject-verb agreement and singular/plural), but not for (subject-verb tense or articles. In all categories, both the intermediate and advanced experimental groups outperformed the respective control groups. The authors concluded that indirect written CF is only effective if the options for a particular error are

(14)

4.2 Comparison and Contrast Based on Feedback

4.2.1 Direct Corrective Feedback

All of the studies that used direct CF, both oral and written, found that it had a significant effect on improving grammar accuracy. Shintani et al. (2014), and Benson and DeKeyser (2019) also found that the effect for the written CF had a long-term effect, even though their target structures differed. Benson and DeKeyser looked at verb tense, while Shintani et al. looked at articles and hypotheticals. In the latter study, only the results related to hypotheticals were statistically significant. However, according to the authors, the most likely reason for this is that hypothetical conditionals are much more complex than articles, resulting in the participants focusing almost solely on the target structure that required more of them. Furthermore, Benson and DeKeyser researched verb tenses since many studies had previously used articles as a target structure. They tried two different tenses, the simple past and the present perfect, finding simple past the only one that facilitated long-term learning. The authors suggested that the present perfect was only beneficial short-term because the learners lacked declarative knowledge about it. Another interesting finding that Benson and

DeKeyser made was how feedback functioned on various skill levels. The more knowledge or language aptitude the learner had before, the more beneficial the written direct CF. Mansourizadeh and Izwan Abdullah (2014) also used written direct feedback, although it was not intended for the test and not mentioned in the aim. Nevertheless, all participants in the study first received written direct CF before applying meta-linguistic explanation (ME). Nassaji (2011) looked at negotiated feedback compared to non-negotiated feedback (which is similar to direct CF). However, he did not use a control group, rendering the results non-applicable. The same case can be made with

Mansourizadeh and Izwan Abdullah, who did not use a control group either.

4.2.2 Meta-linguistic Explanation

Every study on ME, both oral and written, showed improvements on the learners’ language

acquisition. Shintani et al. (2014) found that the effect of ME on the hypothetical conditional proved effective when compared to the control group, but only short-term. As stated above, there was no statistically significant difference on the acquisition of articles. Almost the same results were found

(15)

by Benson and DeKeyser (2019); however, on the target structure of the simple past and the present perfect. Additionally, Bitchener and Knoch (2010) found that written and oral ME had significant improvements on long-term writing accuracy with articles as the target structure. A result in line with this is the findings of Kim (2016), who showed that oral ME provides implicit long-term knowledge for L2 learners, although not explicit. Furthermore, Mansourizadeh and Iswan Abdullah (2014) compared written and oral ME, their findings suggesting that both types of ME improved the grammar accuracy of subject-verb agreement; although, oral being superior to written. They argue that this finding has great pedagogical value since it is possible to have a mini-lesson with the whole class which makes it time-efficient. Even though Shintani et al’s results only were effective short-term, they share this notion of the time efficiency of ME. Bitchener and Knoch also suggest that oral ME can be used in order to save time as a teacher.

4.2.3 Recasts

All studies researching recast found it to be effective for their respective target structures. Both Nassaji (2017b) and Kim (2016) looked at articles, namely first-mention and anaphoric, while Li et al. (2016) investigated the past passive. The procedures of each study differed in relation to each other, but all of them tested the effects of recasts on both implicit and explicit knowledge of the learners. Li et al. compared the timing of the recasts, immediate or delayed, while Nassaji compared extensive and intensive recasts, and Kim looked at the relationship between recasts and ME. The results of both Kim and Nassaji (who both looked at the article system) showed that recasts are effective for both implicit and explicit knowledge. However, Li et al., who looked at past passive on learners with no knowledge of that target structure before the study, found that recasts improved explicit knowledge, but not implicit knowledge.

4.2.4 Indirect Corrective Feedback

Both Wagner and Wulf (2016), and Bitchener and Knoch (2010) showed that indirect CF improves acquisition of written grammar in general. Even though both of them had a control group, Wagner

(16)

Both studies looked at the English article system but found differing results. Wagner and Wulf did not find indirect CF to have a positive effect, whereas Bitchener and Knoch found the feedback to provide short-termed accuracy gain, however not long-termed gain.

4.2.5 Findings not Directly Related to Type of Feedback

Several of the studies mention feedback in relation to language-analytic ability (LAA). Nassaji (2017b) hypothesized that feedback can be more extensive rather than intensive the higher the LAA of the receiver. The same conclusion was drawn by Shintani et al. (2014), as the LAA of the

participants in their study was too low to take advantage of the feedback given on two target structures simultaneously, i.e. the feedback was too extensive for their language proficiency level. There are also similarities to this in Benson and DeKeyser’s study. Their findings on LAA in

correlation with feedback suggested that learners of low LAA benefit more from ME, while learners with high LAA benefit from direct feedback, at least on the simple past tense.

Additionally, most studies that compared the effect of CF on different target structures saw contrasting results. These differences were explained by the target structures being of various degrees of treatability (Nassaji 2011, 2017, Shintani et al. 2014, Wagner and Wulf 2016). Nassaji (2011) hypothesized that his findings of negotiated feedback being more effective than non-negotiated feedback (similar to direct feedback and explicit correction) for articles were a result of the complexity, or degree of treatability, of that target structure. Wagner and Wulf were of a similar opinion, that the effectiveness of direct CF is dependent on the degree of treatability of target structures, which is why they saw positive results for grammatical constructions with binary options, such as subject-verb agreement or singular/plural, as opposed to articles or verb tense, which have more than two options. Shintani et al. had the same hypothesis, hypothetical conditionals being more complex and therefore less treatable than articles. However, their results showed written CF to be more effective on hypothetical conditionals. It should be noted though, that their choice of target structures being too complex for the participants to make use of any feedback rendered the results not valid enough in this matter. Furthermore, Benson and DeKeyser (2019) argued that the reason the present perfect did not have a longitudinal effect might be because it used two different target structures simultaneously, thus overwhelming the participants.

(17)

4.2.6 Result Conclusion

The results of the studies reviewed showed CF to be effective when it comes to grammar acquisition for L2 learners. However, this effect is highly dependent on the target structure and how the studies have been conducted. One single superior type of CF or a ranking of the different types could not be found from the results of the selected studies.

4.3 Discussion

The findings and comparisons of the studies showed that all of the different feedback types are more or less effective for grammar acquisition, which is in alignment with other authors’ findings (Hattie & Timperley 2007, Dabaghi 2006). However, it is quite difficult to prove one feedback being better than the other. This finding is also supported by previous research, which has not been able to rank the different kinds of CF with regards to their effectiveness in relation to each other (Shao, 2015). One reason for this could be that there is a high number of factors or variables present in the studies on CF, which has to be taken into consideration when analyzing and interpreting the results. Some of these factors included the target structure researched, the outcome measure(s) used, the timing of the CF given, and how long after the treatment that the effect was measured. Guenette (2007), who also reviewed a number of studies carried out on various types of CF and their effectiveness, argued that there are difficulties in comparing the results because of differing elicitation tasks, feedback measures, research designs and populations.

Furthermore, the language proficiency level of the participants also affected the results, as the participants’ ability to make use of the feedback given differs depending on how much cognitive processing is needed (Shintani, 2014). This is also supported by Benson and DeKeyser (2019) when they stated that, “Evidence furthermore indicates that learners with higher LAA will have greater success when making the cognitive comparisons that are required for CF to result in learning” (p. 704).

(18)

Additionally, Benson and DeKeyser (2019) provided the only study, among those in this paper, which used LAA in their method to see how feedback affects the grammar acquisition on the learners’ aptitude in the language. They argue that this is one of the reasons why their study showed different results from other studies on the same target structure. Another factor, suggested by Ferris (2006) is the high number of standard deviations on research regarding written feedback, thus making the results differ highly. Therefore, in line with Benson and DeKeyser (2019), one can argue that future research should contain some elements of LAA to get a clearer picture on what type of feedback that functions on certain types of target structures. Additionally, the L1 of the participants can affect their ability to learn different target structures. For example, Li et al. (2016) reasoned that their participants, originating from China, had difficulties with acquiring implicit knowledge on the past passive. The authors argue that the reason for this is because it does not exist in their L1.

Another reason for the lack of consensus regarding which type of CF is most effective is the fact that they are based on different learning theories, of which there are many. The studies reviewed in this paper mentions several of these, which are directly connected to their research and their findings. Shintani et al. (2014), for example, argued that direct CF and ME draws on two entirely different learning theories, Long’s interactionist theory, and the noticing hypothesis. This is explained by direct CF being input-providing and ME being output-prompting.

Recasts, being output-prompting, also draw on the noticing hypothesis, as the verbal remark makes the receiver aware of their potentially erroneous utterance (Nassaji 2017). In addition, Benson and DeKeyser (2019) compared memory-oriented learners and analytical learners. They argued that the learners acquire language differently, depending on their capability to notice it. Direct feedback might function for the memory-oriented learners, while ME might be beneficial for analytical learners. These arguments suggest that learners of English could improve their grammatical

acquisition in one way or another by simply noticing their error through various types of feedback.

Output-prompting CF also draws on the skill acquisition theory according to Shintani et al. (2014). They argued that declarative knowledge becomes proceduralized through the output that CF elicits. Wagner and Wulf (2016) also used skill acquisition theory (SAT) to explain their findings, that the effect of written indirect CF depends on the target structure. According to them, it becomes effective if the recipient of the feedback already has some degree of declarative knowledge of the

(19)

target structure. This enables the recipient to use this knowledge in other instances of the same target structure, assisting in the proceduralization of the declarative knowledge. However, if there is no declarative knowledge of a target structure, the opportunities of proceduralization lose their effect. Li et al. (2016) also discussed the importance of taking SAT into account. Their study showed that the learners gained explicit knowledge, but almost no implicit knowledge about the target structure, a structure that the learners had little or no prior knowledge of. This indicates that before a language has become automated, declarative knowledge is of importance.

Nassaji’s (2011) research, finding negotiated feedback to be effective, is an example of CF based on the interaction hypothesis. As language is acquired through negotiation of meaning, it is

hypothesized that feedback in the same format would contribute to language acquisition. There is a consensus between many of the studies claiming the relevancy of the interaction hypothesis

regarding CF. For example, Mansourizadeh and Izwan Abdullah (2014) stated that they based their research on the interaction hypothesis, and Li et al. (2016) argued that recast enables the learners to compare their erroneous utterance with the correct target structure; therefore, achieving cognitive acquisition through interaction. These points support the interaction hypothesis part in CF and how it improves grammar acquisition.

The findings of this paper support the use of CF when it comes to L2 acquisition of grammar. According to the Swedish curriculum for the lower secondary school, the English education should include: “Language phenomena such as pronunciation, intonation, grammatical structures, sentence structure, words with different registers, as well as fixed language expressions pupils will encounter in the language” (Skolverket, 2018, p. 37). As grammar is directly related to these aims, the findings of this paper are relevant to the Swedish curriculum, and applicable to teachers of the lower

secondary school. As stated in the introduction of this paper, it is mentioned in the curriculum for the upper secondary school that students should know how to create structure in written and oral communication. However, it should be noted that grammar is not explicitly mentioned at all. Thus, one can argue that the aim for grammatical knowledge in the upper secondary school is implicit rather than explicit knowledge, in other words, being able to communicate fluently without any grammatical errors.

(20)

Furthermore, the studies in this paper make it arguable that there is no specific feedback that functions in all kinds of situations. This is an important factor to consider when using CF in the communicative classroom. In other words, there is no feedback that is specifically better suited for raising the grammar accuracy either orally or written. This might be because there is a high number of variables regarding feedback in an ESL or EFL environment, such as the level of the learners, the target structure, the classroom environment and so on. However, since all the studies agree upon feedback being beneficial for second language acquisition, it is arguable that using feedback in grammar teaching is better than not using it. Another factor to consider is the time-consuming nature of CF. The workload of teachers in Sweden nowadays is a hot topic of discussion (Bergling, 2020). Written direct CF, while potentially effective, does take up a lot of time. Oral ME, on the other hand, can be applied to many students simultaneously, making it less time-consuming.

5. Conclusion

There is a consensus among the authors of the studies reviewed in this paper, that CF generally improves grammar acquisition for ESL and EFL learners. However, the results comparing different types of CF differ to such a degree that any ranking of their effectiveness cannot be established through this paper alone. This can be applied to both short-term and long-term knowledge gain, whether or not the aim is to facilitate explicit or implicit knowledge acquisition for the learners. Furthermore, the findings of this paper suggest that the effectiveness of different types of CF is dependent on the focused target structure. This is manifested in different ways, including the complexity of the target structure in relation to the language proficiency of the learner, or how a particular target structure relates to the L1 of the learner. Moreover, this complexity of CF is something which EFL and ESL teachers should be aware of in order to ascertain efficiency and relevancy in the feedback they provide. It can be argued that this is important for teachers to strive for, as the studies show that CF in general is helpful for grammar acquisition, which in turn is an important part of L2 learning. Additionally, the findings of this paper can be applied to the Swedish classroom because of its connection to the Swedish curricula for lower and upper secondary school. It is a well-known fact that Swedish teachers are under stress and burdened by the heavy workload. Therefore, it is a welcomed finding that this review supports the use of oral ME and its time-efficient attributes.

(21)

There are some caveats to this paper that have been recognized and should therefore be mentioned. The small number of studies reviewed compared to the number of different types of CF mentioned makes the basis of which to generalize any assumptions quite thin. Although this seems to be pervasive for most studies in this area of research, the methodologies of the studies reviewed have differed based on various factors. Some of these factors include the number of participants, the outcome measures, and the length of the studies. Furthermore, it should also be noted that some of the studies took place in a classroom setting while some were conducted in a lab environment. Potentially, this could have affected the results since student-teacher as well as intra-student

relationships can affect how the feedback is received, and thus, its effectiveness. Lastly, not all of the six types of CF, according to Lyster and Ranta (1997) have been researched in the selected studies.

Most of the studies reviewed in this paper concluded that more research is needed within the field; however, the researchers present various suggestions on future research based on their findings. The authors of this paper suggest further research to be focused on the relationship between CF and long-term language gain for L2 learners. Therefore, it could be interesting to see if other factors such as repetition will provide a long-lasting effect on the learners’ grammar accuracy.

(22)

References

Babanoğlu, M. P. & Ağçam, R. (2015). Explicit and Implicit Types of Corrective Feedback in Turkish Primary Education. International Journal of Languages’ Education and Teaching, 3(6), 126– 132. https://doi-org.proxy.mau.se/10.18298/ijlet.502

Benson, S., & DeKeyser, R. (2019). Effects of Written Corrective Feedback and Language Aptitude on Verb Tense Accuracy. Language Teaching Research, 23(6), 702–726.

Bergling, M. (2020). Tiden Räcker Inte Till. Skolvärlden, 10(8), 22-28.

Bitchener, J., & Knoch, U. (2010). Raising the Linguistic Accuracy Level of Advanced L2 Writers with Written Corrective Feedback. Journal of Second Language Writing, 19(4), 207–217. Chen, J., Lin, J., & Jiang, L. (2016). Corrective Feedback in SLA: Theoretical Relevance and

Empirical Research. English Language Teaching, 9(11), 85–94.

Dabaghi, A. (2006). Error Correction: Report on a Study. Language Learning Journal, 34(1), 10–13. DeKeyser, R. (2015). Skill Acquisition Theory. In VanPatten, B. & Williams, J. (Eds.), Theories in

Second Language Acquisition: An Introduction (2nd ed., pp. 94–112). New York: Routledge

EF. (2020). EF English Proficiency Index. Retrieved November 24, 2020, from

https://www.ef.se/epi/regions/europe/sweden/

Ellis, R. (2005). Measuring implicit and explicit knowledge of a second language: A psychometric study. Studies in Second Language Acquisition, 27, 141–172.

Ellis, R. (2006). Current Issues in the Teaching of Grammar: An SLA Perspective. TESOL Quarterly: A Journal for Teachers of English to Speakers of Other Languages and of Standard English as a Second Dialect, 40(1), 83–107.

Ellis, R. Loewen, S. & Erlam, R. (2006). Implicit and Explicit Corrective Feedback and the Acquisition of L2 Grammar. Studies in Second Language Acquisition, 28(2), 339.

Ferris, D.R. (2006). Does error feedback help student writers? New evidence on the short- and long-term effects of written error correction. In: K. Hyland, & F. Hyland (Eds.), Feedback in second language writing: Contexts and issues (pp. 81–104). Cambridge: Cambridge University Press. Guenette, D. (2007). Is feedback pedagogically correct?: Research design issues in studies of

feedback on writing. Journal of Second Language Writing, 16, 40-53

Hattie, J., & Timperley, H. (2007). The Power of Feedback. Review of Educational Research, 77(1), 81– 112. https://doi.org/10.3102/003465430298487

(23)

Ji Hyun Kim. (2016). The Effects of Different Types of Oral Corrective Feedback on the

Development of Explicit/Implicit L2 Knowledge of English Articles. English Teaching, 71(3), 53–77. https://doi-org.proxy.mau.se/10.15858/engtea.71.3.201609.53

Li, S., Zhu, Y., & Ellis, R. (2016). The Effects of the Timing of Corrective Feedback on the Acquisition of a New Linguistic Structure. Modern Language Journal, 100(1), 276–295. Lightbown, P., & Spada, N. (2012). How languages are learned. 4th Ed. Oxford: Oxford University

Press.

Lyster, R. & Ranta, L. (1997). Corrective Feedback and Learner Uptake: Negotiation of Form in Communicative Classrooms. Studies in Second Language Acquisition, 19(1), 37.

Mansourizadeh, K., & Izwan Abdullah, K. (2014). The Effects of Oral and Written Meta-Linguistic Feedback on ESL Students Writing. 3L: Southeast Asian Journal of English Language Studies, 20(2), 117–126. https://doi-org.proxy.mau.se/10.17576/3L-2014-2002-10 Nassaji, H. (2011). Correcting Students’ Written Grammatical Errors: The Effects of Negotiated

versus Nonnegotiated Feedback. Studies in Second Language Learning and Teaching, 1(3), 315– 334.

Nassaji, H. (2017a). Grammar Acquisition. In Loewen, S., & Sato M. (Eds.), The Routledge Handbook

of Instructed Second Language Acquisition (pp. 205-223). New York: Routledge.

Nassaji, H. (2017b). The Effectiveness of Extensive versus Intensive Recasts for Learning L2 Grammar. Modern Language Journal, 101(2), 353–368.

Nayar, P.B. (1997), ESL/EFL Dichotomy Today: Language Politics or Pragmatics?. TESOL Quarterly, 31: 9-37. https://doi-org.proxy.mau.se/10.2307/3587973

Shao, X. (2015). On Written Corrective Feedback in L2 Writing. English Language Teaching, 8(3), 155– 168. English as a Second Dialect, 40(1), 83–107.

Shintani, N., Ellis, R., & Suzuki, W. (2014). Effects of Written Feedback and Revision on Learners’ Accuracy in Using Two English Grammatical Structures. Language Learning, 64(1), 103–131. Skolverket. (2011). Ämne - Engelska [Ämnesplan]. Retrieved 2020-11-18, from

https://www.skolverket.se/download/18.4fc05a3f164131a74181056/1535372297288/Engli sh-swedish-school.pdf

Skolverket. (2018). Curriculum for the compulsory school, preschool class and school-age educare 2011. (Rev. 2018). Stockholm: Skolverket.

(24)

Second-Wilson-Gegeo, K. A., & Nielsen, S. (2003). Language Socialization in SLA. In Doughty, C. J., & Long M. H. (Eds.), The Handbook of Second Language Acquisition (pp. 155-177). Malden: Blackwell Publishing Ltd.

References

Related documents

They divided the 53 students into three groups, the different groups were given: feedback with a sit-down with a teacher for revision and time for clarification, direct written

The students were divided into two groups, where one of the groups was a control group and the other group received direct and indirect electronic feedback (the term

Hur kvinnor upplever det att leva med en psykisk ohälsa i form av nedstämdhet framkom nio teman: ständig ensamhet och relationsförlust; hopplöshet och

Brå konstaterar alltså i sin rapport att avsaknaden av aktuella fakta om brottsligheten bland utrikes födda och deras barn underlättar att myter skapas och befästs och att

The main research question was whether developmental life stage at chronic pain onset is related to pain characteristics and psychosocial outcomes (recovery

Konstruktionen av socionom som identitet, i relation till andra grupper och till utbildningen, kan förstås utifrån social identitetsteori genom att hur socionomstudenterna väljer

We introduce the noncontextuality polytope as a generalization of the locality polytope and apply our method to identify two different tight optimal inequalities for the

a) Hur många kast gjorde Jonas? b) Vilket antal prickar var typvärdet? c) Vilken är medianen?. 7 Hur många tändstickor är det i varje ask om det är lika många