EXPLORING UNDERGRADUATE EFL LEARNERS’ BELIEFS ABOUT WCF AND THEIR IMPACT ON IMPROVING WRITING SKILLS IN PAKISTANI UNIVERSITIES

Purpose of the study: This study aims to explore students’ beliefs, which play a vital role in learning a foreign language. The present study investigated this important factor on undergraduate EFL students at Khwaja Fareed University of Engineering & IT, Rhaimyar Khan, Pakistan. Methodology: The study is quasi-experimental and used questionnaires and interviews. Students were divided into two groups: treatment and control groups. Writing accuracy for the pre and two post-delayed tests were checked and statistically analyzed on SPSS. Different repeated measures ANOVA tests were used. Main Findings: Main findings of this study revealed that there were marginal differences in the beliefs of the students from urban and rural areas. Findings also showed that types of the written CF which are the most effective were different according to their beliefs. Besides, beliefs about written CF were also found to impact students' writing accuracy significantly. Students in the control group showed significant accuracy in writing compared to the students in the control group. Applications of this study: The present study will be useful for constructing university students’ existing beliefs about written CF and practices with respect to learning English as a foreign language. Novelty/Originality of this study: This study's most significant research contribution regarding language learning beliefs is that it probes the extent to which differences of beliefs may impact students' development of writing accuracy after they received written CF. As no earlier studies could investigate this issue; hence, this study has attempted to fill an important research gap.


INTRODUCTION
This is one of the most instructional practices in L2 learning classrooms to provide Written Corrective Feedback (henceforth WCF), which most EFL instructors view as a part of their jobs. This is also what their students expect. Nevertheless, the role of WCF as an instructional instrument to facilitate foreign language (henceforth FL) learning has been unclear generally. Moreover, it also investigates the extent to which correction in grammatical comprehension could help EFL learners and notice their errors in writing whether or not that leads to more accuracy in subsequent drafts.
The opening theme of this current pedagogical debate regarding the significance of WCF stemmed from Truscott's (1996) claim that correction of grammatical comprehension was not only ineffective but harmful also, and therefore, might be abandoned. Responding to this argument and defending the case of grammar correction, Rich et al. (2017) contended that claims made by Truscott were impulsive. Ferris (2015) put forth the fast-emerging research evidence to support the effectiveness of WCF. He also elaborated that EFL students require supplementary adjusted intervention from their teachers to compensate for their limitations. Students also need to learn some strategies to assist them in finding out corrections and preventing errors in EFL writing.
In spite of the several research studies conducted over the last two decades, some important issues and criticism on the studies of the effects of WCF can be highlighted, which are: (1) the incompetence to provide evidence in relation to language that EFL learners apply the information obtained from WCF on the earlier written prompts to new writing production and (2) the deficiency of WCF research studies outlined within the Second Language Acquisition (SLA) theory (Shabir, 2017).
Although there is increasing indication of the significant relationship between WCF and development in EFL writing over time, that the basis of research studies has been restricted so far to examine the effectiveness of WCF (i.e., students are provided WCF on one or two types of errors) with particular linguistic categories (e.g., the use of verbs or articles). As Chen et al. (2016) elucidate that the problem with this pedagogical method is that focusing on one or two grammar or linguistic categories might lead EFL learners to be consciously involved in monitoring the use of the targeted linguistic features while overlooking the others. From this perspective, providing focused WCF received criticisms because it did On the other hand, some research studies have probed the impacts of comprehensive WCF that is the most extensively applied in the teaching grammar comprehension in EFL classrooms in which EFL students are provided error corrections on several error types simultaneously. In addition, the research studies have examined the usefulness of comprehensive WCF on newly written drafts, which have evaluated the outcomes of their various treatments by common measures of accuracy, e.g., percentage/ratio of error-free words, error-free sentences and clauses, and error rate in the broader categories such as 'non-grammatical' and 'grammatical' (Han, 2017).
The findings and deductions of these empirical studies revealed that the learners in the treatment groups acquired (or did not acquire) knowledge obtained from WCF that was provided on previous writing tasks (i.e., pre-tests) to the second writing task (i.e., post-tests). Hence, if there is no significant difference in the average error rates between two groups (i.e., treatment group and control group) in the learners' second texts, it is then presumed that the learners of the treatment group used no knowledge gained from the WCF. Nonetheless, as Gries & Deshors (2015) argue, several illustrations in which errors in the subsequent written drafts stand in no relation with the previously corrected errors. So, the application of such metrics provides little evidence on the effects of WCF in subsequent writings. Likewise, the researcher himself contends that global methods of providing accuracy may also run the risk of complicating the cases in which learning occurrs.
This question is vital as finding the evidence of how WCF can affect the specific linguistic features may provide more insight into assessing the effectiveness and pedagogical significance of a certain WCF treatment by addressing Truscott (1996) claims that no WCF is useful to help the learners acquiring lexical and syntactic knowledge. In doing so, the WCF research study requires to be guided by the SLA theory because Rizwan and Akhtar (2016) argue that some theories can be invoked to address the efficacy or lack regarding errors' correction (p. 376). Similarly, within the perspective of learning EFL writing, wherein writing is perceived as a tool for EFL learning, SLA-based research studies on grammar correction in writing are significant to obtain better understandings of the role of writing skill and learning grammar for L2 efficacy.
Comparing to what may happen in corrective feedback, connections between theory and research have remained common practices in any corrective feedback investigations for a long (e.g., Ellis, 2008). Only a few studies conducted on longitudinal design (Holec, 2010;Rummel, 2014) were designed to investigate the efficacy of comprehensive WCF within the SLA approach. These studies were carried out using SAT (Skill Acquisition Theory) principles to frame for providing WCF in EFL classrooms. According to the findings of these studies as mentioned before, WCF should give a reflection of "what is the most needed by an individual learner" and "what the learner shows in producing writing," and both the writing activities and the WCF should be "timely, meaningful, constant and easy to be managed" (Cephe & Yalcin, 2015).
Likewise, past research studies within both sociocultural and interactionists' perspectives established a series of thoughts and interest in the fertile discourse of the WCF studies. They allowed conducting more forceful, empirically-based investigations (Ellis, 2012). Preponderantly, the constructs of uptakes and noticing attracted a good deal of attention in the WCF research (Lyster & Ranta, 2017) as they might establish L2 learners' capabilities to improve the current knowledge and later learning. 'Noticing' refers to denote "the conscious awareness of surface-level linguistic phenomena (Belaid & Murray, 2015). 'Uptake' is learners' response to the teacher's feedback provided on a linguistic feature and is considered effective when the learners use those features correctly or understand them (Belaid & Murray, 2015;Hassan, 2020;Hassan, 2021). Though, the correct use and understanding of L2 forms, as Panova and lyster (2002) denoted, did not define that the features are learned; instead, they claimed that it was necessary to look into whether learners were able to produce the corrections on their later written drafts. But, investigations on written CF that have to see these concepts are still very confined. These researches (e.g., Bitchener & Knoch, 2012; Saeed, 2015; Kerz et al., 2017) especially found the processing of written CF on learners' uptakes and noticing during texts' revision by observing language features without considering students' beliefs about written CF. This would offer a research gap in the current research. Hence, this research first probed EFL students' beliefs about extensive written CF and then analyzed the impact of comprehensive written CF on later revisions of written texts to assist learners in obtaining writing accuracy. This would permit the researcher to examine students' writing accuracy over time.

RESEARCH QUESTIONS
The following research question is addressed in the study: To what extent do the beliefs about written corrective feedback have an impact on students' uptake and retention of linguistic features in two different contexts?

OBJECTIVE OF THE CURRENT STUDY
The research objective in the current study is to: To find out the impact of beliefs about written corrective feedback on students' uptake and retention of linguistic features in two different contexts.

METHODOLOGY
The study is quasi-experimental, which used instruments, questionnaires, and interviews to collect data from EFL students. The population was 163 undergraduate EFL students at Khwaja Fareed UEIT, Rahimyar khan. The sample population was selected by applying the purposive sampling technique. Students were divided into two groups: treatment and control groups. The duration of the study was the whole semester. Therefore, data were collected over a different period to see the accuracy of writing in the pre and two post-delayed tests, and writing accuracy for the pre and two postdelayed tests were checked and statistically analyzed on SPSS.

RESULTS AND DISCUSSION
The students' feedback preferences and beliefs are shown below: Which type of written CF will you prefer to receive in the future? 49 01 0

Source: Authors
During the interviews, when it was asked form the students what they would feel if they were not given any written CF, they responded that they could not improve their writing; however, two of the students said that they would request their teacher and tell, she would have perhaps forgotten to provide feedback on their writing (RS2 & RS20). Bio-data of the rural students revealed that 38 out of 50 students were studied at government higher secondary schools before joining the university as students while other students got education from private higher secondary institutions except one who was made to receive education in a well-reputed private institution. Furthermore, in the interview, one student only agreed on indirect CF. Teachers in a private college (Punjab College and Aspire College) used indirect feedback. He added that students from English medium school systems got a better learning environment, but he preferred to receive indirect CF.
The other students said that in government higher secondary institutions, EFL teachers used mixed types of written CF due to which they could not obtain accuracy in their writing in pre-education. However, they preferred to receive direct CF in the research process because it seemed to them effective.

Source: Authors
The table above depicts no significant links between time and types of corrective feedback given to EFL students. Nevertheless, a significant difference is noted about time and within the subjects. The impact is also observed. To this, one-way ANOVAs tests were carried which reveals that three groups depict significant writing accuracy over time (direct CF group, p value= .00, indirect CF group, p-value = .05 and metalinguistic group, p value= .03) over the time.
As opposed to this, the control group showed no such consistency in writing accuracy (p value= .74). It is also depicted that, although EFL learners who got direct CF, they first depicted no writing accuracy in the immediate post-test. But this appeared no significant at time 2 (post-test). The Control group could depict writing improvement in test 3 (first delayed post-test). Whereas indirect CF treatment and metalinguistic CF groups revealed consistency of significant increase in the accuracy at time 2. They also continued improving accuracy significantly at time 3.
The statistical results also depicted that three treatment groups were observed having a decline in the writing accuracy, which, in fact, was insignificant from time 3 to 4. But, they were able to obtain a higher degree of writing accuracy than they depicted at the beginning of treatment significantly. While control group, in the beginning, started depicting a higher degree of writing accuracy but later could show insignificant variations over the different period of time. Nevertheless, the direct CF treatment group also depicted observable accuracy in the writing between the pre-test and the immediate post-test. Besides, rural EFL learners in the direct CF group also revealed a significant writing accuracy between the immediate post-test and 1 st delayed post-tests. This development was consistent with the 2 nd delayed posttests. Although a slight decline was noted in the indirect treatment group on their 2 nd delayed post-test yet both groups still depicted observable distinction in their accuracy right from their pre-tests.
ANOVA tests were conducted to compare the treatment and control groups' scores in one pre-test, post-tests, and two delayed post-tests. As one-way ANOVA depicted insignificant distinctions between two groups, F (3, 22.59) =.730, p=.73; therefore, a two-way ANOVA measure was carried out. The score of the tests was placed as the dependent variable of time and two written corrective feedback as independent variables.

Findings of rural university students
For the rural student-participants too, the same process was adopted which was applied for the urban students. Data collected from students' surveys and interview questions showed that out of 50 rural students, 49 students preferred to receive direct written CF, only one student preferred to receive indirect CF, and no student preferred to receive

Source: Authors
Out of 50 rural students who got written CF according to their preferences, 34 students were able to eliminate their alltargeted errors on their 2 nd delayed post-test, as the above table shows. Moreover, the 16 students were unable to eradicate all their targeted errors on their last delayed post-test. These are similar findings to the urban students, which is an evidence that students' beliefs have an impact on the uptake and retention of the written CF for improving the targeted linguistic features. However, in contrast to the urban students, less number of rural students in the control group were able to eliminate their targeted errors.
The present study's findings have significant theoretical implications pertaining to the university students' developmental levels and illustrate how these levels may impact the effectiveness of written CF provided by their teachers. When selecting linguistic and grammatical features to provide written feedback, teachers usually select the ones which may hinder the students' ability to communicate (Baker & Burri, 2016). Admissions such as these can support Truscott's (1996) assertion that the ways a teacher provides corrective feedback on L2 learners' writing are not useful. It is because teachers often do not keep in mind the students' developmental levels in the selection of grammatical features. Belaid and Murray (2015) presented their argument that in order to make grammar instruction effective, it should happen when a student is at the stage where he could catch grammatical points quite naturally. In the information processing models, the salience of the written CF is significant. It means that a student's current grammatical knowledge could impact his ability to pay attention to the provided feedback (Bitchener et al., 2014).
Moreover, Krashen's (1985) "i+1" (input hypothesis) and Vygotsky's (1978) ZPD (Zone of Proximal Development) assert that students' developmental levels should be considered when providing any type of corrective feedback. As these theories say something about the significance of considering the students' levels, it looks like a quite relevant issue. However, if the teacher merely provides feedback on the issues that cause problems, he cannot provide salient CF to his students. However, the ability to convey ideas effectively is very significant to students' L2 writing. After consideration of the theory, it becomes clear that simply targeting students' issues that interfere with the effective communication of ideas cannot be the better strategy for providing any feedback. The current study provides theoretical implication here that a teacher must consider students' contextual factors to understand their developmental level so that effective communication may occur for providing preferred CF. According to Plonsky and Mills (2016), even if CF is tailored towards students' errors, it may not be useful and effective if it is not up to their developmental levels.
The present study's findings also support the current empirical research about written CF and students' language learning beliefs. In this respect, this research corroborates prior studies which investigated that the focused written CF cannot improve the longer-lasting acquisition of salient grammatical and linguistic features (Bitchener & Knoch, 2012;Cohen, 2015). Besides, this study added to the current flow of research by focusing on comprehensive written CF, including local and global errors, which have received limited attention. This study also confirms the findings of the Rummel and Bitchener (2015) study by revealing that written CF on local and global errors may help students acquire linguistic accuracy in some students. The study also endeavored to throw light on which type of written CF is most useful and effective. Regarding this question, no difference in the efficacy of three types of CF (direct, indirect & metalinguistic) was found among the treatment groups in urban; however, direct written CF was proved to be the most effective and useful for the students' linguistic accuracy among rural students.
With regard to beliefs, this study lends great support to the earlier research, which investigated differences in the beliefs among the students belonging to two different contexts. It adds to existing knowledge by investigating similarities and differences between university students from urban and rural areas in Pakistan. By employing multiple data collection methods and cases, a complete picture of how university students' previous learning environment and educational experiences may affect their existing beliefs is presented. This study's most significant research contribution regarding language learning beliefs is that it probes the extent to which differences of beliefs may impact students' development of writing accuracy after they receive written CF. As no earlier studies could investigate this issue; hence, this study has attempted to fill the important research gap. The current study has a methodological contribution toward selecting data collection methods to investigate the written CF. Previous studies have mostly looked at this issue of providing written CF to improve the writing accuracy of EFL students by collecting writing samples; however, the current study used questionnaires and semi-structured interviews along with collecting writing samples to present the holistic picture of the ways by which beliefs can have an impact on Pakistani EFL students' uptake and retention of written CF (Hasan & Qureshi, 2020;Hasan, 2020a).

RECOMMENDATIONS
The current study has highlighted several issues that require further investigations. First of all, EFL students from other universities in Pakistan need to be further investigated to determine the extent to which the findings of this research could be generalized. Besides, gender is another possible variable that may explore further differences in foreign language learning beliefs, which may also impact written CF. The current study was carried out on EFL students' beliefs towards written CF, but future researchers may also add EFL teachers of universities in Pakistan to investigate what impact teachers' beliefs may have on students' choice toward written CF. In addition to this, similarities and differences of belief between teachers and students can also be important factors to explore.

CONCLUSION
The study concludes that language learning and its use are integrated into a socially intervened world; hence, those social aspects are required to be examined as a part of the similar cognitive process that underlies L2 learning and development. The present study has revealed how previous social, contextual, and educational experiences may be effective for constructing students' existing beliefs about written CF and practices with respect to learning English as a foreign language. By considering this, future researchers may continue to investigate the environmental factors (present classroom environment, educational backgrounds, etc.) and social factors (cultural expectations, identity, etc.). This may also affect the extent to which provided CF is effective. Moreover, individual factors like personality, age, mental health, etc., should also be probed to examine if they have any impact on students' uptake and retention of written CF.

LIMITATIONS OF THIS STUDY
The study has some limitations. The participants of this study come from two different contexts: urban and rural areas of Pakistan. They have different languages and cultures. Therefore, their beliefs may have some impacts on participants' ways of responding and using WCF. Hence, the study has the first limitation pertaining to generalizability which means that the current study cannot be generalized to the population in other contexts. This has been done because the researcher seeks to provide deeper understandings of the two areas (urban and rural). So, other research is also required to be conducted to compare the inter-university. EFL students with diverse cultures within Pakistan to find out if there is a similar correlation between students' beliefs towards WCF and the impact of those beliefs on students' uptake and retention in EFL writing.
Second, participants of this study are adult learners of advanced level (enrolled for the undergraduate program at FUEIT Yar Khan, Pakistan), which again is a limitation for its generalization because the students with different level and age groups are required to be investigated to explore their beliefs about WCF similar or different and whether other factors too cause any change in students' beliefs over the time. Thirdly, as this study seeks to investigate adult learners within the same university department, the sample size is not large. Again, this study cannot be generalized to a larger sample, and including the students of other departments studying English composition and comprehension may provide different findings and results.