EFL LECTURERS’ PERCEPTION AND PRACTICE OF SCREENCAST FEEDBACK

The current study aims to deepen knowledge on Indonesian English as a foreign language (EFL) lecturers' perception of screencast feedback, their feedback practice, and the consistency between their feedback practice and perception. To investigate the phenomena, five Indonesian university-level EFL lecturers and their EFL students (N=30) were recruited to participate in the study. The researcher employed several data collection techniques including open ended pre-survey, students' essay collection, think-aloud protocol, and semi-structured interviews. The results showed that the EFL teacher participants have positive perception and towards screencast feedback in L2 writing, which influenced their feedback practice. However, the study also discovered two inconsistencies between the teachers' perception of the focus and the type of feedback they provided and their actual practice.


INTRODUCTION
Most L2 writing practitioners (teachers, researchers, and developers) regard feedback as essential to improve students' L2 writing skills. Many believe that feedback may inform students' learning goals, performance, and strategies to improve their learning (Hattie & Timperley, 2007). Written corrective feedback on language may directly or indirectly inform necessary changes to improve students" linguistic accuracy (Biber et al., 2011;Ghosn-Chelala & Al-Chibani, 2018) and feedback on content, organization, and rhetoric may help students in making overall revision (Goldstein, 2004;Hartshorn & Evans, 2015).
Many suggested that feedback provision is difficult task because it is time consuming (e.g., Ferris et al., 1997;Guo, 2015) and findings regarding the effectiveness of written feedback provision or the provision of various types of WCF are inconclusive. Some studies which compared the experimental groups which receive feedback and control group which did not receive feedback reported the positive effects of WCF provision on students" L2 writing (e.g., Bitchener, 2008;Bitchener & Knoch, 2008Chandler, 2003), while some other studies reported that feedback provision is detrimental towards students" L2 writing development (e.g., Truscott, 1996;Truscott & Hsu, 2008).
However, the importance of feedback has been supported by second language acquisition (SLA) theories. Cognitive SLA theorists suggested that feedback provision facilitate learning in form of conscious attention (e.g., Schmidt, 1990), help noticing and realizing learning gaps (e.g., Long, 1996;Swain, 1995). Similarly sociocultural theories argued that feedback facilitate dialogic process and scaffolding within learners' zone of proximal development (ZPD) (e.g., Lantolf, 2006). Thus, L2 writing practitioners focused in finding the best way to provide feedback in L2 writing. The development of audio visual technology and the influence of cognitive and sociocultural theories in second language learning motivated the use of audio visual technology to promote dialogic interaction in feedback. Alternatives include audio feedback (Boswood & Dwyer, 1996;Johanson, 1999) and screencast/video feedback (Bakla, 2020).
Screencast feedback is a version of computer-mediated multimodal video feedback provided by a teacher (human feedback) (see Ware & Warschauer, 2006) by recording the teacher giving feedback on a digital copy of student work using a screen-capture software such as Jing, Screencast-O-Matic, and Google Meet which can record the teacher"s audio/video along with computer screen where student"s writing is seen and commented. The software also record teacher"s mouse movements, annotations, comments, highlighting, direct and indirect error correction, as well as written and verbal commentary (see Bakla, 2020;Ghosn-Chelala & Al-Chibani, 2018). Therefore, it can be positioned between written feedback and face-to-face writing conferences (Lee, 2017, p. 131). Screencast video can be played in a browser (Edwards et al., 2012) or a video player and is usually shared using a web-based tool so the student can view the feedback as often as needed (Bakla, 2020).
However, there have been limited studies focusing teacher as source of feedback, including teacher"s perception, practice, and goals in providing screencast feedback. Moreover, most studies have been conducted in experimental design which many presumed to be less ecologically valid (e.g., Storch, 2010). Some other studies suffered from methodological limitations such as the lack of inter-rater reliability which might undermine the findings (see e.g., Ferris, 2003) and dependence on personal observation techniques rather than meticulous statistical analysis (Hynson, 2012;Séror, 2012). Additionally, more studies should cover new L2 writing context, such as International Program in Indonesia. To date some universities in Indonesia offer international program or class which utilizes English as an instructional language and provide students with international exposure. International program students are expected to have a good command of English, but relatively few studies have investigated international program students" English skills.
In brief, there is still need for studies on teachers" screencast feedback as a computer-mediated human feedback technique in less extensively studied EFL contexts outside the UK and Australia (Silva, 2012), particularly with regard to teacher"s and student"s feedback practices in real classrooms (Lee, 2014;Storch, 2010). Moreover, there is a need to investigate ways to provide more effective feedback to improve the Indonesian EFL students" writing skills which were perceived to be poor (Zubaidi, 2019). In line with this, the present study adopted a mixed method study design to investigate Indonesian EFL lecturers' screencast feedback practice. The questions guiding this mixed method study are as follows: 1. How do Indonesian English as a foreign language (EFL) teachers perceive screencast/video feedback? 2. How do the teachers provide screencast feedback in their EFL class? 3. Is the EFL lecturers" feedback practice consistent with their perception of screencast feedback?

LITERATURE REVIEW
A huge body of research in written feedback in L2 writing has been carried out with most studies reported that written feedback is beneficial to improve students" linguistic accuracy and enhance overall writing skills. With the development of audio visual technology, many L1 and L2 writing researchers search for ways to provide more effective feedback, including using screencast feedback. Depending on the mediating tool used, some main forms of teacher feedback on students" writing are written feedback, audio feedback, screencast, and automated feedback. Teacher"s written feedback refers to teacher"s written comments and corrective feedback to enhance students" language proficiency and analysis (Storch & Wigglesworth, 2010). Audio feedback refers to providing feedback on student's work using audio recording (Fawcett & Oldfield, 2016), and electronic feedback refers to providing feedback using a computer (Ware & Warschauer, 2006), such as computer-mediated communication (also see Section 1 for the definition of screencast feedback).
Much of the literature on screencast feedback in L1 composition has shown that screencast video can enhance teacherstudent relationship (Anson et al., 2016) and provide positive (Warnock, 2008), personal (Anson et al., 2016;Grigoryan, 2017;Sommers, 2013), and explanatory (Thompson & Lee, 2012) feedback conversation. Studies focusing on L2 students" preference highlighted that students receiving screencast feedback prefer screencast feedback and understand the video comments better (Sommers, 2013;Thompson & Lee, 2012;Warnock, 2008), although students may feel awkward or difficult to hear harsh comments (Sommers, 2013). Many studies have also investigated students" perception (Mathieson, 2012;Silva, 2012). For instance, Mathieson (2012) found that health sciences students (N=15) preferred the combination of screencast feedback and written feedback. Similarly, Silva (2012) also noted that engineering students (N=19) considered a combination of written and screencast feedback to be ideal.
The emergence of audio visual and internet technologies motivated the popularity of audio, video, and internet feedback. Previous studies have investigated the effectiveness of screencast feedback in L2 writing (e.g., Ali, 2016;Bakla, 2020;Ducate & Arnold, 2012;Özkul & Ortactepe, 2017;Tseng & Yeh, 2019). Özkul and Ortactepe (2017) investigated two groups of Turkish EFL students in university preparatory classes (N=47) who received different feedback modes: written feedback (control group, n=24) and screencast feedback (experimental group, n=23). The researchers found that students who received screencast feedback could produce better revision than written feedback group. The researchers suggested that screencast feedback generated more correction because it provides richer information, resembling oral conferencing, and containing multimodality (graphic, text, and audio visual). Ali (2016) compared written feedback on local issues (language) with screencast feedback on global issues (content) on Egyptian EFL writing students and found that screencast feedback group outperformed written feedback group in a writing post-test. However, since both feedback targeted different aspects of writing, the findings were hard to compare. Other studies on technology-mediated teacher feedback also have reported that screencast feedback potentially improved students" writing skills (e.g., Zubaidi, 2019).
Studies on students" perception have indicated that L2 students believed that screencast feedback could improve their English speaking proficiency (e.g., Huang & Hung, 2013;Toland et al., 2016;Xu et al., 2017). For example, Xu et al. (2017) examined Chinese EFL students (N=35) who watched five videos, recorded a retelling video, and received screencast feedback through WeChat social media app and found that students with positive attitudes toward screencast feedback gained more speaking confidence.

METHOD
A mixed-method research design was employed to help understand how Indonesian EFL lecturers provided screencast feedback and the processes involved while providing screencast feedback. The main study was carried out in an English II (academic English writing and research) course offered to first-year students of International Program in a private university in Indonesia. English II course run for 16 weeks and each week was eight-hour long. Writing was predominantly taught using a process-oriented writing approach.

Participants
Five Indonesian EFL lecturers (two males and three females, age range between 24 and 33 year olds) teaching an undergraduatelevel EFL course at an International Program in a private university in Indonesia were purposefully selected for the analysis. Each lecturer selected six first-year undergraduate International Program students (N=30; 16 males and 24 females, age range between 18 and 20 year olds) studying in their EFL class to participate in the study. The International Program administered a 200-250 word paragraph writing test to map students' writing skill in the first week. The students' writing were assessed using an analytical scoring rubric adapted from the City University of New York Assessment Test in Writing (CATW) (2012). All five teachers held master degrees in English education or Literature. The teachers had taught in the university for a minimum of two years. The assessment of students" paragraph writing indicated that the students have lowerintermediate to intermediate level writing ability.

Data Collection
The data were collected using quantitative and qualitative techniques. The quantitative data collection was aimed to evaluate the teachers" perception of feedback screencast using open ended presurvey and teachers" written feedback on students" drafts. The presurvey was used to obtain in depth information on the participants" demography (age, gender, technology and Internet access) and teachers" perception of screencast feedback.
The qualitative data collection was intended to allow a deeper insight into the participants" screencast feedback practices and their thinking processes while providing feedback. The researcher employed think-aloud protocol procedures (screen recordings) and semi-structured interview. At the beginning of the study, the researcher invited the lecturers to participate in a two-hour TAP training session to discuss, simulate, and exercise the TAP procedures. However, the researcher did not provide training on feedback provision as to minimize the researcher"s interference.

Feedback procedures
The essay writing processes were parts of assessment of English II course which the students were taking in the semester. After the course introduction and the pre-test (week 1), students learned the basics of writing academic argumentative paragraph and wrote a multi-draft paragraph (Week 2 to Week 7) and a multi-draft essay (week 9-15). Students wrote a multi-draft argumentative essay, submit the drafts to their lecturers via e-mail (week 10, week 12, and week 14), and then revise and resubmit the drafts. The teachers received the essay drafts, and then provided three sessions of screencast feedback while providing written feedback (on students" hard copies or electronically) using Google Hangouts (Google Meet) application. The screencast videos were uploaded automatically into YouTube platform (https://www.youtube.com/) when the lecturer finished providing feedback, as displayed in Figure 1. Afterwards, the lecturers shared the URL link of screencast video recording and returned the essay draft via e-mail. Students were instructed to repeat the drafting phases (drafting-screencast feedback-revision cycle which students should follow) thrice and the duration for each drafting phase was two weeks. Feedback was provided by the authentic EFL lecturers who were familiar with the students" learning styles and progress to establish the authenticity of this study.
The lecturer participants were instructed to record their screen while checking and providing feedback on students' multi-draft essays through a think-aloud procedure (TAP) technique, and then to share their recording with their students. At the end of the semester, the EFL lecturers were interviewed to gain more detailed, specific qualitative information and to triangulate the TAP data (Creswell, 2009).

Figure 1: Screencast feedback recorded using Google Meet and uploaded into YouTube platform
The data collected consisted of students' essay drafts (n=81 drafts), teachers' survey responses (N=5), teacher screencast feedback recordings (N=70, ranging from 2-10 minutes in length, altogether totalling 340 minutes after trimming), teacher interview recordings (totalling 225 minutes). Nine essay drafts and screencast feedback recordings were discarded because the observation showed that the student writers had not revised their drafts. 11 other recordings were discarded because the screencast recordings were not related with the selected student participants.

Data Analysis
The researcher employed an analytical scoring procedure to compute descriptive statistics of teachers" feedback. The number of screencast feedback the teachers provided were calculated. Following Hyland (1997Hyland ( , 1998, Lee (2008Lee ( , 2009) and Zubaidi (2019), each feedback instance which constituted a meaningful unit was called a feedback point, which contain information of the focus and type of feedback. The focus of feedback were then classified into micro level issues (consisting of language use [grammar, lexis and style, morphology, and syntax] and mechanics and referencing) and macro level issues (consisting of content [clarity and ambiguity, idea and logic, lack of elaboration, and terms] and text structure [conciseness, paragraph organization, sentence flow, unity]). The type of teacher"s feedback was categorized into direct and indirect feedback. To triangulate the coding framework, another coder identified and calculated each feedback in students" drafts to calculate the interrater reliability using Cohen"s Kappa. The overall interrater reliability was found to be K= 0.90 which indicates that the raters were in high level agreement. Both the researcher and the intercoder also regularly met to discuss the framework.
The semi-structured interview recordings (15 hours) were transcribed using Easytranscript v. 2.50 software and were then uploaded into Nvivo v. 12 software. Nvivo software was used to organise and code the data using a priori coding (Ferris, 2006;Storch & Tapper, 2009) and emergent coding from the analysis and analyze the data qualitatively.

Quantitative results
The number of feedback points identified in the screencast feedback was 1062 and the average number of feedback points per draft reviewed in screencast recordings was 15.17. Analysis indicated that the lecturers focused on both macro level and micro level issues of writing, as Table 1 displays. Most feedback points in the first feedback session focused on macro level issues, while most feedback points in the second and third feedback sessions focused on micro level issues.
In addition, Table 1 shows that the frequency and percentage of the teachers" screencast feedback increased significantly over three feedback sessions. More specifically, the number of screencast feedback on micro level issues of writing (language use and mechanics and referencing style) increased from 20.94% to 91.02%, while the number of screencast feedback on content and organization decreased drastically from 79.05% in the first feedback session to just 8.96% in the third feedback session. In addition, results showed that the EFL lecturers provided both direct and indirect feedback on all types of writing issues, as displayed in Table 2. Direct feedback was given by explicitly correcting students" errors and indirect feedback was provided by signalling the occurrence, location, or a way to revise the writing problems. The analysis showed that the teachers provided a higher frequency of direct feedback (56.59%) than indirect feedback (43.40%). Of the three feedback sessions, indirect feedback was used as the main strategy in the first feedback session (117:31), while direct feedback was used as the main one in the second and third feedback sessions, respectively (162:139) and (408:205).
Based on the teachers" focus of feedback, the EFL lecturers consistently used direct feedback as the main feedback strategy to deal with language use issues. On the other hand, indirect feedback was consistently used as the main strategy to deal with content and organization and rhetoric issues over three feedback sessions. However, the percentage of direct feedback used to correct content and organization and rhetoric problems increased over three feedback sessions. As Table 2 displays, the percentage of direct feedback increased from 8.57% to 25% in content issues and the percentage of indirect feedback increased from 10.86% to 45.71% in organization and rhetoric. The teachers provided direct feedback by identifying the writing problem, explaining the reason, and suggesting the correct form. In the following Example (4.1), T2 checked S1"s second draft and correct two tense and verb form errors.  [Reading] time to have preparation of the youth ... due to this demographic bonus." (T2 -S1 -draft 2) Indirect feedback was provided by means of codes, highlighting, coloring, and comments in the document margin. In excerpt (4.2), T3 identified two relatively similar unclear expression/word choice problems. For both problems, she wrote " what do you mean by this one?" to request information regarding the student"s word choices: occupation"s chairs and company"s chair.

Qualitative results
The qualitative analysis of the data produced three main themes, producing valuable information regarding the EFL lecturers" perception of screencast feedback and the nature of their screencast feedback.

EFL lecturers' perception of screencast feedback
Four teachers preferred to provide screencast feedback or a combination of written and screencast feedback, while another teacher preferred written feedback only. Most lecturers who preferred screencast feedback or the combination of written and screencast feedback modes suggested that lecturers" explanation in screencast feedback help students to easily comprehend the feedback, so students can revise the draft more efficiently. The EFL lecturers also suggested that both screencast and written feedback modes complement each other. T2 stated in the following excerpt (4.3). (4.3) "Giving feedback using video is beneficial because I can provide both written feedback and relevant and audio visual explanation using screencast. If students cannot understand my points, they can replay the video as many times as they need. They can increase or decrease the video speed too. However, to help them learning better, I think we still need to send back their essay that we have commented so they get more guideline and can make sense better by processing information in the text. I can also provide more detailed explanation to students. It is easier to encourage students and convince them regarding writing issues in their draft by speaking via video; it is more efficient and clearer than when I gave comments in writing." (T2) Another teacher added that lecturers could improve students" comprehension by speaking more clearly, as shown by T1 (excerpt 4.4). (4.4) "As long as we speak clearly and slowly, not too fast, it will be easier for students to understand our video feedback. Moreover, when we record our feedback, students can hear our intonation, stress, so they know which feedback is primary and not. It is also a kind of exercise for students to comprehend feedback in the video." One EFL lecturer preferred providing written feedback because written feedback contains richer information because feedback is organized according to the structure of the essay; students may find such structure to be useful guide of revision. He admitted that he might miss some information when explaining certain feedback points or became repetitive, so some students might get confused.
(4.5) "I think giving written feedback may help students revise the draft with easy because when students see the written feedback on their draft, they see it in an organized way according to the structure of their essay. When I video record my feedback, sometimes I easily get distracted or talk redundant things, jumping up and down between lines. When students see it in a video, they may get confused easily." (T4) The lecturers suggested that screencast feedback seems to contain many potentials to use for EFL lecturers including being direct, rich in information. Although not all lecturers preferred providing screencast feedback, the pre-survey indicated that all lecturers considered using video technology for other teaching activities, such as explaining the subject, interacting with students, and posting information in LMS.

The EFL lecturers' feedback practice
All lecturers reported that they provided comprehensive feedback on all aspects of writing. The lecturers believed that comprehensive feedback is useful because their students" backgrounds were non-English major. All lecturers stated that they prioritized the content of writing because it is more important than other aspects of writing and is one of primary criteria of assessment in the course. Four teachers also reported that they provided more feedback on content.
(4.6) "I provided more feedback on content because content is the most important aspect in writing an essay. The goal of this course is that students can construct ideas in a logical order (. . . ) so we emphasize the content. Grammar is not very important." (T1) The teachers realized that they provided more direct feedback in some feedback sessions and for certain students although they also believed that indirect feedback may help students to be active participants in their study by self-revising their writing problems. T1 suggested that the use of more direct error correction was unavoidable because they had to ensure that students could produce a good piece of writing in a short time.

DISCUSSION
The qualitative data showed that the EFL teacher participants have positive perception and attitudes towards screencast feedback in L2 writing. The lecturers stated that screencast feedback is easier to understand compared to written or audio feedback only. Students who hear the lecturer"s explanation on feedback and simultaneously see the relevant part of writing are expected to understand the feedback better and be able to revise their drafts more efficiently. Another benefit of screencast feedback is that the feedback recording can be played repeatedly (playability) to increase students" understanding of feedback. The findings were in line with Séror"s (2012) reflections on his screencast feedback. Séror wrote his appreciation of the options provided by screencast feedback, including ability to communicate more flexibly with dynamic visual explanation, conversational and personalized feedback.
One problem highlighted in the study is the potential confusion for students due to lecturer"s impromptu feedback. Due to being new to the technology, the lecturers provided impromptu feedback which they themselves believed to be disorganized. In addition, the layout of screencast feedback recording was different with conventional hard copies of essay draft and so it may cause confusion for students. Some solutions offered include requesting the lecturers to give brief overview at the start of the screencasts (Edwards et al., 2012), to plan/structure feedback, and to use a preprepared feedback template (e.g., JISC, 2010) .
The researchers' perception of the efficacy and benefits of screencast feedback was a dominant factor influencing their feedback practices. All lecturer participants in this study provided screencast feedback. The quantitative data indicated that the lecturers provided significantly higher number of feedback on micro level issues of writing. The result corroborates the findings of many studies on feedback in L1 composition (e.g., Stern & Solomon, 2006) and in L2 writing (Diab, 2005;Hyland, 2003;Zubaidi, 2019). For example, Stern and Solomon (2006) analyzed teachers" comments on 598 papers and found that tutors focused more on grammar, mechanics, and style (form) than on idea and content. Similarly, Hyland (2003) examined two New Zealand university-level ESL teachers' feedback and found that the teachers provided more feedback on forms than on content. The quantitative and qualitative data indicated that the lecturers provided comprehensive feedback on all aspects of writing, including language use, mechanics and referencing style, content, and organisation and rhetoric. This finding accords with many previous studies on screencast feedback in L2 writing (e.g., Ali, 2016;Bakla, 2020;Ducate & Arnold, 2012;Özkul & Ortactepe, 2017;Tseng & Yeh, 2019).
To produce more robust findings, the researcher compared the lecturers" perceptions and their actual practice. Analysis showed two inconsistencies between teachers" perception and practice. First, the lecturers preferred providing feedback on content over other aspects of writing, but the quantitative data showed that the teachers provided more feedback on micro level issues (language use and mechanics and referencing). The lecturers suggested that time limitation encouraged them to provide more feedback on forms. Second, the lecturers preferred indirect feedback, but the students" essays indicated that the lecturers provided a higher number of direct feedback especially in the third feedback sessions. The teachers emphasized that indirect feedback was beneficial for their students, but they felt they need to provide direct feedback to help students produce a good piece of writing.
Some previous written feedback studies in L2 writing (Lee, 2008(Lee, , 2009) have reported such inconsistencies. For example, Lee (2008Lee ( , 2009 investigated Hong Kong high school ESL teachers using interview and teachers" comments on students" essay. She found that that the teachers believed that a good writing mainly depended on development of ideas and organization followed by accuracy. However, students" essays indicated that from 5,353 feedback points the teachers provided, 94.1 per cent of feedback addressed form and only 5.9 per cent of feedback addressed content, organization, and other aspects.

CONCLUSION
The present study aimed at gaining insights into the understanding of screencast feedback in university-level EFL course, particularly the EFL lecturers" perceptions and screencast feedback practice. The results showed that almost all teachers preferred providing screencast feedback due to its ease of use and efficiency. Further studies on the effectiveness of screencast feedback in comparison with other feedback modes should be conducted to see if screencast feedback could help students improving their English writing skills better. Some previous studies have shown support for screencast feedback (e.g., Ali, 2016;Bakla, 2020;Ducate & Arnold, 2012;Özkul & Ortactepe, 2017;Tseng & Yeh, 2019).
Some lecturer participants in this study noted that a number of students revised their draft independently and not based on the lecturers' screencast feedback. These students did not check the uploaded screencast videos due to limited internet quota at their home. It is a common problem for many Indonesian students to have limited internet access. They could gain more access internet freely at the university, but they sometimes missed or forgot to check or download the feedback. Moreover, few other students still preferred written feedback compared to screencast feedback. The future lecturers who plan to employ screencast feedback in their courses should note that despite screencast feedback may be easier to prepare and more efficient than written feedback, the instructors should deal with the distribution of video recordings and ensure that students could benefit from the screencast feedback.
The current study also looked at teacher"s perception of screencast feedback. It should be noted that a teacher"s perception and beliefs about students" needs may mismatch students" actual needs. Thus, future studies should investigate teachers" and students" perception and belief regarding screencast feedback in order to see if the perception and practice match with students" needs.
This study looked at a single essay writing assignment over six weeks with the lecturers only providing one feedback mode. In addition, there is also another possibility that the lecturers and students might show a novelty effect, that is giving increased positive response toward a new media/technology (Clark, 1983;Cunningham, 2019). Future studies should consider more longitudinal studies involving other types of writing assignments, comparing various feedback modes, or employing screencast feedback intensively in the learning process in order to gain deeper understanding and mitigate the possible novelty effect. Beyond these, studies investigating the effectiveness of various technology-mediated feedback modes and tools, as well as the use of feedback by lecturers and students from different contexts and backgrounds should be encouraged to expand our understanding of feedback in second language writing.