Unraveling the Lecturer’s feedback quality and the Students’ engagement in Online Learning

*Corresponding Author: ahmadsugianto@upi.edu Abstract: This study aimed to reveal the lecturer’s feedback quality intertwined with students’ engagement in online learning through a WhatsApp group. Also, it unraveled the follow up actions conducted by the students regarding the feedback. 24 graduate students studying in one of the universities in Bandung city involved. A mixed-method, an explanatory sequential design, was used. Utilizing questionnaires, the study reveals that with the mean of 82.38 and standard deviation of 6.51, the lecturer’s feedback was deemed good with some qualities emanated through the in-depth interviews comprising ‘timely’, detail and relevant to the materials learned, facilitative, supportive, and objective. Meanwhile, with the mean of 73.43 and standard deviation of 10.92, the students’ engagement is considered good, yet challenging issues were found concerning the learner autonomy and the students’ motivation. Moreover, it was found that with the 95% level of confidence the p value obtained was higher than the 95% level of significance, i.e. p>.05=.98>.05, hence the regression model between the two variables was considered linear. However, at the level of 95% level of significance, it was found that there was no significant relationship between the two variables, i.e. .097 >.050. Moreover, consulting with friends and the lecturer and reading relevant materials constituted the alternatives the students did to handle issues concerning the feedback.


INTRODUCTION
Feedback constitutes one of the essential and extricable parts in the teaching and learning process, and particularly in terms of the assessment conducted. In order that the all the teaching and learning process as well as the assessment conducted meet their objectives, the quality of feedback given should be taken into account by the teacher or lecturer. Hattie and Timperley (2007) assert that feedback is delineated as information vis-à-vis with individual performance and understanding. Thus, it is regarded at the heart of assessment since one's performance can be illuminated through it.
Concerning the use of feedback in AFL, some judicious considerations should be taken into account. Brown (2001) emanates that it is a teacher's task to provide learners with 'positive affective' and 'appropriate' feedback. Such a condition is important since as students are learning, they are not only involving their thoughts but also their emotion (Brown, 2000).
By having affective and appropriate feedback, it is expected that the students will be motivated to learn (Raihany, 2014). Also, it was found that having good quality feedback significantly relates to the way they engage in assessment tasks given (Bahati et al., 2019).
Moreover, Gibbs and Simpson (2005) point out some considerations required to be taken into account to generate effective feedback on learning.
Here are some of the utmost ones: 1) sufficiency which has to do with specific, regular, and useful; 2) students' performance in focus instead of their characteristics; 3) timing, i.e. needed in the right time; 4) suitability with the objective of the assignment; 5) usefulness, i.e. giving information about what to do.
Meanwhile, Newmann, Wehlage, and Lamborn in Fred (1993) emanate engagement as 'the student psychological investment and endeavour' concerning understanding, knowledge mastery, and skills. Based on Newmann, Wehlage, and Lamborn's notion, it can be indicated that engagement constitutes concomitant of psychological aspect. Likewise, Dixson (2015) asserts that engagement is constructed from psychological components such as attitudes, thoughts, and behaviours with another additional component, i.e. communication. Meanwhile, Kuh (2003) stipulates that student engagement is defined as all the endeavors including time and energy students have to do the activities either in or out the classroom. By contrast, Finn and Rock, (1997) argue that engagement has to do with activities manifested in a program through which students may prevent them from risks such as dropout from school. Concerning the issue related to dropout, Finn and Voelkl (1993) bear witness that characteristics that should be taken into account by school that underpin student engagement particularly those who are at risk comprise structural environment and regulatory environment. In this regard, the former delineates the school population consisting of the size and racial ethnic aspect, whereas the latter has to do with the statutory obligation or system at school (p. 252).
In the meantime, regarding online learning, Lee et al. (2019) reveal six factors influencing the student engagement in online learning comprising as follows: 1) psychological motivation regarding psychological aspect of learning such as enjoyment of learning, stimulation of interest, usefulness of the course, course satisfaction, learning expectations, and motivation; 2) peer collaboration regarding the collaborative learning activities with peers, such as requesting for help, solving problems collaboratively, responding to question, learning collaboratively, and doing assignments collaboratively; 3) cognitive problem solving regarding the internalization of cognitive tasks such as having an idea, knowledge application, knowledge analysis, judgement of information, and approach with novel perspective; 4) Interactions with lecturers concerning the communication activities between the learner and lecturer; 5) community support concerning the psychological factors such as perceived bonds and relationship with other learners; 6) learning management concerning the learner willing to have an active participation in learning Those six factors above are the constructs that are employed to assess student engagement in online learning. Concerning this, Dixson (2015) advocates four constructs which constitutes affective and behavioral components which derive and are modified from the constructs advocated by Handelsman et al. (2005) that comprise some factors such as skill concerning the things that the students can do, emotion concerning the way the students feel their connection to the course/content, participation/interaction concerning the way they connect with others and enjoy the course/content, and performance concerning the students' desire/goal of viability of the course.
In regard to online learning environment or distance learning, Hyland (2001) accentuates the importance of the role of feedback particularly in distance learning for it is the only opportunity students' get information about their performance students' engagement constitutes one of the essential elements that teachers should take into account as well. Concerning this, Martin and Bolliger (2018) explicate that satisfaction, motivation, sense of isolation, and performance is influenced by the way the students get engaged in online learning. Based on Hyland and Martin and Bolliger's views above, it can be indicated that both feedback given by a teacher and students' engagement are considered two essential elements in online learning.
In the today's English learning context, online learning becomes inextricable from students' learning. It is due to the current condition, Covid-19 pandemic, every schools all over the world should be closed down for a while, included in Indonesia. Officially, in his speech dated March 15, 2020, the president of Republic Indonesia, Joko Widodo asserts "Dengan kondisi ini saatnya kita kerja dari rumah, belajar dari rumah, ibadah di rumah (due to the recent condition, it is time for us to work from home, learn from home, and pray from home)" in Minta Masyarakat Tenang, Presiden: Saatnya Bekerja, Belajar, Dan Beribadah Dari Rumah (2020) from https://setkab.go.id/mintamasyarakat-tenang-presiden-saatnyabekerja-belajar-dan-beribadah-darirumah/).
This study attempts to extends the work reported in Bahati et al. (2019) by using some open-ended questions that are expected to assist the students to articulate the subjects' voices besides it aims to scrutinize the follow up actions conducted by them when they have already obtained the teacher's feedback. Also, it specifically aimed to investigate the relationship between the concerned variables, namely the lecturer's feedback and students' engagement in online learning. Hence, some research questions are posed as follows: 1. To what extend do the quality of the lecturer's feedback in online learning? 2. To what extend do the students' engagement in online learning? 3. Is there a significant relationship between the lecturer's quality feedback and the students' engagement in online learning? 4. How do the students act on the lecturer's feedback?

METHOD
A mixed-method was employed in this study. An explanatory sequential design was used. In this regard, the quantitative was garnered and corroborated with the qualitative data. Moreover, the rationale for using such a combination method was based on the research questions proposed. In this case, the first question concerned the quality of the teacher's feedback followed with the second one, i.e. the students' engagement in online learning, the third question that dealt with the relationship between the quality of the teacher's feedback and the students' engagement in online learning, and the fourth one was the students' actions after receiving the teacher's feedback. These were garnered through questionnaire. These were deemed as quantitative data. These quantitative data were corroborated by the interview instrument which was included as qualitative data.
The subjects participating in this study were the students taking Master's in the English Education Study Program in one of the universities in Bandung city.
There were sixty five students involved. A half of them were involved for the instrument try out. Meanwhile, only twenty three subjects studying in a certain online course were involved. They were selected and determined through a purposive sampling technique. Some considerations for selecting them as the subjects were taken into account. First, the ease of access to the subjects constituted the main reason for involving them in this study. Second, the course was not a compulsory one, i.e. the students chose by themselves, thus it arose this study to be conducted to see whether the course chosen by their own interest would be in line with their engagement in their learning in the online course. Another consideration was their cooperativeness to participate in this study. To keep the confidentiality of the subjects' identities, some codes were used, i.e. S1, S2, etc.
Two types of instruments, i.e. questionnaire and interview were employed. In this case, the questionnaires used in this study comprised the demographic questionnaire, the questionnaire assessing the quality of the teacher's feedback, actions after receiving feedback, and the students' engagement in online learning. The first one was developed from the constructs advocated by Gibbs and Simpson (2005), the second one was adapted from Hyland's work (2001), the third one was adapted from Lee, Song, and Hong's Engagement in E-Learning Scale (2019) and Dixson's Online Student Scale (2015). Meanwhile, the interview was utilized to ensure the data obtained from the questionnaire.
The data of this study were analysed by using two types of approaches. The first one was quantitative approach, which in this case, concerning the data taken through questionnaire. All the students' responses concerning the quality of the teacher's feedback were scored by using the following rules: strongly agree (5), agree (4), uncertain (3), disagree (2), strongly disagree (1) for the items assessing the quality of the teacher's feedback. Likewise, the items assessing the students' engagement in online learning were scored based on the following rules: always or almost always true of me (5), usually true of me (4), somewhat true of me (3), usually not true of me (2), never or almost never true of me (1). Also, a descriptive statistics (by utilizing PASW Statistics 18.0 software) was used to depict the quantitative data gleaned from questionnaire concerning the quality of teacher's feedback and students' engagement in online learning. Besides, to assess the relationship between the two variables, a correlational analysis was used. Testing of the linearity and normality distribution of the data was conducted to determine the type of the hypotheses testing used, whether parametric test (i.e. through Pearson's Product Moment correlation) or nonparametric test (Spearman's rho). Next, the value of the correlation coefficient was employed to challenge the research hypotheses. Again, in this regard, PASW Statistics 18.0 software was used. Meanwhile, the data assessing actions after receiving the feedback were analyzed and reported in percentage. Furthermore, the qualitative approach was conducted through interview. In this case the data gained was transcribed and put into texts to be coded. Lastly, all the data, both quantitative and qualitative, were analyzed to lead to a conclusion drawing.
Regarding the relationship between the quality of the teacher's feedback and the students' engagement in online learning, the statistical hypotheses were proposed. They consisted of: Null Hypothesis (i.e. there is no significant relationship between the quality of the lecturer's feedback and the students' engagement in online learning) Ha: Alternative hypothesis (i.e. there is a significant relationship between the quality of the lecturer's feedback and the students' engagement in online learning).

RESULT AND DISCUSSION Instrument Validity and Reliability
Based on the instrument try out with 29 students, the validity of the questionnaire gauging the quality of the lecturer's feedback, most of the items, i.e. six out of seven items were found to be valid, only item no. 3 was found to have a lower validity with r_counted value of .27 with rt (α=.05)= .433. Meanwhile, concerning the questionnaire assessing the students' engagement in online learning, it was found that nineteen out of twenty one items were found to be valid, three items, i.e. no. 8 and 21 to have a lower validity statistic with r_counted values of .31 and .32 consecutively. Based on the findings, dropping the items with lower values of validity was preferable. Therefore, the items no. 3, 8, and 21 were not included to the instrument disseminated to the target participants.
Regarding the reliability, based on the instrument try-out with the 29 students, the questionnaire measuring the lecturer's feedback quality was found to be good or high, i.e. with the Cronbach's alpha value of .808 before the item no. 3 was deleted and .833 after it was deleted. Meanwhile, it was found that the questionnaire gauging the students' engagement in online learning had the Crobanch's alpha value of .929 before the items no. 8 and 21 were deleted and after they were deleted the instrument's reliability registered a rise in value, i.e. .932. It was indicated and interpreted that the questionnaire to have an excellent or very high reliability.

Credibility, Dependability, and Confirmability
These three terms, credibility and dependability, dependability, and conformability, were employed for the qualitative data issue. In this regard, concerning the credibility, member checking was employed, i.e. the data gained from the interview were transcribed which then were communicated to the subjects of this study to ensure whether the transcript had already suited their report based on the interview conducted with them. Also, to inquire the objectivity of the findings, peer debriefing with reliable colleagues was employed. Besides, the findings were corroborated with the results obtained from the questionnaire conducted. Meanwhile, regarding confirmability, the researcher tried to be as objective as possible and hence report based on what he saw, heard, and observed. Besides, the findings were also scrutinized and corroborated with the theory and relevant studies.

The Quality of the Teacher's Feedback
The quality of the teacher's feedback was assessed by the students. In this regard, the information related to the quality of the teacher's feedback was obtained from the questionnaire created in Google Form disseminated online to the subjects through the following link https://docs.google.com/forms/d/e/1FAIp QLSc3CzHO_OHXK6-GTLPPYAyCyR6P9T8elZr-YbJazGT2iUHDNA/viewform?usp=sf_li nk. These were distributed to 23 students studying English Education Study Program at a Master degree Program of one of the universities in Bandung city. However, unfortunately only 20 questionnaires returned. Table 1 shows the detail results: Based on the maximum and minimum scores obtained, the values of ideal mean (Mi) and ideal standard deviation (SDi) is gained and consulted with the following criteria shown in Table 2 below:  Table 3 shows the results of the calculation for the quality of the lecturer's feedback. x < 76.66 Poor By taking account of the statistics above, it can be clearly seen that based on the students' assessment conducted, in average with mean of 82.38 and standard deviation of 6.51, the quality of the teacher's feedback was deemed good. This result seems in line with the result of the interview conducted with some of the students. For instance, based on S1's view the teacher was perceived to react responsively to the students' work and students' questions had been clearly answered but S1 preferred to choose direct feedback. Akin to what S1 said, S5 also stated that the lecturer frequently did peer feedback during the online learning instead of direct feedback. Meanwhile, S2 states, "… saya suka feedback yang bapak (dosen) berikan apalagi saat Mid tes. Semuanya jelas dan terpahami (I like the feedback the lecturer gave, even for the mid test, everything was clear and understandable)." Likewise, S3 perceive positively the teacher's feedback, S3 perceived to be objective based on the students' performance but S3 also revealed that it still had a dearth of personal feedback for the students. Similarly, S4 also articulates the same thing, 'Yang selama ini bapak sudah melakukan beberapa macam feedback ya…semuanya bagus menurut saya, gak ada masalah sih… nerima-nerima saja gitu selama itu bisa improve tugas kita atau pemahaman kita (So far, he has already conducted various feedback … all of them are good, in my view, there is no problem… we accept them as long as they can improve our work or understanding)'. Also, S6 perceived that the lecturer's feedback had been very good and responsive whenever the students met difficulties, and detail as well, hence it was considered to be very useful for her.

The Students' Engagement in Online Learning
The results of the students' responses in relation to their engagement in online learning obtained through the questionnaire disseminated. Likewise, the questionnaire was distributed to the students through the following link link https://docs.google.com/forms/d/e/1FAIp QLSc3CzHO_OHXK6-GTLPPYAyCyR6P9T8elZr-YbJazGT2iUHDNA/viewform?usp=sf_li nk. The results are shown in Table 4 below. Akin to the data of the lecturer's feedback quality, using Table 2 the categories of the interval values of the students' engagement are obtained. These are shown in Table 5 as follows:

65.53< x < 75.27
Good 4 x < 65.53 Poor By taking account of Table 5 above, it can be interpreted that the students' engagement with the mean of 73.43 and standard deviation of 10.92, the students' engagement in online learning can be considered good.
Moreover, based on the interview conducted, it was found that they would get more engaged to act on the feedback given if there was reinforcement from the lecturer, i.e. in the form of a deadline for every corrected task. Concerning this, S9 reported that "Segera... langsung difollow-up, apalagi jika menggunakan deadline… Jika sudah menggunakan deadline itu pastinya cepat sekali mengerjakannya (as soon as possible.. it was followed up, moreover if the lecturer gave a deadline… if the deadline was given, it would be conducted very immediately)" Based on this finding, it can be indicated that the lecturer power still predominates the students' learning albeit it was in the context of higher education. It is in line with Schebbuenner (2019) who revealed that "teachers still hold the central power in classroom while students themselves are aware of that." Hence, it also can be considered that students' engagement in doing the tasks given still rely on the encouragement given by teachers (Balçıkanlı, 2010).

The Quality of the Lecturer's Feedback vis-à-vis the Students' Engagement in Online Learning
To investigate whether the quality of the lecturer's feedback correlates to the students' engagement in online learning, a correlational analysis was conducted. To conduct the correlational analysis, the normality distribution and linearity of the data of the two variables were tested first.

Normality Distribution Testing
To ensure whether the data met the normal distribution, the two data sets were tested. The assumptions for the data normality distribution comprised the following hypotheses: -Null hypothesis (H0): the data set was normally distributed. -Alternative hypothesis (Ha): the data set was not normally distributed. If the PASW statistics is utilized, in this regard Saphiro-Wilk, the hypthotheses were tested based on the following criteria based on 95% level of confidence: -H0 was rejected, hence Ha was accepted, if p (sig. value) > .05 -H0 was accepted, hence Ha was rejected, if p (sig. value) < .05 The hypotheses above were tested based on the values represented in Table 6 below: Based on Table 6, it is clearly seen that the p value of the data set of students' engagement in online learning was found to be higher than the 95% level of confidence, i.e. p>.05=.813>.050. Thereby, H0 was accepted, thus the data set of students' engagement in online learning was considered to be normally distributed. Likewise, concerning the data of quality of the lecturer's feedback indicates that the p value obtained was higher than the 95% level of confidence, i.e. p>.05=.25>.05, hence H0 was accepted and the data was considered to be normally distributed.

Linearity and Heteroskedasticity Testing
To find out the linearity of the two variables, ANOVA was employed. The detail resul of ANOVA was presented in Table 7 below:  Table 7 above, it could be indicated that the 95% level of confidence the p value obtained was higher than the 95% level of significance, i.e. p>.05=.98>.05, thus H0 was rejected. Thereby, it was interpreted that the regression model between the two variables was considered to be not linear. Meanwhile, the heteroskedasticity model was tested by using Glejser test. It was tested based on the following criteria: -If p (sig. value) > .05, the heteroscedasticity was not found in the regression model -If p (sig. value) < .05, the heteroscedasticity was found in the regression model The assumptions above ascertained by using the statistics shown in Table 8 Based on statistics in Table 8 above, because the p value obtained was higher than the 95% level of confidence, i.e. .097>.050, it was interpreted that the heteroskedascity was not found in the regression model.

Relationship between the lecturer's feedback and the students' engagement in online learning
To find out whether there is significant relationship between the lecturer's feedback and the students' engagement in online learning, the PASW statistics was employed. Based on the sig. (2-tailed) value at 95% level of shown in Table 9 above, .097 >.050, it can be indicated that there is no significant relationship between the lecturer's feedback and the students' engagement in online learning. This finding is in contrast with the findings in a study conducted by Bahati et al. (2019).

Students' Actions on the Feedback Given
Concerning the feedback given, most of the students did some actions on it. In this regard, Figure 1 revealed that correction the errors made was conducted by most of the students, i.e. 45% of the students did it. Next, taking some notes constituted another most frequently, 23% of the students did it. Nevertheless, to bear in mind, an interesting finding was obtained, namely, some students, i.e. 16% of them, would only take pay attention if they were required to conduct further assignment.

Figure 1. Actions on the Feedback Received
Meanwhile, Figure 2 illustrates that to cope with the problems concerning with feedback given, most of students, 44% of them would consult to their friends.
While, reading books independently constituted the second alternatives that they would conducted followed with direct inquiry to the lecturer during the online course took place. This finding seemed to be in contradiction to Hyland's study (2001) study who revealed that students tended to use their own references instead of searching for help from others. In addition, the findings of the interview conducted concerning the actions on the lecturer feedback were also in line with the findings from the questionnaire disseminated. In this regard, the students tended to consult with their friends first instead of consulting directly to the lecturer when they found some issues regarding the feedback given by the lecturer. For instance, one the student S8 stated that "Aku coba konsultasi ke teman yang lain… selama mereka bisa menjawab atau menjelaskan, ya aku bakalan tanyain ke temen dulu (I tried to consult to my friends… as long as they could answer or explained it, I would ask them first). Again, this finding was in line with the result of the questionnaire and it seemed to be in contradiction to Hyland's work (2001) who revealed that students tended to use their own references instead of searching for help from others.

CONCLUSION
Based on the elaborations revealed in the previous sections, it is cocluded that the quality of the lecturer's feedback and the students' engagement during the online learning taking place is categorised as the good one. However, some challenges were still found, particularly those associated with the learner autonmy that goes hand in hand with motivation. Thus, even though the engagement was considered good, utilising reinforcement should be taken into account by the lecturer. Meanwhile, no significant relationship betweenn the lecturer's feedback quality and the students' engagement especially in online learning was found. In addition, student-teacher rapport was deemed to be in surface for the students tend to consult with their friends or books first before they consult directly with the lecturer if they met some challenges during the learning process took place.
Investigating the present areas of research by using more massive participants and other instruments gauging the feedback and online learning engagement are preferable and expected for future studies.

ACKNOWLEDGMENT
Best and deepest gratitudes are served and adressed to all the colleagues and participants involving remarkably in this study. Without any of their help and sincere endeavour, this study will remain unfinished. Kuh, G. D. (2003). A comparative study of student engagement, satisfaction, and academic success among international and american students.