Effects of synchronized and asynchronized e-feedback interactions on academic writing, achievement motivation and critical thinking

: We conducted a comparative study to evaluate the efficacy of synchronous and asynchronous interaction modes when providing feedback for improving academic writing, achievement motivation and critical thinking. The participants undertook an academic writing task, and were given feedback by a mentor using one of three interaction modes: asynchronous, synchronous e-feedback, and face to face. Data was collected from the participants via an assessment rubric applied to their writing after each stage of interaction, along with pretest and posttest questionnaires, and a semi-structured interview at the end. We found that synchronous feedback is more effective in increasing the quality of academic writing and achievement motivation, but for critical thinking we did not find any significant difference. We suggest that a strategy combining the advantages of each mode of interaction, considering the writers’ experience, may be the most effective way to promote academic writing, achievement motivation and critical thinking.


Introduction
University students and researchers need to improve their academic writing skills to prepare technical reports, write research proposals, apply for scientific scholarships, and publish their scientific results in conference proceedings and journals. Feedback on writing is the comments, questions, and suggestions from a reviewer to a writer for improving the quality of the writing (Keh, 1990). The reviewer can be a professor who is an expert co-author, or an English language reviewer and the writer is a researcher who needs to write a scientific manuscript. Feedback interaction on academic writing refers to the way that the reviewer prepares and delivers the comments, questions, and suggestions on the written academic manuscript to the writer besides the way that the writer can receive, discuss, and ask for help about what he/she receives from the reviewer. The mentioned process can be enhanced within synchronous activities through Skype or asynchronous activities through email (electronic feedback) or traditional feedback through face-to-face meetings. One often faces pressure and feels stress when having to produce academic publications required for granting an academic degree or obtaining a scientific position (Herbert et al., 2014). People who are good researchers are not always good writers, and a majority of students face difficulty at various times when reporting about their research (Glasman-Deal, 2010). With an increasing use of IT in education that can be effectively employed to improve academic writing for university students and researchers. Also, it also provides both synchronous and asynchronous interaction applications that support collaborative academic writing and help the writer to receive feedback from senior researchers or professors.
The importance of improving university students and researchers' academic writing (Sachs, 2002), critical thinking skills (Zhang & Toker, 2011;Borglin & Fagerström, 2012;Hall, 2017) and achievement motivation (Chiang et al., 2014;Liao, Ferdenzi, & Edlin, 2012), has been acknowledged. In addition, McClelland (2005) found that feedback is essential for academic achievement and e-feedback can also lead to improving learners' academic competencies. Although the importance of achievement motivation has been recognized, there are only a few studies on how it can be improved among university students and researchers (Zhang et al., 2015). Besides, there is little research on how different interaction synchronizations between a teacher and a student support the construction of arguments, which is applied in critical academic writing to enable students to edit their scientific documents. In particular, there is no study for distinguishing the efficacy of different e-feedback methods, which include face-to-face, synchronous and asynchronous interaction.
Therefore, the goal of our research is to obtain a deeper understanding of how different modes of feedback (synchronous and asynchronous e-feedback and traditional face-to-face feedback) affect the improvement of academic writing, achievement motivation and critical thinking among university students and researchers. Specifically, we seek to answer the following research questions: 1. How do modes of interaction (asynchronous e-feedback vs. synchronous efeedback vs. face-to-face feedback), affect the improvement of academic writing skills for university students and researchers?
2. How do modes of interaction (asynchronous e-feedback vs. synchronous efeedback vs. face-to-face feedback), affect the improvement of university students and researchers' motivation?
3. How do modes of interaction (asynchronous e-feedback vs. synchronous efeedback vs. face-to-face feedback), affect the improvement of critical thinking skills for university students and researchers?

Synchronous and asynchronous e-feedback
In synchronous e-feedback interaction, students and teachers interact in real time through digital technology. In asynchronous e-feedback interaction, after a teacher has given material and after a student has accessed material, completed and submitted an assignment, feedback is given in writing and the student accesses it at a different time.
Each of these modes has pros and cons with respect to their effectiveness in learning. Abrams (2003) noted that both synchronous and asynchronous interactions enable more extensive students-teacher interactions and provide more time for discussion thereby leading to improved proficiency in language skills.
Synchronous e-feedback can support learning by creating a positive social experience and engaging learners in cooperative learning activities (Boling et al., 2012). It can also increase the learning output compared with asynchronous interactions because learners may take asynchronous learning activities less seriously (Abrams, 2003). Synchronous interaction decreases ambiguity with learning experiences because it provides immediate e-feedback. Synchronous systems yield better improvement of the learning motivation compared with asynchronous interactions (Hrastinski, 2008). Synchronous audio interaction helps teachers to provide direct instructions to the students to support quality learning, but requires finding a meeting time that is convenient both for the teacher and the student, thereby lacking flexibility, which may negatively affect the student's motivation toward learning activities. (Murphy et al., 2011). Thus, providing synchronous e-feedback seems more effective than asynchronous feedback for improving learning motivation.
Asynchronous interaction is less effective than face-to-face interaction in improving writing and the dropout rate is higher (Carpenter et al., 2004). However, asynchronous interactions compared with synchronous have a more positive effect on the long-term retention of the learning information (Roblyer et al., 2007). Asynchronous interactions lead to an increased reflection on the learning activities because they have better flexibility of scheduling and gives an opportunity for a prolonged individual interaction with the learning resources. Thus, although asynchronous e-feedback has some drawbacks, it can lead to improving long-term learning and enhancing reflection, thereby leading to improved achievement motivation.
Given this controversy, we chose to investigate how these two different modes of interaction affect the improvement of academic writing, critical thinking and achievement motivation for university students and researchers.

Impact of feedback on students' achievement motivation
Achievement motivation is a student's inner drive to desire and work to accomplish the learning goals. It has many dimensions such as the learner's willingness to expand their learning effort, a tendency to take a learning initiative, spend more time and effort in learning by combining engagement with flow, to become better and faster in learning, flexibility in accepting learning challenges, and to take the responsibility of learning without fail (Basaran & Yalman, 2020;Lang & Fries, 2006;Hermans, 1970). Achievement motivation is a criterion for the formation of critical thinking because it helps learners to exert more effort to search for the right solution for the problems, they are facing in learning activities (Averkieva et al., 2015;McClelland, 1961, p. 42).
Asynchronous written e-feedback helps students to invest their time and effort in studying the received feedback, and make a plan for improving their writing abilities. It can be suitable for solving the problem of individual differences among students and can engage students in their writing task (Waller & Papi, 2017). Also, written comments (asynchronous interaction) can be provided as direct, indirect, explicit, or implicit feedback. Different types of feedback impact the accuracy of the writing efficiency and students' motivation differently (Papi et al., 2020). Sometimes, written corrective feedback is not preferred as it makes the student a passive recipient, which can lead to decreased motivation, rather than as an active participant in synchronous interactions mediated via online tools or face-to-face interaction (Hyland, 2011). Thus, asynchronous e-feedback can cater to individual differences among students, thereby improving their writing skills, but it may also lead to decreased student motivation.
On the other hand, using synchronous e-feedback in on-line writing sessions, compared with face-to-face interactions, helps students to achieve higher scores in constructing arguments, because in this kind of feedback the students' questions can be answered in real time. Also, synchronous interactions facilitate more discussion with the teacher, which provides opportunities to practice brainstorming, thereby leading to improved writing ability (Heift & Caws, 2000). Thus, synchronous e-feedback helps in improving the students' critical thinking. Only a few studies have studied the effect of feedback timing on the improvement of achievement motivation among university students and researchers (Zhang et al., 2015), though the need for such a study was recognized awhile back by VandeWalle (2003). Our goal is to address this problem by investigating how synchronous and asynchronous e-feedback can improve university students and researchers' abilities on academic writing, achievement motivation, and critical thinking.

Improving critical thinking through academic writing
Critical thinking is a higher-order thinking skill that is essential to lifelong learning and to dealing effectively with a fast changing world. It is one of the required skills for university students and researchers in the 21 st century (Kong, 2014;Dwyer & Walsh, 2020), and there is an interest for improving critical thinking among students of all ages (Alexander, 2014). It also includes assessment of the data, and linking new ideas and solutions with the students' prior knowledge. Critical thinking skills are also required in academic writing (Newman et al, 1997). Critical thinking is an intellectual process for generating ideas through applying, analyzing, and evaluating information. Editing written ideas with providing feedback is a technique for improving critical thinking (Muthmainnah, 2019). While practicing academic writing, the students should also generate new ideas, organize them, and construct convincing arguments. Critical thinking is an effective tool for improving and assessing students (Carter et al., 2017).
Although, academic writing and critical thinking are considered to be key objectives of students in higher education, and there is a need to enhance it among university students and researchers (Borglin & Fagerström, 2012), still, there are many obstacles in achieving this objective (Borglin & Fagerström, 2012;Sachs, 2002). Critical thinkers should be able to understand other viewpoints before reaching a conclusion (Thompson et al., 2003;Zhang & Toker, 2011). Practicing critical thinking entails effective communication and problem-solving abilities that are used through the discussion amongst the researchers and the authors (Paul & Elder, 2002). As the quality of communication for constructive discussion is a key factor, finding a suitable platform that provides feedback when the writer needs it and that gives adequate time for reflecting on the received feedback for solving the mistakes on academic writing, asking more details, adding valuable comments, and expressing writers' points of view related to the received feedback can help in improving critical thinking skills. Providing timely synchronous e-feedback often can increase the students critical thinking (Stein et al., 2013). Feedback is required in online learning, and is also key to improving critical thinking skills. Towards this goal, synchronous feedback can maintain students' motivation and keep them focused on learning from their mistakes (Tallent-Runnels et al., 2005). Immediate feedback can help the writer to concentrate on their writing problems and argue the reasons behind some written ideas. In addition, it can also lead to fruitful discussions related to the received feedback. Immediate feedback may help the writer to construct logical relations between their mistakes in writing and the methods of solving them.
Using asynchronous feedback within forums and chat software helps students to propose solutions and evaluate them critically. This improves their critical thinking (Kew & Tasir, 2021;Chiang & Fung, 2004). Also, both synchronous and asynchronous interactions can help in improving critical thinking, but it is necessary to provide students adequate time for reflecting on the feedback received. Osborne (2018) noted that using asynchronous discussion facilitates students' engagement and promotes their critical thinking. Asynchronous feedback supports improving critical thinking through providing adequate time for receiving, interpreting, reflecting on, searching for more details related to the received feedback. Writers when using asynchronous interaction can prepare the arguments and support their viewpoints and can also prepare more data for evaluating the received feedback that can help the writer to accept or reject the suggested feedback with strong evidence. As a result, asynchronous feedback can be a suitable interaction for improving writer' academic writing skills and motivation toward academic writing.
Synchronous interactions can also be used for improving critical thinking skills. Jacob (2012) noted that critical thinking skills in mathematics can be facilitated through problem-solving sessions in synchronous online discussion forums. We argue that such online forums can also be considered an asynchronous interaction. Moreover, the relationship between improving critical thinking in online discussion systems and academic achievement has not received much attention in literature; hence a detailed discussion is not possible (Jacob, 2012). However, the widespread and rapid shift to online teaching during the COVID-19 pandemic created unprecedented demand for turning to remote learning and using technology for improving learners' skills and knowledge (Carpenter, Krutka & Kimmons, 2020;Weldon et al., 2021).
Critical thinking has many dimensions like induction, deduction, interpretation, and evaluation. Prior studies show that improving critical thinking requires more than four weeks of focused practice (Osborne et al., 2018), with discussions and debate to make the learning activities more interesting and engaging (Thompson et al., 2003). Although there is a positive correlation between improving academic writing and critical thinking, this connection needs to be further studied (Borglin & Fagerström, 2012): there is a lack of studies that explore the improvement of university students' and researchers' critical thinking through academic writing (Preiss et al., 2013). Little is known about using synchronous interactions to support academic writing through online virtual space (Sun et al., 2017). The current study aims to fill this gap by investigating different effects of using synchronous and asynchronous feedback in improving critical thinking, academic writing, and how it may lead to support university students and researchers' achievement motivation.

Method
A quasi-experimental research design was used to study how different modes of efeedbacksynchronous (audio feedback through Skype meetings combined with MS Word comments) and asynchronous (text feedback through email combined with MS Word comments) -and traditional feedback (face-to-face meetings combined with MS Word comments) affect the improvement of academic writing, critical thinking and achievement motivation among university students and researchers. The study design was reviewed and approved by the Ethics Review Committee on Human Research at Tokyo University of Agriculture and Technology, Japan (No.191206-3151).

Participants
There were 19 initial participants, but three of them withdrew later, leaving 16 active participants: 7 Vietnamese, 3 Bengalis, 2 Egyptians, 2 Japanese, 1 Iraqi, and 1 Irani (Table 1). All participants had completed undergraduate degrees and were at a graduate institute. One participant was enrolled at South Valley University in Egypt, and all others at Tokyo University of Agriculture and Technology in Japan. The participation was voluntary: all the participants signed an informed consent. The participants were divided into three groups: Group 1 (Experimental 1), Group 2 (Experimental 2), and Group 3 (Control). There was no significant age difference between the groups (χ2 = 1.57, df = 2, Sig. = 0.46). To avoid bias in dividing participants into the three research groups and for enhancing balance among these groups the authors assigned participants based on their research experience which was reflected by the participants' current position, participants were divided also based on their English proficiency which can be understood from their grade on an international English certificate such as TOEFL and IELTS, their scientific experiences in academic writing is also important; therefore each participant was assigned to a research group based on his/her international publications that were written in English, and each group had only one female at most. Moreover, participants' grades in the pretest were a critical criterion for dividing participants into research groups.

Methodology
The participants were given the task of writing an academic paper: it could be for a journal article or a conference paper or just an internal report meant for their lab or professor using a scientific publication format with a suggested length of 2 to 10 pages. A reviewer, who is a domain expert with more than 13 years of experience as a researcher/reviewer, was assigned to each participant and worked with the participant in a one-on-one fashion using one of the interaction modes. There were four reviewers, and each reviewer worked with at least two participants each in a different group. Each participant received feedback on their writing using the mode of interaction (synchronous, asynchronous or face-to-face) according to their group. Each participant had five interaction sessions with the reviewer within a four-week period. The entire study was conducted from 14 December, 2019 through 7 March, 2020. The participants discussed the received comments on their writing, using Google Docs or MS Word, individually with the reviewer by one of the following methods: (1) Asynchronous e-feedback text: the reviewer used email for discussing the feedback in delayed interaction with the participant (Experimental Group 1). (2) Synchronous efeedback audio: the reviewer used Skype for discussing the feedback in real-time interaction with the participant (Experimental Group 2). (3) Face-to-face meeting: the reviewer discussed the feedback in face-to-face interaction with the participant (Control group 1).

Academic writing assessment rubrics
The reviewers rated the quality of academic writing at each stage using the Academic Writing Assessment Rubrics that was designed based on prior studies (Moriarty, 1997;Newman et al., 1997;Swales & Feak, 2004, 2012Derish et al., 2007;Bauer et al., 2009;Glasman-Deal, 2010;Biber et al., 2011;Bailey, 2014;Ecarnot et al., 2015;Razi, 2015;Wallwork, 2016). It consisted of three main criteria: Mechanics of academic writing, Content Structure and Language of academic writing with a total of 16 indicators. Each indicator was graded on a scale of 0-10, with 0-2 meaning inadequate ability and 8-10 meaning excellent ability. We revised this rubric slightly based on the advice of three professors at our university.

Achievement motives scale
We used Achievement Motives Scale (Hermans, 1970;Lang & Fries, 2006) which is applicable for university students and researchers because its reliability coefficient using the Cronbach Alpha Test was measured to be 0.79 (Demirel & Arslan Turan, 2010). It consists of 12 items, each being answerable on a five-point scale from strongly disagree (1) to strongly agree (5); except for two negative questions (4 and 5), the scales are inverted.

Critical thinking questionnaire
We asked the participants to answer the Critical Thinking Questionnaire, based on the Watson-Glaser Critical Thinking Appraisal, which has a reasonably high level of statistical reliability (Behrens, 1996;Bernard et al., 2008). It has four main criteria: recognition of assumption, interpretation, deduction, induction, and evaluation of arguments. It consists of 22 items, each being answerable on a five-point scale from strongly disagree (1) to strongly agree (5).

Semi-structured interview
We also conducted a semi-structured interview to get in-depth subjective feedback on the pros and cons of each interaction method. The interview consisted of 22 questions, half of which were closed questions, with the possible answers being Yes (2), I don't know (1), or No (0); these questions were based on the Acceptance Technology Model (ATM) (Scherer et al., 2019). The remaining questions were free-response type.
The interviews were conducted orally: the participants' responses were recorded using a microphone and later transcribed. We used a Laptops Lenovo ThinkPad, free voice recorder software on Win 10, and a ZealSound portable microphone.

Procedure
The experiment was conducted over three months, but for each participant it ran for about four weeks. Each participant had three interaction sessions, one each week, to receive feedback on their writing. The mode of interaction -synchronous, asynchronous, or faceto-face -depended on the group to which the participant was assigned. A flowchart of the experiment is shown in Fig. 1, and the tasks for each week are described below.
First week: The participants were called for a 30-minute meeting to explain the experiment, in which they were given a brief presentation about the experiment and were asked to read and sign the informed consent. They also filled out a background questionnaire, the achievement motives scale, and a critical thinking questionnaire. Then they were asked to prepare the first version of their academic writing (2-10 pages) within one week. They could base this paper on their previous or ongoing research, which needed to be written in English. A reviewer was assigned to each participant, as explained in Sec. 3.2, who applied the academic writing assessment rubrics to rate the submitted writing sample. Each participant was assigned to one of the three interaction groupssynchronous, asynchronous, or face-to-face (control) -based on their background and their score on the academic writing assessment rubrics. The participant's responses to both the questionnaires and their score of the academic writing assessment rubrics was combined into the pretest score.
Second week: The reviewer sent an email with an attached Google Docs or MS Word file that included the reviewers' comments on the participants' academic writing. Two days after sending the comments, the interaction between the participant and the reviewer proceeded according to the participant's group as follows: Group 1 (Synchronous): A 15-minute Skype meeting between the participant and the reviewer to discuss the comments on the submitted writing. Group 2 (Asynchronous): The participant asked questions to discuss comments on the writing by sending an email to the reviewer, who responded within two days. Group 3 (Face-to-face): A 15-minute face-to-face meeting between the participants and the reviewer to discuss the comments on the submitted writing. After receiving the comments, the participant was asked to revise her or his academic writings, and submit the next version within two days.
Third week: The same procedure as the 2nd week was repeated.
Fourth week: First, the same procedure as the 2nd week was repeated, after which the participant produced the 4th version of their academic writing. This was rated by the reviewer through applying the writing assessment rubrics as before. Then the participant gave their feedback on the mode of interaction through a semi-structured interview, and filled out again the questionnaires for both the achievement-motives and critical thinking scales. The responses of these two questionnaires were combined with the academic writing assessment rubrics grade for the 4th version of academic writing into the posttest score.

Pretest: Academic writing, achievement motivation, critical thinking skills
As the number of participants in each group was less than seven, we applied the Kruskal-Wallis nonparametric statistical test (Table 2) to confirm that there was no significant difference among the three groups in the academic writing assessment rubrics, achievement motivation, or critical thinking skills scales. The distribution of participants in the three groups was considered to be balanced. A semi-structured interview was conducted with each participant immediately after submitting the fourth version of their academic writing. The interview questions were based on the technology acceptance model (TAM). The participants' consent was obtained to record the interview. Each interview took an average of 16 minutes, which was audio recorded and transcribed for analysis. We recorded 192 minutes of interviews and collected more than 432 comments. Excerpted parts of these comments are grouped together to reflect advantages and disadvantages of various interaction modes.

Academic writing skills -Quantitative and qualitative analysis
In the first stage of the posttest interviews, the participants responded to closed questions on a 3-point Likert scale (Yes = 2, I don't know = 1, and No = 0). No significant difference (χ2 = 0.76, p = 0.68) was found among the three groups, or in any of the TAM dimensions of perceived ease of use, perceived usefulness, attitude to use, and intention to use (Fig. 2).

Fig. 2. Participants' responses correlated to the technology acceptance model (TAM)
In the second stage of the posttest interviews, the participants, having received feedback on four versions of academic writing, were asked to explain the reasons for their posttest answers. We wanted to know their suggestions for improvements, and to understand why certain aspects of the interactions were considered positive. However, the participants were asked to explain their answers without labelling them positive or negative. Their comments were classified as advantages and disadvantages in a later analysis. We present in Table 4 the key relevant concepts from the participants extracted from their posttest interviews. To maintain anonymity, participants' self-chosen IDs are used which follow concepts extracted from their comments.
We also quantitatively analyzed the improvement in academic writing skills by comparing pretests and posttests between versions 1 and 3 of the academic writings in all the experimental and control groups. The results of applying the Wilcoxon Signed Ranks Test are shown in Table 3 for the asynchronous group (Z = 3.20, P = 0.03), the synchronous group (Z = 2.02, P = 0.04), and the face-to-face group (Z = 2.02, P = 0.04). Overall, both e-feedback and face-to-face feedback are found to improve academic writing.
To investigate whether there is a significant difference among the three modalities of interaction, we applied the Kruskal-Wallis test to the participants' academic writing assessment rubric. The results revealed a significant difference between the three groups (χ2 = 8.74, p = 0.01). The synchronous e-feedback group gained a significantly higher score than the other two groups (Fig. 3). To better understand this difference, we further analyzed the writing assessment rubric with respect to its three components: content structure, language of writing, and mechanics of writing. The results indicate that the synchronous e-feedback group gained a significantly higher score on the content structure (χ2 = 6.99, p = 0.03), and the language of academic writing (χ2 = 8.08, p = 0.02), compared to the other two groups. However, there is no significant difference in improving the mechanics of academic writing between the three groups (χ2 = 3.20, p = 0.20). In the first research question, we try to investigate how various modes of interaction (asynchronous e-feedback vs. synchronous e-feedback vs. face-to-face feedback), affect the improvement of academic writing skills for university students and researchers. In order to answer the mentioned question, we analyzed Academic Writing Assessment Rubrics that were used for measuring the 4 versions of academic writing manuscripts that were submitted by participants. For understanding the acceptance level of using the technology for providing asynchronous or synchronous e-feedback. We also analyzed the data of the questionnaire that is designed based on the TAM model. The result refers to no significant difference being found among the three groups in accepting the applied technology for providing feedback in terms of ease of use, perceived usefulness, attitude to use, and intention to use.

Fig. 3. Improving academic writing through various interaction groups
The results indicate that providing e-feedback through the Skype applicationsynchronous group -has a significantly higher score than the other two groups. The results can be supported by prior research results that indicate synchronous interaction can provide direct instruction (Murphy et al., 2011) and decreases ambiguity with learning experiences because it provides immediate e-feedback (Basaran & Yalman, 2020;Cha & Park, 2010;Hrastinski, 2008). Consequently, when the professor would like to contact their students for providing feedback on their writing, it is recommended that they use synchronous interaction especially in online interaction. Software developers who seek to develop a collaborative writing system should add more features for improving the quality of synchronous interaction.

Achievement motivation -Quantitative and qualitative analysis
The results of analyzing the achievement motivation questionnaire indicate that the synchronous e-feedback group gained a significantly higher score on achievement motivation than the other two groups: Kruskal-Wallis test result is (χ2 = 8.63, p = 0.01). For comparing the improvement in achievement motivation across the three groups, we analyzed the pretest and posttest data. The Wilcoxon Signed Ranks test results for the asynchronous group are (Z = 0.54, P = 0.59), for the synchronous group they are (Z = 1.75, P = 0.08), and for the face-to-face feedback group they are (Z = 1.62, P = 0.10) as shown in Table 5. Though there is no significant difference among the three groups, the level of improvement among participants in the synchronous e-feedback group is slightly higher than the other groups (Fig. 4). -helps me to achieve my writing goals (A00, N11, A94).

Fig. 4. Improving achievement motivation
-I believe that face-to-face provides me with more details (N11).
-helpful for me to accomplish my writing (D13, S31) -provides for receiving valuable comments for improving my writing (D13) -For non-native English speaker, using email is better than oral interaction (S18, S11).
-more convenient (D13, S18) -providing more freedom to discuss writing with the reviewer (S31) -Receiving feedback face-toface is harder than email because the reviewer in the email can specify the problem and explain how to correct it (Q36, S11). -Skype is not uncomfortable to use, yet, one has to adapt to the condition of the internet and PC audio quality (N11).
-It is not easy because I have to adjust meeting appointments (Q36) and it is more time consuming (S31).
In relation to RQ2 (how various modes of interaction affect the improvement of the achievement motivation of the university students and researchers), the results have clearly shown there is no significant difference among the three groups in terms of improving achievement motivation. At the same time, as shown in Fig. 4, compared with using email or face-to-face meetings for providing feedback, using synchronous efeedback through Skype can support the participants to achieve better improvement in their achievement motivation. Because sometimes delayed written feedback decreased students' motivation as a result of making the student a passive recipient for the feedback, in contrast with immediate feedback (Hyland, 2011;Schillings et al., 2021); this is consistent with the participants' viewpoints ( Table 6) which indicate that Email provides complicated feedback which is not easy for discussing details (N12 & N99). Moreover, using Skype as a synchronous interaction technique can improve the students' motivation because it helps students to engage in effective discussions, which provides opportunities to practice brainstorming, thereby leading to improved engagement in learning skills which can lead to improve the achievement motivation (Heift & Caws, 2000). Participants' viewpoints also support that when they mentioned receiving e-feedback through Skype application is easy (A00, M19, N11, A94) and much more flexible (N11).
From Fig. 4 it is clear that the achievement motivation decreased for the participants who used face to face interaction. The achievement motivation can be affected negatively by a number of factors like the conflict of styles between the reviewer and the writer (Zeynali, Pishghadam, & Fatemi, 2019): our participants declared that face-to-face interaction for getting the feedback is harder than email and Skype meetings. Also, in almost all the sessions of providing feedback, the reviewers provided conflicting ideas. Although providing feedback through face-to-face interaction may lead to better management of the learning workload because of the immediate and detailed feedback that learners can get, it curtails the freedom of the participant, which may negatively affect the participants' motivation. This is consistent with Bouwmeester et al. (2019), who found that the student's motivation can be enhanced in the flipped classroom because they can practice learning autonomy, which is supported by the freedom in studying according to the students' own time and pace.
As a result, participants became demotivated while using face-to-face interaction. In addition, enhancing participants' motivation is required to avoid negative emotion while providing the feedback and achieving this is difficult, which leads to demotivating students (Filgona et al., 2020). Moreover, providing feedback on academic writing should balance between the writers' needs and the interaction modes to keep a suitable level of motivation that can support students' learning. Developing academic writing systems should provide learners and professors with various tools that make students more active through practicing brainstorming and presenting their viewpoints in the received efeedback. Therefore, it can support improving learners' achievement motivation.

Critical thinking -Quantitative and qualitative analysis
The quantitative data of applying a Kruskal-Wallis test showed that there is no significant difference between the synchronous e-feedback group, the asynchronous e-feedback group, and the face-to-face feedback group with respect to critical thinking (χ2 = 0.79, p = 0.67). The pretests and posttests were analyzed to assess the participants' critical thinking skills across the experimental and control groups. As shown in Table 7   Prior studies (Hall, 2017) have indicated a correlation between improving critical thinking and practicing academic writing. However, comments from our participants (Table 8) in the posttest interview are not so clear in this regard as shown below (Fig. 5). Overall, in relation to RQ3, as shown in Fig. 5., various modes of interaction for providing e-feedback or conventional feedback can slightly improve the participants' critical thinking skills. Although there is no significant difference between the three groups with respect to improving critical thinking skills, the results of the current research can highlight the importance of using effective interaction modes for providing feedback on academic writings. The duration of conducting the experiment needs to be extended for getting better results. In addition, it was recommended for improving critical thinking to ask participants to focus on writing the most critical sections in the academic manuscripts such as the introduction and the discussion section that can easily reflect the improvement of critical thinking skills for the participants.

Conclusion and implications
For improving academic writing skills, critical thinking and achievement motivation, we compared the efficacy of two different interaction modes -synchronous e-feedback (in real-time, as with Skype) and asynchronous (based on emails) -and contrasted them with each other and with face-to-face interaction. The results reveal advantages and disadvantages of each interaction mode.

•
Although feedback can enhance the improvement of university students and researchers' academic writing skills, synchronous e-feedback compared with other feedback modes supports writers to gain significantly higher scores on academic writing assessment rubrics.
• Achievement motivation can be improved by providing synchronous e-feedback and was found to be better than asynchronous e-feedback or face-to-face feedback. This can lead to a decrease in the dropout rate of learners from collaborative writing applications.
• There is no significant difference between the three interaction modes for critical thinking, and with respect to the technology acceptance model.
The following additional observations were made: • Novice researchers, in contrast with senior researchers, prefer synchronous interactions, with or without technology, because it supports finding an immediate response to their questions, decreases learning time, and provides more details as needed.
• We need to embed synchronous interaction tools with collaborative writing systems by incorporating audio and visual interactions besides text-based ones. Based on the pros of using synchronous interaction through Skype we can conclude that incorporating audio and visual interaction in collaborative writing systems would enhance the accuracy of academic writing.
• Synchronous e-feedback interaction is easier compared to the other modes with respect to the usefulness of and the motivation for using technology. However, asynchronous interaction is considered better than the other two interaction modes with respect to the technology acceptance model. This suggests that it will be useful to develop a strategy combining the benefits of these two interaction modes.
Improving the content structure and the language of academic writing is more complex than the mechanics of writing. Both the reviewer and the writer need to exchange their ideas on the manuscript to achieve an acceptable structure and language for their paper. Based on the participants' comments, it is not easy to figure out how to enhance various dimensions of academic writing. Sometimes the writer's personality affects the quality of discussion; sometimes the workload, for instance when the writer receives hundreds of emails every day, makes using asynchronous interaction more difficult than other types of interaction; sometimes the effort required to make extra arrangements, as in face-to-face meetings, adds more cognitive load on the writer. All these factors affect the improvement of a writer's academic writing.
In conclusion, academic writing is complex and different writers show different effects of feedback depending on the mode of interaction due to different preferences. Using synchronous e-feedback interaction can help participants to enhance their achievement motivation because it is interesting, easy to carry out, and can provide immediate feedback on participants' inquiries and questions. It has no psychological burden and is flexible in timing and workplace. It may be that because of the short duration of our study, no significant difference in improvement was observed between university students and researchers' critical thinking. To remedy this, a study over a longer period is necessary.
Our results suggest that as novice researchers have different needs than senior researchers, the effectiveness of academic writing systems and applications could be enhanced by incorporating a dynamic understanding of the researchers' needs. Providing both asynchronized and synchronized e-feedback along with face-to-face feedback can improve the quality of academic writing and achievement motivation, while using inappropriate e-feedback modes decreases the users' motivation to write academic publications. Inappropriate modes include complicated or ambiguous feedback without providing a chance for interpreting the received feedback, delays on urgently needed feedback, and feedback with negative emotions that decrease the accuracy of academic writing, writers' critical thinking abilities, and their achievement motivation toward academic writing.
We used non-parametric statistical tests for analyzing our data because our sample size was small. Hence there is a need to replicate the study with more participants. Also, the current study was conducted as one-to-one interaction between the reviewer and a participant (the writer), However, sometimes the reviewer in collaborative writing needs to provide feedback to more than one participant who works on the same project. Therefore, we need to conduct more studies to understand how various feedback interactions may affect the improvement of academic writing skills. Nonetheless, within one month the participants showed improvement in their critical thinking skills. We plan to conduct experiments of longer durations, with follow-up studies to better understand how various modes of interaction for providing the e-feedback affect the improvement of critical thinking skills.

Author Statement
The authors declare that there is no conflict of interest.