1.1. Oral presentation

In an EFL context, an OP refers to a form of public speaking where students, individually or in groups, present a prepared talk on a certain topic in English to the class with the help of visual aids (^{Chen, 2009}; ^{Sundrarajun & Kiely, 2010}). It occurs in a specific physical, social, and cultural context, with a certain theme and purpose of communication. The message is delivered from the speaker to the audience through linguistic, paralinguistic (intonation, volume, etc.), and extra-linguistic (gestures and facial expressions, etc.) elements. The intended meaning and its interpretation rely on the interaction between the production context and participants. It demands a combination of qualities: good knowledge of the topic of presentation, audience-friendly organization of content, high levels of language proficiency, adequate online language-processing skills, ability to talk extemporaneously, effective delivery skills, appropriate display of paralinguistic elements (for example, body language, eye contact, etc.), engaging the audience, an element of performance (appropriate voice manipulation and bodily action), good psychological qualities (overcoming shyness, nervousness, and sense of insecurity and inferiority), sensitivity to the context (audience’s response, etc.), register and discourse skills, a lucid presentation, adherence to specific behaviors valued by each discipline, and multimedia management (^{Morita, 2000}; ^{Tuan & Neomy, 2007}; ^{Zappa-Hollman, 2007}; ^{Sundrarajun & Kiely, 2010}; ^{Tsai, 2010}; ^{Chou, 2011}).

Given the demanding and comprehensive nature of the OP task, it constitutes a powerful situated learning experience that leads to deeper understanding and better learning (^{Joughin, 2007}; ^{Sundrarajun & Kiely, 2010}). It contributes to improving linguistic ability, building confidence, acquiring subject-specific knowledge, developing interpersonal and collaborative skills, refining presentation skills, socializing students into the academic context, and increasing creativity and problem-solving ability (^{Morita, 2000}; ^{Usher, 2002}; ^{Kobayashi, 2003}; ^{Zappa-Hollman, 2007}; ^{Sundrarajun & Kiely, 2010}; ^{Chou, 2011}).

In teaching EFL students to make OPs as a situated task, the above skills serve as a reference for formulating teaching objectives. To scaffold students’ skill acquisition, SR is one of the most important steps in instructional design.

1.2. Situated learning and self-reflection

Experiential learning theory (^{Schön, 1983}; ^{Kolb, 1984}) highlights that effective learning lies in the process of grasping experience of accomplishing a meaningful goal and transforming this experience into knowledge. The experiential learning model goes through a cyclical pattern of four phases: concrete experience during which learners complete a situated task, reflective observation in which learners reflect on what is done and consider what is working and what is not, abstract conceptualization where they think of ways to improve, and active experimentation where they make new attempts to try it again based on previous experience, reflection, and new conceptions.

Learners grasp the information from the situated learning experience phase which forms the basis for reflection. Reflection serves as the basis for conceptualization, which then leads to growth in new knowledge about the world. Learners then experiment with the new knowledge and the new cycle of experiential learning starts all over again. ^{Schön (1983}) defines professionals as those who link knowing and doing through reflection. ^{Kolb (1984}) argues for the importance of reflection by contending that only when professionals reflect on their learning can they develop theoretical understanding of it, apply new knowledge to new situations, and keep reframing and modifying their knowledge. It is learners’ consecutive SRs that lead them to generate a change in conceptual perspectives (^{Schön, 1983}; ^{Kolb, 1984}).

The pedagogical implication of the experiential learning theory is that instructors should provide situated learning opportunities for learners by creating real-world scenarios (^{Kolb, 1984}), and guide and encourage them to make reflections to support their growth of knowledge (^{Joyce, Calhoun & Hopkins, 2002}; ^{Smith, 2011}).

In teaching OPs, learners’ SRs can take various forms. Viewing video recordings of their own OPs is one of them.

1.3. Video recording as a pedagogical tool

Video recording is a widely-used pedagogical tool in teacher education. Teacher education scholars see video recordings of students’ teaching practice as reliable sources for students to make SRs (^{Robinson & Kelley, 2007}; ^{van Es, Tunney, Goldsmith & Seago, 2014}).

Literature on teacher education indicates that the support of video recordings of their own teaching practice helps improve students’ performance. Video recordings can help students analyze and evaluate their teaching practice from an observer’s stance (^{Kong, Shroff & Hung, 2009}), stimulate them to externalize their reflective ideas on their teaching practice with evidence (^{Calandra, Brantley-Dias, Lee & Fox, 2009}; ^{Kong et al. 2009}; ^{Rich & Hannafin, 2009}), contribute to the quantity, depth, and quality of their SRs (^{Robinson & Kelley, 2007}; ^{Sönmez & Hakverdi-Can, 2012}; ^{Baecher, Kung, Jewkes & Rosalia, 2013}; ^{van Es et al., 2014}), and allow them to actively construct new insights and be responsible for themselves in learning (^{Cherrington & Loveridge, 2014}). As a result, student teachers make improvement in their practice (^{Dawson et al., 2001}; ^{Shepherd & Hannafin, 2008}; ^{Tripp & Rich, 2012}a; ^{McFadden et al., 2014}).

Although video recording is widely used in teacher education, it is a rare pedagogical practice in teaching OPs, as is evidenced by a recent state of the art review (^{van Ginkel, Gulikers, Biemans & Mulder 2015}). ^{Barry (2012}) and ^{Murphy and Barry (2016}) investigated students’ perception of video recording and viewing of presentations within an Australian university and found that viewing and then reflecting on videos of group presentations was an effective way of feedback and could improve students’ performance. So far, the use of videos to assist students’ SRs on their OP experiences in EFL contexts is still an under-researched area. The effectiveness and students’ perceptions of this pedagogical intervention are not known yet. This study is concerned with the use of VASRs in teaching OP skills to first-year undergraduates in an EFL context.

2.1. Research design

Using a quasi-experimental two-group pre-test post-test comparison design, this study aims at answering two questions: ‘how effective is VASR in teaching EFL students to make OPs?’ and ‘how do students perceive the pedagogical use of VASRs?’

A self-assessment (SA) questionnaire was used to test the two groups’ baseline competence as the pre-test at the very beginning of OP teaching. Differentiated instructional designs were then applied to the control and experimental group. While the two groups received instructions regarding the same content from the same instructor (researcher) in the same classroom, they differed in one important procedure. The experimental group made SRs by viewing video recordings of their OPs for about 14 weeks, while the control group did not. The same SA questionnaire was used to test the two group’s OP skills as the post-test at the end of OP teaching. The gains between the pre-test and post-test offered evidence to research question one.

Another questionnaire with open-ended questions was completed only by the experimental group towards the end of OP teaching for students to reflect on their experiences of VASRs. Data from this questionnaire were used to answer research question two.

2.2. Context and participants

This study was conducted at a university in China. In the local context, all learners are non-native speakers of English. On the one hand, they need to use English to communicate in public, which requires them to be active and sociable learners. On the other hand, given the impact of Confucian traditions, they favor a teacher-centered classroom and are more silent and passive. Therefore, OP is a very challenging task for them since it may cause anxiety, nervousness, and lack of confidence. However, it is also a way to improve their linguistic ability, build confidence, and develop interpersonal skills.

Participants were 29 undergraduates. They were assigned randomly into the experimental group (n = 14) and control group (n = 15). There were 6 males and 23 females with an average age of 19. As first-year English learners, they were learning English as a foreign language within a Bachelor of Translation degree program. They were all admitted based on their academic achievement in university entrance examination and took courses within the same curriculum. When the current research was administered, they were undertaking a comprehensive English course aimed at improving their overall communicative competence. Before taking the course, most of them had never had any prior OP experiences.

2.3. Instructional design

OP was not a separate course in the context concerned, but one of the most important oral activities in an English course to enhance students’ communicative competence. In the first semester of their first year, the course consisted of six units on six topics related to international and domestic current affairs. In each unit, the students were asked to make group OPs towards the conclusion of the unit. The OPs were designed as situated tasks because the students went through an experiential learning cycle. They gained concrete learning experiences in real-world OP scenarios, made VASRs, and thought of ways to improve before taking new action to try the next OP task. The seven-step instructional procedures, including pre-, during- and post-tasks, are displayed in Table 1. As shown in the table, the control group and experimental group received differentiated instructions because they differed in the last step, VASRs, which was the focus of the current research. The instructional design involved six OPs, one in each unit. Five of all students’ six OPs (except the first one) were recorded.

Although OP demands a lot of skill from presenters, as mentioned previously, only some of the skills were involved in this study. OP was designed as an instructional activity to help the students achieve seven objectives, all relating to students’ ability to perform certain functions in English in specific communicative contexts. Before the participants made their first OPs at the beginning of the semester, they attended an orientation class to make sure that they had a clear understanding of what each objective meant. Those objectives were concerned with seven OP skills:

2.4. Instruments

Two questionnaires were used in the present study: ‘Questionnaire A. SA of OP Skills’ and ‘Questionnaire B. Perceptions on the Use of VASRs’. Both were developed and then revised for comprehension and clarity, following suggestions from colleagues.

Questionnaire A consisted of seven statements, each corresponding to one of the seven targeted skills: research and documentation, oral expression, logical thinking, use of non-verbal elements, knowledge, psychological quality, and team work. Students rated their achievement on each of the skills on a four-point Likert scale (1 strongly disagree, 2 disagree, 3 agree, and 4 strongly agree). The ‘forced choice’ scaling method was used, with the ‘neutral’ option removed. This is because the neutral option is an easy option to take when a student is not sure about which one to choose. When a ‘neutral’ option is selected, it may mean ‘undecided’ instead of ‘neither agree nor disagree’.

Questionnaire B included three open-ended questions. The first one asked what students focused on when they were viewing video recordings of their OPs. The second was concerned with two issues: if viewing recordings of their OPs was helpful, and in what aspects they made progress because of viewing recordings of their OPs. The third one asked which was more helpful in improving their OP skills -the instructor’s feedback in the debriefing session or viewing recordings of their OPs, and in what ways they complemented each other.

It needs to be pointed out that SA has been identified as a valid indirect measure of students’ achievement, although researchers are unsure of its suitability as a proxy for objective standardized measures. Learners’ SA of skills reflects their self-efficacy (beliefs in their ability to accomplish a task) which guides their efforts and persistence in learning (^{Bandura, 1993}). Research in many subject areas suggests that results of students’ SA of learning outcomes correlate positively with those of instructors’ direct external measures and that students’ self-reported rating of learning is an important indicant of learning achievement (^{Falchikov & Boud, 1989}; ^{Chesebro & McCroskey, 2000}; ^{Shen, 2002}; ^{Marks, Fairris & Beleche, 2012}; ^{Benton, Duchon & Pallett, 2013}). In a recent study on the validity of SA in language teaching, ^{Brown, Dewey and Cox (2014}) conclude that students’ SA of learning outcomes can be high in reliability and offer valuable information on their perceived achievement, supporting findings of previous language assessment research that SA and standardized tests are positively correlated (see ^{Bachman & Palmer, 1989}; ^{Alderson, 2005}; ^{Alexandria, 2009}; ^{Brantmeier, Vanderplank & Strube, 2012}; ^{Préfontaine, 2013}; ^{Lappin-Fortin & Rye, 2014}). Though students’ learning outcomes can be ideally assessed by combining SA and direct measures, composing and administrating direct measures for each of the seven teaching objectives is time-consuming, costly, and impractical for the current study. Therefore, students’ progress in each objective was assessed through SA questionnaires.

2.5. Procedure

Before involving students in the current research project, the researcher obtained their consent to be participants. They were told that they would participate in a research project on a voluntary basis, that their responses would not have any effect on their grades, that their identities would remain anonymous, and that results may help them improve their learning efficiency.

The present study consisted of four phases. Questionnaire A was administered as the pre-test for the participants to rate their competence in each of the seven OP skills when they had immediately finished their first OPs at the beginning of the semester. Then the differentiated instruction was implemented to the control and experimental group as shown in Table 1, for about 14 weeks. Subsequently, questionnaire A was administered again as the post-test when the participants had immediately finished their last OPs, towards the conclusion of the course. After the post-test, questionnaire B was completed only by the experimental group.

Data collected through questionnaire A were analyzed using SPSS 16.0 for Windows. With Cronbach’s alpha (α) values of .796 and .918 (for the pre-test and post-test respectively), it has good internal consistency reliability. SPSS GLM was used to perform a univariate analysis of covariance (ANCOVA). The differentiated instructions received by the two groups were used as an independent variable. Students’ post-test on their seven OP skills was used as the dependent variable. The ANCOVA was run to examine if the independent variable impacted the independent variable, using the pre-test results as a covariate to control initial differences in the seven skills. Preliminary checks were performed to determine if statistical assumptions were violated in the dataset for ANCOVA, such as normality, reliable measurement of covariance, and homogeneity of variance. The results of between-subjects effect tests showed that the assumptions were satisfied (F = .033, p = .857; F = .711, p = .407; F = .272, p = .607; F= 2.936, p = .099; F= .000, p = 1.000; F = .468, p = .500; F = .010, p = .921).

Data from questionnaire B were analyzed in accordance with the steps recommended by ^{Creswell (2014}): organizing and preparing the data, reading through the information, coding the data, developing a thematic analysis from the codes, and presenting findings in tables before interpreting them. Two broad categories of themes were identified, one related to use of non-verbal elements including eye contact, facial expression, hand movement, and body posture, while the other concerned oral expression (pronunciation, voice projection, fluency, and intonation).

3.1. The effectiveness of VASR

The first research question was how effective VASRs were in teaching EFL students to make OPs.

Table 2 shows the number of participants, mean, and standard deviation of the two groups’ scores on the seven OP skills in the pre-test and post-test. As displayed, in the post-test both groups obtained higher scores than they did in the pre-test, suggesting that both groups made progress in the seven skills, although the two groups received differentiated instructions.

Table 3 presents the ANCOVA results on the post-test with the pre-test scores as the covariate. As is evidenced, for the use of non-verbal elements, the two groups showed a statistically significant difference (F = 4.894, ^_η2 = .157, p = .037), suggesting that VASRs resulted in the variance in the use of non-verbal elements between the two groups. For the other six skills, the two classes did not demonstrate statistically significant differences, suggesting that the effect of VASRs was not evident.

In sum, after running descriptive and inferential statistics, two findings were noted. Firstly, although only the experimental group viewed recordings of their own OPs and made SRs, both groups made progress in all of the seven skills. Secondly, VASRs was effective in developing students’ skill in using non-verbal elements. Its effect in the other six skills was not evident.

The findings contribute to knowledge on the pedagogical value of OPs. The finding that both groups made progress in the seven skills supports the literature that OPs help students improve linguistic ability, build confidence, acquire subject-specific knowledge, and develop interpersonal and collaborative skills (^{Morita, 2000}; ^{Usher, 2002}; ^{Kobayashi, 2003}; ^{Zappa-Hollman, 2007}; ^{Sundrarajun & Kiely, 2010}; ^{Chou, 2011}). Although only the experimental group made VASRs, both groups followed the same procedures including starter, input, mind mapping, briefing, OP, and debriefing. In the debriefing session, the instructor provided feedback to both groups, highlighting their strengths and weaknesses. Therefore, both groups might have made progress because of their own efforts in preparing and delivering OPs and the help of the instructor’s feedback. Students’ improvement in skill may have been influenced more by obtaining feedback from their instructors than by VASRs. This is consistent with previous research which supports students’ preference and confidence concerning the usefulness of instructors’ feedback (^{Halter, 2006}; ^{Rich & Hannafin, 2008}; ^{Tripp & Rich, 2012}b).

The current findings also add knowledge to the literature that VASRs is effective in teaching EFL students to make OPs, though its effect is limited to non-verbal skills. The effect was very limited, compared with findings of research on teacher education that it contributes to student teachers’ improvement in their teaching practice (^{Dawson et al., 2001}; ^{Shepherd & Hannafin, 2008}; ^{Tripp & Rich, 2012}a; ^{McFadden et al., 2014}).

What might have been responsible for the limited effects of VASR? One possible explanation is that a well-established protocol was not provided to the students in the experimental group to make SRs on their recorded OPs. On the one hand, the students in the current study were in a teacher-centered culture where learners are passive and silent and expect the instructor to shoulder more responsibilities (^{Watkins & Biggs, 2001}; ^{Woodrow, 2006}; ^{Bankowski, 2010}). On the other hand, the instructor did not provide any protocol for the students to follow in their SRs. Research on teacher education finds that students observing their own videos of performance require more scaffolding than observing videos of others’ performance (^{Kleinknecht & Schneider, 2013}). Prior research has identified the benefits of a systematic SR protocol to facilitate students to notice certain aspects of their performances (^{Zeichner & Tabachnick, 1991}; ^{Collins, Cook-Cottone, Robinson & Sullivan, 2004}; ^{van Es et al., 2014}). With a protocol, students’ SRs are of better quality (^{Fox, Brantley-Dias & Calandra, 2007}), while without it students tend to concentrate on non-verbal elements (^{Calandra, Gurvitch & Lund, 2008}; ^{McFadden et al., 2014}; ^{van Es et al., 2014}) and lack a critical stance (^{Cherrington & Loveridge, 2014}). It has also been indicated that, besides learners’ observation of their own behaviors, their progress requires the assistance of an external facilitator who interprets their current behaviors and provides alternatives in terms of better practice (^{Timperley & Robinson, 2001}). Since the students in the experimental group were not provided with any protocol to assist their SRs, their concentration might have been primarily placed on non-verbal elements -for example, how they behaved in the video (non-verbal elements)- and ignored other aspects such as oral expression, logical thinking, knowledge, and so on.

Another factor that might have limited the effect of VASRs is the nature of some of the skills. Some skills, such as research, documentation, and team work, though they can be reflected in the final product of their OPs, took place in the preparation phase before their OPs. Problems in those regards are much more difficult to trace in the videos. Therefore, students’ SR might ignore those areas.

One implication of the current findings is that the instructor should provide a collaborative reflection protocol to prescribe what students and their instructor are supposed to focus on in making SRs and giving feedback respectively. The protocol can take various forms. For example, checklists can help students narrow their focus and gain more knowledge (^{Prusak, Dye, Graham & Graser, 2010}). Essays, journal writing, notes, questionnaire responses and other forms of written reflection also help students produce more focused and accurate reflection (^{Welsch & Devlin, 2004}; ^{Rich & Hannafin, 2008}; ^{Shepherd & Hannafin, 2009}). Additionally, video editing, conferences, and interviews are useful tasks that students can engage in to aid their SRs (^{Tripp & Rich, 2012}a). However, the instructor should be cautious in formulating the protocol. Although a checklist facilitates reflection, instructors may have difficulties in deciding how many items should be listed in the checklist (^{Prusak et al., 2010}). It is more effective for students to focus on a subset of the protocol, because they may feel overwhelmed and find it less helpful in VASR when the checklist includes too many items (^{Tripp & Rich, 2012}a). Students prefer to decide and concentrate on their own list of criteria in their SRs (^{Rich & Hannafin, 2008}). It is advisable for the instructor to provide a well-structured protocol to facilitate students and to allow them to have their own foci of reflection (^{Tripp & Rich, 2012}a).

3.2. Students’ perceptions of VASR

The second research question was about students’ perception of VASRs. Data from questionnaire B were analyzed and classified under different headings, along with their frequencies. Students’ answers were quoted as evidence.

What did the students in the experimental group focus on when they made VASRs? Table 4 presents the categories of students’ foci and their frequencies. The 14 students in the experimental group reflected on two broad categories of problems. The non-verbal element category was mentioned 15 times, while oral expression was mentioned 12 times. These can be evidenced from students’ answers:

Was viewing videos of their OPs helpful? In what aspects had they made progress because of the VASR experience? All 14 students expressed the view that viewing videos of their own OPs was helpful. As shown in Table 5, they attributed their progress in the use of non-verbal elements (mentioned 13 times) and oral expression (mentioned 7 times) to their experience of viewing videos of their OPs, as can be seen from their answers to the question:

Based on analysis of the students’ answers, it can be concluded that students primarily focused on their use of non-verbal elements and oral expression in OPs while making VASRs, and that it was in the former area that VASRs were effective. The results indicated that VASRs were effective in helping students make progress in the use of non-verbal elements, which is consistent with the results of previous research where no protocols were provided to guide students’ SRs (^{Timperley & Robinson, 2001}; ^{Calandra et al., 2008}; ^{Cherrington & Loveridge, 2014}; ^{McFadden et al., 2014}). This finding concurs with the results of quantitative analysis in the previous section that the use of video recordings was only effective in developing students’ skill in using non-verbal elements. The agreement between the quantitative and qualitative data pointed to the necessity of a systematic protocol to scaffold students to ensure the quality of their SRs (^{Zeichner & Tabachnick, 1991}; ^{Collins et al., 2004}).

Which was more helpful in improving their OP skills, the instructor’s feedback in the debriefing session or viewing videos of their OPs? In what ways can they complement each other? Students’ responses indicated that they benefited from both the instructor’s feedback and VASRs, which complemented each other.

Of the 14 students, seven indicated that feedback and VASRs were equally helpful; four thought the instructor’s feedback was more helpful; and three rated VASRs as more helpful. All students suggested that the instructor’s feedback and VASRs had different advantages. The instructor’s feedback was more systematic and in-depth because it covered all presentation-related aspects, particularly weaknesses that students could hardly see by themselves, such as logic and content, and gave suggestions for further improvement. By contrast, in VASR, students tended to have a narrowed focus, concentrating on non-verbal skills, such as hand movement, eye contact, and body posture.

According to students’ responses, SRs were very good supplementary sources of information for two reasons. Firstly, although the instructor gave feedback concerning the non-verbal skills of their OPs, the feedback was not impressive because they could not directly see their weaknesses at the moment of being given feedback. The use of VASRs, however, allowed students to view their performance from the perspective of the audience. Such a perspective was more direct and impressive for them, helping them to better understand the non-verbal elements of their presentations. This was evidenced by students’ answers:

Secondly, the use of VASRs provided an unlimited access for reflection. The instructor’s oral feedback did not allow for future review. However, videos could be reviewed for as many times as wanted, as was pointed out by one student:

“I learned from both the teacher’s feedback and viewing videos. The feedback was more inclusive. Viewing videos of my presentations was more individual-specific. I could view them repeatedly and freely” (Student GMZ).

The results presented here support the literature that video recordings can help students analyze and evaluate their own performance as observers (^{Kong et al., 2009}) and externalize ideas on their competence (^{Calandra et al., 2009}; Kong et al., 2009; ^{Rich & Hannafin, 2009}), and that students may have a preference for and confidence in instructors’ feedback (^{Halter, 2006}; ^{Rich & Hannafin, 2008}; ^{Tripp & Rich, 2012}b).

The findings of the current study are meaningful in that so far there have been very few contributions on the effect of VASRs in teaching OPs in EFL contexts. This study shows that videos contribute to the quality of students’ SRs (^{Robinson & Kelley, 2007}; ^{Sönmez & Hakverdi-Can, 2012}; ^{Baecher et al., 2013}) by providing students with information to evaluate their English OPs from the perspective of observers (^{Kong et al., 2009}) and externalizing evidence of their competence (^{Calandra et al., 2009}; ^{Kong et al., 2009}; ^{Rich & Hannafin, 2009}). This study thus shows the effectiveness of VASRs in teaching EFL students to make OPs. Moreover, it reveals the weaknesses of VASRs without external guidance and points to the necessity of a well-structured protocol to increase the quality of students’ SRs (^{Zeichner & Tabachnick, 1991}; ^{Collins et al., 2004}). Additionally, it shows the complimentary role of VASRs to instructors’ feedback which students still prefer and trust (^{Halter, 2006}; ^{Rich & Hannafin, 2008}; ^{Tripp & Rich, 2012}b). The finding, that students’ limited progress in their English OPs as a result of VASRs could be improved by means of a well-structured protocol and the instructor’s feedback, also provides evidence to the effectiveness of teaching English OP according to the philosophy of experiential learning: that learners’ growth of knowledge depends on the instructor’s guidance to reflect on their experience (^{Joyce et al., 2002}).

Based on the advantages and complimentary role of VASRs, the pedagogical implication is that both the instructor’s feedback and VASRs are necessary in teaching OPs. When students finish their OPs, the instructor should provide feedback during debriefing and guide students to make VASRs with the help of a focused protocol as discussed above.