Abstract
While the flipped learning approach has been widely implemented in science-relevant subjects, little empirical research has been conducted in language learning, particularly among low-proficiency learners. The present study aims to examine the pedagogical effects of flipped learning on low-proficiency students’ learning outcomes and attitudes (motivation and self-efficacy). An eight-week experiment was conducted on 52 low-proficiency learners who attended intensive English grammar classes in a national high school. The experimental group (EG) (27 students) learnt with the flipped approach whereas the contrast group (CG) (25 students) was taught through a conventional approach. The performance by both groups in grammar, writing and learning motivation and self-efficacy were collected for analysis. The EG’s learning experience was also qualitatively explored. The results show that the EG significantly outperformed the CG in terms of overall learning achievement, motivation and self-efficacy. Additionally, despite facing challenges at some points, the EG gave positive comments on the pedagogical effects of flipped learning, judging it to be interesting and leading to active engagement. This article concludes with a discussion of the effects of the flipped learning approach on low-proficiency language learners and with instructional implications for future empirical research.
I Introduction
Over the past decade, the flipped classroom has grown in popularity in educational settings with particular reference to foreign language teaching (Turan & Akdag-Cimen, 2020). Its endorsement derives from the flipped approach, which reverses the roles of conventional course content and take-home assignment. That is, knowledge originally delivered in class becomes the material for preview tasks at home to facilitate students’ independent learning of content before class, whereas homework is shifted to in-class activities consisting of, for example, discussion, exercises and other tasks that help scaffold students’ learning (Akçayır & Akçayır, 2018; Vitta & Al-Hoorie, 2020).
As the flipped classroom evolves, the concept of flipped learning emerges. Whereas the former focuses on reversing the instructional locales (Sams & Bergmann, 2012), the latter aims at developing pedagogical practices in group-learning spaces that stress the inclusion of higher-order thinking skills, particularly creativity (Flipped Learning Network [FLN], 2014), for instance in collaboration or student-centred learning tasks (Sams & Bergmann, 2014). FLN (2014) also proposes a four-pillar framework for implementing flipped learning, consisting of a flexible environment, learning culture, intentional content and a professional educator. These components serve as useful guidance for designing suitable instruction methods and a dynamic and engaging classroom environment for the flipped learning approach.
The pedagogical strengths of flipped learning help create fertile ground for active learning (Chuang et al., 2018; Låg & Sæle, 2019), a dynamic and interactive learning environment (FLN, 2014) and collaborative learning (DeLozier & Rhodes, 2017). Many potential pedagogical advantages of the flipped approach described here may explain why most empirical studies in the field also found empirical evidence in favour of the flipped approach (e.g. Chen Hsieh et al., 2017; Hung, 2017; Turan & Akdag-Cimen, 2020; Vitta & Al-Hoorie, 2020). Although the flipped approach seems to have gained a foothold in the field, more empirical research on underrepresented language learner groups is required (Vitta & Al-Hoorie, 2020). One group that demands urgent investigation is English as a foreign language (EFL) low-proficiency learners. This is because most of the previous studies involved students with average or high proficiency who demonstrated self-regulation for finishing pre-class materials (e.g. Lee & Wallace, 2018; Lo & Hew, 2017). In contrast, low-proficiency students are relatively poorly self-regulated and may thus procrastinate in previewing materials when learning with the flipped approach (Filiz & Kurt, 2015; Vitta & Al-Hoorie, 2020). Additionally, it is essential in the flipped learning setting to investigate English grammar learning outcomes (Turan & Akdag-Cimen, 2020) because grammar is found challenging to most EFL learners who have been observed (Liu et al., 2019), and especially so to low-proficiency students. The significance of English grammar instruction should thus be stressed in the flipped learning context in order to explore its pedagogical effectiveness on EFL low-proficiency students. Lastly, while most experiments have quantitatively examined the effects of flipped learning in EFL settings, more qualitative approaches are needed to understand learners’ individual differences in their perceptions and perspectives.
Taken together, these concerns over the implementation of the flipped approach endorse the legitimacy of assessing the pedagogical effects of flipped learning on EFL low-proficiency grammar students. Hence, the present study sets out to examine the pedagogical effects of flipped learning on EFL low-proficiency learners. An empirical study was thus conducted on two groups of low-proficiency grammar students, with one taught through the flipped learning approach and the other through conventional teaching. Three research questions are addressed:
Research question 1: Do the two approaches (i.e. flipped learning and conventional teaching) significantly improve EFL low-proficiency students’ performance in grammar learning outcomes and in learning attitudes (motivation and self-efficacy)?
Research question 2: How great are the differences between the two instructional methods as seen in EFL low-proficiency students’ grammar learning outcomes and learning attitudes?
Research question 3: What are the EFL low-proficiency grammar students’ overall learning experiences with the flipped learning approach?
1 Fundamental concepts of flipped learning
The widespread success of flipped learning has foregrounded educational technology, collaborative learning, and the development of higher-order thinking. To begin with, educational technology helps facilitate the pedagogical shift where the course content is presented to students outside of the classroom for preview activities before each class. Learning resources and processes can be presented in various ways, including tasks on social media (Chen Hsieh et al., 2017), the use of teaching platforms (Hsiao et al., 2021), and the implementation of a student response system during class (Liu et al., 2019). Online pre-recorded instructional videos are a particularly common and effective measure for the flipped classroom because they primarily allow students to watch and study at their own pace (Chuang et al., 2018). These technological aids help a flipped class to create flexible instructional periods, optimize the use of class time (FLN, 2014), attend to students’ individual learning needs (Chuang et al., 2018), and eventually improve their learning outcomes and positive learning attitudes (e.g. Amiryousefi, 2019; Boyraz & Ocak, 2017; Ekmekci, 2017; Hung, 2015, 2017; Kang, 2015; Turan & Akdag-Cimen, 2020; Vitta & Al-Hoorie, 2020).
Collaborative learning is another important pedagogical element in flipped learning. The change to the learning sequence produces more class time (Akçayır & Akçayır, 2018) in which to implement collaborative learning (Francl, 2014). With it, a flipped class can afford multifaceted learning input and output as well as a range of group tasks and activities that stimulate students’ interaction between peers and with the instructor (DeLozier & Rhodes, 2017; Turan & Akdag-Cimen, 2020). Note that these collaborative in-class interactions largely rely on learners’ active preparation beforehand. Students who fail to prepare may fall behind in class activities and discussions (Y. Chen et al., 2014), which in turn makes it difficult for a flipped lesson to be successful. To avoid this, instructors are advised to prepare, for example, check quizzes to accompany the preview materials. A teaching intervention as such is reported to help students complete their preview (DeLozier & Rhodes, 2017) and should lay a foundation for success in implementing collaborative learning in the context of flipped learning.
The reinforcement of technological assistance and collaborative learning may explain why it is possible for flipped learning to develop students’ higher-order thinking skills (see Alsowat, 2016; Anderson & Krathwohl, 2001; Hung, 2015). As discussed above, flipped classes have various opportunities that allow learners to engage in meaningful preview activities and collaborative in-class tasks. Many of these ask students to do team work on a project, cooperate to solve problems, or discuss questions pertaining to course content (Hsiao et al., 2021; Liu et al., 2019). These venues help students to jointly scaffold knowledge with their peers (Mehring, 2016), resulting in shared, enhanced understanding of the target subject. Learners’ skills in analysing, evaluating and creating on the basis of their subject knowledge input are also expected to improve over the process. From this line of discussion, it is reasonable that scholars consider that flipped learning will improve learners’ higher-order thinking skills.
2 Flipping language learning classrooms
The evidence of flipping language learning classrooms has been extensively examined for its potential advantages in language learning (Akçayır & Akçayır, 2018; Amiryousefi, 2019; Boyraz & Ocak, 2017; Chuang et al., 2018; Hung, 2015; Lee & Wallace, 2018; Turan & Akdag-Cimen, 2020; Vitta & Al-Hoorie, 2020). To begin with, although some similar levels of performance were observed between students involved in flipped learning and their counterparts (e.g. Jensen et al., 2015), students’ learning performance and attitudes mostly significantly improved with the help of flipped learning, compared with those in conventional teaching contexts (e.g. Boyraz & Ocak, 2017; Lo & Hew, 2017; Turan & Akdag-Cimen, 2020). For example, Amiryousefi (2019) found that the flipped learning approach had positive effects on EFL students’ listening and speaking skills, class engagement and out-of-class participation. The findings of Chen Hsieh et al. (2017) echo this, revealing that the approach significantly improved students’ use of idiom and learning motivation.
Specifically investigating grammar learning outcomes, Liu et al. (2019) flipped their English grammar course at the tertiary level by implementing the student response system. In their study, the students in the flipped group increased their motivation and engagement, but their grammar learning achievement did not improve. One of the explanations for this that Liu et al. gave was that such an interactive device and system might distract students’ learning. This is because they spent too much time and effort adapting to this learning device during the limited class hours, which may have reduced their capacity to obtain as much grammar knowledge as they should have. Furthermore, when Al-Harbi and Alshumaimeri (2016) targeted an EFL grammar course in a private secondary school, they reported that the mean scores of the flipped group outperformed those of its counterpart, despite no statistical evidence. However, a methodological concern might be raised that Al-Harbi and Alshumaimeri (2016) had a non-equivalent, post-test only research design. A pre- and post-test design should have generated more complete and comparable research findings. These studies indeed offer valuable insight and show that further investigation is warranted.
It must also be mentioned that the body of flipped learning research has relied heavily on average or high-proficiency learners. This obliges the research community to shift the study focus to other groups of learners (Vitta & Al-Hoorie, 2020). One group worth attention is low-proficiency students. As Filiz and Kurt (2015) warned, they are more likely to procrastinate in completing pre-class tasks because of their low self-regulation and motivation, especially when compared with advanced and average learners. Hitherto, very little research has been conducted on EFL low-proficiency students’ engaging with the flipped learning approach. Such little empirical evidence thus drives the present study to examine how the flipped learning approach affects this specific cohort.
3 Motivation and self-efficacy in language learning
Learning motivation and self-efficacy are important affective factors that have been investigated in the field of foreign language acquisition. The former determines why learners have a desire to work towards a goal, how hard they pursue it, and how long their action lasts (Dörnyei, 1998, 2001). The latter refers to learners’ judgement of their capabilities and their self-perceived confidence in these capabilities for completing a certain task (Bandura, 1997). Many researchers have found correlations between motivation and self-efficacy and students’ performance in language skills (e.g. Dewaele, 2009; Dörnyei & Ushioda, 2013; Mills et al., 2007; Pajares, 2003; Saito et al., 2018; Wang et al., 2021). Others have even found evidence suggesting that the developing learners’ motivation and self-efficacy led to improved language performance (see Raoofi et al., 2012). For example, enhanced motivation was found to improve students’ performance in such areas as writing (Bai & Guo, 2021; Lin, 2015), reading (Chuang et al., 2018; Moskovsky et al., 2016), grammar (Lin, 2016) and overall competence (Lasagabaster, 2011). Similarly, stronger self-efficacy brought out significantly more gains in reading (Heydarnejad et al., 2022), writing (Lin, 2015; Zhou et al., 2022) and speaking (Asakereh & Dehghannezhad, 2015; Y. Y. Chen et al., 2021). Clearly, assessing students’ performance in both motivation and self-efficacy should offer valuable evidence for judging the pedagogical effects of flipped learning.
However, perhaps because flipped learning in foreign language classrooms is ‘still in its infancy’ (Amiryousefi, 2019, p. 150), only a few scholars in the field have addressed motivation and self-efficacy. Examples of these are Huang and Hong (2016), who examined 40 EFL junior high school student readers’ performance in a flipped intervention and compared it to 37 other students’ performance in a conventional setting. After 12 weeks’ treatment, the participants qualitatively reported enhanced learning motivation in English learning. In addition, Namaziandost and Çakmak (2020) investigated the self-efficacy of 58 EFL students in general English classes over a 14-week flipped treatment that emphasized active learning, group discussions, collaborations, problem-solving tasks, and other activities. Their study results showed that the participants studying general English through flipped learning developed higher levels of self-efficacy than their counterparts in the traditional classroom. Similarly, Hsiao et al. (2021) explored how the flipped classroom approach might affect EFL college students’ self-efficacy in an English for specific purposes class (i.e. English for Tourism). They also found that the experimental group significantly outperformed the control group in terms of self-efficacy development. Last but not least, the study by Chuang et al. (2018) was one of the very few studies that looked at both motivation and self-efficacy, among other factors. A seven-week experiment was undertaken involving 85 undergraduates who studied English in a vocational educational context. The study showed mixed results: the participants’ motivation had significantly positive effects on their reading performance, but their self-efficacy had none. Further investigation is urged to shed light on reasons why Chuang et al.’s findings differed from most studies, which claimed that motivation and self-efficacy both affected learning achievements (e.g. Asakereh & Dehghannezhad, 2015; Y. Y. Chen et al., 2021; Heydarnejad et al., 2022; Lin, 2015, 2016; Zhou et al., 2022).
In a nutshell, the discussion above highlights the need to examine also language learners’ motivation and self-efficacy in flipped learning contexts. This is especially important with low-proficiency EFL learners, since the empirical research in the flipped classes discussed above has always involved only learners with average or higher English proficiency. The fact that no consistent or sufficient evidence is yet in the field makes this an even more pressing issue to confront.
II The present study
The research design is presented in the flowchart (see Figure 1).
Flowchart of the research design.
1 The preliminary study: Targeting common grammatical errors
To determine what grammar should be taught in the experiment, a preliminary study was conducted. It used Burt’s (1975) error analysis on 250 writing samples collected at a national high school in East Asia, which was also the experimental site for the main study. The samples included students’ homework, in-class practices, and writing tests. The analysis results revealed the following as the most prevalent grammar errors that need improving: tenses, pronoun references, adjective clauses, sentence structure (including sentence fragments, comma splices, and run-on sentences), and noun clauses. In the main study, course materials along with pedagogical treatments were thus prepared to cope with these grammatical areas.
2 Participants
Convenience sampling was employed and two intact classes were chosen. Each offered to students whose performance in their previous year’s English grammar classes had failed to meet the benchmark set by the school. After consenting to participate in the study, the classes were randomly assigned into an experimental group (EG, 27 students) learning with the flipped learning method and a contrast group (CG, 25 students) taught by the conventional approach. Overall, the two groups had similar English proficiency at the A2 level in the Common European Framework of Reference for Languages.
3 Treatment
The EG and the CG underwent a two-month teaching treatment with equivalent week arrangements, targeted grammar points and the instructor. The groups received a 120-minute session per week. Each session dealt with one grammar point: orientation and a pre-test (Week 1), tenses (Week 2), pronoun references (Week 3), adjective clauses (Week 4), sentence structure (Weeks 5 and 6), noun clauses (Week 7), and a review and post-test (Week 8). The main difference between the groups was the teaching treatment. The details below that were presented to the groups were also different, due to the nature of the two different teaching approaches used.
a The experimental group
The EG was taught through the FLIP model (Hamdan et al., 2013; FLN, 2014), which features four pillars: flexible environment, learning culture, intentional content, and professional educators. First of all, a flexible environment allowed students to learn the grammar units both online and in the physical classroom. This involved three main stages. First, the EG students had to finish the pre-recorded clips on the teaching platform. The clips were created by the instructor; they mainly consisted of an explanation and examples of English grammar as well as the instruction for each pre-class task. In view of the attention span of the low-proficiency participants (Ryan & Reid, 2016), each clip was about 15 to 20 minutes on average. Second, after watching a clip, they had to complete a corresponding pre-task that helped them examine and develop their understanding of each grammar unit. For example, in one of the pre-tasks, the students had to practice using adjective clauses to describe the features of an unforgettable gift. They then posted their descriptions in the forum section online for peer comment. The third stage was mostly reflected in the learning culture of the flipped learning approach. The instructional approach and scaffolding activities engaged the EG students so that they could collaborate with one another in various classroom-based activities, including pair or group discussion, individual or group writing tasks, and even oral presentations. At this stage, the interactive feedback from both peers and the instructor was important because this would stimulate the students to express their views. The instructor’s feedback and instruction should also lead the learners to self-scaffold their grammatical knowledge so that the students could observe and categorize the use of the tenses. This manifested the intentional content. Class discussion, writing projects, and oral presentations were then implemented to develop students’ higher-order thinking skills. Throughout these activities, they were expected to analyse and evaluate others’ written products and even create their own work orally or in writing. Take the unit on noun clauses, for instance. At the end of this unit, the students read the given news articles, wrote a short summary using reported speech, and finally gave a one-minute presentation. Lastly, throughout the treatment, all the instructional design, course content and classroom activities necessitated the presence of the professional educator, who was a certified high-school teacher with experience of running flipped classes. This background enabled him to facilitate the students’ learning online and in-class, provide personal and group feedback, and assess their learning process as well as the outcomes.
b The contrast group
The instructor deductively taught all the grammar points to the CG. Most of the teaching was lecture-based. He would usually begin by writing a grammatical rule or point on a blackboard and then orally explain it to the class. Then he led the class through drill practices presented in handouts to familiarize them with the corresponding sentence pattern. One of such examples called for supplying the missing words in a sentence. The teacher clarified the grammar used in the sentences to the students and guided them in filling the blanks to preserve the correct grammar patterns. The class concluded by students doing other similar drills, such as answering multiple-choice grammar questions, filling the blanks in short paragraphs, or doing translation: typical tasks administered in grammar classes at the research site.
4 Instruments
The following instruments triangulated the findings (Turner, 2014): the grammar tests, the writing tests, the questionnaire and the semi-structured interviews.
a The grammar tests
Two self-developed grammar tests were administered to both groups before and after the treatment. The tests were created to test students’ understanding of the five target grammar areas selected for the experiment. Each test consisted of 25 four-option multiple choice questions. To maintain the test validity and reliability, two English teachers at the experimental site and a professor specialized in language teaching helped evaluate the test items. Afterwards, 50 other students were recruited for a pilot study on the tests. The results showed that the tests had good item difficulty (averagely at 0.6) and item discrimination (higher than 0.3) (Boopathiraj & Chellamani, 2013). The final versions were verified and randomly assigned to a pre-test and a post-test, each requiring 30 minutes for answering.
b The writing tests
In both pre-test and post-test, the groups were asked to write an English paragraph within 40 minutes. This examined whether the participants could apply the grammar learnt to their writing by creating meaningful and correct sentences. This is important because creating a piece of writing using learnt grammar should reflect the participants’ higher-order thinking skills, the results of which can complement the outcomes of their lower order thinking (i.e. the multiple-choice tests).
Two raters (the instructor and an experienced secondary-school English teacher) marked all the writing samples, with a specific focus on the five grammar points taught in the study. They marked them by judging whether grammar points had been correctly applied and whether the ideas were fluent or logical in the sentence and context. The scores were grammatical corrective ratios. For example, if a sample included 8 uses of tenses and 5 of them were correct, then the score for tenses was 62.5%. The average of the corrective ratios for the five types of grammar represented the overall scores of a writing sample. To ensure reliability, the raters’ scores were examined against one another, and a strong inter-rater reliability was obtained (kappa coefficient = .91, p < .001).
c Questionnaire: Motivation and self-efficacy in grammar learning
A 5-point Likert scale questionnaire with 17 items assessed both groups’ learning attitudes. The questionnaire was adapted from Lin’s (2016) study, which also investigated EFL students’ motivation and self-efficacy in grammar learning. A pilot study involving 50 other students was performed to safeguard the reliability and validity of the questionnaire. The results show that, overall, it reached high reliability (Cronbach’s α = .91) and construct validity (64.63% variance explained) (Table 1). The questionnaire encompassed three underlying factors, looking into self-efficacy in learning grammar (Items 1–7) (Factor 1: Cronbach’s α = .910; 28.87% variance explained), self-efficacy in applying specific grammar (Items 8–14) (Factor 2: Cronbach’s α = .884; 23.57% variance explained), and learning motivation with regard to grammar (Items 15–17) (Factor 3: Cronbach’s α = .709; 12.19% variance explained). The final questionnaire was administered to both groups before and again after the treatment as entry and exit questionnaires respectively.
Validity and reliability of the questionnaire.
d Semi-structured interviews
One-on-one interviews were conducted to obtain qualitative accounts of the participants’ flipped learning experience throughout the sessions. Eighteen EG students agreed to take part in the interviews. All the interviews followed the same interview guidelines. Each interview took about 40 minutes, and the participants were first asked the five same questions that led them to share information about:
their general learning experience or perceptions of flipped learning;
their judgements on the pedagogical effects of the approach;
any difficulties or challenges they encountered;
their willingness to attend similar courses; and
any other relevant comments or feedback they might have had about the treatment.
Follow-up questions were also asked where necessary, to clarify any ambiguity in the students’ responses.
5 Data analysis
To address the study’s objective, methods of both quantitative and qualitative analysis were applied. Quantitatively, paired-sample t-tests were performed on the scores of grammar tests, the writing tests, and the learning attitude questionnaire, to examine whether the groups were at all significantly different after the treatment. Second, independent t-tests examined the same variables to test any significant differences between the groups. Third, effect sizes were calculated using the formula d = (M1 – M2)/SD (Plonsky & Oswald, 2014) to determine the magnitude of the effect observed; a d value of .80 indicates a large effect, .50 a medium effect, and .20 a small effect.
Qualitatively, to portray the nature of the flipped learning experience of English grammar in the high-school setting, the interview data were processed under the phenomenological analysis approach (Moustakas, 1994). Briefly speaking, after producing a complete list of interviewees’ responses, any common, relevant statements were kept while ambiguous information was eliminated. The remaining statements were thematically developed into groups of invariant constituents. Finally, the corresponding themes and descriptions generated a textual-structural account, presenting the overall flipped learning experience of the EFL low-proficiency participants.
III Quantitative results
1 Entry behaviours
Table 2 reveals no statistical differences between the groups in their grammar test (t = −.141; p = .888), writing test (t = .146; p = .884) and questionnaire outcomes (t = −.582; p = .563). This indicates that before the treatment the groups had similar grammar proficiency and learning attitudes, verifying the prerequisite for the subsequent analysis.
The entry behaviours of the groups.
Notes. EG = experimental group. CG = contrast group.
2 Changes in grammar performance
Paired sample t-tests examined the participants’ performance in grammar after the experiment. As Table 3 shows, both the EG (t = −9.20; p < .001; d = 2.57) and the CG (t = −8.47; p < .001; d = 1.38) significantly improved in their grammar test, with a large effect size. This means that both approaches had significant pedagogical effects on EFL low-proficiency students’ learning of grammar. However, independent t-tests revealed that the EG significantly outperformed the CG (t = 3.24; p = .002), with a strong effect size (d = .910) (Table 4). This further suggests that the flipped learning treatment had significantly greater effects on the participants’ grammar learning than the conventional approach did. It is also worth noting that the SDs (standard deviations) of the EG decreased by 6.89 while those of the CG slightly increased. This may suggest that the flipped learning model could help enhance homogeneity among the participants, in comparison with the conventional approach.
Paired sample t-test comparing the grammar pre- and post-test.
Notes. EG = experimental group. CG = contrast group.
Independent t-tests comparing the grammar post-test.
Notes. EG = experimental group. CG = contrast group.
3 Changes in writing performance
A battery of paired sample t-tests examined whether the groups improved their performance in writing. First, as Table 5 shows, compared to their overall performance in the pre-test writing, the EG gained significantly higher scores in the post-test (t = −10.6; p < .001), with a strong effect size (d = 2.79). This means that flipped learning had a significant effect on helping the participants to apply grammar to correct and meaningful sentences created in writing. The same effects were also found in the five grammar points at issue because the EG also made significant gains in them all: tenses (t = −5.91; p < .001; d = 1.29), pronoun references (t = −4.77; p < .001; d = 1.18), adjective clauses (t = −5.20; p < .001; d = 1.25), sentence structure (t = −5.13; p < .001; d = 1.59) and noun clauses (t = −8.51; p < .001; d = 2.16). Although the conventional teaching approach was also found to be significantly effective, its effects were only on the CG students’ overall writing performance (t = −4.67; p < .001; d = .928) and two specific grammar points: sentence structure (t = −2.56; p = .017; d = .606) and noun clauses (t = −3.62; p = .001; d = 1.13).
Paired sample t-tests for the writing pre- and post-tests of the EG and CG.
Notes. EG = experimental group. CG = contrast group. T = tenses. PR = pronoun references. AC = adjective clauses. ST = sentence structure. NC = noun clauses.
Independent t-tests showed further evidence that, under comparison, the flipped learning approach was indeed a significantly more effective approach than conventional teaching. As Table 6 reveals, an overall, significant difference was found between the post-test writing scores of the EG and the CG (t = 4.21; p < .001), with a strong effect size (d = 1.18). Moreover, except for noun clauses, the groups statistically differed in their scores for tenses (t = 3.21; p < .001; d = .893), pronoun references (t = 4.53; p < .001; d = 1.27), adjective clauses (t = 3.02; p = .004; d = .837) and sentence structure (t = 3.76; p < .001; d = 1.04). Taken together, the flipped learning treatment positively affected the EG’s use of the targeted grammar skills in their writing.
Independent t-tests for the writing post-test between the experimental group (EG) and contrast group (CG).
Notes. T = tenses. PR = pronoun references. AC = adjective clauses. ST = sentence structure. NC = noun clauses.
4 Changes in learning attitudes
As with the results of the grammar and writing tests, whereas both groups improved their overall scores in the exit questionnaire and those of its underlying factor (Table 7), the EG (M = 4.10; SD = .28) outperformed the CG (M = 3.48; SD = .45) in their overall learning attitudes (t = 5.89; p < .001), with a large effect size (d = 1.65) (Table 8). Such statistically significant differences were also observed with Factor 1 (t = 5.55; p < .001; d = 1.57), Factor 2 (t = 4.63; p < .001; d = 1.31) and Factor 3 (t = 2.34; p < .001; d = .655). The results show that flipped learning also had greater effects on the motivation and self-efficacy in grammar learning of EFL low-proficiency students. Likewise, it should be noted that the SDs of the overall questionnaire results and those of each factor of the questionnaire were almost all halved in the EG, but those of the CG remained similar or even increased (i.e. Factor 2). This suggests that flipped learning could lead to a more homogenous learning group.
Paired sample t-tests on the questionnaire.
Notes. EG = experimental group. CG = contrast group.
Independent t-test comparing the groups’ questionnaire outcomes.
Notes. EG = experimental group. CG = contrast group.
IV Qualitative results
This section presents the results of the participants’ flipped learning experience. Four main thematic results were found and are presented below.
1 Positive changes to low-proficiency students’ learning behaviours by flipped learning
Most of the EG participants endorsed their new learning experience, and conversely considered their past grammar learning experiences to be ‘boring’, ‘prosaic’, ‘full of practices and drills’. They favourably described their grammar learning in the flipped learning method, calling it ‘interesting’, ‘innovative’, ‘inspiring’ and ‘motivating’ (S1, S2, S3, S4, S7, S9, S13, S15, S17, S18). Some also submitted the view that such learning helped them ‘change [their] way of studying’ (S3, S15), ‘develop the previewing habit’ (S2, S12) and even ‘regain [their] passion for learning grammar’ (S9, S11). These changes were also reflected in their in-class performance, for several recalled that they were now ‘more willing to take part in both class tasks and group discussion’ (S6, S12, S17). Such positive changes in learning behaviours may be ascribed to ‘preview videos’ (S6, S8, S17) and ‘the guidance worksheet’ (S3, S5, S18), which jointly ‘engage [students] in learning activities’ (S4, S7, S12, S13).
2 A learner-friendly environment for low-proficiency students to engage themselves in learning
A ringing endorsement of flipped learning derived from the ‘new friendly learning environment’ (S2, S16, S18) where the participants found themselves at ‘the centre of the classroom’ (S3, S7, S15, S18). This new setting helped motivate the EG participants to learn. As several students commented, they ‘liked the teaching method’ because they ‘learnt from . . . peers through various in-class activities’ (S2, S8, S13, S16), ‘enjoyed working with them to observe grammar rules’ (S8, S12, S14), and ‘found presentation and writing projects interesting’ (S6, S17). As some elaborated, ‘the collaborative learning mode gives me somehow a sense of achievement’ (S15); ‘I was more comfortable when it was my classmates who corrected me, and I felt I could freely share my thoughts with them’ (S17). Students found themselves ‘willing to learn’ (S5, S9) because such a learning environment gave them ‘comfortable’ (S4, S13, S14) and ‘confident’ (S1, S7, S18) feelings. As S12 concluded, ‘once I felt supported in class, I want to learn and to think about the grammar use’.
3 Challenging but effective flipped learning experiences
Although at first the EG students considered their learning ‘challenging’ (S3, S9, S10, S11, S18) and ‘tough’ (S5, S12, S17), they eventually complimented the flipped learning approach on the ‘effective learning experience’ provided (S2, S3, S9, S11, S12, S16, S18). For example, S9 and S12 complained that they ‘had to watch the preview videos . . . at least two times so as to finish the worksheet and pre-tasks’. However, many of the students commented later. It was work of this kind that stealthily helped them ‘dominate [their] learning in advance’ (S11) and ‘make [their] learning effective’ (S7, S8, S14, S18). As S13 shared, ‘I replayed when I felt confused about what the teacher said’; S2 added that ‘the questions asked in the video and worksheet somehow make me find the answer by myself’. This pre-learning helps the students ‘control . . . [their] learning’ (S2) and ‘get prepared before the class’ (S4). As some collectively commented, the preview tasks could ‘strengthen [their] understanding of each grammar point’ (S1, S5, S6, S12). After the students familiarized themselves with the preview content, they felt they ‘learnt better’ during the class (S6, S9, S14, S17). S10 expressed the view that ‘knowing the basic concept of the grammar beforehand would assist me in discussing and practicing it with my group’. S14 further supported this point: ‘getting understanding at home really makes it easier to finish all the in-class work’.
4 The low-proficiency learners’ changes in feelings from anxious to self-assured
The flipped learning approach positively changed students’ feelings about learning. When the course started, most of the students had feelings of ‘stress’, ‘high-pressure’ and ‘fear’ (S3, S7, S8, S11, S13, S16, S17, S18). S8, S10 and S13 shared the complaint that ‘this class is a heavy burden for . . . [them] especially in the first two weeks’. This is because they either felt ‘anxious about watching the pre-class clips’ (S10, S16) or ‘nervous about so . . . many in-class activities’ (S8). In particular, many others also find it stressful to have to ‘keep listening, reading, writing and speaking English’ (S2, S5, S6, S11, S12, S13, S16, S17). However, as each session progressed, their attitudes changed. S14 stated ‘I found myself getting used to this way and felt less anxious’. This self-assured feeling is favoured in tandem with some reasons addressed. S8 and S17 jointly observed that ‘the percentage of Chinese and English is changing’; S17 realized that ‘gradually increasing the use of English is a good way for me’. S15 thought that ‘preview tasks really help with my readiness in class’. S2 excitedly explained that ‘I feel confident when the teacher moves onto some advanced usages and even when we [students] are doing practicing and discussion’. S6 ‘felt happy because I could catch up with what the teacher and my classmates said’. As S17 asserted, ‘I feel that I can pass the course this time’.
V Discussion
This mixed methods study explored the effects in East Asia of flipped learning on low-proficiency students’ English grammar learning outcomes and perceptions of this learning approach, in response to urgent calls from Turan and Akdag-Cimen (2020), Lee and Wallace (2018), and Lo and Hew (2017). Quantitatively, grammar tests were employed to examine two groups’ performance after the treatment and writing tests were used to investigate whether they could effectively produce correct and meaningful sentences with the learnt grammar concepts in writing. Additionally, a questionnaire was administered to identify the groups’ affective changes. Qualitatively, the interviews collected the EG’s perceptions of and experiences in learning grammar within the flipped learning context. The findings were mixed, thus meriting discussion below.
This empirical assessment of the flipped classroom contributes to the literature by expanding the effects of flipped learning to EFL low-proficiency learners. First, the results of the grammar tests indicate that the EG significantly outperformed the CG. This concurs with the pedagogical effect of flipped learning on English learning in general (Hung, 2015; Lee & Wallace, 2018) and on grammar learning specifically (Liu et al., 2019). Second, the results of the writing tests indicate that the EG and the CG improved in their overall scores, suggesting that both approaches could develop students’ higher order thinking skills. However, the EG had significant gains in all five grammar points while the CG had gains only in sentence structure and noun clauses. Furthermore, the EG achieved statistically higher scores on average and on the four other grammar points. Given the findings, it does not seem premature to claim that flipped learning can more effectively help low-proficiency students develop lower- and higher-order thinking skills than the other method can. Moreover, it is worth noting that the standard deviations (SDs) of the EG’s overall performance in grammar and writing greatly declined whilst those of the CG did not. This means that flipped learning may enable learners to mitigate their proficiency differences in English grammar, particularly low-proficiency learners like those in the present study. This echoes the research in both science-based fields (Chu et al., 2019) and language classes (Hsiao et al., 2021). The nature of flipped learning may explain the positive findings. In the EG’s interview accounts, they acknowledged that the approach created a learner-friendly, interactive learning context. Additionally, the qualitative findings showed that the EG capitalized on cooperative learning, which may explain the improved homogeneity in the EG. Conversely, the intensified differences in grammar performance in the CG may imply that the conventional teaching approach may not equally suit all low-proficiency students. Reasons for the CG’s wider gap need more investigation, but this present result may back up previous claims that traditional teaching needs to be changed, especially in East Asia (Lee & Wallace, 2018; Lin & Lee, 2019).
Correspondingly, the results for motivation and self-efficacy in grammar learning are in line with those for learning achievement. Although both groups improved, the EG significantly nurtured stronger learning attitudes than the CG did. This confirms the positive impact of the flipped learning approach on motivating students in the EFL context, as reported by previous scholars. For example, Liu et al. (2019) claimed that flipped students enhance their motivation and self-efficacy in English grammar. Hsiao et al. (2021) also found their flipped group revealing higher English learning self-efficacy. Both studies attributed their findings to the high-level engaging activities presented, which also may be seen in the current treatment. Although the EG students engaged in challenging activities, they held positive learning attitudes. Such results in particular lend further support to the recommendation that flipped learning should be effective for the cohort of low-proficiency students in tertiary (Nouri, 2016) and secondary sites (Yang et al., 2019). However, the fact must not be ignored that the conventional teaching approach still had significant effects on low-proficiency grammar students. While this improvement may be treated as evidence that the present researcher was not biased when teaching the CG, it is also necessary to explain why the approach was effective. A plausible explanation may be ascribed to the purpose of the class. The experimental treatment by definition offered intensive training; within the 8-week course, the participants spent more learning hours over a semester than they had usually done. This probably enhanced their learning and helped them develop positive learning attitudes even though the approach was as usual. The fact that the focus of the treatment was on their usual grammatical mistakes might also have given them better motivation to remedy their weaknesses than that in a regular English class. Nonetheless, readers are still advised to be cautious when thinking of practicing the conventional approach in an intensive manner. This is not only because the approach was less effective than flipped learning, but because it caused individual differences to increase.
Qualitatively, the participants made an overall favourable judgement on the pedagogical effects of flipped learning. They approved of their gratifying learning processes, especially in comparison with their past traditional learning experience. Such findings corroborate those by Chen Hsieh et al. (2017) and Doman and Webb (2017). They jointly reported that the flipped approach is applicable to Asian contexts because its nature provides learners with varied opportunities to interact with learning materials, peers, and the instructor. Moreover, the fact that the current EG participants reported themselves engaged in the various pre- and in-class activities echoes the finding by Hung (2015); her participants also recognized the nature of flipped learning and showed high participation levels. Lastly, it is true that the EG participants in the study commented that some coursework was challenging and thus made them anxious at first. However, they gradually began to acknowledge the process and judged it to be effective in learning. Over time, they eventually developed even self-assured attitudes over the process.
Taken together, the present empirical evidence concurs with the four flipped pillars for all subjects (Hamdan et al., 2013; FLN, 2014) and Hung’s (2017) adjusted model for language learning. With this, the present study pedagogically proposes the following expanded flipped model for EFL low-proficiency learners:
Flexible environment for low-proficiency learners prioritizes learners’ attention span and proficiency levels by establishing pre-class videos and providing clear guidance when preparing materials.
Learning culture for low-proficiency learners actively engages them in a wide variety of opportunities to maximize their use of the core content in class by means of peer-based or team-based discussion and interaction.
Intentional content for low-proficiency learners is constructed with purposeful linguistics and content input through varied language learning theoretical frameworks. This aims to trigger learners’ lower- and higher-order thinking skills, from observing, understanding, practicing, creating to mastering.
Professional educator in EFL contexts meticulously designs courses involving the meaningful connection of all teaching elements; attends to learners’ differences during study; facilitates their learning path; and eventually transforms them into active and self-assured learners.
The expanded flipped model proposed here may serve as a curriculum design framework for language educators and researchers who would like to work with EFL low-proficiency students.
VI Conclusions
This study empirically assessed the pedagogical suitability of the flipped learning approach and broke through the earlier limitations by focusing on EFL low-proficiency students. Overall, the present study elicits positive evidence in favour of the flipped learning effects, making it possible to recommend the approach for EFL grammar students in general and particularly for low-proficiency students. Whereas the current study validates itself and contributes to the literature, it has some limitations that require future reinforcement. It should first be made clear that the teaching treatment was comparatively short in length (two months), although some previous studies had incorporated similar or even shorter time controls in experiments. Moreover, a sample drawn from one research site at a certain period may seem – to rigorous scholars – insufficient for generalization. A larger, longitudinal study is thus suggested. Additionally, the present study focused on learners’ experience only. Future researchers may like to consider also the perspectives and perceptions of the teachers who work with low-proficiency learners in the flipped learning context.
