ArticlePDF Available

Abstract and Figures

Flipped learning has become a popular approach in various educational fields, including second language teaching. In this approach, the conventional educational process is reversed so that learners do their homework and prepare the material before going to class. Class time is then devoted to practice, discussion, and higher-order thinking tasks in order to consolidate learning. In this article, we meta-analysed 56 language learning reports involving 61 unique samples and 4,220 participants. Our results showed that flipped classrooms outperformed traditional classrooms, g = 0.99, 95% CI (0.81, 1.17), z = 10.90, p < .001. However, this effect had high heterogeneity (about 86%), while applying the Trim and Fill method for publication bias made it shrink to g = 0.58, 95% CI (0.37, 0.78). Moderator analysis also showed that reports published in non-SSCI-indexed journals tended to find larger effects compared to indexed ones, conference proceedings, and university theses. The effect of flipped learning did not seem to vary by age, but it did vary by proficiency level in that the higher proficiency the higher the effects. Flipped learning also had a clear and substantial effect on most language outcomes. In contrast, whether the intervention used videos and whether the platform was interactive did not turn out to be significant moderators. Meta-regression showed that longer interventions resulted in only a slight reduction in the effectiveness of this approach. We discuss the implications of these findings and recommend that future research moves beyond asking whether flipped learning is effective to when and how its effectiveness is maximized.
Content may be subject to copyright.
https://doi.org/10.1177/1362168820981403
Language Teaching Research
1 –25
© The Author(s) 2020
Article reuse guidelines:
sagepub.com/journals-permissions
DOI: 10.1177/1362168820981403
journals.sagepub.com/home/ltr
LANGUAGE
TEACHING
RESEARCH
The flipped classroom in
second language learning:
A meta-analysis
Joseph P. Vitta
Rikkyo University, Japan
Ali H. Al-Hoorie
Royal Commission for Jubail and Yanbu, Saudi Arabia
Abstract
Flipped learning has become a popular approach in various educational fields, including second
language teaching. In this approach, the conventional educational process is reversed so that
learners do their homework and prepare the material before going to class. Class time is then
devoted to practice, discussion, and higher-order thinking tasks in order to consolidate learning.
In this article, we meta-analysed 56 language learning reports involving 61 unique samples
and 4,220 participants. Our results showed that flipped classrooms outperformed traditional
classrooms, g = 0.99, 95% CI (0.81, 1.17), z = 10.90, p < .001. However, this effect had high
heterogeneity (about 86%), while applying the Trim and Fill method for publication bias made it
shrink to g = 0.58, 95% CI (0.37, 0.78). Moderator analysis also showed that reports published
in non-SSCI-indexed journals tended to find larger effects compared to indexed ones, conference
proceedings, and university theses. The effect of flipped learning did not seem to vary by age,
but it did vary by proficiency level in that the higher proficiency the higher the effects. Flipped
learning also had a clear and substantial effect on most language outcomes. In contrast, whether
the intervention used videos and whether the platform was interactive did not turn out to be
significant moderators. Meta-regression showed that longer interventions resulted in only a slight
reduction in the effectiveness of this approach. We discuss the implications of these findings and
recommend that future research moves beyond asking whether flipped learning is effective to
when and how its effectiveness is maximized.
Keywords
CALL, flipped learning, foreign language learning, research synthesis, second language learning
Corresponding author:
Joseph P. Vitta, Rikkyo University, Nishi-Ikebukuro, Toshima-ku, 171-8501, Japan
Email: vittajp@rikkyo.ac.jp
981403LTR0010.1177/1362168820981403Language Teaching ResearchVitta and Al-Hoorie
research-article2020
Article
2 Language Teaching Research 00(0)
I Introduction
Education has traditionally been viewed as the transfer of information from the teacher
to learners within the context of the classroom, though the desire to move away from
this paradigm has existed for some time (e.g., Freire, 1968/1970). Although flipped
learning was not the first paradigm to challenge this traditional model, it has recently
emerged as a popular and topical alternative to teacher-dominated instruction across
various educational domains (van Alten, Phielix, Janssen, & Kester, 2019) and espe-
cially in the second language (L2) field (Mehring & Leis, 2018). Flipped learning (or
flipped classrooms) is colloquially described as a process of ‘flipping’ what has tradi-
tionally been done inside the classroom to independent homework activities preced-
ing the lesson. Thus, the lesson involves problem-solving and higher-order thinking
tasks traditionally assigned to subsequent homework activities (Låg & Sæle, 2019;
Mehring, 2016, 2018). Over the past several decades, flipped learning has become
one of the most discussed trends in education for both practitioners and researchers.
Consider that a non-profit organization, the Flipped Learning NetworkTM (www.
flippedlearning.org; see Hamdan, McKnight, McKnight, & Arfstrom, 2013), has also
been established to help teachers flip their classrooms more effectively, while confer-
ences are regularly taking place around the globe for teachers to share techniques and
tips on this approach.
Another clear indication of the interest flipped learning has generated is the amount
of research conducted on it. Meta-analyses and systematic reviews subsequently appeared
across varied domains such as higher education (Lundin, Bergviken Rensfeldt, Hillman,
Lantz-Andersson, & Peterson, 2018), engineering education (Lo & Hew, 2019), health
professions education (Hew & Lo, 2018), nursing education (Xu et al., 2019), and L2
learning (Turan & Akdag-Cimen, 2019). In all of these review studies, the number of
flipped reports has increased dramatically over time. The same trend is observed in com-
prehensive reviews comparing the effectiveness of flipped learning across educational
domains (e.g. Cheng, Ritzhaupt, & Antonenko, 2019; Låg & Sæle, 2019; Shi, Ma,
MacLeod, & Yang, 2020). In sum, flipped learning has grown to be one of the most
influential phenomena within the broad educational arena.
While flipped classroom research has grown exponentially in recent years and has
been the focus of several meta-analyses and systematic reviews, the effectiveness of the
flipped classroom for L2 learning has admittedly been under-researched. Consider a
recent synthesis by Turan and Akdag-Cimen (2019), who conducted a systematic review
of 43 published L2 reports. Because their work was a systematic review of published
reports, their article does not present summary effect size estimates or moderator analy-
ses, nor does its scope cover unpublished reports, thus raising the risk of publication bias.
In their recent meta-analysis of flipped classroom interventions across educational
domains, Shi et al. (2020) included only six L2 reports that were subsumed under a gen-
eral ‘social sciences’ label. Strelan, Osborn, and Palmer (2020) in a similar vein sub-
sumed second language flipped reports under a broader ‘humanities’ label. L2 reports
were also subsumed under humanities in another comprehensive recent flipped meta-
analysis (Låg & Sæle, 2019) with the authors noting that about 70% (k = 23) of these
humanities reports were L2-focused.
Vitta and Al-Hoorie 3
Considering the above, there appears to be a lack of focus on meta-analytic work on
L2 flipped learning. This is problematic given the diverse range of skills (e.g. writing and
reading) and underlying competencies (e.g. vocabulary and pragmatics) that underpin L2
proficiency theory and frameworks (Council of Europe, 2001, 2011; Green, 2012;
Halliday & Matthiessen, 2014) which often act as L2 learning outcomes as highlighted
in Green’s (2012) presentation of English Profile Programme, an L2 curriculum plan
operationalizing the Common European Framework of Reference for languages (CEFR;
Council of Europe, 2001, 2011) guidelines. The subsuming of L2 studies under ‘humani-
ties’ without further exploration (e.g. Strelan et al., 2020) does not fully account for the
effectiveness of L2 flipped learning given the range of skills and competencies that
underpin language learning. In other words, an L2-focused meta-analysis of flipped
learning interventions is needed to better understand how effective flipped learning is
within the multidimensional L2 space, and we address this gap in this current article. As
with recent meta-analyses of L2 classroom pedagogy, such as Bryfonski and McKay’s
(2019) task-based learning meta-analysis, the current meta-analysis has implications for
both future research and practice, and this has been considered in its construction and
analysis of findings.
II Flipped learning
1 Definition of flipped learning
Despite its popularity, flipped learning has been somewhat inconsistently defined by edu-
cational researchers and practitioners (Mehring & Leis, 2018; van Alten et al., 2019). In a
general sense, there is agreement on the ‘flipped’ or ‘inverted’ aspect of the approach,
where classroom teaching and independent learning are switched. The disagreement is on
what exactly flipping a classroom means. For some (e.g. Bergmann & Sams, 2012;
Mehring, 2018), the essence of the flipped approach is this pedagogical shift to presenting
new content before class, allowing the teacher and students to apply this new content in
meaningful ways during class time. The manner in which the content is presented to stu-
dents outside of the class (i.e. via technology or not) is assumed to be inconsequential.
From this perspective, flipped learning has been described as having its roots in the 1980s
when active learning emerged within educational circles emphasizing learning by doing
(Ryback & Sanders, 1980). For other flipped theorists (e.g. Adnan, 2017; Evseeva &
Solozhenko, 2015), flipped learning is heavily dependent on (digital) technology to allow
students to engage with new content. This latter definition appears to be especially popular
in recent L2 flipped learning scholarship, where both primary studies (e.g. Chen Hsieh,
Wu, & Marek, 2017; Hung, 2015, 2017) and research syntheses (e.g. Turan & Akdag-
Cimen, 2019) have included or emphasized technology in their definitions of flipped appli-
cations. In the context of the present study, we have followed the example of recent flipped
learning meta-analyses (e.g. Låg & Sæle, 2019) and adopted a general definition where: “A
flipped intervention first involves presentation of new content to learners to be indepen-
dently studied before class, and then class time is devoted to reinforcing and engaging
with the ‘flipped’ content”.
4 Language Teaching Research 00(0)
2 Anticipated effects of flipped learning
In addition to disagreement on certain definitional specifics, the case has been made both
for and against flipped learning. The case for flipped learning is grounded in its optimiza-
tion of class time use (Mehring & Leis, 2018; Voss & Kostka, 2019). Compared to tradi-
tional lecturing, flipped learning pushes learners toward developing ‘the upper cognitive
levels of the taxonomy where knowledge application and skill building are happening’
(Davis, 2016, p. 2). Concerns about flipped classrooms has been expressed along at least
three lines.
The first concern is that learners have to be able to comprehend the flipped content
independently. This can be challenging for learners with lower proficiency levels, espe-
cially when such material is in the target language. In this vein, Milman (2012) suggested
that flipped learning might not be ideal for such learners as they may not have the chance
to ask for clarifications in real time. Milman also posited that procedural, factual, con-
ceptual, and metacognitive knowledge were best for flipped interventions, while L2
learning outcomes such as initial vocabulary learning (meaning to form mapping; Nation,
2013) may not always clearly fit into these areas.
The second concern is that the demand of flipped learning can be impractical. Mehring
(2016), for instance, argues that flipping classes requires a substantial amount of effort
and planning from the teacher. On the student side, because this approach requires each
student to show a level of proactivity and self-directedness in learning the new content
before class, flipped lessons could fail if the students, in the aggregate, do not effectively
perform the activities assigned to them (Mehring, 2018). Language learners, especially
those with lower proficiency levels, may not be able to benefit fully from independent
study outside of class.
Finally, there is also the argument, as summarized by Webb and Doman (2016), that
modern language teaching approaches such as communicative language teaching and
thereby task-based language teaching (see Ellis, 2009) are essentially flipped learning by
another name. The argument here is students prepare for the communicative activity and/
or task before engaging with it (e.g., pre-task; see Ellis, 2009) and thus the ‘flip’ has already
been baked into these approaches. Mehring (2016, 2018) noted however that flipped learn-
ing is not defined by interaction on even pre-task planning but by the agency that students
are provided while studying content on their own and then by the active engagement with
that content during class time. Overall, the above concerns make it essential to find out how
effective flipped learning is when it comes to the long and arduous process of L2 learning,
and whether its effectiveness varies by learner level and target L2 outcome.
3 Applications of L2 flipped learning
Considering the various skills involved in L2 learning, research has examined the effec-
tiveness of the flipped approach on different learning outcomes. For example, some
interventions targeted writing performance (e.g. Leis, Cooke, & Tohei, 2015), vocabu-
lary development (e.g. Oh, 2017), and standardized tests such as TOEIC (e.g. Ishikawa
et al., 2015). These investigations were implemented on a range of age groups, from
elementary (e.g. Baş & Kuzucu, 2009) to adult learners (e.g. Karimi & Hamzavi, 2017).
Vitta and Al-Hoorie 5
This body of research has also relied heavily on technology. Some research reflected
the argument that flipped learning went hand in hand with technology by emphasizing
the use of technology, such videos and apps, to deliver the content outside of the class-
room (Alnuhayt, 2018; Chen Hsieh et al., 2017), though little space is usually devoted to
explaining how extra class time was used. Some of these applications adopted a Web 1.0
framework (Lomicka & Lord, 2016). Mori, Omori, and Sato (2016), for instance, used
PowerPoint and other one-way technology to flip their teaching of Japanese writing
characters, kanji. On the other hand, technology was also used to facilitate student inter-
action outside of the class via learning management systems and well-known Web 2.0
applications such as chat boards and blogs (Lin & Hwang, 2018; Lin, Hwang, Fu, &
Chen, 2018).
Given the flexibility of this approach, this wide range manner in which flipped learn-
ing has been applied should not be surprising. Theorists such as Bergmann and Sams
(2012) and Mehring (2016, 2018) emphasized the need for flipped applications to maxi-
mize class time for higher-order thinking activities, while investigators such as Hung
(2015) and AlJaser (2017) detailed how the lesson was used to facilitate cognitively
engaging and student-centered tasks when describing their flipped interventions. On the
other hand, Chen Hsieh et al. (2017) and Alnuhayt (2018) focused more on how the fea-
tures of technology were used to ‘flip’ the content. L2 flipped learning applications have
thus varied in their contexts, learning outcomes, use and engagement with technology,
and focus on class time use.
III Past non-L2 flipped learning meta-analyses
While there have been arguments at the theoretical level for and against flipping class-
rooms in the context of L2 learning, the actual empirical evidence from these diverse
applications points to the approach being effective across many contexts and domains.
The results of a number of non-L2 meta-analyses conducted recently (see Table 1) show
that effect sizes tend to range from around 0.20 to just over 0.50. In one case (Xu et al.,
2019), the average effect size approached 1.80, which is substantial. This ‘extra’ large
effect might be attributed to the fact that that meta-analysis was limited to 22 published
reports related to the Chinese context, and so this magnitude may not be representative
of unpublished research and research published elsewhere.
Table 1. Examples of recent flipped learning meta-analyses in different disciplines.
Effect size (95% CI) kDomain
Strelan etal. (2020) g = 0.50 (0.42, 0.57) 198 Cross-disciplinary
Cheng etal. (2019) g = 0.19 (0.11, 0.27) 55 Cross-disciplinary
Låg and Sæle (2019) g = 0.35 (0.31, 0.40) 272 Cross-disciplinary
Lo and Hew (2019) g = 0.29 (0.17, 0.41) 29 Engineering education
Shi etal. (2020) g = 0.53 (0.36, 0.70) 60 Cross-disciplinary
van Alten etal. (2019) g = 0.36 (0.28, 0.44) 114 Cross-disciplinary
Xu etal. (2019) d = 1.79 (1.32, 2.27) 22 Nursing education in China
6 Language Teaching Research 00(0)
Furthermore, all these meta-analyses found high heterogeneity in their effects. In an
attempt to explain this heterogeneity, the researchers employed various moderators.
Examples of commonly used moderators are age of learners (or educational level), sub-
ject of study, and duration of the intervention. Låg and Sæle (2019) found that subject of
study did not predict the effectiveness of flipped learning significantly. In contrast,
Cheng et al. (2019) found that it did vary by discipline. The largest effect was for arts and
humanities, g = 0.63, 95% CI (0.16, 1.10), but the effect was non-significant for engi-
neering education, g = −0.08, 95% CI (−0.25, 0.08). Cheng et al. (2019) also investi-
gated the role of study duration. Interventions that were less than one semester tended to
obtain larger effects, though this was not significantly different from interventions last-
ing one semester or longer. Similarly, whether the intervention used videos or not did not
seem to have an effect (Lo & Hew, 2019).
Some meta-analysts examined study quality as part of moderator analysis. van Alten
et al. (2019), for example, investigated three aspects related to study quality, all of
which turned out to be non-significant. They compared allocation type (i.e. non-ran-
dom, pre-existing groups, and individual allocation), group equivalence test (tested–
equal, tested–not equal, not tested–descriptive statement, and not tested–no descriptive
statement), and report source (journal article, conference proceeding, and university
thesis). Thus, heterogeneity has been consistently obtained in flipped learning meta-
analyses, though moderators used to date tend to either fail to explain it or explain it
inconsistently.
IV The present study
As reviewed above, past meta-analyses show that the effectiveness of flipped learning
can vary by discipline, and in some cases its effectiveness is non-significant (e.g. engi-
neering education; Cheng et al., 2019). This underscores the need for an L2-focused
meta-analysis, especially since within language learning there are different skills requir-
ing different learning and teaching strategies. The present study therefore aimed to con-
tribute to the research synthesis work on flipped learning by meta-analysing L2 flipped
learning interventions. As explained above, we adopted a broad definition of flipped
learning that includes, first, learners studying the material before class whether technol-
ogy-supported or not and, second, class time is then spent on learners engaging with that
material (Låg & Sæle, 2019). In cases where technology is employed to present new
content, flipped learning, according to our definition, also becomes a specific type of
blended learning as technology is being fused with face-to-face instruction (Mahmud,
2018; Zarrinabadi & Ebrahimi, 2019). Teng (2017) captured the intersection between
flipped and blended learning when stating that flipped learning is ‘a pedagogical method
to blended learning’ (p. 114).
To be more specific, we attempted to answer the following research questions:
1. To what extent does the flipped learning approach improve L2 learning compared
to traditional classroom teaching?
2. To what extent does the effectiveness of the flipped learning approach vary by L2
learning outcome?
Vitta and Al-Hoorie 7
3. To what extent do learner characteristics (educational level and L2 proficiency
level), report characteristics (peer review and journal indexing), flipped applica-
tion characteristics (use of videos and interactive technology), and methodologi-
cal characteristics (reliability, pretesting, and duration of the intervention)
account for the observed variation in the effectiveness of flipped learning in L2
settings?
V Method
1 Inclusion criteria
In order to qualify for inclusion in the present meta-analysis, the report had to satisfy the
following inclusion criteria:
1. The report must apply a (quasi-)experimental design, whether between- or
within-group. A between-group design must involve at least one group of learn-
ers learning the material outside of class time and a comparison group learning
the same material via a traditional face-to-face approach during class time. A
within-group design must include a comparison of these two approaches alternat-
ing on the same learners.1
2. The report must establish pre-treatment equivalency between experimental
groups/conditions by either an empirical measurement (i.e. a researcher-adminis-
tered pretest relative to the outcome variable) or an argument referencing stu-
dents’ L2 ability/performance in relation to the outcome variable vis-à-vis a
standard proficiency scale, such as the CEFR.
3. The participants must be learning a language, whether English or another lan-
guage, whether as a second or additional language, and whether in a second or
foreign language context.
4. The report must include a quantitative dependent variable measuring gains in the
target language learning outcomes (e.g. vocabulary and writing proficiency).
5. The report must provide an effect size or sufficient statistics to calculate it.
6. The language of the report must be English.
2 Literature search
Following standard practice in meta-analyses, we conducted a keyword-driven database
search to build our report pool. However, given the particular features of L2 research
(discussed below), we commenced our search at the journal level and then moved to the
database level. In total, our literature search process had four stages.
Stage 1. As our meta-analysis was L2-specific, we expected the bulk of L2 flipped learn-
ing studies to be found in L2 journals. We therefore focused the initial stage of our search
on these journals (the following stages expanded this scope). We first created a list of 73
L2 and educational technology journals adapted from previous bibliometric work and
relevant flipped literature (Al-Hoorie & Vitta, 2019; Mehring, 2016; Vitta & Al-Hoorie,
8 Language Teaching Research 00(0)
2017; Zhang, 2020; for the complete list; see Appendix A). Considering the inconsist-
ency in author-supplied keywords in L2 journals (see Lei & Liu, 2019), which could
limit our ability to obtain a comprehensive list of flipped learning reports, we then uti-
lized the Scopus search engine to search articles in these journals. The Scopus search
engine permits searching the title, abstract, keyword list, and other meta-data of each
article (Burnham, 2006). We used the keywords flip*, invert*, and blend*. We included
blend* because L2 researchers tend to view flipped learning as a pedagogic approach to
blended learning (Chen Hsieh et al., 2017; Hung, 2015; Teng, 2017). Journals were
searched with an ‘all time’ parameter where each journal was searched comprehensively
without time range limitations.
Although this step helped us avoid relying on author-supplied keywords, we still
wanted to ensure that our Scopus search was indeed comprehensive. We manually
inspected all articles in all issues of eight relevant journals (CALL-EJ, Computer Assisted
Language Learning, ReCALL, Language Learning & Technology, CALICO Journal,
Teaching English with Technology, JALTCALL, International Journal of Computer-
Assisted Language Learning and Teaching). Each journal was inspected after its auto-
mated processing, and this manual search did not uncover additional reports not captured
by the automated Scopus search, thus raising confidence in our search protocols.
Stage 2. We then expanded the search to EBSCO and ProQuest. Within EBSCO data-
base, our search covered OpenDissertations, Academic Search Ultimate, ERIC, and Edu-
cation Research Complete. Within ProQuest, our search covered Educational Database,
Linguistics Database, Psychology Database, and Social Science Database, as well as
ProQuest Thesis and Dissertation Global. In addition to the search keywords above, we
further limited the search at this stage by adding L2-specific keywords (second language
or foreign language or L2 or ESL or EFL) to filter out research conducted on other par-
ticipants. As with Stage 1, there were no time constraints, and the search was performed
at the ‘full text’ level with subsequent relevance ordering to facilitate a quicker screening
of false negatives.
Stage 3. In an attempt to minimize publication bias, we issued a call for papers request-
ing reports meeting our inclusion criteria. This call for papers was announced in various
L2 outlets including Linguist List, BAALmail, Korea TESOL, and IATEFL Research
SIG, as well as social media.
Stage 4. We finally conducted a saturation search to ensure our search was comprehen-
sive. We performed an ancestry search in three recent L2 flipped learning syntheses
(Filiz & Benzet, 2018; Mahmud, 2018; Turan & Akdag-Cimen, 2019) to find out whether
they included reports not captured by our search. We also searched two generic data-
bases: Google Scholar and AskZad. These two databases contain reports from non-
indexed journals as well as theses and dissertations not found in ProQuest.
Our literature search concluded in August 2019, resulting in 56 unique reports satisfying
our inclusion criteria (for the complete list; see Appendix B). Comparing the number of
reports in our pool to the domain-specific meta-analyses in Table 1, we note that it was
larger than that by Lo and Hew (2019, k = 29), Låg and Sæle (2019, k = 23), and Xu et al.
Vitta and Al-Hoorie 9
(2019, k = 22). It was also larger than the number of quantitative reports found in L2
flipped learning systematic reviews, including Turan and Akdag-Cimen (2019, k = 21) and
Filiz and Benzet (2018, k = 25). Figure 1 presents a flow diagram of our search process.
3 Moderators
To operationalize research questions 2 and 3, we coded the reports for three groups of
moderators related to learners, report source, and design characteristics, the latter sub-
suming flipped application and methodological design features.
Regarding learner characteristics, we coded for educational stage: elementary, inter-
mediate, secondary, and university. We coded adult learners as university learners (k = 2).
Identification:
Total initial
reports
Scopus
n = 54,630
EBSCO
n =
1,355
ProQuest
n =
20,600
ProQuest
Theses
n =
148,202
Call
for
Papers
n = 4
Screening:
Reports retained
after inspection of
titles andabstracts
n = 1,781n = 719 n =
1,700
n =
1,300 n = 4
Eligibility:
Reports retained
after inspection
of full texts
n = 41 n = 23 n = 7 n = 13 n = 4
Saturation Search
(Ancestry Search,
Google Scholar,
AskZad)
n = 14
Included: Total
reports retained
after removing
duplicates
n = 56
Figure 1. Flow diagram of the search process.
Notes. † Includes all articles in the 73 journals at Stage 1.
10 Language Teaching Research 00(0)
We eventually compared secondary and university learners only due to the small number
of reports on the other educational stages (k = 3 combined). In previous L2 meta-analyses
(e.g. Bryfonski & McKay, 2019), proficiency was omitted because of the inherent diffi-
culty of standardized proficiency judgments across reports. In light of this, we imple-
mented a three-category proficiency moderator: 1) below intermediate, 2) intermediate,
and 3) above intermediate. Intermediate was anchored to B1 according to the CERF. As
an illustration, Ishikawa et al. (2015) was coded as ‘below intermediate’ as the reported
TOEIC scores were within the A2 range of 250 to 550; Karimi and Hamzavi (2017) was
coded as intermediate since reported Cambridge PET scores established a B1 level. The
remaining studies were coded in the same manner where either empirical evidence or an
argument anchoring the learners’ proficiency (e.g. to the CEFR) was presented as evi-
dence of the learners’ proficiency. Reports spanning multiple levels or omitting such pro-
ficiency evidence were not coded, and those reporting proficiency in a manner that makes
such anchoring not possible were likewise not coded.
As for report source, some reports did not undergo conventional editorial-driven peer
review (e.g. conference proceedings and university theses). Some methodologists rec-
ommend including such reports for comprehensiveness (e.g. Norris & Ortega, 2000), as
they may contain a higher proportion of statistically non-significant results (Dickersin,
2005). Similarly, it has been argued that reports published in journals have a higher like-
lihood of publication bias as significant results with noteworthy effects tend to be favored
(Fanelli, 2010). We therefore coded whether the report was published in a peer-reviewed
journal. Since there is also evidence suggesting that report quality can vary depending on
the indexing of the journal (Al-Hoorie & Vitta, 2019), we also compared these journals
in relation to their indexing in SSCI, Scopus, and other indices. Table 2 presents a break-
down of the report types in our pool.
We also examined the effect of certain design characteristics in two areas: flipped
application features and report methodological features. In relation to flipped applica-
tions, we examined the effect of whether the intervention utilized videos, and whether
the technology employed was interactive (Lo & Hew, 2019). An example of an ‘interac-
tive’ flipped intervention was Lin and Hwang (2018) where the content was presented
via Facebook, and students used the platform to discuss it with their peers and with the
instructor. In relation to methodological features, we examined whether the design
included an empirical pretest before the implementation of the treatment or relied on
Table 2. Types of reports satisfying our inclusion criteria.
Report type k
Journal:
SSCI and Scopus 14
Scopus only 12
Neither Scopus nor SSCI 19
Other:
Conference proceeding 4
Thesis/dissertation 7
Vitta and Al-Hoorie 11
pre-existing holistic judgements, whether the reliability of dependent variable scores was
reported (Al-Hoorie & Vitta, 2019; Brown, Plonsky, & Teimouri, 2018), and how long
the intervention lasted (Cheng et al., 2019).
Finally, we tested whether the effectiveness of the flipped approach was related to the
L2 outcome targeted in the report. We compared the effectiveness of flipped learning on
the four skills (listening, speaking, reading, and writing) and two competencies (vocabu-
lary and grammar). When scores were combined across two or more L2 outcomes, we
coded the report as ‘multi-outcome’. Four reports had outcomes targeting performance
on standardized tests combining reading and listening scores (e.g. TOEIC Listening and
Reading). We coded these as ‘standardized tests’.
4 Data analysis
a Software. We used Comprehensive Meta Analysis 3.3 (Borenstein, Hedges, Higgins,
& Rothstein, 2014) for all analyses. We applied a random-effects model as we had no
reason to assume one common effect size underlying all reports (see Borenstein, Hedges,
Higgins, & Rothstein, 2009). We also examined heterogeneity using the I2-statistic and
its significance value. Significant heterogeneity suggests that the effect highly varies
from report to report, and this variability could potentially be explained through modera-
tor analysis of certain report characteristics.
b Publication bias. Publication bias can occur because of the tendency of journals to
favor significant results over non-significant ones. As a result, some non-significant
findings may not find their way to the research community, leading to what is commonly
known as the file-drawer problem (Rosenthal, 1979). We tested publication bias using
the Trim and Fill method (Duval & Tweedie, 2000a, 2000b). We also examined the
results of the classic fail-safe N test (Rosenthal, 1979), Orwin’s fail-safe N test (Orwin,
1983), and the p-curve (Simonsohn, Nelson, & Simmons, 2014) to further shed light on
potential bias.
c Coding. Initially, 40 reports were coded independently by two coders against our
inclusion criteria. This procedure resulted in 85% agreement (Cohen’s = .70, p <
.001). All discrepancies were subsequently resolved by discussion until 100% agreement
was reached. The two coders then independently coded the effects of 16 reports (approxi-
mately 30%), resulting in 88% inter-coder agreement ( = .86, p < .001). All discrepan-
cies were also resolved by discussion until 100% agreement was reached. When a study
had multiple data collection points (e.g. several quizzes and a final exam), we used the
last test for the analysis (k = 5). If the report had multiple assessments for one dependent
variable (e.g. essay subdomains and an overall score), we used the most comprehensive
measure (k = 7). In one case, a report had two outcome variables; we selected the one
with the best construct validity corresponding to modern ‘complexity–accuracy–fluency’
theory (Pallotti, 2009).
d Effect size computation. Effect sizes were computed using Comprehensive Meta
Analysis software with Hedge’s g being the effect size metric employed to correct for
12 Language Teaching Research 00(0)
smaller sample sizes. Each report was weighted by the inverse of its variance including
the estimated between-studies variance. Most effect sizes were directly estimated from
the means, standard deviations, and sample sizes. In cases where these data were una-
vailable, test statistics or other effect size metrics were used in tandem with sample size
to estimate g (for detailed formulae; see Borenstein et al., 2009). Thus, all selected
reports provided enough information to estimate effects. A small number of the reports
(k = 3) used within-participant designs. According to Lakens (2013), such effect sizes
are best estimated with gav when meta-analysing them with between-participant effects.
Nevertheless, gav values are always nearly identical to gs (for between-participant effects;
Lakens, 2013), and this was the case with our data. Therefore, g has been employed sub-
suming gs and gav.
VI Results
The reports included in our pool were interventions conducted in different parts of the
world, though the target language in almost all of these reports was English. Only a
minority of reports tested the effectiveness of the flipped approach on learning other
languages, such as Chinese (k = 2), Japanese (k = 2), and Korean (k = 1). Only a few
studies, also, reported the results for each gender separately (kfemale = 5, kmale = 2)
whereas the remainder reported the results for the two genders combined. Some of these
reports were unpublished university theses/dissertations (k = 7). As mentioned above,
most of these reports adopted a between-participant design, whereas a few were within-
participant (k = 3). These reports involved 61 unique samples and 4,220 learners.
Using a random-effects model, the results showed that groups receiving the flipped
intervention achieved significantly better than those receiving traditional face-to-face
teaching, g = 0.99, 95% CI (0.81, 1.16), z = 10.90, p < .001. This average effect size
exhibited substantial heterogeneity, Q(60) = 432.82, I2 = 86.14, p < .001. These results
indicate that around 86% of the dispersion of the true effect is over and above sampling
error and is potentially explainable by certain moderator variables.
In relation to publication bias, the classic fail-safe N test showed that 694 missing
reports would be required to bring the effect size down to zero, z = 26.02, p < .001. The
Orwin’s fail-safe N also showed that 58 additional reports are needed to reduce the effect
size to below 0.40, the generally recognized threshold for effective educational interven-
tions (Hattie, 2009). These results provide strong evidence of a non-zero effect size.
Similarly, the p-curve did not indicate evidence of questionable research practices such
as p-hacking (Figure 2). The p-curve included 45 statistically significant results (p <
.05), of which 38 were significant at p < .025.
However, the Trim and Fill method did suggest the possibility of publication bias. As
Figure 3 shows, reports with smaller samples tended to find larger effect sizes. This
analysis showed that there could be at least 17 missing reports. Adjusting for these miss-
ing reports made the average effect size shrink, g = 0.58, 95% CI (0.37, 0.78). This sug-
gests that the 0.99 effect size originally obtained might be inflated.2
In relation to research question 2, the moderator analysis revealed some interesting
results in relation to the target L2 outcomes investigated (Table 3). The findings showed
that flipped learning had a non-significant effect on reading and standardized tests, as
Vitta and Al-Hoorie 13
Figure 2. Results of p-curve analysis.
-4 -3 -2 -1 01234
0.0
0.1
0.2
0.3
0.4
0.5
Standard Error
Hedges's g
Funnel Plot of Standard Error by Hedges's g
Figure 3. Funnel plot showing publication bias based on the fixed-effects model.
Note. Imputed results are filled dots.
the 95% confidence intervals overlapped with zero. The confidence intervals for the
reading outcome were also so wide that they were hardly informative, underscoring the
need for more research on reading. Vocabulary did show a significant effect, though
14 Language Teaching Research 00(0)
lower confidence interval was barely above zero. In contrast, the effects were substan-
tial for writing, listening, grammar, speaking as well as assessments comprising multi-
ple outcomes.
Table 3. Results of moderator analyses.
Subgroup k g Lower CI Upper CI Q(df) p
L2 outcome: 43.70(7) < .001
Writing 13 1.50 1.00 1.99
Listening 4 1.42 0.62 2.21
Speaking 8 1.14 0.81 1.48
Multi-outcome 14 1.03 0.65 1.41
Grammar 5 1.01 0.38 1.63
Vocabulary 9 0.25 0.03 0.47
Standardized tests 4 0.33 –0.07 0.72
Reading 3 1.25 –0.09 2.59
Educational level:
Secondary 10 1.21 0.65 1.77 1.07(1) .302
University 48 0.90 0.72 1.08
Proficiency level:
Below intermediate 13 0.65 0.34 0.96
Intermediate 16 0.89 0.58 1.20 7.12(2) .028
Above intermediate 11 1.45 0.95 1.96
Peer review status:
Journal article 50 1.07 0.85 1.28 5.27(1) .022
Other 11 0.64 0.35 0.93
Report source:
Not Scopus or SSCI 19 1.18 0.78 1.57 10.76(3) .013
Scopus only 13 1.39 0.94 1.84
Scopus and SSCI 18 0.73 0.47 1.00
Thesis/conference 11 0.64 0.35 0.93
Technology type:
Video 47 0.97 0.76 1.18 < 0.001(1) .987
No Video 13 0.97 0.62 1.31
Interactive technology:
Yes 23 1.01 0.78 1.24 0.05(1) .831
No 38 0.97 0.71 1.23
Reliability of DV:
Reported 31 1.07 0.82 1.32 0.82(1) .364
Not reported 30 0.90 0.64 1.15
Pre-test:
Empirical Pre-test 39 1.07 0.85 1.29 1.49(1) .223
Pre-existing
evaluation
22 0.84 0.53 1.14
Note. k = unique samples.
Vitta and Al-Hoorie 15
When it comes to research question 3, the results did not provide evidence that learner
age, specifically whether they are in secondary or university level, was related to how
effective the flipped intervention was. In contrast, the effectiveness of flipped learning
varied significantly in relation to proficiency level. As the post hoc results in Table 4
show, learners with the higher proficiency were the ones exhibiting the larger effect sizes.
Regarding the type of the report itself, the analysis showed that peer reviewed journal
articles reported significantly larger effects than other types of reports such as conference
proceedings and unpublished theses. Furthermore, comparison by report source sug-
gested that the largest effect sizes came from journals not indexed in the SSCI (Table 4).
Analysis of whether the intervention used videos or not, and whether the technology
was interactive, did not result in a significant difference. Similarly, whether the research-
ers reported the reliability of their dependent variables did not seem to have an effect on
the results. The same applied to whether the researchers administered their own pretest
or relied on a pre-existing judgement or evaluation reported by learners.
Finally, we examined the relationship between the length of the intervention and its
effectiveness. Meta-regression analysis showed that there was a small negative effect of
duration of the study (see Figure 4 and Table 5), suggesting that the novelty of the
approach might slightly wane with time. One report lasted for 60 weeks, which was the
longest duration in our pool. Excluding that report led only a minor decrease in the coef-
ficient from −0.02 to −0.03 (see Figure 5).
Table 4. Q-values in post hoc analyses showing whether differences in moderator levels are
significant.
Proficiency: 1 2 3
1. Below intermediate
2. Intermediate 1.20
3. Above intermediate 7.11* 3.47
Report source: 1 2 3 4
1. Neither SSCI nor
Scopus
2. Scopus only 0.50
3. SSCI and Scopus 3.246.09*
4. Thesis/conference 4.46* 7.52** 0.21
L2 outcome: 1 2 3 4 5 6 7
1. Writing
2. Listening 0.03
3. Multi-outcome 2.14 0.73
4. Grammar 1.47 0.64 0.005
5. Speaking 0.53 0.05 0.14 0.46
6. Standardized tests 13.16*** 5.78* 6.33* 3.2510.58**
7. Vocabulary 20.50*** 7.65** 12.12*** 5.01* 18.79*** 0.10
8. Reading 0.11 0.04 0.09 0.10 0.0003 1.66 2.06
Notes. p < .10, *p < .05, **p < .01, ***p < .001.
16 Language Teaching Research 00(0)
VII Discussion
The purpose of the present meta-analysis was to extend existing research synthesis work
on the effectiveness of flipped learning in the context of L2 learning. We aggregated
effect sizes in reports located through a broad literature search process that included dif-
ferent report types. In this section, we discuss the following three notable findings, in
relation to both research and practice, emerging from this meta-analysis:
There is clear evidence that flipped learning is effective for L2 learning overall
(research question 1).
Flipped learning seems more effective under certain conditions and for certain L2
outcomes (research questions 2 and 3).
Publication bias and methodological issues seem to have impeded accurate esti-
mation of the effect of flipped learning (research question 3).
Regression of Hedges's g on Duration in weeks
Duration in weeks
-20.0 -10.0 0.0 10.0 20.0 30.0 40.0 50.0 60.0 70.0 80.0
Hedges's g
6.00
5.00
4.00
3.00
2.00
1.00
0.00
-1.00
-2.00
-3.00
-4.00
Figure 4. Meta-regression of the relationship between effect size and duration of the
intervention in weeks.
Table 5. Results of meta-regression.
Coefficient SE Lower
95%
Upper
95%
Z p VIF
Intercept 1.25 0.16 0.94 1.55 8.97 < .001 2.95
Duration –0.02 0.01 –0.04 –0.0008 2.04 .041 1.00
Vitta and Al-Hoorie 17
1 Overall effectiveness of flipped learning
The main finding in this meta-analysis is that flipped learning seems to be an effective
approach for L2 learning. The overall effect size was substantial (g = 0.99), though this
magnitude might be somewhat inflated as discussed in more detail below. Virtually all
moderator analyses on L2 outcome displayed a positive effect size point estimate, with
the lowest being 0.25 in the case of vocabulary learning. These findings echo what many
teachers have probably noticed: Students who prepare for the lesson before class tend to
find it easier to understand the lesson during class. Flipped learning provides a more
systematic approach so that all learners prepare for the lesson, and then consolidate what
they have learned during class. These results, therefore, indicate that the flipped approach
is strongly recommended for language teachers.
As reviewed above, one potential argument is that the flipped approach is simply
communicative language teaching by another name (see Webb & Doman, 2016). While
the present meta-analysis was not designed to specifically engage with this debate,
closer examination of the reports in our pool supports Webb and Doman’s (2016) posi-
tion that the flipped approach and the communicative approach are not interchangeable.
Comparison groups in some studies tended to engage in communicative language
teaching-governed learning activities. As a demonstration, in Hung (2017) communica-
tive features were observed in both the flipped and non-flipped groups. Chen Hsieh
Regression of Hedges's g on Duration in weeks
Duration in weeks
-10.0 -5.0 0.0 5.0 10.0 15.0 20.0 25.0 30.0 35.0 40.0
Hedges's g
6.00
5.00
4.00
3.00
2.00
1.00
0.00
-1.00
-2.00
-3.00
Figure 5. Meta-regression of the relationship between effect size and duration of the
intervention in weeks after excluding one potentially outlier report.
18 Language Teaching Research 00(0)
et al. (2017) also had students draft ‘the final dialog collaboratively’ (p. 4) under the
conventional learning condition. Hung (2015) and Ishikawa et al. (2015) likewise had
their non-flipped learning groups engage in classroom discussions about the content
presented in class. While not all researchers intentionally used communicative activities
for their comparison groups, the fact that communicative features were observed in both
the flipped and non-flipped groups makes it unlikely that the results of the present meta-
analysis are attributable simply to communicative activities. Thus, flipped learning
applications in our report pool do not appear to be communicative language teaching by
another name, and are possibly superior in that additional, structured out-of-class activi-
ties are involved. Again, as this issue was not within the scope of the present meta-
analysis, direct comparative analysis between these two approaches seems an interesting
future direction.
2 Effectiveness according to learning outcomes and learner characteristics
As explained above, a clear implication of the results of this and previous meta-analyses
is that flipped learning is effective. The moderator analyses addressing research ques-
tions 2 and 3, however, uncovered variation in its effectiveness in relation to learning
outcomes and learner characteristics. In sum, flipped learning appears to be most effec-
tive for intermediate to above-intermediate proficiency learners where the learning out-
come is skill-based and procedural (e.g., writing).
The results showed that the lowest effects were obtained for reading, vocabulary, and
standardized test performance. One explanation for this finding is the small number of
reports addressing these areas (see Table 3), underscoring the need for future research to
fill in this gap. We might additionally speculate that flipped learning may work best for
procedural and conceptual learning, where the teacher can lead the students through
higher-order thinking activities with the extra class time that flipping facilitates (Milman,
2012). With vocabulary, there may be less utility in this regard as vocabulary begins with
form to function mapping (Nation, 2013). There may be less room for the teacher to add
to this process when new lexical items are being learned. If this is true, then the effect of
flipped learning on vocabulary might be smaller (even if positive) than the effect on other
language outcomes. With writing and speaking, for example, a flipped classroom could
optimize students’ learning experience as the teacher has time to engage them in thinking
about processes involved and reinforce their learning utilizing the extra class time. At the
assessment level, furthermore, competence measurements such as speaking and writing
are often criterion-referenced where the content taught is assessed. Thus, students have a
better chance of demonstrating improvement. Standardized tests, on the other hand, may
not map as closely to content taught as does teacher-made course tests. It is an open
empirical question as to the extent to which flipped learning applications can be adapted
to maximize impact on vocabulary, reading, and performance on standardized tests.
Proficiency was the only learner characteristic to significantly moderate effect sizes
where the post-hoc comparison between below-intermediate (g = 0.65) and above-inter-
mediate (g = 1.45) was significant. Although the effect in both cases was large, the sig-
nificant difference might indicate that low-proficiency learners have less ability to sustain
student-centered engagement with material in the target language, which corresponds to
Vitta and Al-Hoorie 19
the concern voiced by Milman (2012). Willis and Willis (2019) in a similar vein posited
that beginner students might have trouble engaging with student-centered task-based
language teaching (see also Vitta, Jost, & Pusina, 2019). Thus, there seems to be a theo-
retical basis for the positive association between proficiency and the effectiveness of
flipped learning. Should teachers seek to implement flipped learning with low-profi-
ciency learners, then extra care may need to be taken in preparing accessible and appeal-
ing content so that these learners can remain engaged with it outside of class time.
Alternatively, flipped learning applications to low-proficiency learners could require
greater first language (L1) and extra-linguistic support.
3 Accurate estimation of the effect of flipped learning
Our findings correspond to the large effects observed in other L2 meta-analyses investi-
gating trends in instructed second language learning. Consider that both Zhang and
Zhang (2020) and Bryfonski and McKay (2020) observed large effects in meta-analyses
of the association between vocabulary and reading (r = .57) and the effects of task-based
language teaching interventions (.93 d .95), respectively. The effect size obtained in
the present meta-analysis was likewise substantial, g = 0.99. Such magnitudes exceed
Cohen’s (1992) classic threshold for a large effect (d = 0.80) and approach Plonsky and
Oswald’s (2014) empirically derived large effect size for between-group differences in
L2 outcomes (d = 1.0). These magnitudes also substantially exceed the large benchmark
for individual difference research (d = 0.60; Gignac & Szodorai, 2016) and the typical
effects from teachers in longitudinal studies (d = 0.15–0.40; Hattie, 2009).
However, the Trim and Fill method suggested the presence of publication bias. Smaller
studies tended to report larger effects, suggesting the possibility of a file-drawer effect
(Rosenthal, 1979) in our field. With the Trim and Fill method, we obtained an effect size
that is almost half of the original one (g = 0.58), which is closer to what Cheng et al.
(2019) found for the arts and humanities (g = 0.63). Moderator analysis further supported
the possibility of publication bias, showing that peer-reviewed articles – particularly in
non-SSCI-indexed journals – have larger effects than conference proceedings and unpub-
lished theses. Still, as can be seen from the p-curve analysis, this pattern does not seem to
have resulted from questionable research practices. Instead, one plausible explanation is
that some researchers publishing in low-impact journals might lack experience and/or
resources to conduct well-controlled interventions (for a discussion; see Paiva et al.,
2017). Conducting an educational intervention is no easy task. Classroom research is
fraught with obstacles and challenges due to the complexity of classroom realities (Hiver
& Al-Hoorie, 2020b; Rounds, 1996). Some unexpected factors that influence how valid
the results are might go unreported, ‘not out of any willful malfeasance, but because we
have been so conditioned to preserve methodological purism, however unrealistic a goal
that might have been’ (Larsen-Freeman, 1996, p. 157).
4 Future directions
As we mentioned above, the status of scholarship on flipped learning indicates that
researchers should move from the question of whether flipped learning is effective to
when and how it is so. To address these questions, we suggest two main future directions
20 Language Teaching Research 00(0)
for the field. First, research should target different underrepresented L2 learners. Just like
it is the case in various L2 subdisciplines (Dörnyei & Al-Hoorie, 2017), L2 flipped learn-
ing research has been English-biased in that learners of languages other than English
have seldom been investigated. As with Lundin et al. (2018), our report pool was domi-
nated by university-level learners by ratio approaching 5:1. Younger learners were espe-
cially underrepresented, making it unclear to what extent flipped learning is effective
with younger learners considering that this approach presupposes a level of commitment
and self-directedness without the teacher’s direct supervision. It is likely that the type of
content that can attract this type of learners will be very different, and possibly more
demanding to prepare. In addition to young learners, older learners and those not suffi-
ciently skilled in or familiar with technology, including those based outside the devel-
oped world (e.g. only one of 56 reports was situated within an African context; Hassan,
2018) might also require different applications of the flipped approach.
A second future direction we recommend for flipped learning research has to do with
intervention quality. Part of understanding when and how flipped learning is effective is
to understand what features maximize its effectiveness. Little comparative analysis has
been conducted to investigate the various online platforms available to L2 teachers and
how their features influence learning (e.g. for lower proficiency learners). Another aspect
of the quality of flipped learning interventions is the teacher’s skill in preparing and
handling online materials. We suspect that teachers who can create custom materials on
demand to suit emerging needs of their particular classes will most likely be more effec-
tive. Investigation of these aspects requires a more micro-analysis of intervention qual-
ity. A further aspect of study quality is rigor in design and statistical analysis (Al-Hoorie,
2018; Hiver & Al-Hoorie, 2020a, 2020b). While Al-Hoorie and Vitta’s (2019) systematic
review found that the statistical quality varies based on the impact of the journal, the
present meta-analysis additionally showed that the actual results also vary. Further
research is needed to understand why the findings of high- and low-impact journals can
be discrepant (see Paiva et al., 2017). Tips and strategies for effective flipped learning
implementation can be found in Mehring and Leis (2018).
Next, research on flipped learning should therefore move to what Zanna and Fazio
(1982) called second-generation and third-generation questions. According to Zanna and
Fazio’s (1982) classification, first-generation research simply asks ‘is’ questions (e.g. is
flipped learning effective?). Second-generation questions move beyond this yes–no ques-
tion to ‘when’ questions (e.g. under what conditions does flipped learning become more
effective?) (see also Al-Hoorie & Al Shlowiy, 2020). Third-generation research asks ‘how’
questions (e.g. how is flipped learning effective?). This last type of questions inquires after
the mechanism, or mediators, making flipped learning effective. While this type of ques-
tions is described as third-generation, thus implying a temporal lag, in reality second- and
third-generation questions are ‘linked inextricably’ (Zanna & Fazio, 1982, p. 284).
Understanding under what conditions a treatment is effective might shed light on why it is
effective, and vice versa. It is at this point that practitioners-as-researchers can contribute
to the future directions of flipped research in L2 contexts as localized studies will be essen-
tial in addressing the second- and third-generation questions. To provide a specific exam-
ple, our findings highlight the need for frontline teachers to pilot and report flipped
approaches that focus on vocabulary outcomes and with lower proficiency learners.
Vitta and Al-Hoorie 21
Finally, as part of intervention quality, researchers should investigate innovative adap-
tations of flipped learning. The prototypical design is that students engage with the mate-
rial before class independently and then have agency to further engage with and explore
the content during subsequent class time. Little research has examined whether and to
what extent group work before class can make flipped learning more effective. Little
research has also examined what we might describe as ‘interval flipping’, the process of
alternating between the flipped and the traditional approaches in order to prevent flipped
learning from losing its novelty over time. Indeed, even long-term retention of learning
from the flipped approach has hardly been compared to that from the traditional approach.
VIII Conclusions
The present study meta-analysed the effects of L2 flipped learning interventions.
Extending past flipped meta-analyses on flipped learning, our literature search was able
to locate about double the number of L2 experimental reports analysed in past syntheses.
Future endeavors could add to our approach by considering gray literature, however. Our
results also clearly demonstrate the effectiveness of this approach over the traditional
face-to-face approach. Still, there was also wide heterogeneity in the results that could be
partially explained by certain moderators, including learner proficiency, study type, and
target L2 outcome. Future research should shift focus from whether flipped learning is
effective to when and how its effectiveness can be maximized.
Acknowledgements
We would like to thank Dr. Jeffrey G. Mehring for his comments on our literature search protocols.
We are also grateful to Alex Sutton and Daniël Lakens for comments on the analysis.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this
article.
ORCID iDs
Joseph P. Vitta https://orcid.org/0000-0002-5711-969X
Ali H. Al-Hoorie https://orcid.org/0000-0003-3810-5978
Supplemental material
Supplemental material for this article is available online.
Note
1. The exclusion of true control (no learning intervention) comparisons was in line with the
methodologies of recent L2 (e.g. Bryfonski & McKay, 2019) and flipped learning (e.g.
Strelan et al., 2020) meta-analyses.
2. The Trim and Fill method was calculated using the fixed-effects model. The random-effects
model (not reported here) showed the opposite pattern, indicating that bias might be resulting
from reports with larger samples—which does not seem likely (see also Shi & Lin, 2019). The
funnel plot based on the random-effects model may be obtained from the authors.
22 Language Teaching Research 00(0)
References
Adnan, M. (2017). Perceptions of senior-year ELT students for flipped classroom: A materials
development course. Computer Assisted Language Learning, 30, 204–222.
Al-Hoorie, A.H. (2018). The L2 motivational self system: A meta-analysis. Studies in Second
Language Learning and Teaching, 8, 721–754.
Al-Hoorie, A.H., & Al Shlowiy, A.S. (2020). Vision theory vs. goal-setting theory: A critical
analysis. Porta Linguarum, 33, 217–229.
Al-Hoorie, A.H., & Vitta, J.P. (2019). The seven sins of L2 research: A review of 30 journals’
statistical quality and their CiteScore, SJR, SNIP, JCR Impact Factors. Language Teaching
Research, 23, 727–744.
AlJaser, A.M. (2017). Effectiveness of using flipped classroom strategy in academic achieve-
ment and self-efficacy among education students of Princess Nourah Bint Abdulrahman
University. English Language Teaching, 10, 67–77.
Alnuhayt, S.S. (2018). Investigating the use of the flipped classroom method in an EFL vocabulary
course. Journal of Language Teaching and Research, 9, 236–242.
Baş, G., & Kuzucu, O. (2009). Effects of CALL method and DynED language programme on stu-
dents’ achievement levels and attitudes towards the lesson in English classes. International
Journal of Instructional Technology and Distance Learning, 6, 31–44.
Bergmann, J., & Sams, A. (2012). Flip your classroom: Reach every student in every class every
day. Eugene, OR: International Society for Technology in Education.
Borenstein, M., Hedges, L.V., Higgins, J.P.T., & Rothstein, H.R. (2009). Introduction to meta-
analysis. Oxford: Wiley.
Borenstein, M., Hedges, L.V., Higgins, J.P., & Rothstein, H.R. (2014). Comprehensive meta anal-
ysis: Version 3.3. Englewood, NJ: Biostat.
Brown, A.V., Plonsky, L., & Teimouri, Y. (2018). The use of course grades as metrics in L2
research: A systematic review. Foreign Language Annals, 51, 763–778.
Bryfonski, L., & Mckay, T.H. (2019). TBLT implementation and evaluation: A meta-analysis.
Language Teaching Research, 23, 603–632.
Burnham, J.F. (2006). Scopus database: A review. Biomedical Digital Libraries, 3(1).
Chen Hsieh, J.S., Wu, W.-C.V., & Marek, M.W. (2017). Using the flipped classroom to enhance
EFL learning. Computer Assisted Language Learning, 30, 1–21.
Cheng, L., Ritzhaupt, A.D., & Antonenko, P. (2019). Effects of the flipped classroom instructional
strategy on students’ learning outcomes: A meta-analysis. Educational Technology Research
and Development, 67, 793–824.
Cohen, J. (1992). A power primer. Psychological Bulletin, 112, 155–159.
Council of Europe. (2001). Common European framework of reference for languages. Strasbourg:
Council of Europe.
Council of Europe. (2011). Common European framework of reference for languages: Learning,
teaching, assessment. Strasbourg: Council of Europe.
Davis, N.L. (2016). Anatomy of a flipped classroom. Journal of Teaching in Travel & Tourism,
16, 228–232.
Dickersin, K. (2005). Publication bias: Recognizing the problem, understanding its origins and
scope, and preventing harm. In Rothstein, H.R., Sutton, A.J., & M. Borenstein (Eds.),
Publication bias in meta-analysis: Prevention, assessment and adjustments (pp. 11–33).
Chichester: Wiley.
Dörnyei, Z., & Al-Hoorie, A.H. (2017). The motivational foundation of learning languages other
than Global English. The Modern Language Journal, 101, 455–468.
Vitta and Al-Hoorie 23
Duval, S., & Tweedie, R. (2000a). A nonparametric ‘trim and fill’ method of accounting for pub-
lication bias in meta-analysis. Journal of the American Statistical Association, 95, 89–98.
Duval, S., & Tweedie, R. (2000b). Trim and fill: A simple funnel-plot–based method of testing and
adjusting for publication bias in meta-analysis. Biometrics, 56, 455–463.
Ellis, R. (2009). Task-based language teaching: Sorting out the misunderstandings. International
Journal of Applied Linguistics, 19, 221–246.
Evseeva, A., & Solozhenko, A. (2015). Use of flipped classroom technology in language learning.
Procedia – Social and Behavioral Sciences, 206, 205–209.
Fanelli, D. (2010). Do pressures to publish increase scientists’ bias? An empirical support from US
States Data. PLoS One, 5(4), e10271.
Filiz, S., & Benzet, A. (2018). A content analysis of the studies on the use of flipped classrooms in
foreign language education. World Journal of Education, 8, 72–86.
Freire, P. (1968/1970). Pedagogy of the oppressed. New York: Herder and Herder.
Gignac, G.E., & Szodorai, E.T. (2016). Effect size guidelines for individual differences research-
ers. Personality and Individual Differences, 102, 74–78.
Green, A. (2012). Language functions revisited: Theoretical and empirical bases for language
construct definition across the ability range. Cambridge: Cambridge University Press.
Halliday, M.A.K., & Matthiessen, C. (2014). Halliday’s introduction to functional grammar. 4th
edition. New York: Routledge.
Hamdan, M., McKnight, P.E., McKnight, K., & Arfstrom, K.M. (2013). A review of flipped learn-
ing. Flipped Learning Network. Available at: https://www.flippedlearning.org/wp-content/
uploads/2016/07/LitReview_FlippedLearning.pdf (accessed December 2020).
Hassan, S.R.R. (2018). Using the flipped learning model to develop EFL argumentative writing
skills of STEM secondary school students. Majalat Kuliyat Altarbiah (Education College
Journal), 70, 24–74.
Hattie, J.A.C. (2009). Visible learning: A synthesis of over 800 meta-analyses relating to achieve-
ment. New York: Routledge.
Hew, K.F., & Lo, C.K. (2018). Flipped classroom improves student learning in health professions
education: A meta-analysis. BMC Medical Education, 18, 38.
Hiver, P., & Al-Hoorie, A.H. (2020a). Reexamining the role of vision in second language motiva-
tion: A preregistered conceptual replication of You, Dörnyei, & Csizér (2016). Language
Learning, 70, 48–102.
Hiver, P., & Al-Hoorie, A.H. (2020b). Research methods for complexity theory in applied linguis-
tics. Bristol: Multilingual Matters.
Hung, H.-T. (2015). Flipping the classroom for English language learners to foster active learning.
Computer Assisted Language Learning, 28, 81–96.
Hung, H.-T. (2017). Design-based research: Redesign of an English language course using a
flipped classroom approach. TESOL Quarterly, 51, 180–192.
Ishikawa, Y., Akahane-Yamada, R., Smith, C., et al. (2015). An EFL flipped learning course
design: Utilizing students’ mobile online devices. In Helm, F., Bradley, L., Guarda, M., & S.
Thouësny (Eds.), Critical CALL – Proceedings of the 2015 EUROCALL Conference, Padova,
Italy (pp. 261–267). Dublin: Research-publishing.net.
Karimi, M., & Hamzavi, R. (2017). The effect of flipped model of instruction on EFL learners’
reading comprehension: Learners’ attitudes in focus. Advances in Language and Literary
Studies, 8, 95–103.
Låg, T., & Sæle, R.G. (2019). Does the flipped classroom improve student learning and satisfac-
tion? A systematic review and meta-analysis. AERA Open, 5, 3.
Lakens, D. (2013). Calculating and reporting effect sizes to facilitate cumulative science: A practi-
cal primer for t-tests and ANOVAs. Frontiers in Psychology, 4, 863.
24 Language Teaching Research 00(0)
Larsen-Freeman, D. (1996). The changing nature of second language classroom research. In
Schachter, J., & S. Gass (Eds.), Second language classroom research: Issues and opportuni-
ties (pp. 157–170). Mahwah, NJ: Lawrence Erlbaum.
Lei, L., & Liu, D. (2019). Research trends in applied linguistics from 2005 to 2016: A bibliometric
analysis and its implications. Applied Linguistics, 40, 540–561.
Leis, A., Cooke, S., & Tohei, A. (2015). The effects of flipped classrooms on English composi-
tion writing in an EFL environment. International Journal of Computer-Assisted Language
Learning and Teaching (IJCALLT), 5, 37–51.
Lin, C.-J., & Hwang, G.-J. (2018). A learning analytics approach to investigating factors affecting
EFL students’ oral performance in a flipped classroom. Journal of Educational Technology
& Society, 21, 205–219.
Lin, C.-J., Hwang, G.-J., Fu, Q.-K., & Chen, J.-F. (2018). A flipped contextual game-based learn-
ing approach to enhancing EFL students’ English business writing performance and reflective
behaviors. Journal of Educational Technology & Society, 21, 117–131.
Lo, C.K., & Hew, K.F. (2019). The impact of flipped classrooms on student achievement in
engineering education: A meta-analysis of 10 years of research. Journal of Engineering
Education, 108, 523–546.
Lomicka, L., & Lord, G. (2016). Social networking in language learning. In Farr, F., & L. Murray
(Eds.), The Routledge handbook of language learning and technology (pp. 225–268). New
York: Routledge.
Lundin, M., Bergviken Rensfeldt, A., Hillman, T., Lantz-Andersson, A., & Peterson, L. (2018).
Higher education dominance and siloed knowledge: a systematic review of flipped classroom
research. International Journal of Educational Technology in Higher Education, 15, 20.
Mahmud, M.M. (2018). Technology and language: What works and what does not: A meta-
analysis of blended learning research. Journal of Asia TEFL, 15, 365–382.
Mehring, J. (2016). Present research on the flipped classroom and potential tools for the EFL class-
room. Computers in the Schools, 33, 1–10.
Mehring, J. (2018). The flipped classroom. In Mehring, J., & A. Leis (Eds.), Innovations in flip-
ping the language classroom: Theories and practices (pp. 1–10). New York: Springer Berlin
Heidelberg.
Mehring, J., & Leis, A. (Eds.). (2018). Innovations in flipping the language classroom: Theories
and practices. New York: Springer Berlin Heidelberg.
Milman, N.B. (2012). The flipped classroom strategy: What is it and how can it best be used?
Distance Learning, 9, 85–87.
Mori, Y., Omori, M., & Sato, K. (2016). The impact of flipped online Kanji instruction on written
vocabulary learning for introductory and intermediate Japanese language students. Foreign
Language Annals, 49, 729–749.
Nation, I.S.P. (2013). Learning vocabulary in another language. 2nd edition. Cambridge:
Cambridge University Press.
Norris, J.M., & Ortega, L. (2000). Effectiveness of L2 instruction: A research synthesis and quan-
titative meta-analysis. Language Learning, 50, 417–528.
Oh, E. (2017). The effect of peer teaching via flipped vocabulary learning on class engagement and
learning achievements. Multimedia-Assisted Language Learning, 20, 105–127.
Orwin, R.G. (1983). A fail-safe N for effect size in meta-analysis. Journal of Educational Statistics,
8, 157–159.
Paiva, C.E., Araujo, R.L.C., Paiva, B.S.R., et al. (2017). What are the personal and professional
characteristics that distinguish the researchers who publish in high- and low-impact journals?
A multi-national web-based survey. ecancermedicalscience, 11, 718.
Pallotti, G. (2009). CAF: Defining, refining and differentiating constructs. Applied Linguistics,
30, 590–601.
Vitta and Al-Hoorie 25
Plonsky, L., & Oswald, F.L. (2014). How big is ‘big’? Interpreting effect sizes in L2 research.
Language Learning, 64, 878–912.
Rosenthal, R. (1979). The file drawer problem and tolerance for null results. Psychological
Bulletin, 86, 638–641.
Rounds, P.L. (1996). The classroom-based researcher as fieldworker: Strangers in a strange land.
In Schachter, J., & S. Gass (Eds.), Second language classroom research: Issues and opportu-
nities (pp. 45–59). Mahwah, NJ: Lawrence Erlbaum.
Ryback, D., & Sanders, J.J. (1980). Humanistic versus traditional teaching styles and student sat-
isfaction. Journal of Humanistic Psychology, 20, 87–90.
Shi, L., & Lin, L. (2019). The trim-and-fill method for publication bias. Medicine, 98(23), e15987.
Shi, Y., Ma, Y., MacLeod, J., & Yang, H.H. (2020). College students’ cognitive learning out-
comes in flipped classroom instruction: A meta-analysis of the empirical literature. Journal
of Computers in Education, 7, 79–103.
Simonsohn, U., Nelson, L.D., & Simmons, J.P. (2014). P-curve: A key to the file-drawer. Journal
of Experimental Psychology: General, 143, 534–547.
Strelan, P., Osborn, A., & Palmer, E. (2020). The flipped classroom: A meta-analysis of effects on
student performance across disciplines and education levels. Educational Research Review,
30, 100314.
Teng, M.F. (2017). Flipping the classroom and tertiary level EFL students’ academic performance
and satisfaction. Journal of Asia TEFL, 14, 605–620.
Turan, Z., & Akdag-Cimen, B. (2019). Flipped classroom in English language teaching: A system-
atic review. Computer Assisted Language Learning, 33, 590–606.
van Alten, D.C.D., Phielix, C., Janssen, J., & Kester, L. (2019). Effects of flipping the classroom
on learning outcomes and satisfaction: A meta-analysis. Educational Research Review, 28,
100281.
Vitta, J.P., & Al-Hoorie, A.H. (2017). Scopus- and SSCI-indexed L2 journals: A list for the Asia
TEFL community. The Journal of Asia TEFL, 14, 784–792.
Vitta, J.P., Jost, D., & Pusina, A. (2019). A case study inquiry into the efficacy of four East Asian
EAP writing programmes: Presenting the emergent themes. RELC Journal, 50, 71–85.
Voss, E., & Kostka, I. (2019). Flipping academic English language learning: Experiences from an
American university. Singapore: Springer Nature Singapore.
Webb, M., & Doman, E. (2016). Does the flipped classroom lead to increased gains on learning
outcomes in ESL/EFL contexts? CATESOL Journal, 28, 39–67.
Willis, D., & Willis, J. (2019). Doing task-based teaching. Oxford: Oxford University Press.
Xu, P., Chen, Y., Nie, W., et al. (2019). The effectiveness of a flipped classroom on the develop-
ment of Chinese nursing students’ skill competence: A systematic review and meta-analysis.
Nurse Education Today, 80, 67–77.
Zanna, M.P., & Fazio, R.H. (1982). The attitude-behavior relation: Moving toward a third gen-
eration of research. In Zanna, M.P., Higgins, E.T., & C.P. Herman (Eds.), Consistency in
social behavior: The Ontario symposium: Volume 2 (pp. 283–301). Hillsdale, NJ: Lawrence
Erlbaum.
Zarrinabadi, N., & Ebrahimi, A. (2019). Increasing peer collaborative dialogue using a flipped
classroom strategy. Innovation in Language Learning and Teaching, 13, 267–276.
Zhang, X. (2020). A bibliometric analysis of second language acquisition between 1997 and 2018.
Studies in Second Language Acquisition, 42, 199–222.
Zhang, S., & Zhang, X. (2020). The relationship between vocabulary knowledge and L2 read-
ing/listening comprehension: A meta-analysis. Language Teaching Research. Epub ahead of
print 31 March 2020. DOI: 10.1177/1362168820913998.
... Though some of the studies so far have focused on the validation of the proposed defining characteristics and understanding the triggers of the DMC phenomena (e.g., Ibrahim, 2017;Muir, 2020), none has focused on the real-time initial conditions, triggers inducing DMCs, and essential features of tertiary level EFL learners' DMCs, namely the whole-DMC journey of those who are currently experiencing DMCs in educational settings. Thus, the implicit assumptions in the DMC literature indicate a need for real-time analysis of DMCs since it is crucial to uncover the whole process leading to such motivational phenomenon as it is in order to investigate the desired outcome (the DMC state), (Al-Hoorie & Al Shlowiy, 2020;Vitta & Al-Hoorie, 2020). Also, there is a further need to develop understanding of the defining characteristics/essential features of DMCs without preconceived notions about this construct in order to provide the inclusion criteria for future research on this topic (e.g., whether the goal should be accompanied by vision, or what the features of the goal should be, what kind and intensity of the emotion during the DMC journey should be, what sub-elements should accompany to the salient facilitative structure or what other defining and non-defining characteristics should be) (Jahedizadeh & Al-Hoorie, 2021). ...
... Though, in the above studies, the factors that trigger the launch of DMCs have been investigated, they have retrospectively explored the triggers of DMCs occurring in different contexts in the past. However, in order to explore the DMC state, it is highly crucial to investigate the process leading to the final DMC state, namely immediately during the DMC journey (Al-Hoorie & Al Shlowiy, 2020; Vitta & Al-Hoorie, 2020). Of all the studies above, only Ibrahim (2017) explored the initial conditions of DMCs in that it was found out that the main conditions necessary for the launch of DMCs are goal/ambitions and perceived feasibility. ...
Article
Full-text available
Recent trends in the field of L2 learning motivation have led to a proliferation of studies that examine Directed Motivational Currents (DMCs). However, there remains a paucity of evidence on the factors that trigger the launch of DMCs and previous research findings into the central components of DMCs have been contradictory. This qualitative study addresses these gaps by looking at the possible parameters that trigger DMCs, and examining the initial conditions necessary for the launch of DMCs, and the (non)defining core features of it. Semi-structured interviews were conducted with 27 EFL learners pursuing a bachelor's degree in Turkey. The factors triggering learners' DMCs were negative emotion, emergent opportunity, single and explicit goal, good teacher/classmates/motivating class group, moments of realization, a sense of achievement, a passion/interest for something, element of competition, external pressure, a big life decision, vision, outcome led, belief that the work had real value, and inspiring experience. Perceived behavioural control and challenge-skill balance were two initial conditions necessary for DMCs to begin. Goal orientedness, salient facilitative structure, and positive emotionality were found to be the core features of DMCs. The findings offer significant implications for how teachers could exploit these motivational surges to promote learning in EFL settings.
... When instructors would prefer to avoid spending too much in-class time on explicit grammar pint, while still addressing students' preference for the inclusion of explicit grammar, one possible solution is to employ a flipped classroom model (Vitta & Al-Hoorie, 2020). In this L2 Journal Vol. 14 Issue 3 (2022) 7 model, students engage with rich input, models of the tasks to perform and relevant vocabulary and grammar content, on their own as assignments and then use what they learned at home to complete specific tasks in class (González-Lloret & Ortega, 2014). ...
... SPA 1Y follows a flipped classroom model (Vitta & Al-Hoorie, 2020) in that students complete the pre-task activities online, at their own pace, in preparation for the face-to-face class meetings. These activities include oral, written and audiovisual input related to the target tasks of each unit, as well as explicit instructional materials to help them understand that input. ...
Article
Task-based language teaching (TBLT) has been at the center of the debates on which approaches are most effective for structuring, planning, and implementing language courses. Several articles have focused on its effectiveness (Bryfonsky & McKay, 2017; Long, 2016; González-Lloret & Nielson, 2015), but few have shared specific implementation experiences in the classroom (Long, 2015; Torres & Seratini, 2016). The limited array of articles that address TBLT course creation focus on language courses for specific purposes, but little is known about the challenges and solutions found when designing and implementing a TBLT course that is part of a large general education language program. This article shares the authors’ experience in developing and teaching such a course. Based on these experiences, realistic and actionable examples are offered of how to surmount the challenges encountered when developing, integrating, and teaching a TBLT course in an otherwise traditional grammar-centric First-Year university language program at a large US university.
... The higher the learners' proficiency, the higher the effects of flipped learning (Vitta & Al-Hoorie, 2020). The cluster centers around the effect of individual differences. ...
... The positive results of flipped learning in EFL contexts may be because learners are exposed to pre-class learning materials so that they are more prepared to participate in interactivebased language tasks (Lee and Wallace, 2018;Li and Li, 2022). Besides, the increased opportunities for interaction between instructor and learner may lead to closer teacherstudent rapport and more peer trust (Vitta and Al-Hoorie, 2020). ...
Article
Full-text available
Flipped classrooms have attracted widespread attention and interest from English for foreign language (EFL) practitioners and researchers and are regarded as a promising pedagogical approach to increase learning outcomes and facilitate learner engagement. This article takes stock of the publications on learner engagement in flipped EFL classrooms. In so doing, we aim to clarify the engagement construct specific to the EFL classroom setting, summarize the effect that flipped classrooms have on EFL learners’ engagement, and identify factors affecting engagement. We find that there has been a range of debates and tensions regarding flipped classrooms and learner engagement in EFL settings, and that more theory-grounded empirical studies are needed to delineate learner engagement in localized flipped learning and teaching contexts. We would argue that in future research, multiple variables, including learner dynamics and social-psychological factors, should be considered while flipping the EFL course so that practitioners can provide tailor-made support to improve learners’ engagement levels.
... Qassem (2020) mentioned that as learning writing in English language learning, technology devices play a crucial role in enhancing English writing training. Another reason may relate to that English language is based on numerus skills (reading, speaking, writing etc..) and that evidenced flipped learning as effective tool in enhancing English language skills (Vitta & Al-Hoorie, 2020). Furthermore, flipped learning technique provide students with suitable space for the exchanging the ideas more than another course (Hamdani, 2019). ...
... The emergence of technology has arguably been a force for good in the classroom and nowhere is this more evident than with second language writing, where translation and other eLearning tools help students produce original content and navigate second language communication (Vitta & Al-Hoori, 2020). Several noteworthy findings pertaining to the frequency and type of writing strategies used during emergency remote online courses emanated from this study. ...
Article
Full-text available
Second language (L2) writing strategies are essential for successful learning outcomes in courses with a substantial writing component and this is especially true during emergency remote teaching (ERT) when online writing tasks help compensate for the missing offline communication. Online writing tasks are multimodal and interactive, and widely delivered through assignment modules, discussion forums, social media, and other online channels, yet little is known pertaining to L2 writing strategies and online writing beliefs. The current study investigates the types of L2 writing strategies students employed during ERT in the midst of Covid-19 and then used structural modeling to understand how strategies relate to online writing task perceptions, L2 writing anxiety, and L2 writing proficiency. The four L2 writing strategy categories are related to planning, monitoring, reviewing, and translating. Following a cross-sectional survey design, a total of 256 South Korean EFL students completed the study’s questionnaire administered during their second semester of ERT. Overall, students reported using planning and monitoring strategies the most while using moderate levels of translation and review strategies. Translation strategies produced a significant positive relationship with L2 writing anxiety, indicating effort-avoidance behavior among apprehensive writers. Contrarily, planning and review strategies were positively associated with increased levels of perceived benefits with online writing tasks. When added to the model, the statistically significant correlations among monitoring category and outcome variables vanished, indicating a mediation effect. Along with addressing future directions in L2 writing strategy research in the post-Covid era, some pedagogical implications for the evolving application of translation strategies are discussed.
Article
Full-text available
The current levels of enthusiasm for flipped learning are not commensurate with and far exceed the vast variability of scientific evidence in its favor. We examined 46 meta-analyses only to find remarkably different overall effects, raising the question about possible moderators and confounds, showing the need to control for the nature of the intervention. We then conducted a meta-analysis of 173 studies specifically coding the nature of the flipped implementation. In contrast to many claims, most in-class sessions are not modified based on the flipped implementation. Furthermore, it was flipping followed by a more traditional class and not active learning that was more effective. Drawing on related research, we proposed a more specific model for flipping, “Fail, Flip, Fix, and Feed” whereby students are asked to first engage in generating solutions to novel problems even if they fail to generate the correct solutions, before receiving instructions.
Article
Full-text available
Few studies have investigated learners’ emotional experiences and the interactions between emotion, cognition, and activity in the flipped foreign language classroom (i.e., a mixed teaching mode that combines in-class teaching and off-class self-directed learning). This study, from the perspective of perezhivanie (a concept from sociocultural theory), addressed these research gaps by exercising a longitudinal narrative study on a total of 32 Chinese-as-the-first-language university students of English who attended a 15-week English course with this teaching design. Among them, eight focal students were randomly selected for further evidence of the characteristics of the interactions between emotion, cognition, and activity. The results showed that the participants experienced more of positive emotions than negative emotions in the flipped classroom (FC) context, which supports the efficacy of the pedagogy. But most importantly, complex interactions between emotion, cognition, and activity were revealed. Generally, (negative) positive emotion, cognition, and activity were interconnected; however, what is also evident is that learners’ emotions either promoted or inhibited their cognitive functions, and positive and negative emotions did not necessarily correspond to positive and negative activities, respectively. This is due to the presence of dynamic, developmental, and historical sociocultural mediators in learners' perezhivanija, be it teacher, peers, technology, teaching materials, teaching activities in an FC, or the learners’ previous English learning anecdotes, etc.
Preprint
Full-text available
Self-determination theory is one of the most established motivational theories both within second language learning and beyond. This theory has generated several mini-theories, namely: organismic integration theory, cognitive evaluation theory, basic psychological needs theory, goal contents theory, causality orientations theory, and relationships motivation theory. After providing an up-to-date account of these mini-theories, we present the results of a systematic review of empirical second language research into self-determination theory over a 30-year period (k = 111). Our analysis of studies in this report pool showed that some mini-theories were well-represented while others were underrepresented or absent from the literature. We also examined this report pool to note trends in research design, operationalization, measurement, and application of self-determination theory constructs. Based on our results, we highlight directions for future research in relation to theory and practice.
Article
Full-text available
Self-determination theory is one of the most established motivational theories both within second language learning and beyond. This theory has generated several mini-theories, namely: organismic integration theory, cognitive evaluation theory, basic psychological needs theory, goal contents theory, causality orientations theory, and relationships motivation theory. After providing an up-to-date account of these mini-theories, we present the results of a systematic review of empirical second language research into self-determination theory over a 30-year period (k = 111). Our analysis of studies in this report pool showed that some mini-theories were well-represented while others were underrepresented or absent from the literature. We also examined this report pool to note trends in research design, operationalization, measurement, and application of self-determination theory constructs. Based on our results, we highlight directions for future research in relation to theory and practice.
Article
Full-text available
In recent years, vision has become a major theme in language motivation research, capturing a core feature of modern theories of language motivation. However, empirical investigations have mostly followed the prototypical design of administering self-report questionnaires and examining correlations among a handful of variables. At the same time, substantial overlap can be found between the current conceptualization of vision theory and the long-standing tradition of goal-setting theory. After demonstrating this substantial overlap, and taking our cues from goal-setting theory, this paper highlights critical gaps in current research into vision, including its sensory element, characteristics of effective vision, vision evolution over time, vision multiplicity and potential inter-vision conflict, vision mediators, collective vision, and the role of emotions and self-satisfaction at the end. The ultimate aim of this article is to propose a research agenda to examine the extent to which vision can be meaningfully distinguished from goal.
Article
Full-text available
Hamdan, N., McKnight, P., McKnight, K., & Arfstrom, K. A This article provides a brief review of the flipped learning pedagogical model. It has become widely cited in peer reviewed journals in articles on the topic.
Article
Full-text available
This article reports on a study which employed a flipped classroom strategy to increase the amount of peer dialogue among a group of Iranian EFL learners. The flipped strategy used in this study required the students to study some online materials provided for students via a social network called Telegram. Before attending the class, the participants watched videos related to the topic of each lesson, studied online texts that included the bolded vocabulary, and listened to audio clips related to the focus of every session's discussion. The analysis of audio-recordings of peer collaborative talk indicated that the flipped strategy, as compared to the traditional teaching, significantly increased collaborative peer dialogue among the learners. Finally, implications of the study for language teachers and researchers interested in this area of inquiry are discussed.
Article
Full-text available
We searched and meta-analyzed studies comparing flipped classroom teaching with traditional, lecture-based teaching to evaluate the evidence for the flipped classroom’s influence on continuous-learning measures, pass/fail rates, and student evaluations of teaching. Eight electronic reference databases were searched to retrieve relevant studies. Our results indicate a small effect in favor of the flipped classroom on learning (Hedges’ g = 0.35, 95% confidence interval [CI] [0.31, 0.40], k = 272). However, analyses restricted to studies with sufficient power resulted in an estimate of 0.24 (95% CI [0.18, 0.31], k = 90). Effects on pass rates (odds ratio = 1.55, 95% CI [1.34, 1.78], k = 45) and student satisfaction (Hedges’ g = 0.16, 95% CI [0.06, 0.26], k = 69) were small and also likely influenced by publication bias. There is some support for the notion that the positive impact on learning may increase slightly if testing student preparation is part of the implementation.
Article
This study set out to investigate the relationship between L2 vocabulary knowledge (VK) and second-language (L2) reading/listening comprehension. More than 100 individual studies were included in this meta-analysis, which generated 276 effect sizes from a sample of almost 21,000 learners. The current meta-analysis had several major findings. First, the overall correlation between VK and L2 reading comprehension was .57 ( p < .01) and that between VK and L2 listening was .56 ( p < .01). If the attenuation effect due to reliability of measures was taken into consideration, the ‘true’ correlation between VK and L2 reading/listening comprehension may likely fall within the range of .56–.67, accounting for 31%–45% variance in L2 comprehension. Second, all three mastery levels of form–meaning knowledge (meaning recognition, meaning recall, form recall) had moderate to high correlations with L2 reading and L2 listening. However, meaning recall knowledge had the strongest correlation with L2 reading comprehension and form recall had the strongest correlation with L2 listening comprehension, suggesting that different mastery levels of VK may contribute differently to L2 comprehension in different modalities. Third, both word association knowledge and morphological awareness (two aspects of vocabulary depth knowledge) had significant correlations with L2 reading and L2 listening. Fourth, the modality of VK measure was found to have a significant moderating effect on the correlation between VK and L2 text comprehension: orthographical VK measures had stronger correlations with L2 reading comprehension as compared to auditory VK measures. Auditory VK measures, however, were better predictors of L2 listening comprehension. Fifth, studies with a shorter script distance between L1 and L2 yielded higher correlations between VK and L2 reading. Sixth, the number of items in vocabulary depth measures had a positive predictive power on the correlation between VK and L2 comprehension. Finally, correlations between VK and L2 reading/listening comprehension was found to be associated with two types of publication factors: year-of-publication and publication type. Implications of the findings were discussed.
Article
The Flipped Classroom design has arguably revolutionized teaching across the globe. In this article we report the first comprehensive meta-analysis of its effects on student performance, relative to traditional teaching models, across disciplines and education level. We included 198 studies comprising 33,678 students in our meta-analysis. There were 174 studies conducted at the tertiary level, 21 at secondary, and three at primary school level. Overall, the flipped classroom had a moderate positive effect (g = .50) on student performance. The flipped classroom was beneficial regardless of discipline, with effect sizes ranging from weak (i.e., for IT, g = 0.30; k = 14) to strong (i.e., for humanities, g = 0.98; k = 34). Testing of other moderators relating to group sizes, in-class activities, pre-class testing, and the sophistication of a flipped classroom design provide evidence to suggest that the primary contributing factor to the flipped classroom effect is the opportunity it provides for structured, active learning and problem-solving.
The aim of this study is to examine the trends and main findings of the studies concerning the flipped classroom method in the field of English language teaching (ELT). For this purpose, databases including Web of Science, Eric, Taylor & Francis and the Educational full text EBSCO were reviewed, and a total of 43 articles were analysed. Systematic review was used as the research methodology. The articles were analysed utilising a content analysis method. The findings of the study revealed that the flipped classroom method in ELT gained popularity among researchers after 2014, and the number of the studies in the field rapidly increased in the last two years (2016–2017). In addition, the most commonly used research methods in flipped classroom in ELT studies were found to be mixed and quantitative methods. In the examined studies, speaking and writing abilities were the most commonly studied language skills. Further analysis revealed challenges, as well as benefits related to the use of the flipped classroom method in English as a foreign language (EFL) classroom. Additionally, in studies reviewed concerning the effectiveness of the flipped classroom methods, the findings mostly pointed to the benefits of the flipped classroom method. On the basis of the review, various suggestions are made for practitioners and future research.
Article
Background: The flipped classroom has become more widely used in engineering education. However, a systematic and quantitative assessment of its achievement outcomes has not been conducted to date. Purpose: To address this gap, we examined the findings from comparative articles published between 2008 and 2017 through a meta-analysis to summarize the overall effects of the flipped classroom on student achievement in engineering education. Scope/Method: We searched and analyzed journal and conference publications on flipped classroom studies in engineering education in K-12 and higher education contexts. Twenty-nine comparative interventions were included in a meta-analysis involving 2,590 students exposed to flipped classroom and 2,739 students exposed to traditional lectures. A content analysis was also conducted to determine how the flipped engineering classroom benefits student learning. Conclusions: The meta-analysis comparing these 29 traditional—flipped interventions in relation to student achievement showed an overall significant effect in favor of the flipped classroom over traditional lecturing (Hedges' g = 0.289, 95% CI [0.165, 0.414], p<.001). A moderator analysis showed that the effect of the flipped classroom was further enhanced when instructors offered a brief review at the start of face-to-face classes. Our qualitative findings suggest that self-paced learning and more problem-solving activities were the two most frequently reported benefits that promoted student learning. Based on quantitative and qualitative support, several implications are identified for future practice, such as offering a brief in-class review of preclass materials. Some recommendations for future research are also provided.
Article
This study used the bibliometric method to examine the field of second language acquisition (SLA) between 1997 and 2018 to provide a systematic overview of the field and to discover major trends in SLA. Based on citation/cocitation information and keywords retrieved from Web of Science, this study performed three types of bibliometric analyses to identify the prominent scholarly documents, authors, research institutions, geographic regions, and research topics that have been highly influential in the field of SLA over the last two decades. The scientific network maps and the keyword analysis revealed a number of significant changes as well as new trends in the field.