Sung 2014 How Effective Are Mobile Devices For Language Learning PDF

Educational Research Review 16 (2015) 68e84
Contents lists available at ScienceDirect
Educational Research Review

journal homepage: www.elsevier.com/locate/edurev
Review
How effective are mobile devices for language learning? A

meta-analysis
Yao-Ting Sung*, Kuo-En Chang, Je-Ming Yang
National Taiwan Normal University, Taipei, Taiwan
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 16 September 2014
Received in revised form 12 September 2015
Accepted 25 September 2015
Available online 30 September 2015
Language learning has undergone rapid changes over the past several years, from
computer-assisted learning to the more recent mobile-device-assisted learning. Although
mobile devices have become valuable language-learning tools, the evident substantial
contribution of mobile devices to language learning have not yet been investigated. The
present meta-analysis of 44 peer-reviewed journal articles and doctoral dissertations that
were written over a 20-year period (1993e2013), with 9154 participants, revealed that
mobile-device-assisted language instruction has produced a meaningful improvement
with an overall mean effect size of 0.55. Different effect sizes for moderator variables, such
as learning stages, hardware use, software used, intervention settings, teaching methods,
intervention durations, learning skills, target languages, and L1/L2, were also reported. The
results are discussed, together with their implications for future research and practices on
the use of mobile devices in language learning.
2015 Elsevier Ltd. All rights reserved.
Keywords:
Mobile-device-assisted language learning
Meta-analysis
Contents
1.
2.
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
1.1.
Integrating computers with language learning and instruction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
1.2.
The features of mobile devices and their applications in language learning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
1.3.
The activity-theory based framework for MALL . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
1.4.
Goals of the current meta-analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
2.1.
Data sources and search strategy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
2.2.
Search results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
2.2.1.
The initial screening stage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
2.2.2.
Screening for experimental and quasi-experimental studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
2.2.3.
Screening for inclusion in/exclusion from the meta-analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
2.3.
Selection and coding of variables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
2.3.1.
Research name . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
2.3.2.
Research participants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
2.3.3.
Treatments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
69
69
69
70
71
71
71
72
72
72
72
72
72
72
72
* Corresponding author. Department of Educational Psychology and Counseling National Taiwan Normal University 162, Sec. 1, Ho-Ping E. Rd., 10610
Taipei, Taiwan.
E-mail address: ntnusung@gmail.com (Y.-T. Sung).
http://dx.doi.org/10.1016/j.edurev.2015.09.001
1747-938X/ 2015 Elsevier Ltd. All rights reserved.
Y.-T. Sung et al. / Educational Research Review 16 (2015) 68e84
2.4.
3.
4.
Data analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
2.4.1.
Calculating effect sizes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
2.4.2.
Evaluating publication bias . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Results and discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
3.1.
Descriptive information . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
3.2.
The overall effect size for learning achievement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
3.3.
The effect size of learning achievement for moderator variables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
3.3.1.
Learning stage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
3.3.2.
Hardware used . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
3.3.3.
Software used . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
3.3.4.
Intervention setting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
3.3.5.
Teaching method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
3.3.6.
Intervention duration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
3.3.7.
Learning skills . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
3.3.8.
Target languages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
3.3.9.
L1/L2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
3.4.
Evaluation of the publication bias . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Conclusions and implications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
4.1.
Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
4.2.
Implications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
4.2.1.
Appropriately incorporating mobile devices into language learning/teaching activities . . . . . . . . . . . . . . . . . . . . . . . . . .
4.2.2.
Using a quality research design to empower effective MALL interventions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
4.2.3.
Substantiating the effects and enriching the diversity of MALL . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
References11References marked with an asterisk indicate studies included in the meta-analysis. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
69
73
73
73
73
73
76
76
77
78
78
78
78
79
79
79
79
80
80
80
81
81
81
82
82
1. Introduction
1.1. Integrating computers with language learning and instruction
Computer-assisted learning has been a focus of educational research for years, and computer-assisted language learning
(CALL) has been a major topic of research within the eld of computer technology since the 1960 s. Despite claims of the
benets of CALL (Gamper & Knapp, 2002; Hwu, 2013), some researchers (Garrett, 2009; Golonka, Bowles, Frank, Richardson,
& Freynik, 2012; Warschauer, 2004) have proposed that CALL remains limited in its ability to assist in language learning and
teaching, including problems such as shallow interactions, inaccurate feedback, distraction from learning tasks, overemphasis
on the delivery modality, extra workload, inadequate teacher training for developing quality CALL programs, and insufcient
software available for effectively training language skills. Mobile technologies offer a potential solution to aforementioned
limitations of CALL (Kukulska-Hulme, 2009; Stockwell, 2013).
1.2. The features of mobile devices and their applications in language learning
In recent years, a large body of literature has documented attempts to develop alternative learning tools for computerassisted learning. The emergence of wireless technology and a variety of mobile-device innovations have received a great
deal of attention in the eld of education. Mobile devices offer features of portability, social connectivity, context sensitivity,
and individuality, which desktop computers might not offer (Chinnery, 2006). Mobile devices have made learning movable,
real-time, collaborative, and seamless (Kukulska-Hulme, 2009; Wong & Looi, 2011), and the use of these devices may be
called mobile learning in general. Our research adopted a broader denition of mobile learning (Burston, 2014; Sharples,
Taylor, & Vavoula, 2007), which focused on the mobility of learners or learning and proposed that the features making
mobile learning distinctive from traditional learning are its integration of both movable and embedded technologies, its
ability to function in both formal (e.g., classroom) and informal (e.g., zoo) settings, its enhancement of both individualized and
collaborative/networked learning, and its capability to transform teacher-centered instruction into learner-center learning.
The unique properties of mobile devices have also been incorporated into language learning and teaching, forming the
emerging research eld of mobile-device-assisted language learning (MALL). The main features of mobile devices and their
applications in language learning are briey described below.
Mobility and portability. Mobile devices are small and lightweight, and are easily carried. Several researchers (e.g.,
Thornton & Houser, 2005; Wood, Jackson, Hart, Plester, & Wilde, 2011) used the portability of mobile phones and text
messaging during after-school hours to facilitate students English reading, spelling, and phonological awareness. These
features not only enable learning/teaching to happen anytime and anywhere, but also stimulate the needs of new teaching/
learning styles for settings/situations different from traditional classrooms.
70
Social connectivity/interaction. Mobile devices allow users to share information, collaborate, and communicate with others
with ease, either synchronously or asynchronously. Researchers have applied the social connectivity/interaction feature of mobile
devices to language leaning. The features of shared tasks and real time interaction not only enhance the efciency of group
learning, but also improve the quality of interaction during language learning (Lan, Sung, & Chang, 2007; Zurita & Nussbaum,
2004a). For example, Zurita and Nussbaum (2004a, 2004b) developed a constructivist learning environment based on wireless interconnected handhelds to facilitate young children's collaborative dialogues about learning Spanish syllables.
Context sensitivity. Mobile devices can be used to collect data specic to a particular location, environment, and time, or can
connect language learning across different settings, times, and locations. Users can track and review learning experiences
across various sites, and access learning resources that are relevant to their language skills and preferences. In contrast with
desktop computers, mobile devices possess functions that allow them to offer more exibility and accessibility to users to
record and deliver timely experiences and events for learning anytime and anywhere, such as location awareness (e.g., GPS)
and video capture. Researchers such as Sandberg, Maris, and De Geus (2011) highlighted the portability and contextsensitivity features of mobile devices to help elementary-school learners with their English reading and writing skills at
the zoo. Other studies did similarly for English listening and speaking (Chen & Li, 2010; Liu, 2009).
Individuality. Mobile devices can be customized and personalized for individual use according to different learning needs,
styles, and interests. Users' learning behavior should be taken into account while designing learning materials for mobile
devices. In both formal and informal language-learning contexts, learning activities designed for mobile devices are tailored
to meet the students' learning needs and pace, and to empower them in their learning by using authentic learning materials.
The features of individualized learning and feedback enhance learning when using mobile devices, provide opportunities for
monitoring and regulating learners' learning process, and also could increase the on-task behaviors. Researchers have used
the individuality of mobile devices to facilitate language learning. For example, Chang, Lan, Chang, and Sung (2010) conducted
a study using a mobile-device-assisted Chinese reading system for sharing individual learner's thoughts and products and
supported discussions in cooperative learning.
It should be noted that although mobile devices have possible advantages for language learning as mentioned above,
mobile devices may be treated as a two-edged sword, which may also produce negative learning effects, such as encouraging
distracting behaviors or offering irrelevant materials during learning periods (e.g., Gauerdau, Miranda, & Gareau, 2014; Handal,
MacNish, & Petocz, 2013). Therefore, instructional design of mobile devices plays a critical role in MALL (Burston, 2014).
1.3. The activity-theory based framework for MALL
As an emerging eld, mobile learning has rich content and diverse forms. To analyze the learning process and effects of
mobile learning, researchers have proposed various theories or frameworks. One of the most used is the activity theory (AT),
which uses activity as a unit for analyzing human practices (Bakhurst, 2009). The AT was originally introduced by Vygotsky
(1978) to interpret the interactive relationships of human being's behaviors and tools. Engestrom (1987; 1999) further
expanded Vygotsky's original ideas by broadening the components of the AT system and elaborating on the dynamic relationships among the components by proposing a visualization system with triangles. Recently, several researchers have
used the AT as a theoretical basis for analyzing mobile learning studies (e.g., Frohberg, Goth, & Schwabe, 2009; Sharples et al.,
2007) or for designing mobile learning scenarios (e.g., Zurita & Nussbaum, 2007).
Based on Engestrom (1987) original AT system and Sharples et al.(2007) revised AT for mobile learning, this study proposed an AT conceptual framework for MALL to analyze MALL studies (Fig. 1). The AT-MALL has six major components: (a)
Subjects in the MALL activity, which involve all the people who may be involved in learning languages through mobile devices, such as students of different age levels or teachers of different levels of teaching expertise. (b) Objects (or objectives) of
the MALL activity, which focus on the goal of MALL, such as acquiring language skills or enhancing learning motivation
through mobile devices. (c) Tools/instruments in the activity, which may be artifacts (e.g., hardware and software) or learning
resources (e.g., tutors) for MALL. (d) Rules/control for the activity, which are norms or regulations that circumscribe the MALL
activity, such as the procedure in teaching scenarios designed for MALL or the learning pace or styles designated in MALL
platforms. (e) Context of the activity, which refers to the physical (e.g., classroom or museum) or social (e.g., ambience of
Fig. 1. The Action Theory based Mobile Assisted Language Learning Framework.
71
learning in a group) environments for conducting MALL. (f) Communication/interaction, which refers to the method of
interaction between users and MALL technologies (such as the process teachers' adaption to MALL) or the communications
styles among MALL learners (e.g., face-to-face vs. computer-mediated-messaging).
The six components in the framework can be used as a base for describing the content of MALL or general mobile learning
studies (e.g., Frohberg et al., 2009), furthermore, the interactive relationships among the six components also construct a
dynamic system which enables the analysis of MALL activities and studies. For example, the upper part of the triangle of Fig. 1
contains the subject, object, and tools components and is useful to describe the mediating effects of mobile devices on
different subjects and learning goals. Similarly, the Tools-Context-Object triangle may be used for the description of the
mediating effects of learning environments on MALL learning (e.g., Wong, Chen, & Jan, 2012). In this study, we used the ATMALL as a framework for selecting moderating variables (the subject, tools, rules/control, and context), which may affect the
achievement variable (i.e., the object) of MALL studies (Section 2.3).
1.4. Goals of the current meta-analysis
Despite the popular use of mobile devices in language learning, there is a dearth of review research into the overall
effectiveness of MALL and mobile-device-assisted language teaching. Of the related reviews (e.g., AbuSaaleek, 2014; Burston,
2013; 2014; Golonka et al., 2012; Godwin-Jones, 2011; Stockwell, 2013; Wong & Looi, 2011), most have been conducted using
a qualitative approach to describe and summarize the ways in which mobile devices have been used in language teaching and
learning, or noted the problems encountered.
It is difcult to evaluate the actual effectiveness of mobile devices overall and how specic moderator variables inuence
effects through qualitative approaches. Thus we conducted a meta-analysis to organize and quantify studies on the effectiveness of mobile devices in language learning to answer the following questions:
1. In terms of quantitative analysis, what is the status quo of MALL learning? Specically, what kind of participants, types of
hardware and software, teaching/learning methods, implementation settings, language skills, target languages and
duration of the intervention were involved?
2. What is the overall effectiveness of employing mobile technology in education on the language achievement by students?
3. What kinds of moderator variables inuence the effects of mobile devices on language learning?
Based on these research questions, this study proposed three purposes:
1. Provide an overview of the use of mobile devices in language learning, including the target population, types of hardware
and software, teaching/learning methods, the setting in which the teaching/learning, language skills were implemented
(hereafter referred to as intervention setting), target languages, rst language (L1) vs. second/foreign language (L2), and
duration of the intervention.
2. Determine the overall effectiveness of employing mobile technology in education on the language achievement by
students.
3. Investigate whether moderator variables inuence the effects of mobile devices on language learning.
2. Methods
2.1. Data sources and search strategy
The reference search procedure proposed by Cooper (2010) was used to collected MALL-related references. To solicit
comprehensive and sufcient literature of MALL, this study searched published journal articles, unpublished conference
papers and doctoral dissertations from the period 1993e2013. Electronic searches, manual searches, and manual reference
list checking to retrieve the relevant literature. Because the limitation of the authors language competence, only materials
written in English were used. The main databases used for the electronic searches were the Education Resources Information
Center (ERIC) and the Social Sciences Citation Index database of the Institute of Science Index (ISI). Based on the key concepts
and key words of mobile learning in previous reviews (e.g., Burston, 2013; 2014; Frohberg et al., 2009; Sharples et al. 2007;
Stockwell, 2013), the three following sets of keywords, and combinations thereof, were used to search these databases: (1)
mobile-device-related keywords, including mobile, wireless, ubiquitous, wearable, portable, handhelds, mobile/cell phone,
personal digital assistant (PDA), palmtop, pad, web pad, tablet PC, tablet computer, laptop, e-book, digital pen, pocket dictionary, and classroom response system; (2) learning-related keywords, including teaching, learning, training, and lecture;
and (3) language-related keywords, including literacy, language, listening, speaking, reading, writing, grammar, phonemic
awareness, and vocabulary. The three sets of keywords were integrated with Boolean operators (Cooper, 2010), specically
using the OR operator within the set and the AND operator between the sets. Manual searches included the major journals
of educational technology and e-learning, such as the Australia Journal of Educational Technology, British Journal of
Educational Technology, Computer Assisted Language Learning, Computers & Education, Journal of Computer Assisted
Learning, Language Learning & Technology, and ReCALL.
72
After harvesting all of the related literature, another round of searches was conducted using the reference lists found in the
literature to try to nd any omitted but relevant works.
2.2. Search results
2.2.1. The initial screening stage
The literature searches yielded 721 journal article abstracts related to mobile learning and language that were published
between 1993 and 2013 (303 and 418 journal articles in ERIC and ISI, respectively), 287 conference papers (162 and 125 in
ERIC and ISI, respectively), and 68 doctoral dissertations (68 and 0 in ERIC and ISI, respectively). Three authors read each
article abstract and then judged whether or not the article was related to using mobile devices in language education; ultimately, 288 articles, 264 conference papers, and 56 doctoral dissertations were selected during this initial screen process.
2.2.2. Screening for experimental and quasi-experimental studies
In the second stage, the studies located during the initial stage were further screened according to the research method
used. Only experimental studies, including those using designs with pretest-posttest equivalent groups, posttest-only
equivalent groups, randomized matched subjects, and posttest-only control groups, and quasi-experimental studies,
including the pretest-posttest nonequivalent groups and counterbalanced designs (Ary, Jacobs, & Razavieh, 2002), were
included. Conceptual analysis of research reviews, case studies and qualitative research, survey research, and preexperimental studies were all excluded at this stage. After this second-stage screening, 80 journal articles, 29 conference
papers, and 10 doctoral dissertations remained.
2.2.3. Screening for inclusion in/exclusion from the meta-analysis
The following three criteria were used to determine whether a study could be included in the meta-analysis:
1. The applications of mobile devices were the key variables of the study. The experimental group had an intervention that
used a mobile device, and the ndings were compared to control groups that used traditional learning (paper-based or
desktop computers). The studies that used the term e-book to describe the treatment tools but actually used
multimedia-based text with no real physical devices were excluded.
2. The studies presented sufcient information to calculate effect sizes, such as means, standard deviations, t or F values, chisquare values, or the number of people in each group. Studies for which the sample sizes of each group were not provided,
which lacked any inferential statistical results, or for which there were inferential statistical results that were inadequate
for calculating an effect size according to Lipsey and Wilson (2001) were excluded.
3. The experimental results are presented in terms of two categories of learning performance, the rst category was learning
achievement, measured by standardized or researcher-constructed tests, and peer interactions during the learning processes. The second category was affect, studies in which the results were related to affective variables (e.g., learning attitude,
learning motivation, learning engagement, preference, self-efcacy) were included. After this stage, only 43 journal articles
and 2 doctoral dissertations were found to be eligible for inclusion in further meta-analysis. Among the 45 articles, one article
used only affective dependent variables, and 4 articles used dependent variables with both achievement and affection.
2.3. Selection and coding of variables
The codebook included four main entries: research name, research participants, research treatments, and research
outcome/dependent variables. Except for the research name, all the other three entries and their sub-level of moderator
variables corresponded to the components of the AT-MALL framework, as described below.
2.3.1. Research name
This refers to the rst author's name, the year of publication, and the article title.
2.3.2. Research participants
In this review, for all the reviewed articles, the research participant corresponded to the subject of the AT-MALL
framework, and was coded by their learning stages, including kindergarten, elementary school, middle school, (senior)
high school, university, graduate school, teachers, adults, and mixed.
2.3.3. Treatments
The treatments of the reviewed articles corresponded to the tools component (e.g., the hardware and software), the
rules/control component (e.g., the teaching methods), and the context component (e.g., intervention settings, intervention duration, target languages, L1/L2). The description for each of these treatment variables are as follows:
1. Hardware: Different types of mobile hardware, which comprised PDAs, laptops, tablet PCs, cell phones, iPods, MP3 players,
e-book readers, pads, digital pens, pocket dictionaries, and classroom response systems (CRSs), or any mixture of thereof.
73
2. Software: Different types of software, which encompassed general-purpose software and learning-oriented software
(Sung & Lesgold, 2007), the former referring to commercial software currently in circulation that was not designed
especially for teaching and learning (e.g., word processors or spreadsheets), and the latter having been designed specifically for educational programs or goals.
3. Intervention setting: Implementation settings were included to establish whether the impact of mobile devices on language learning differed according to the environment in which they were used, which included classrooms, outdoors (e.g.,
zoo or campus gardens), museum, workplaces, and unrestricted settings (devices may be used anywhere).
4. Teaching method: Different teaching methods, including lectures, cooperative learning (students were divided into groups
and completed language learning tasks collaboratively, e.g., Chang et al., 2010; Huang, Liang, Su, & Chen, 2012), inquiryoriented learning (using problem-, project-, or inquiry-based methods with mobile devices for language learning, e.g.,
Chen, 2010; Lowther, Ross, & Marrison, 2003), self-directed study (teachers/researchers did not designate or implement
specic teaching scenarios for students to follow, students use mobile devices for self-paced learning, e.g., Chen & Li, 2010;
Chen, Tan, & Lo, 2013), computer-assisted testing/assessment (using mobile devices for formative assessment or quizzes in
classroom or outdoors, e.g., Agbatogun, 2012), and mixed methods thereof.
5. Intervention duration: Different periods of time for the intervention, including periods no more than 4 h (4 h), between ve
and 24 h (>4 and 24 h), between one day and seven days (>1 day and 7 days), between one week and four weeks (>1 week
and 4 weeks), between one month and six months (>1 month and 6 months), and more than six months (>6 months).
6. Learning skills: Learning skills comprised listening, speaking, reading, writing, pronunciation, vocabulary, and mixed
thereof.
7. Target languages: Different languages using mobile devices, such as English, Spanish, Chinese, Japanese, etc.
8. L1/L2: using a mobile device for L1 or L2 learning.
Three of the researchers for this study coded the 44 articles according to this framework. Missing values were coded as 0,
indicating it was not mentioned in the literature. The coding process was based on the procedure suggested by Cooper (2010),
the three researchers (coders) rstly built consensus on the denitions the entries and their related variables by selecting and
discussing two exemplar articles. Secondly, the coders independently coded 10 articles, discussed their differences in coding
results, and negotiated for consensus. Finally the coders coded all the remaining articles, discussed the differences of codes,
and negotiated until they got a consensus of all the codes. The software Comprehensive Meta-analysis (Borenstein, Hedges,
Higgins, & Rothstein, 2005) was used for coding.
2.4. Data analysis
2.4.1. Calculating effect sizes
The following meta-analysis steps recommended by Borenstein, Hedges, Higgins, and Rothstein (2009) were employed:
(a) create the effect sizes of each article, (b) calculate the weighted mean effect size across articles, (c) calculate the condence
interval for the average effect size, and (d) determine whether the effect size of any particular group was inuenced by a
moderator variable based on a heterogeneity analysis (using the test statistic QB).
Two formulae were used to calculate the effect sizes of the study results. For experimental research with random
assignment and without a pretest, Cohen's d (Lipsey & Wilson, 2001, p. 48) was used to determine the effect sizes. For
experimental or quasi-experimental research with pretests, it has been proposed that the pretest should be considered rather
than the posttest to mitigate possible selection bias (Furtak, Seidel, Iverson, & Briggs, 2012). Therefore, the formula of Furtak
et al. (2012, p. 311) was used to obtain effect sizes for research with pre- and posttests. Both types of effect size were calibrated
using sample weights to calculate a Hedge's g (Lipsey & Wilson, 2001, p. 49).
In this study, only one effect size was determined for each article; when studies had multiple effect sizes, they were integrated into a single value as per Borenstein et al. (2009).
2.4.2. Evaluating publication bias
The fail-safe N of Rosenthal (1979) (i.e., classic fail-safe N) was used to estimate how many insignicant effect sizes
(unpublished data) would be necessary to reduce the overall effect size to an insignicant level. The comparison criterion was
5n 10 (where n is the number of studies included in the meta-analysis). If the fail-safe N is larger than 5n 10, then the
estimated effect size of unpublished research is unlikely to affect the effect size of the meta-analysis. Moreover, we also
adopted the fail-safe N of Orwin (1983) to estimate the number of missing null studies that would be required to bring the
mean effect size to a trivial level.
3. Results and discussion
3.1. Descriptive information
Table 1 presents the distribution of moderator variables and their corresponding percentages. In total, there were 45
articles, 223 effect sizes (before integration), and 9204 participants. The largest proportion of studies in the learning stage
74
Table 1
Categories of the 45 included articles.
Variable
Category
Number of studies (k)
Number of effect sizes
Proportion of studies (%)
Learning stage
1. Kindergarten
2. Elementary school
3. Middle school
4. High school
5. College
6. Adults
7. Mixed
1. Handheld
2. Laptop
3. Tablet PC
4. Cell phone
5. iPod or MP3 player
6. Mixed
7. E-book reader
8. Digital pen
9. Pocket dictionary
10. CRS
1. General-purpose
2. Learning-oriented
1. Classroom
2. Outdoors
3. Unrestricted
0. Not mentioned
1. Lecture
2. Cooperative learning
3. Inquiry-oriented
4. Self-directed study
5. Mixed
6. Computer-assisted testing
0. Not mentioned
1. 4 h
2. > 4, 24 h
3. > 1, 7 days
4. > 1 week, 4 weeks
5. > 1 month, 6 months
6. > 6 months
1. Listening
2. Reading
3. Writing
4. Pronunciation
5. Vocabulary
6. Mixed
1. English
2. Chinese
3. Hebrew
4. Spanish
1. L1
2. L2
1
18
9
3
12
1
3
14
8
2
10
4
1
2
1
2
1
19
26
26
2
17
6
4
7
2
21
4
1
1
5
0
1
14
18
6
3
14
7
2
18
9
38
3
1
3
19
26
1
62
62
11
33
1
53
52
50
17
27
13
2
41
9
11
1
116
107
142
5
76
37
10
68
13
65
29
1
8
46
0
6
36
88
39
5
87
28
9
47
47
208
10
1
4
125
98
0.022
0.400
0.200
0.067
0.267
0.022
0.067
0.311
0.178
0.044
0.222
0.089
0.022
0.044
0.022
0.044
0.022
0.422
0.578
0.578
0.044
0.378
0.133
0.089
0.156
0.044
0.467
0.089
0.022
0.022
0.111
0.000
0.022
0.311
0.400
0.133
0.067
0.311
0.156
0.044
0.400
0.200
0.844
0.067
0.022
0.067
0.422
0.578
Hardware used
Software used
Intervention setting
Teaching method
Intervention duration
Learning skill
Language
L1/L2
included elementary-school students (40.0%), and the next largest group used college students as subjects (26.7%). Regarding
software used, learning-oriented software was used more often (57.8%) than general-purpose software (42.2%). As for
hardware, most studies (73.3%) used handhelds (including cell phones, iPods or MP3 players, digital pens, pocket dictionaries,
and CRSs), followed by laptops (26.7%; including laptops, tablet PCs, and e-book readers). The setting for more than half of the
studies was a classroom (57.8%), followed by unrestricted (37.8%). The most frequently used teaching method was selfdirected study (46.7%), and the most frequent intervention duration was 1e6 months (40.0%), followed by 2e4 weeks
(31.1%) and >6 months (13.3%). With respect to learning skills, the most frequently studied was vocabulary skills (40.0%),
followed by reading (31.1%) and mixed skills (20.0%). Finally, the language being learned via the intervention was mostly
English (84.4%), followed by Chinese (6.7%) and Spanish (6.7%), and in more than half of the studies (57.8%) L2 learning, with
the remainder being L1 learning (42.2%).
Among those moderating variables, the evolution of hardware and software may have seen the greatest amount of change
during the two decades. Fig. 2 shows the evolution of the use of different mobile devices. Although smaller handheld devices
(e.g., cell phones) were used more in recent years, the use of larger devices, such as laptops and tablet PCs, has remained
steady. One of the reasons for the laptops being used in MALL may be that, although handheld devices (e.g., cell phones) have
their advantages in mobility and portability, the disadvantages inherent to smaller devices (small screens, lower processing
75
Fig. 2. Histogram of the hardware used in mobile assisted language learning across time.
power, etc.) may hinder the use of handhelds in regular classroom learning. Instead, when equipped with wireless technologies, laptops can not only be more comfortable for the use in classrooms for regular reading and writing tasks, they can
also be used in informal settings for novel activities such as communication and data collection. The comparison of effects of
using different types of hardware is shown in Section 3.3.
Another concern of previous researchers was if there was sufcient software for learning/teaching activities. Scholars have
proposed that it is thus difcult for teachersdwho often experience severe restrictions on time availabilitydto achieve
effective and efcient teaching styles with general-purpose software (Sung & Lesgold, 2007; Weston & Bain, 2010). From
Fig. 3, we are pleased to learn that the learning-oriented software for MALL increased during the two decades, and the amount
of learning-oriented software has surpassed that of general-purpose software, which means researchers are more aware of
Fig. 3. Histogram of the software used in mobile assisted language learning across time.
76
the importance of software tailored for specic teaching/learning purposes. The comparison of the effects of the two types of
software is shown in Section 3.3.
3.2. The overall effect size for learning achievement

There were 223 effect sizes in the 45 articles. If there was more than one effect size in any article, they were integrated into
a single effect size for the entire article as per Borenstein et al. (2009). The distribution of the effect sizes of the 45 articles is
shown in Fig. 4. An unusually large effect size (g 4.045) was found for the study of Hsu and Lee (2011), and this was much
larger than the average effect size (g 0.611) of the 45 articles; therefore, this effect size was not included in further analyses
(Lipsey & Wilson, 2001). Using the procedure of Lipsey and Wilson (2001) with a random-effect model to integrate the effect
sizes of the 44 included articles, the overall mean effect size of the 44 articles was a moderate one of 0.546, with a 95%
condence interval of 0.387e0.705. Q statistics revealed that the effect sizes in the meta-analysis were heterogeneous
(Qtotal 338.210, z 10.541, p < 0.001), which indicates that there were differences among the effect sizes that were
attributable to sources other than subject-level sampling error. According to Cohen (1988), effect sizes of 0.80 and 0.20
are considered large and small, respectively, with intermediate values considered moderate. Thus, the present ndings
suggest that MALL has a moderate effect size for learning achievement; in other words, 70.7% of the learners using a mobile
device performed signicantly better than those learning who were not using one.
Furthermore, to understand if a difference of effect sizes exists in the achievement-related (such as test scores) and
affective-related dependent variables (such as motivation, engagement, attitude, satisfaction, preference), we conducted an
analysis for the studies related to the two categories of dependent variables, the results are shown in Table 2. MALL studies
with achievement and affective variables had effect sizes of 0.531 (p < 0.001) and 0.550 (p < 0.001), respectively, which
suggest that MALL has a similar moderate effect on students achievement and affective in language learning.
The ndings above provide solid evidence for the advantages of using mobile devices in language learning and teaching
(Kukulska-Hulme, 2009; Stockwell, 2013), and complements the qualitative descriptions of previous review research on
mobile-device-based language learning (e.g., Burston, 2013; 2014). Previous studies of desktop CALL revealed that computer
technology was associated with a smaller effect size (g 0.24e0.35) for language learning (Grgurovic, Chapelle, & Shelley,
2013); it therefore seems that mobile devices generally generate larger effects than desktop computers in supporting language learning and teaching. However, more research is needed to establish unequivocally the different effect sizes of using
desktop computers and mobile devices for language education.
3.3. The effect size of learning achievement for moderator variables

To learn more about the effects of moderating variables on MALL, this study conducted analyses for the effects of learning
achievement with moderator variables. Because there were only 5 studies related to affective dependent variables, which is
not comprehensive enough to cover different levels of moderating variables, the moderator analyses did not include the
affective effects.
Fig. 4. Histogram of the effect sizes for the 45 included articles of this meta-analysis.
77
Table 2
The effect sizes of achievement- and affective-related studies.
Category
Dependent Variable
1. Achievement
2. Affective
95% CI
QB
df
0.011
43
5
0.531
0.550
6.404***
3.450**
dfdf
1
[0.369e0.694]
[0.238e0.862]
Some moderators had small samples (see Table 1), and in these several levels were merged for some moderator variables. For
learning stage, kindergarten and elementary school were combined into a young children category; middle school and high
school were combined into secondary schoolers and college and graduate students, teachers, and adults were combined into
adult users. For hardware used, laptops, tablet PCs, and e-book readers were combined into a laptop computer category, and
PDAs, iPods or MP3 players, cell phones, digital pens, dictionaries, and CRSs were bundled together into one handheld category.
Some intervention durations were also combined, with 4 h, > 4 h and 24 h, and >1 day and 7 days becoming 1 week. The
learning skills of pronunciation and vocabulary were also combined. Table 3 lists the effect sizes for the moderator variables.
3.3.1. Learning stage
The effect sizes for mobile device usage for learning achievement in different learning stages were examined. Adult usage
had the largest effect on learning achievement (g 0.948, p < 0.001), followed by young children (g 0.508, p < 0.001). QB
reached signicance (QB 9.457, p 0.024), indicating that the mean effect sizes differed between the categories.
Table 3
The learning-achievement effect sizes of categories and their related moderator variables.
Category
Learning stage
1. Young children
2. Secondary schoolers
3. Adults
4. Mixed
Hardware used
1. Handheld
2. Laptop
Software Used
1. General-purpose
2. Learning-oriented
Intervention setting
1. Classroom
2. Outdoors
3. Unrestricted
Teaching method
0. Not mentioned
1. Lecture
2. Cooperative learning
3. Inquiry-oriented
4. Self-directed study
5. Mixed
6.Computer-assisted testing
Intervention duration
0. Not mentioned
1. 1 week
2. > 1, 4 weeks
3. > 1 month, 6 months
4. > 6 months
Learning skill
1. Listening
2. Reading
3. Writing
4. Pronunciation & vocabulary
5. Mixed
Language
1. English
2. Chinese
3. Hebrew
4. Spanish
L1/L2
1. L1
2. L2
Note. CI condence interval.
*p < 0.05; **p < 0.01; ***p < 0.001.
95% CI
18
12
12
3
0.508
0.488
0.948
0.457
4.344***
5.270***
3.805***
1.189
[0.279e0.738]
[0.307e0.670]
[0.459e1.436]
[1.209e0.296]
31
12
0.729
0.151
5.769***
1.506
[0.481e0.976]
[0.046e0.348]
18
25
0.479
0.578
3.802***
4.782***
[0.232e0.726]
[0.341e0.815]
25
2
16
0.367
0.498
0.797
3.606***
2.303*
5.102***
[0.167e0.566]
[0.074e0.923]
[0.491e1.103]
6
2
7
2
21
4
1
0.132
1.285
0.268
0.376
0.581
1.127
1.769
1.950
1.207
0.805
0.651
4.083***
2.618**
6.068***
[0.001e0.264]
[0.801e3.370]
[0.384e0.920]
[0.756e1.508]
[0.302e0.859]
[0.283e1.970]
[1.198e2.340]
1
6
13
17
6
0.949
0.231
0.622
0.772
0.130
4.934***
0.560
3.333**
4.371***
2.158*
[0.572e1.326]
[0.578e1.040]
[0.256e0.988]
[0.426e1.119]
[0.012e0.248]
3
14
7
19
8
0.733
0.280
0.253
0.446
0.812
1.606
1.468
2.364*
3.969***
4.409***
[0.161e1.627]
[0.094e0.653]
[0.043e0.463]
[0.226e0.666]
[0.451e1.173]
36
3
1
3
0.548
0.426
0.366
0.557
5.940***
2.322*
1.745
2.021*
[0.367e0.729]
[0.067e0.786]
[0.045e0.777]
[0.017e1.097]
18
25
0.181
0.837
2.174*
5.583***
[0.018e0.344]
[0.543e1.131]
QB
df
9.457*
12.834***
0.322
5.334
40.080***
29.461***
7.976
0.875
14.623***
78
3.3.2. Hardware used

Handhelds (g 0.729, p < 0.001) achieved a moderate-to-high effect size, while there was no signicant effect size for
laptops (g 0.151, p 0.132). QB was signicant (QB 12.834, p < 0.001), indicating that the effect sizes differed signicantly
among the various categories.
Handhelds appear to be more effective than portable computers. They provide the opportunity to access target languages,
interact with others, both give and receive feedback, and are able to increase learner interest in language learning compared
to desktop computers (Chinnery, 2006; Golonka et al., 2012). Handhelds have been used to organize specic learning topics,
clarify learning goals, and develop innovative teaching methods (Golonka et al., 2012). The present ndings reveal that
handhelds produced signicant moderate-to-high effects in language learning, in contrast to portable computers (including
laptops and tablet PCs). The portability, context-sensitivity, and easy interaction offered by handhelds appears to provide
more opportunities for language learning without time and space restrictions, and ultimately produces a positive learning
outcome.
It is important to note that while our analysis showed handhelds outperforming mobile computers, some of this difference
may be because most of the research on laptops included in this study used learning activities that did not deal directly with
knowledge or standards closely related to their measurement tool, and this obfuscates the effect of laptops on learning (Bebell
& Kay, 2010; Gulek & Demirtas, 2005). Research on handhelds did not often encounter this issue.
3.3.3. Software used
The effect sizes for learning-oriented software (g 0.578, p < 0.001) and for general-purpose software (g 0.479,
p < 0.001) approached a moderate level. However, QB did not reach the 0.05 signicance level (QB 0.322, p 0.571), which
means that the average effect size did not differ signicantly between the various categories of software.
3.3.4. Intervention setting
Unrestricted settings had the largest effect size (g 0.797, p < 0.001), followed by informal settings such as outdoors
(g 0.498, p 0.021) and formal settings such as classrooms (g 0.367, p < 0.001). QB did not reach statistical signicance at
the 0.05 level (QB 5.334, p 0.069), showing that the effect size did not differ signicantly between the different categories
of intervention settings.
One of the reasons why mobile devices created more marked learning effects in multiple and unrestricted settings may be
that integrating mobile devices with language-learning contexts permits students to conduct learning tasks beyond the
formal learning environment and with fewer restrictions. Formal and classroom learning can be complemented by outsidethe-classroom or informal learning (Sandberg et al., 2011). Learning that takes place in multiple and unrestricted settings
exerts a maximal learning effect by connecting formal (e.g., classroom) and informal learning (e.g., home or real-life situations). Moreover, learners are given abundant opportunities to practice their language learning through well-dened and
structured curricula in formal learning environments, which continues through the use of mobile applications in informal
settings (Chen & Li, 2010; Saran, Seferoglu, & Cagiltay, 2009).
3.3.5. Teaching method
The effect size was largest for computer-assisted testing (g 1.769, p < 0.001), followed by mixed methods (g 1.127,
p 0.009) and self-directed study (g 0.581, p < 0.001). The teaching methods of lectures (g 1.285, p 0.227), cooperative
learning (g 0.268, p 0.421), and inquiry-oriented learning (g 0.376, p 0.515) did not yield signicant effect sizes. QB
reached statistical signicance (QB 40.080, p < 0.001), indicating that the effect sizes differed signicantly among the
various categories of teaching.
Most studies using mixed teaching strategies integrated multiple teaching/learning scenarios that might have greatly
enhanced the language skills of the students. For example, Hwang, Chen, Shadiev, Huang, and Chen (2012) integrated
computer-based writing support along with peer feedback for students English writing. Liu and Chu (2010) also designed
multiple learning scenarios with game-based, self-directed study, and collaborative learning for practicing English speaking
and listening. These learning scenarios provided ample opportunities for practicing, resulting in the integration and
strengthening of individual separate skills, which may be the main reason for the sound effects produced by mixed-learning/
teaching scenarios.
Since only one study comparing the use of mobile devices for language assessment with traditional learning/teaching
methods has been reported, more research is needed to conrm the reliability of the effect sizes for mobile devices in this
study. The features of real-time feedback and individualized user interface make mobile devices unique tools for formative
assessment, which may need instant individualized feedback and be important for self-paced learning. However, these
functions seem under-recognized in MALL, because among the 119 experimental studies screened in the second stage, only
two studies (Agbatogun, 2012; Chen & Chung, 2008) took advantage of those functions to implement assessment tools for
learning languages, and only one study was eligible for inclusion in this study. The positive effects of mobile devices for selfdirected study provide supporting evidence for both the feasibility of ubiquitous access and the importance of practice for
language learning. For example, several researchers (e.g., Lu, 2008; Thornton & Houser, 2005; Wood et al., 2011) used text
messaging as a strategy for helping students to learn vocabulary-related knowledge in their spare time during out-of-school
hours; Brooks, Miles, Torgerson, and Torgerson (2006) used phone-based recordings, and Chen, Tan, and Lo (2013) provided
students with digital pens for practicing, monitoring, and feedback of their oral reading.
79
The features of individuality, portability, and social connectivity have rendered mobile devices a potential tool for
cooperative or collaborative learning. Many studies in elds other than language learning (e.g., Jones, Antonenko, &
Greenwood, 2012; Baloian & Zurita, 2012) have exploited these features to enhance the effectiveness of group learning.
Although individual language-learning studies have found cooperative learning and problem-solving methods effective for
producing positive learning outcomes and motivating learning (Chang et al., 2010; Lan et al., 2007, 2009; Zurita & Nussbaum,
2004a), our ndings indicate a statistically insignicant and small overall effect size for cooperative learning. This may be
related to the criticism of Golonka et al. (2012) that the use of computers can result in shallow interactions in language
learning. Since most of the research included in the present study did not designate clear steps and mechanisms for stimulating learners interactions in their collaborative tasks, it is unknown whether their communication initiations, responses,
negotiations, and discussions were sufciently intensive as to enhance the communication competence of the students. It
seems that more cooperative teaching/learning scenarios that are specically designed to increase the frequency and depth of
interactions among learners are needed to illuminate the features and functions of mobile devices with respect to cooperative/collaborative learning.
3.3.6. Intervention duration
When the not-mentioned category was ignored, interventions of 1e6 months had the largest effect size (g 0.772,
p < 0.001), followed by 2e4weeks (g 0.622, p 0.001) and >6 months (g 0.130, p 0.031). No signicant effect size was
found for interventions lasting 1 week (g 0.231, p 0.575). QB reached statistical signicance (QB 29.461, p < 0.001),
indicating that the effect sizes differed signicantly among the various categories.
The lower effect sizes found for short-duration interventions may be because the participantsdand especially young
childrendneeded more time to become familiar with using mobile devices and with the teaching/learning scenarios. The
lower effect sizes for long-term interventions may be due to a loss of the sense of novelty in the devices themselves or a loss of
interest in the repetitive content by the students. Furthermore, teachers might be not able to provide innovative teaching
methods and might even reduce the amount of time allocated for students to spend with mobile devices if there is no
appropriate logistical support for hardware maintenance, software supplementation, and curriculum design during a longterm program (Sandberg et al., 2011). Cross-analysis of intervention durations with other moderator variables provides
some supporting information for these arguments. For example, most research that took place over a 6-month period was
conducted in classrooms or formal settings (83%, due to space limitations, the cross-analysis between variables is presented at
the URL http://140.122.96.188/MV), used general-purpose software (67%), and employed nonspecically designed teaching
methods (67%). The monotony of teaching scenarios and learning software may result in the learners not maintaining their
motivation to use mobile devices for learning, and reduced frequency of using mobile devices in the classrooms, which then
ultimately made the mobile use in teaching supercial (Fleischer, 2012). However, most studies in which the intervention was
implemented for 2 weeks to 6 months used specically designed teaching methods (93%) and teaching-oriented software
tailored for specic learning activities (67%).
3.3.7. Learning skills
The effect size was largest for mixed skills (g 0.812, p < 0.001), followed by vocabulary and pronunciation (g 0.446,
p < 0.001), and writing skills (g 0.253, p 0.018). The effect sizes for listening (g 0.733, z 1.606, p 0.108) and reading
skills (g 0.280, p 0.142) were not signicant. QB did not reach statistical signicance (QB 7.976, p 0.092), indicating that
the effect size did not differ signicantly between the different categories.
The features of mobile devices might play particularly appropriate roles for facilitating the learning performance for skills
that can be enhanced over a short period of time, such as vocabulary learning. For example, Sandberg et al. (2011) found the
portability of smartphones useful for extending children's learning of English vocabulary from classrooms to informal environments (e.g., at home and at a zoo).
When coupled with an appropriate intervention duration and opportunities for practice, mobile devices seem able to
produce an overall benecial effect for language skills for which intensive practice is required, such as writing. For example,
most of the studies (71.4%) involved with writing tasks were conducted for at least 6 months. In contrast, for reading and
listening skills, which require long-term training, less than 10% of the teaching interventions were longer than 6 months. This
may be one of the main reasons why the effects of mobile devices on listening and reading skills were not signicant.
3.3.8. Target languages
When integrated with mobile devices, Spanish (g 0.557, p 0.043), English (g 0.548, p < 0.001), and Chinese
(g 0.426, p 0.020), had moderate effect sizes. However, there was no signicant effect size for Hebrew (g 0.366,
p 0.081). QB did not reach statistical signicance (QB 0.875, p 0.831), which shows that the effect size did not differ
signicantly between the categories.
3.3.9. L1/L2
The effect size for L2 approached a large level (g 0.837, p < 0.001), while that for L1 was small (g 0.181, p 0.030). QB
reached statistical signicance (QB 14.623, p < 0.001), indicating that the effect size differed signicantly between the two
categories.
80
One possible explanation for mobile devices produced better effects for L2 learning may be that learners are more familiar
with and skillful at L1 skills compared with L2 skills, and so it may be more difcult to enhance the learning effects of L1
through mobile devices. The second reason may be that most studies on L1 learning using a mobile device were conducted for
either very short (i.e., 1 week; 22%) or very long durations (i.e., >6 months; 28%); as discussed above, these two intervention
durations did not readily produce intervention effects.
3.4. Evaluation of the publication bias
The classic fail-safe N and Orwin's fail-safe N were adopted in this study to demonstrate the publication bias for the 43
selected studies. As suggested by the data given in Table 4, the classic fail-safe N test determined that a total of 1960 studies
with null results would be needed in order to nullify the effect size. Moreover, the results for Orwin's fail-safe N (see Table 4)
show that the number of missing null studies required to bring the existing overall mean effect size to a trivial level (g 0.01)
was 1041. Both tests therefore suggest that publication bias cannot explain the signicant positive effects observed across all
of the studies (Table 5).
4. Conclusions and implications
4.1. Conclusions
By investigating the empirical research that has been conducted into mobile language learning in published journal articles and unpublished dissertations and conference papers, our research provides concrete evidence regarding the overall
effects of using mobile devices in language education and how those effects vary between moderator variables. We found that
learning/teaching with mobile devices produced a moderate mean effect size of 0.531 and 0.550, for the achievement and
affective effects, respectively. These mean that around 70% of students in the experimental groups who were learning with
the aid of mobile devices would outperform their counterparts who learned languages without such devices. Furthermore,
analyses of effect sizes with different moderator variables revealed the following interesting pictures of mobile language
learning:
1. Adults and school children had similar benecial effects from MALL.
2. Handhelds had a larger effect size than laptops for language learning.
3. Learning-oriented software and general-purpose software rendered similar positive effects/learning achievement in
MALL.
4. The functionalities of mobile devices in multiple learning settings generated more marked effects/learning achievement
than the more restricted settings in classrooms or outdoors.
5. Integrating mobiles with multiple teaching/learning strategies produced better effects/learning achievement than with
only lectures, or inquiry-oriented or cooperative learning.
6. Midterm interventions (1e6 months) produced better effects/learning achievement than very-short-term (1 week) and
long-term (>6 months) interventions.
7. Using mobile devices for vocabulary or mixed language skills produced better effects/learning achievement than for single
skills such as listening and reading.
8. Using mobiles produced better effects/learning achievement for L2 learning than for L1 learning.
Are the limitations of CALL mentioned in previous research (e.g., Golonka, et al., 2012), such as distraction, lack of effective
software, shallow interaction, and increased workload, eliminated in MALL? Our ndings provide both pros and cons for the
claims of advantages in using mobile devices for language learning. Integrating mobile devices with language learning
increased more learning-oriented software which may be less emphasized in CALL studies; mobile devices also brought more
diverse and innovative activities, scenarios, and situations for learning and teaching. However, the diverse MALL activities did
not necessarily bring out higher learning performance: for example, MALL did not overcome the critiques of shallow interaction because the overall effects of collaborative MALL were not signicant, which perhaps resulted from the lack of elaborate interaction/communication designs for MALL activities. Researchers (e.g., Stockwell, 2013; Lan, 2013, 2015; Lan, Fang,
Table 4
Results of the classic fail-safe N.
Z value for observed studies
p value for observed studies
Alpha
Tail
Z for alpha
Number of observed studies
Number of missing studies that would bring the p value to > alpha
13.38
0.00
0.05
2.00
1.96
43
1960
81
Table 5
Results of Orwin's fail-safe N.
Hedge's g in observed studies (xed effect)
Criterion for a trivial Hedge's g
Mean Hedge's g in missing studies
Number of missing studies needed to bring Hedge's g to under 0.01
0.25
0.01
0.00
1041
Legault & Li, 2015) proposed that the portability and connectivity of mobile devices may lead to more diverse and innovative
learning activities and situations, especially outdoor environments. However, except for text messaging, MALL activities for
informal settings and inquiry-oriented activities were scarce and their effects were very limited.
Our quantitative review also concurs with some ndings of qualitative reviews. For example, Abusaaleek (2014) and
Burston (2014) found that MALL studies suffered from the lack of oral, communication, and interaction studies, both in terms
of quality and quantity. Our research also found that there are scarce experimental studies regarding expressive skills like
speaking and writing, and the MALL effect on writing skills is not obvious. Furthermore, both Godwin-Jones (2011) and
Burston (2014) proposed that MALL studies needed to enhance the quantity and quality of collaborative MALL activities, and
our research also indicated that the effects of collaborative learning are obscure and more powerful mechanisms for
enhancing peer interaction are needed. The ndings above provide insights and implications for further MALL research and
practice, which are elaborated below.
4.2. Implications
4.2.1. Appropriately incorporating mobile devices into language learning/teaching activities
The features of mobile devices are not sufcient for positive language learning effects; instead, researchers must determine the key points for embedding those features into teaching scenarios and to maximize the functions of those features.
For example, researchers may use the portability of mobile devices to design anywhere language learning. Our research
conrmed that using the portability feature of mobile devices facilitated multiple-settings learning and created a seamless
learning environment that provided learners with distributed, continued, and intensive opportunities for practicing their
language skills or knowledge. Nevertheless, the feature of portability did not produce an overall positive effect when the
researchers tried to use mobile devices for learning languages in the outdoor setting only, which may induce temporary
feelings of novelty but does not maintain learning effects.
4.2.2. Using a quality research design to empower effective MALL interventions
The ndings of previous research have highlighted several common problems that are encountered in CALL studies (Felix,
2005; Golonka et al., 2012; Hubbard, 2005, Liu, Lin, Gao, Yeh, & Kalyuga, 2015; Hong, Hwang, Liu, Lin, & Chen, 2013), such as a
non-rigorous research design, inappropriately chosen variables, and a lack of systematic investigation of the key variables.
These issues also applied to MALL research. Several noteworthy design factors will inuence the effects of MALL. Firstly, the
targeted dependent variables (e.g., language skills) should be closely related to the features emphasized in the teaching
intervention. For example, the features of real-time sharing, communication, and feedback in mobiles are benecial for
cooperative learning, and therefore researchers may be interested in enhancing language-skills-related communication
through those features. Interestingly, among the ten studies looking at the utility of mobile devices for cooperative learning
that we analyzed, only one (Lan et al., 2007) chose communication-related skills as dependent variables.
Secondly, the use of an appropriate intervention duration is important for MALL programs to produce positive effects.
Different language skills may need different intervention durations; mobile language programs thus require more discretion
for locating time slots for different skills and maximizing the functionalities of mobile devices. For example, vocabularyrelated skills may be achieved with bite-sized materials and very short-term activities. However, more complicated skills
such as reading, listening, and writing may need a fuller range of time slots, and the features of mobile devices (e.g., screen
size) should be considered for suitability. Furthermore, both short- and long-term programs deserve special consideration in
terms of research methodology and teaching practices to make their effect obvious. For example, short interventions (e.g., <1
week) may be confounded by technology-novelty effects, and their effects risk insufcient reliability and ecological validity;
however, long interventions (e.g., >6 months) need to be combined with logistical support, such as elaborate teaching materials, software for teaching/learning, and sufcient access to professional development or consultation for technology use.
Thirdly, researchers may consider the possible inadequacy of certain research topics, and try to deepen and broaden their
designs in order to produce more innovative teaching scenarios and to strengthen their effects. For example, all of the
research related to text messaging in the present review used a similar design of sending messages to students for studying
vocabulary when they were out of the classroom. This treats students as passive receivers and message processors, and none
investigated how the students reacted to the messages or how they learned from them. Further research should consider how
to design appropriate functions to record reactions from the learners during messaging, and guide them to engage in active
learning through appropriate feedback instead of receiving messages passively.
Fourthly, researchers may pay more attention to the effect of MALL on affective variables, such as attitude, motivation,
engagement, preferencesetc. In our research, among the 119 experimental studies, there were only ve studies involved
82
with affective variables, the lack of such studies is a missing link of MALL studies, and should be compensated for the understanding of the whole picture of implementing MALL programs.
4.2.3. Substantiating the effects and enriching the diversity of MALL
The essence of MALL is using mobile devices to enhance the effectiveness of language teaching/learning methods. Most
current mobile-device-based teaching methods are focused on self-directed study, with very little in the way of specic
scenarios involved. Although some research has incorporated strategies proposed in the literature, such as cooperative
learning or problem solving, unfortunately these methods have not shown the overall positive effects evident in our research.
These ndings indicate that more proven effective teaching methods need to be elaborated and then integrated into learning
scenarios. Some feasible candidates are annotation (e.g., Hwang, Shadiev, & Huang, 2011), speech-recognition technologies
(e.g., Tanner & Landon, 2009), and language-task-based role-playing (e.g., Bygate, Swain, & Skehan, 2013). Furthermore, the
features of individuality and real-time feedback in mobile devices are especially useful for real-time assessment, and have
been exploited in other domain subjects such as science and mathematics (e.g., Shih, Kuo, & Liu, 2012; Siozos, Palaigeorgiou,
Triantafyllakos, & Despotakis, 2009). However, among the 44 articles included in our research, only 1 study (Agbatogun, 2012)
used these functions. To exploit the interactive/individualized learning functions of mobile devices, future research may
incorporate more validated language assessment methods into MALL programs for feedback; using tailored L2 text for
adaptive L2 reading/learning (Sung et al., 2014; Sung, Lin, Dyson, Chang, & Chen, 2015) may also be considered in MALL
designs.
Regarding the target languages, the present study found that the language skills investigated through experimental
studies are limited. Most of the studies focused on receptive skills such as reading (33%), while ignoring communication skills
such as listening, speaking, and writing. Furthermore, most studies emphasized vocabulary-related knowledge, such as
vocabulary and pronunciation (44%), and none of them addressed the issue of grammar knowledge. Similar to the issue of
limited skills, the number of target languages for mobile learning is also limited. Most of the experimental language learning
programs involve English, with other target languages being scarce. It therefore appears that the problem of imbalanced
development of CALL programs (Golonka et al., 2012; Warschauer, 2004) is also present in MALL programs. Future research,
including meta-analyses, needs to extend mobile-device interventions to different linguistic systems along with different
skills in order to determine whether the effectiveness of language learning with mobile devices is inuenced by specic
linguistic systems and dimensions of language competence.
Finally, it is important to understand how language teachers adapt to MALL activities, especially how mobile devices
hardware and software enable teachers to reduce workload and innovate teaching practices. Despite the importance of those
issues, none of the 44 quantitative studies reviewed in our research addressed those questions. More empirical research of
teacher behavior will be of great value for improving MALL research and practices.
References1
AbuSaaleek, O. A. (2014). A review of emerging technologies: mobile assisted language learning (MALL). Asian Journal of Education and e-Learning, 2(6),
469e475.
*Agbatogun, A. O. (2012). Exploring the efcacy of student response system in a sub-Saharan African country: a sociocultural perspective. Journal of Information Technology Education, 11(1), 249e267.
Ary, D., Jacobs, L. C., & Razavieh, A. (2002). Introduction to research in education (6th ed.). Belmont, CA: Wadsworth.
Bakhurst, D. (2009). Reections on activity theory. Educational Review, 61(2), 197e210.
Baloian, N., & Zurita, G. (2012). Ubiquitous mobile knowledge construction in collaborative learning environments. Sensors, 12(6), 6995e7014.
*Bebell, D., & Kay, R. (2010). One to one computing: a summary of the quantitative results from the Berkshire wireless learning initiative. The Journal of
Technology, Learning, and Assessment, 9(2), 1e60.
Borenstein, M., Hedges, L., Higgins, J., & Rothstein, H. (2005). Comprehensive meta-analysis (Version 2) [Computer software]. Englewood, NJ: Biostat.
Borenstein, M., Hedges, L. V., Higgins, J. P. T., & Rothstein, H. R. (2009). Introduction to meta-analysis. Chichester, UK: Wiley.
*Brooks, G., Miles, J. N. V., Torgerson, C. J., & Torgerson, D. J. (2006). Is an intervention using computer software effective in literacy learning? A randomised
controlled trial. Educational Studies, 32(2), 133e143.
Burston, J. (2013). Mobile-assisted language learning: a selected annotated bibliography of implementations studies 1994-2012. Language Learning &
Technology, 17(3), 157e225.
Burston, J. (2014). MALL: the pedagogical challenges. Computer Assisted Language Learning, 27(4), 344e357.
Bygate, M., Swain, M., & Skehan, P. (2013). Researching pedagogic tasks: Second language teaching, learning, and testing. New York, NY: Routledge.
*Chang, K. E., Lan, Y. J., Chang, C. M., & Sung, Y. T. (2010). Mobile-device-supported strategy for Chinese reading comprehension. Innovations in Education and
Teaching International, 47(1), 69e84.
*Chen, Y. (2010). Dictionary use and EFL learning. A contrastive study of pocket electronic dictionaries and paper dictionaries. International Journal of
Lexicography, 23(3), 275e306.
Chen, C. M., & Chung, C. J. (2008). Personalized mobile English vocabulary learning system based on item-response theory and learning memory cycle.
Computers & Education, 51(2), 624e645.
*Chen, C. M., & Li, Y. L. (2010). Personalised context-aware ubiquitous learning system for supporting effective English vocabulary learning. Interactive
Learning Environments, 18(4), 341e364.
*Chen, C. M., Tan, C. C., & Lo, B. J. (2013). Facilitating English-language learners' oral reading uency with digital pen technology. Interactive Learning Environments, 1e23.
Chinnery, G. M. (2006). Emerging technologies going to the mall: Mobile assisted language learning. Language Learning & Technology, 10(1), 9e16.
Cohen, J. (1988). Statistical power analysis for the behavioral sciences (2nd ed.). Hillsdale, NJ: Erlbaum.
Cooper, H. (2010). Research synthesis and meta-analysis: A step-by-step approach (4th ed.). London, UK: Sage Publications.
References marked with an asterisk indicate studies included in the meta-analysis.
83
Engestrom, Y. (1987). Learning by expanding. Helsinki, Finland: Prienta-Konsultit Oy.

Engestrom, Y. (1999). Activity theory and individual and social transformation. In Y. Engestrom, R. Miettinen, & R.-L. Punamaki (Eds.), Perspectives on activity
theory (pp. 19e38). Cambridge: Cambridge University Press.
Felix, U. (2005). Analyzing recent call effectiveness researchdTowards a common agenda. Computer Assisted language Learning, 18(1e2), 1e32.
Fleischer, H. (2012). What is our current understanding of one-to-one computer projects: a systematic narrative research review. Educational Research
Review, 7(2), 107e122.
Frohberg, D., Goth, C., & Schwabe, G. (2009). Mobile learning projects e a critical analysis of the state of the art. Journal of Computer Assisted Learning, 25(4),
307e331.
Furtak, E. M., Seidel, T., Iverson, H., & Briggs, D. C. (2012). Experimental and quasi-experimental studies of inquiry-based science teaching: a meta-analysis.
Review of Educational Research, 82(3), 300e329.
Gamper, J., & Knapp, J. (2002). A review of intelligent call systems. Computer Assisted Language Learning, 15(4), 329e342.
Garrett, N. (2009). Computer-assisted language learning trends and issues revisited: Integrating innovation. The Modern Language Journal, 93(s1), 719e740.
Gaudreau, P., Miranda, D., & Gareau, A. (2014). Canadian university students in wireless classrooms: What do they do on their laptops and does it really
matter? Computers & Education, 70, 245e255.
Godwin-Jones, R. (2011). Emerging technologies: mobile apps for language learning. Language Learning and Technology, 15(2), 2e11.
Golonka, E. M., Bowles, A. R., Frank, V. M., Richardson, D. L., & Freynik, S. (2012). Technologies for foreign language learning: a review of technology types
and their effectiveness. Computer Assisted Language Learning, 27(1), 70e105.
Grgurovic, M., Chapelle, C., & Shelley, M. (2013). A meta-analysis of effectiveness studies on computer technology-supported language learning. ReCALL,
25(2), 165e198.
*Gulek, J. C., & Demirtas, H. (2005). Learning with Technology: the impact of laptop use on student achievement. The Journal of Technology, Learning, and
Assessment, 3(2), 1e38.
Handal, B., MacNish, J., & Petocz, P. (2013). Adopting mobile learning in tertiary environments: Instructional, curricular and organizational matters. Education Sciences, 3(4), 359e374.
Hong, J. C., Hwang, M. Y., Liu, Y. T., Lin, P. H., & Chen, Y. L. (2013). The role of pre-game learning attitude in the prediction to competitive anxiety, perceived utility
of pre-game learning of game, and gameplay interest, Interactive Learning Environments (pp. 1e13).
Hsu, L., & Lee, S. N. (2011). Learning tourism English on mobile phones: how does it work? Journal of Hospitality, Leisure, Sport and Tourism Education, 10(2),
85e94.
*Huang, Y. M., Liang, T. H., Su, Y. N., & Chen, N. S. (2012). Empowering personalized learning with an interactive e-book learning system for elementary
school students. Educational Technology Research and Development, 60(4), 703e722.
Hubbard, P. (2005). A review of subject characteristics in call research. Computer Assisted Language Learning, 18(5), 351e368.
*Hwang, W. Y., Chen, H. S. L., Shadiev, R., Huang, R. Y. M., & Chen, C. Y. (2012). Improving English as a foreign language writing in elementary schools using
mobile devices in familiar situational contexts. Computer Assisted Language Learning, 27(5), 359e378.
Hwang, W. Y., Shadiev, R., & Huang, S. M. (2011). A study of a multimedia web annotation system and its effect on the EFL writing and speaking performance
of junior high school students. ReCALL, 23(2), 160e180.
Hwu, F. (2013). A review of the use of script-based tracking in call research for data sharing: applications providing meaning aids. Computer Assisted
Language Learning, 26(4), 350e370.
Jones, M. E., Antonenko, P. D., & Greenwood, C. M. (2012). The impact of collaborative and individualized student response system strategies on learner
motivation, metacognition, and knowledge transfer. Journal of Computer Assisted Learning, 28(5), 477e487.
Kukulska-Hulme, A. (2009). Will mobile learning change language learning? ReCALL, 21(2), 157e165.
Lan, Y. (2013). J, The effect of technology-supported co-sharing on L2 vocabulary strategy development,. Educational Technology & Society, 16(4), 1e16.
Lan, Y. J. (2015). Contextual EFL learning in a 3D virtual environment. Language Learning & Technology, 19(2), 16e31.
Lan, Y. J., Fang, S. Y., Legault, J., & Li, P. (2015). Second language acquisition of Mandarin Chinese vocabulary: context of learning effects. Educational
Technology Research & Development, 63(5), 671e690.
*Lan, Y. J., Sung, Y. T., & Chang, K. E. (2007). A mobile-device-supported peer-assisted learning system for collaborative early EFL reading. Language Learning
& Technology, 11(3), 130e151.
*Lan, Y. J., Sung, Y. T., & Chang, K. E. (2009). Let us read together: development and evaluation of a computer-assisted reciprocal early English reading
system. Computers & Education, 53(4), 1188e1198.
Lipsey, M. W., & Wilson, D. B. (2001). Practical meta-analysis. Thousand Oaks, CA: Sage Publications.
*Liu, T. Y. (2009). A context-aware ubiquitous learning environment for language listening and speaking. Journal of Computer Assisted Learning, 25(6),
515e527.
*Liu, T. Y., & Chu, Y. L. (2010). Using ubiquitous games in an English listening and speaking course: Impact on learning outcomes and motivation. Computers
& Education, 55(2), 630e643.
Liu, T. C., Lin, Y. C., Gao, Y., Yeh, S. C., & Kalyuga, S. (2015). Does the redundancy effect exist in electronic slideshow assisted lecturing? Computers & Education,
88, 303e314.
*Lowther, D. L., Ross, S. M., & Marrison, G. M. (2003). When each one has one: the inuences on teaching strategies and student achievement of using
laptops in the classroom. Educational Technology Research and Development, 51(3), 23e44.
*Lu, M. (2008). Effectiveness of vocabulary learning via mobile phone. Journal of Computer Assisted Learning, 24(6), 515e525.
Orwin, R. G. (1983). A fail-safe N for effect size in meta-analysis. Journal of Educational Statistics, 8(2), 157e159.
Rosenthal, R. (1979). The le drawer problem and tolerance for null results. Psychological Bulletin, 86(3), 638e641.
*Sandberg, J., Maris, M., & De Geus, K. (2011). Mobile English learning: an evidence-based study with fth graders. Computers & Education, 57(1), 1334e1347.
*Saran, M., Seferoglu, G., & Cagiltay, K. (2009). Mobile assisted language learning: English pronunciation at learners' ngertips. Eurasian Journal of
Educational Research, 34(1), 97e114.
Sharples, M., Taylor, J., & Vavoula, G. (2007). A theory of learning for the mobile age. In R. Andrews, & C. Haythornthwaite (Eds.), The sage handbook of
elearning research (pp. 221e247). London, UK: Sage Publications.
Shih, S. C., Kuo, B. C., & Liu, Y. L. (2012). Adaptively ubiquitous learning in campus math path. Educational Technology & Society, 15(2), 298e308.
Siozos, P., Palaigeorgiou, G., Triantafyllakos, G., & Despotakis, T. (2009). Computer- based testing using digital ink: Participatory design of a table PC based
assessment application for secondary education. Computers & Education, 52(4), 811e819.
Stockwell, G. (2013). Mobile-assisted language learning. In M. Thomas, H. Reinders, & M. Warschauer (Eds.), Contemporary computer-assisted language
learning (pp. 201e216). Sydney, Australia: Bloomsbury Academic.
Sung, Y. T., Chen, J. L., Cha, J. H., Tseng, H. C., Chang, T. H., & Changt, K. E. (2014). Constructing and validating readability models: the method of integrating
multilevel linguistic features with machine learning. Behavior Research Methods, 47(2), 340e354.
Sung, Y. T., & Lesgold, A. (2007). Software infrastructure for teachers: a missing link in integrating technology with instruction. Teachers College Record,
109(11), 2541e2575.
Sung, Y. T., Lin, W. C., Dyson, S. B., Chang, K. E., & Chen, Y. C. (2015). Leveling L2 texts through readability: combining multilevel linguistic features with the
CEFR. Modern Language Journal, 99(2), 371e391.
Tanner, M. W., & Landon, M. M. (2009). The effects of computer-assisted pronunciation readings on ESL learners use of pausing, stress, intonation, and
overall comprehensibility,. Language Learning & Technology, 13(3), 51e65.
*Thornton, P., & Houser, C. (2005). Using mobile phones in English education in Japan. Journal of Computer Assisted Learning, 21(3), 217e228.
Vygotsky, L. S. (1978). Mind in society: The development of the higher psychological process. Cambridge, MA: Harvard University Press.
84
Warschauer, M. (2004). Technological change and the future of CALL. In S. Fotos, & C. Brown (Eds.), New perspectives on CALL for second and foreign language
classrooms (pp. 15e25). Mahwah, NJ: Lawrence Erlbaum Associates.
Weston, M., & Bain, A. (2010). The end of techno-critique: The naked truth about 1:1 laptop initiatives and educational change, Journal of Technology.
Learning, and Assessment, 9(6), 1e26.
Wong, L. H., Chen, W., & Jan, M. (2012). How artefacts mediate small-group co-creation activities in a mobile-assisted seamless language learning environment? Journal of Computer Assisted Learning, 28(5), 411e424.
Wong, L. H., & Looi, C. K. (2011). What seams do we remove in mobile assisted seamless learning? A critical review of the literature. Computers & Education,
57(4), 2364e2381.
*Wood, C., Jackson, E., Hart, L., Plester, B., & Wilde, L. (2011). The effect of text messaging on 9- and 10-year-old children's reading, spelling and phonological
processing skills. Journal of Computer Assisted Learning, 27(1), 28e36.
*Zurita, G., & Nussbaum, M. (2004a). Computer supported collaborative learning using wirelessly interconnected handheld computers. Computers & Education, 42(3), 289e314.
*Zurita, G., & Nussbaum, M. (2004b). A constructivist mobile learning environment supported by a wireless handheld network. Journal of Computer Assisted
Learning, 20(4), 235e243.
Zurita, G., & Nussbaum, M. (2007). A conceptual framework based on activity theory for mobile CSCL. British Journal of Educational Technology, 38(2),
211e235.

Sung 2014 How Effective Are Mobile Devices For Language Learning PDF

Загружено:

Сведения о документе

Оригинальное название

Авторское право

Доступные форматы

Поделиться этим документом

Поделиться или встроить документ

Параметры публикации

Этот документ был вам полезен?

Это неприемлемый материал?

Авторское право:

Доступные форматы

Sung 2014 How Effective Are Mobile Devices For Language Learning PDF

Загружено:

Авторское право:

Доступные форматы

Educational Research Review 16 (2015) 68e84

Contents lists available at ScienceDirect

Educational Research Review

How effective are mobile devices for language learning? A

Y.-T. Sung et al. / Educational Research Review 16 (2015) 68e84

Y.-T. Sung et al. / Educational Research Review 16 (2015) 68e84

Y.-T. Sung et al. / Educational Research Review 16 (2015) 68e84

Y.-T. Sung et al. / Educational Research Review 16 (2015) 68e84

Y.-T. Sung et al. / Educational Research Review 16 (2015) 68e84

Y.-T. Sung et al. / Educational Research Review 16 (2015) 68e84

Number of studies (k)

Number of effect sizes

Proportion of studies (%)

Y.-T. Sung et al. / Educational Research Review 16 (2015) 68e84

Y.-T. Sung et al. / Educational Research Review 16 (2015) 68e84

3.2. The overall effect size for learning achievement

3.3. The effect size of learning achievement for moderator variables

Y.-T. Sung et al. / Educational Research Review 16 (2015) 68e84

Y.-T. Sung et al. / Educational Research Review 16 (2015) 68e84

3.3.2. Hardware used

Y.-T. Sung et al. / Educational Research Review 16 (2015) 68e84

Y.-T. Sung et al. / Educational Research Review 16 (2015) 68e84

Y.-T. Sung et al. / Educational Research Review 16 (2015) 68e84

Y.-T. Sung et al. / Educational Research Review 16 (2015) 68e84

References marked with an asterisk indicate studies included in the meta-analysis.

Y.-T. Sung et al. / Educational Research Review 16 (2015) 68e84

Engestrom, Y. (1987). Learning by expanding. Helsinki, Finland: Prienta-Konsultit Oy.

Y.-T. Sung et al. / Educational Research Review 16 (2015) 68e84

Вам также может понравиться