An online adaptive learning environment for ... - Wiley Online Library

6 downloads 15004 Views 2MB Size Report
Critical-thinking-infused language learning activities and online discussion .... created by a postdoctoral researcher with a PhD in Educational Technology.
British Journal of Educational Technology doi:10.1111/bjet.12080

Vol 45 No 4 2014

723–747

An online adaptive learning environment for critical-thinking-infused English literacy instruction Ya-Ting Carolyn Yang, Jeffrey Hugh Gamble, Yu-Wan Hung and Tzu-Yun Lin Ya-Ting C. Yang is an associate professor at the Institute of Education and Centre for Teacher Education, National Cheng Kung University, Taiwan. Her research interests are in the field of educational research and the role of information and communication technologies for improving teaching and learning processes. Jeffrey Gamble is Dr. Yang’s doctoral advisee. His research interests include innovative education and the integration of higher order thinking skills with language learning. Yu-Wan Hung is a post-doctoral researcher at National Cheng Kung University. Her research interests are in the field of second language acquisition and Computer-Assisted Language Learning (CALL). Tzu-Yun Lin is a junior high school language teacher in Taiwan. Address for correspondence: Prof Ya-Ting Carolyn Yang, Institute of Education, National Cheng Kung University, No. 1 University Rd., Institute of Education, Tainan, 701, Taiwan. Email: [email protected]

Abstract Critical thinking (CT) and English literacy are two essential 21st century competencies that are a priority for teaching and learning in an increasingly digital learning environment. Taking advantage of innovations in educational technology, this study empirically investigates the effectiveness of CT-infused adaptive English literacy instruction using a Moodle system. A one-group pretest–posttest design was employed to evaluate the effect of the treatment on students’ acquisition of CT skills (CTS) and English literacy. A total of 83 students enrolled in two sections of a general studies course at a large university in Taiwan participated in the semester-long experiment. Adaptive learning was achieved through the use of an online Moodle system for (1) online grouping (based on pretest English literacy scores), (2) delivery of specifically designed adaptive learning materials for each group and (3) provision of individualised feedback. CT-infused language activities based on social constructivist principles were designed for each level of adaptive instruction, whereas direct instruction for fostering CTS was provided in class and practiced or reflected upon in groups. Empirical results demonstrate that CT-enhanced adaptive English literacy instruction simultaneously improved students’ CTS and English literacy and that students’ online discussions developed towards higher levels of interaction. This paper illustrates an effective blended learning model for adaptive instruction and offers recommendations for designing CT-infused language learning activities that can successfully foster both CT and English literacy outcomes. Introduction Critical thinking Due to the rapid pace of technological development, information diffusion and digitisation of learning, critical thinking skills (CTS) have been trumpeted as one of the key competencies for success in contemporary society. According to Facione (1990), CTS involve reflective thinking leading to a decision of what to do or believe. CTS encapsulate the cognitive mechanisms that allow the critical thinker to engage in problem solving and decision making. Based on a systematic inquiry into the construct of CTS and based on the Delphi method, a set of six cognitive skills have been proposed: interpretation, analysis, evaluation, explanation, inference and selfregulation (Facione, 1990). Definitions of these elements of CTS are provided in Figure 1. © 2013 British Educational Research Association

724

British Journal of Educational Technology

Vol 45 No 4 2014

Practitioner Notes What is already known about this topic • Online learning environments can provide opportunities for social interaction that are beneficial to language learning. • The use of adaptive learning techniques has been shown to assist instructors in managing learners of different proficiency levels, which shows promise for development of effective technologies for language learning. • Asynchronous online Socratic questioning and debate activities are two effective strategies for fostering both language and thinking skills. What this paper adds • This paper designs, implements and empirically evaluates an online adaptive learning environment that incorporates proficiency-level grouping, adaptive instructional materials and adaptive feedback. • The experimental design incorporated critical-thinking-infused English reading and writing instruction through an online platform, leading to statistically significant improvements in critical thinking skills and English literacy scores for basic, intermediate and advanced English learners. • Qualitative analysis of online discussion posts reveals different levels of interaction and negotiation for Socratic questioning and debate activities but supports the appropriateness of these activities for learners of all proficiency levels. Implications for practice and/or policy • Adaptive learning through the use of student proficiency level grouping can effectively account for differences in learner ability when used in conjunction with online or blended learning. • Explicit critical thinking instruction coupled with practical application of these skills in online learning modules and collaborative online discussions is successful for students to learn and apply critical thinking skills effectively. • Critical-thinking-infused language learning activities and online discussion activities provide a promising avenue for improving students’ ability to write critically and negotiate meaning through asynchronous online interactions.

Concerning guidelines for CT-based instruction, studies have emphasised the importance of learner interaction, open-ended discussion, adequate time for reflection, and encouragement of learning transfer to other domains. As learner acquisition of intellectual resources, such as strategies, background knowledge and thinking patterns, is also necessary for CT, Bailin, Case, Coombs and Daniels (1999) suggested that instructors should assist learners in acquiring intellectual resources by designing engaging learning activities for CT and creating an environment conducive to CT that supports learners in critical dialogue. Specific strategies for CT instruction have also been proposed. Simpson and Courtney (2008) recommend incorporating activities such as Socratic questioning (SQ), debates, role play and small group activities, which may be suitable across disciplines. Among the different instructional methods, SQ is one of the most popular and powerful teaching tools (Hernandez, Kaplan & Schwartz, 2006). Because the level of thinking that occurs is influenced by the level of questions asked (King, 1995), SQ promotes reflection and discovery of solutions rather than requiring the acceptance of answer provided directly by the teacher (Neenan, 2009). Aukerman (2006) and © 2013 British Educational Research Association

Online adaptive CT-infused English literacy

725

Figure 1: Definition of elements of critical thinking skills (Facione, 1990, 2007)

Kovalik and Kovalik (2007) also encourage debates between students wherein teachers served as facilitators rather than evaluators, in order to assist freshmen with both writing and CTS. English language learning In tandem with an increasing global emphasis on CT, the importance of English literacy has inspired educational research internationally. Due to the convention of English as a universal language in a global economy, there is a growing awareness in higher education—particularly in Asia—that English literacy is synonymous with competitive advantage in the workplace of the future (Nunan, 2003). The language one learns has a profound influence on how one thinks. Vygotsky (1962) described language as a medium through which thoughts are organised, the abstract and symbolic nature of which allows the individual to draw inferences from and categorise elements of the external world. Based on Vygotsky’s work, research in second language acquisition (SLA) has demonstrated the effect of scaffolding, collaboration and social interaction on promoting SLA within the zone of proximal development (ZPD; Ohta, 1995; Swain & Lapkin, 2002). Consideration of the ZPD can assist in creating optimal learning conditions (Lantolf & Thorne, 2006) by guiding instructional design in the selection of task type and difficulty level and amount and type of support provided to learners based upon their level of experience. Within the adaptive learning framework implemented in this study, the design of learning materials based on learner proficiency level is emphasised. While language development involves a continuous process of constructing new knowledge of that language, Watson-Gegeo and Nielsen (2003) point out it is both a cognitive and social process, as cognition originates in social interaction (p. 156). In the same vein, Tarone (2010) also suggested that social settings affect learners’ sensitivity to language form and negotiation of meaning and thereby affect their SLA. A formal learning setting in this study was likely to © 2013 British Educational Research Association

726

British Journal of Educational Technology

Vol 45 No 4 2014

encourage students to work on language form and thereby increased the chance to acquire target-like language form. Moreover, participating in computer-mediated communication (CMC) provides students more opportunities to develop their communicative competence, particularly sociolinguistic aspects of communicative competence (Darhower, 2002). Integrating critical thinking with English literacy learning Because thinking skills are essential elements for literacy development, educators have argued for the importance of promoting CT in English as a Second or Foreign Language (ESL/EFL) instruction. The influence of CTS on literacy has been explored, with studies finding a strong correlation between CT and reading (Fahim, Bagherkazemi & Alemi, 2010), reporting that students who possess higher levels of CTS demonstrate a greater ability to express themselves clearly and integrate their viewpoints in writing, rather than just copying from external sources without elaboration (Alagozlu, 2007). Moreover, studies have shown that CTS assist learners in seeking alternatives, making inferences, performing analyses and solving problems, thereby understanding English more deeply and using the language more proficiently (Liaw, 2007). Given both theoretical and practical reasons for fostering CT, Ennis (1987) proposed four possible approaches to teaching CT: (1) a general approach (context-free), (2) an infusion approach (infused; context-based), (3) an immersion approach (immersed; context-based) and (4) a mixed approach (a combination of the general approach with either infusion or immersion). Both the infusion and immersion approach aim to encourage students to think critically using domainspecific content (such as English literacy); however, whereas the former provides direct instruction in the general principles of CT, the latter does not do so. Researchers have discovered that students who receive CTS instruction and then apply it to course content achieve better learning outcomes than those who receive no direct CT instruction (Abrami et al, 2008). Moreover, the infusion approach is necessary for transferring CTS across domains (Zohar, 1994). Adaptive instruction One widely accepted definition of adaptive instruction is “the use of alternative instructional strategies and school resources to provide learning experiences that meet the different needs of individual students” (Wang & Walberg, 1983, p. 603). Adaptive instruction can be characterised by three elements: (1) selection: diagnostic tests are administered before the instruction, then students are provided learning materials based on their assessed capabilities, (2) enrichment: alternative activities and materials are given to aid students in pursuing individual goals and doing cooperative learning and (3) acceleration: students learn to master the instructional content at a pace suited to their own abilities and interests, and periodic evaluations are provided to inform students on their progress (Wang & Lindvall, 1984). Adaptive instruction is indeed central to social constructivist pedagogy. In order for learners to develop cognitively, they must be provided with materials that provide an appropriate level of challenge, within their ZPD (Vygotsky, 1962). If the materials are too difficult or fast-paced, the learner may become frustrated, fail to acquire key contextual knowledge and subsequently lack the knowledge required to think critically about the subject. On the other hand, materials that are too easy or slow-paced will fail to challenge learners, provide little motivation for improvement and may foster a shallow approach to CT. From this perspective, an adaptive instructional approach is necessary. Adaptive instruction, according to the model developed by Wang and Lindvall (1984) is not synonymous with completely personalised instruction, such as that provided through intelligent tutoring systems. In fact, a core emphasis on collaboration and the social context of learning is critical for adaptive instruction, particularly when applied to language learning and the fostering of critical thinking. In light of the above discussion of social constructivist theory, an approach to © 2013 British Educational Research Association

Online adaptive CT-infused English literacy

727

adaptive instruction that accounts for and integrates social interaction and collaborative learning will be presented. Technology for adaptive learning Researchers have implemented adaptive instructional strategies to promote reading comprehension and fluency (eg, Reis, Eckert, McCoach, Jacobs & Coyne, 2008) and writing accuracy (Chandler, 2003; Jaehnig & Miller, 2007). Adaptive programs make use of a variety of techniques that have been found to be effective in different classroom settings, including, but not limited to, a computer-assisted instruction approach (Wang & Lindvall, 1984). Because course materials must be adapted for different learners simultaneously, our study adopts an online environment, taking advantage of the adaptive elements of selection, enrichment and acceleration for CT-infused English instruction. In selecting an appropriate environment for adaptive learning, we took into account limitations on time and computational resources as well as the limited programming skills of many instructors. By adopting an adaptive grouping strategy, based upon Wang and Lindvall’s (1984) element of “selection,” course materials could be designed and online access provided for three proficiency levels, allowing a pragmatic, yet truly adaptive, learning environment. Moodle, as a popular Learning Management System, was selected for this study because it provides a variety of features suitable for adaptive instruction and was designed based on social constructivist pedagogy, emphasising collaboration, inquiry and discovery-based learning (Brandl, 2005). Through the online learning environment, students can be divided into groups and provided with learning materials, such as articles, questions and follow-up activities, adapted to their level (Limongelli, Sciarrone & Vaste, 2011). Although second language is best learnt through social interaction with others (Ellis, 1999), students complete asynchronous learning tasks with their partners and receive individualised feedback from their instructor as well as systematic feedback from the Moodle system (Brandl, 2005; Oertig, 2010). Moreover, to foster a rich and varied educational experience, additional learning resources may be provided, allowing students to pursue individual goals and discuss the topic on Moodle’s asynchronous online discussion boards. Although some research has specifically adopted Moodle for language learning (eg, Wu, 2008), few empirical studies have evaluated the adaptive capabilities of Moodle for literacy learning, particular for EFL. Purpose of the study The purpose of this study was to examine the effectiveness of CT-infused adaptive English literacy instruction utilising a Moodle system. Specifically, the study attempted to answer the following three questions: 1. Will students who receive CT-infused online adaptive English instruction demonstrate improvement in CTS, including the subscales of analysis, inference, evaluation, induction and deduction? 2. Will students who receive different levels of CT-infused online adaptive English instruction (advanced, intermediate and basic levels) demonstrate improvement in English literacy? 3. How will students apply CTS to construct knowledge through different types of online collaborative activities? Method Research design and participants Participants in this experiment were 83 undergraduate students (44 males and 39 females) enrolled in a general studies course entitled “English language learning and thinking training” at a large university in Taiwan. Participants were full-time freshmen, all reporting over 10 years of English language learning, which include EFL courses, beginning in elementary school, and extra © 2013 British Educational Research Association

728

British Journal of Educational Technology

Vol 45 No 4 2014

hours of English learning in cram schools. In addition, the university’s entrance policy requires English proficiency scores of greater than 70% on the Joint College Entrance Examination. As with most Taiwanese freshmen, participants had previous computer and discussion board experience (such as Yahoo, MSN and Skype synchronous and asynchronous discussion board experience) and had access to the Internet from home or on campus. The research team included a native-speaking instructor and graduate student who assisted in designing the course, monitoring students’ learning and providing individualised feedback. A professor at the Institute of Education and a graduate teaching assistant developed Socratic dialogues modelling CT, implemented face-to-face workshops with students and facilitated students in applying CTS through asynchronous online discussions. The online environment was created by a postdoctoral researcher with a PhD in Educational Technology. As this study was partially funded by a federal grant, the postdoctoral researcher was an additional resource for this study. The regular student/staff ratios are 45:1 for a general studies course; if the number of students surpasses 70, a teaching assistant will be provided to assist the instructor. The study was implemented over a period of 10 weeks, including two weeks for pretests and posttests. Each week, students spent a total of 150 minutes learning English in a blended environment. At the beginning of the semester, the students provided informed consent and completed a series of pretests. Each week, after logging into the instructional website, students accessed content, activities and support materials via the user interface (Figure 2). Using a username and password, participants were provided with adaptive instructional learning tailored to their proficiency level. Each week included face-to-face instruction and in-class discussion, with completion of the online learning package at students’ own pace. On the main instructional page, access was provided to additional resources, including reading websites and vocabulary files, free for students to download. Variables Instructional treatment Our treatment was CT-infused online adaptive English instruction. CT concepts were integrated into English learning through direct instruction and CT-infused practice activities based on weekly topics, combining CTS instruction and adaptive English instruction (adaptive grouping, adaptive learning materials and individualised feedback). Individuals were grouped according to their pretest reading and writing scores. Table 1 provides participants information and scoring criteria for assignment to basic, intermediate and advanced groups. Classes were held in a technology-enhanced classroom with individual computer stations for each student, Internet access and an integrated network for audio-visual presentations. The classroom was booked for two 50-minute periods each week and included 20 minutes of face-toface instruction and 80 minutes of supervised online learning in the classroom. Although homework assignments were estimated by the instructional designer to require approximately 50 minutes of work, actual time spent completing the tasks may have varied among participants. Free computer access was provided for students who may not have had Internet access at home. Outcome variables The outcome variables evaluated in this study include critical thinking, English literacy and social construction of knowledge in online activities. CTS. CTS was measured by the Mandarin version of the California Critical Thinking Skills Test (CCTST) and Holistic Critical Thinking Scoring Rubric (HCTSR). The CCTST consists of 34 multiple-choice questions and requires 45 minutes to complete. The test results in a total score and five subscale scores for analysis, evaluation, inference, deductive reasoning and inductive reasoning (Facione, 1990). The CCTST was designed for college students, with a reliability of test KR-20 = 0.70. © 2013 British Educational Research Association

Online adaptive CT-infused English literacy

729

Figure 2: Moodle user interface, highlighting certain features (A: the course content; B: a bulletin board; C: a list of online users; D: Frequently Asked Questions; E: participant’s proficiency group assignment; F: user access statistics; G: contacts for teaching assistants; H: sample tests; and I: supplementary materials)

Table 1: Grouping of participants according to pretest TOEIC scores Reading

Basic Intermediate Advanced Total

Writing

n

criteria

n

criteria

5 39 39 83

295 points

15 34 34 83

130 points

In order to assess CTS in students’ writing, student essays were evaluated by the HCTSR, with scores ranging from a minimum of zero to a maximum of four. Two raters, a research assistant and the instructor, followed the instruction manual (Facione & Facione, 1994), which details scoring criteria and procedures. After a training session, discussion of the instrument and prac© 2013 British Educational Research Association

730

British Journal of Educational Technology

Vol 45 No 4 2014

Figure 3: Criteria for a full score of 4 on the HCTSR

tice with several samples, the raters separately evaluated a random selection of 20% of the students’ writing pretests. Interrater reliability, defined as the number of agreements divided by the total number of messages coded, achieved 90% agreement. For a full score of 4 on the HCTSR, an essay must consistently demonstrate all or almost all of the criteria shown in Figure 3 (Facione & Facione, 1994). English proficiency. The Test of English for International Communication (TOEIC) measures the English proficiency of non-native speakers with test questions based on real-life settings in an international business environment. The Reading component contains 100 multiple-choice reading comprehension questions, whereas the Writing component includes five questions involving writing a sentence based on a picture, two questions involving responding to a written request, and one opinion essay. The responses are scored by rubrics that evaluate factors such as grammar, vocabulary, organisation and relevance (Educational Testing Service, 2009). Social construction of knowledge in online discussions. The Interaction Analysis Model (IAM) was developed by Gunawardena, Lowe and Anderson (1997) for evaluating the construction of knowledge through social negotiation in CMC. They focused on assessing the quality of interactions in a CMC environment from both the “interactive” and “cognitive” dimensions. As shown in Figure 4, their model includes five phases of interactions and 21 operations (subcategories) to study critical discussions in electronic mail. This model was designed to assess the exchanges made among class members and how these exchanges moved from the lower to the higher phases of CT, with movement from one phase to the next showing that knowledge is constructed by the process of social negotiation. Two raters discussed the operationalisation of IAM to reach an agreement about the interpretation of the five phases and their subscales, then collaboratively analysed several posts from each proficiency level. After training, the two raters worked independently before collaboratively discussing any differences in their analytic results, finally reaching consensus. Procedures The instructional design for CT-infused instruction was based on an extensive review of the literature. Our approach involved (1) weekly face-to-face CTS workshops providing modelling of CT concepts, instructor-led class discussions, small group discussions and opportunities for reflection on previous CT-infused learning activities, (2) online adaptive CT-infused English literacy © 2013 British Educational Research Association

Online adaptive CT-infused English literacy

731

Figure 4: Interaction analysis model for examining social construction of knowledge (Gunawardena et al, 1997)

training and (3) collaborative online activities in which students used English reading and writing to practice CTS. Each week, the instructor introduced CTS concepts and led CTS warm-up activities and reflection activities on previously completed activities. After class, students completed the online weekly learning modules and collaborated on online projects that served as their © 2013 British Educational Research Association

732

British Journal of Educational Technology

Vol 45 No 4 2014

homework. Each week’s online activities were organised into a single module, including sections entitled “focus on the topic,” “focus on English skill (reading or writing),” “focus on TOEIC” and “focus on critical thinking” that offered a CT-infused English language activity. Every two weeks, students completed a reflective feedback form evaluating their progress in the course and offering comments for improving the course. An outline of the topics and goals for each week are provided in Table 2. Three characteristics of adaptive instructional strategies were adopted: An online adaptive grouping system, adaptive learning materials and adaptive individualised feedback. Adaptive grouping system For adaptive grouping, students were divided into three groups (advanced, intermediate and basic) for both reading and writing based on TOEIC scores. Proficiency level grouping was selected, rather than a completely individualised approach, in order to (1) take into account students’ learning experiences and elements of collaborative learning associated with social constructivist learning theory and (2) practically achieve differentiation of learning materials and activities through the design of course content for three distinct levels of proficiency. Proficiency levels established and published by the creators of the TOEIC test were used to divide students into levels, based upon pretest scores. Because students could be assigned to different levels for reading and writing, this also provides another level of adaptivity, as evidenced in Figure 5. Adaptive learning materials Students received individualised learning materials based on their English proficiency level. For example, Figure 6 illustrates individualised instructional activities for the Reading II module under the topic of travel. For basic level students, the research team provided readings such as tables, memos and advertisements that were shorter in length (no more than 300 words), suitable for a learner with basic English proficiency. Activities including matching and multiple-choice practices. For intermediate level students, longer (approximately 500 words) authentic reading materials such as letters, announcements and notices were utilised, and activities required synthesis and organisation of information. Advanced level students were provided with single passages, reports and double passages (using two sources) between 500 and 700 words, with activities that required understanding context and using judgement to answer more challenging questions. Figure 7 illustrates individualised instructional activities for the Writing II module under the topic of travel. In a task requiring participants to introduce a travel destination, we designed a step-by-step guide for scaffolding learners in writing a paragraph. Students in the basic level group were asked to write individual sentences using given words, whereas intermediate level participants wrote sentences using prepositions and conjunctions in order to construct coherent paragraphs. Adaptive learning activities were designed to encourage advanced level students to compose multiple sentences and organise their writing in the form of a letter or essay. Adaptive individualised feedback Adaptive individualised feedback was provided through both automatic systematic feedback and instructor/peer feedback. Students received simultaneous feedback from the online system, including immediate grading of online multiple-choice or cloze activities. For multiple-choice questions, for example, individualised responses were provided for incorrect answers, leading learners to return to the reading to find the correct response. Similarly, they received individual feedback from instructors and teaching assistants for writing assignments and asynchronous discussions. An example of individualised feedback is provided for both reading comprehension activities and for writing assignments (see Figure 8). © 2013 British Educational Research Association

1 2 3 4 5 6 7 8

Week

Business Life

Travel

Theme

Writing

Reading

Writing

Reading

Skill

Topic

Travel preferences Group tours Travel memories Taking a trip Job hunting Business communication Workplace life English in daily life

English instruction

If P, then Q. Logical syllogism Thought-provoking questions Socratic questioning Analog Assumptions Tautology Syntax Making a clear statement

Direct instruction

Deduction Deduction Analysis, Inference Evaluation Deduction, Induction Evaluation, Induction Analysis, Inference Evaluation

CTS objectives

Critical thinking instruction

Table 2: Topics and goals for each week

Debate

Socratic Questioning

CT-integrated language activity

Online adaptive CT-infused English literacy 733

© 2013 British Educational Research Association

734

British Journal of Educational Technology

Vol 45 No 4 2014

Figure 5: Adaptive grouping for two students based on reading and writing pretest scores

Figure 6: Snippets from reading activities illustrating differences among the three proficiency groups (A) Basic level reading activity: a shorter article with definition matching. (B) Intermediate level reading activity: a longer and more complex article with questions that address the importance and structure of key ideas. (C) Advanced level reading activity: a longer and more lexically complex reading with questions that ask students to select appropriate vocabulary from a list of similar words © 2013 British Educational Research Association

Online adaptive CT-infused English literacy

735

Figure 6: Continued

Asynchronous online discussions The course included two 4-week themes: travel and business life. For each theme-based instruction, reading was taught before writing to ensure sufficient input before learners began producing output in the form of written assignments. At the end of each module, an asynchronous online discussion allowed learners to apply CTS in an online collaborative environment. © 2013 British Educational Research Association

736

British Journal of Educational Technology

Vol 45 No 4 2014

Figure 7: Snippets from writing activities illustrating differences among the three proficiency groups (A) Basic level writing activity: students are asked to write a single sentence using provided words. Several such sentences were then combined to create a paragraph. (B) Intermediate level writing activity: students are guided in writing multiple sentences using specific grammatical elements. (C) Advanced level writing activity: students complete a chart comparing the relative advantages and disadvantages or three modes of travel, and later write a paragraph to summarise their decision

The two CT-integrated online discussions were SQ and asynchronous online debates. For the first theme, travel, students practiced use of SQ to probe for deeper meaning by discussing accommodation preferences. For the second theme, business, students engaged in an asynchronous online debate (see Figure 9 for an example), which encouraged probing for deeper meaning, investigating justifications and challenging the logical strength of a claim, position or line of reasoning. Data analyses Descriptive statistics were performed to describe the basic features of the data. For English literacy, TOEIC scores were analysed using a two-way mixed design analysis of covariance (ANCOVA), © 2013 British Educational Research Association

Online adaptive CT-infused English literacy

737

Figure 7: Continued

in which group level was a between-subjects factor and measurement occasion was a withinsubjects factor. Proficiency (advanced, intermediate and basic) serves as group level and the measurement occasion consisted of the pretest and posttest. Due to sample size limitations, the basic level was excluded from ANCOVA of TOEIC scores. HCTSR and CCTST scores, including the results for each subscale, were analysed using t-tests. Furthermore, in order to provide a comparison to a nonadaptive methodology, the quantitative results from this intervention were compared with another experiment conducted by the same lead researchers that did not include an online learning environment or provide adaptive grouping or adaptive learning materials. Content analysis was conducted to investigate the data from two asynchronous online discussion activities, namely an SQ discussion (Writing 2) and a debate (Writing 4), to evaluate the cognitive processes of participants from the three proficiency levels. The IAM, proposed by Gunawardena et al (1997), was selected for provided coding categories for interaction patterns because that model provides “a more holistic view of discussion flow and knowledge construction” from a social constructionist perspective (De Wever, Schellens, Valcke & Van Keer, 2006). Results CTS Mean scores and standard deviations for HCTSR and CCTST scores are presented in Table 3, along with the p-values from t-tests. Significant increases in CTS were observed for CCTST (p = 0.00) and HCTSR (p = 0.00). Moreover, the results of CCTST indicate a significant increase in overall CTS on each of the subscales: analysis, inference, evaluation, induction and deduction. These results can be compared to a nonadaptive methodology for infusing CT into EFL instruction, in which we found an insignificant difference between the pretest and posttest for scores on the CCTST (p = 0.3 overall and p > 0.05 on each of the subscales). English literacy: reading Mean scores and standard deviations for the three proficiency groups are reported in Table 4. Substantial improvements in reading proficiency appear to have resulted from the treatment. Excluding the basic group (due to a small sample size), overall English reading performance was analysed using a 2 × 2 mixed design ANCOVA. Including students’ additional reading hours as a covariate, the results show a significant main effect for both occasion and proficiency level, as well as a significant interaction effect for occasion × group (Table 5). The effect sizes for the main effects were large, whereas the interaction effect size was moderate (Cohen, 1988). © 2013 British Educational Research Association

738

British Journal of Educational Technology

Vol 45 No 4 2014

Figure 8: Feedback from peers, the instructor and the Moodle system (A) Feedback from peers and the instructor: students were able to raise and discuss specific questions with each other and the instructor via Forums on the Moodle system. (B) Feedback from the instructor: after students uploaded files for assignments or completed online discussion through the Moodle system, the instructor was able to provide individualised feedback on students’ writing. (C) Systematic feedback: students received immediate systematic feedback for both correct and incorrect responses

© 2013 British Educational Research Association

Online adaptive CT-infused English literacy

739

Figure 8: Continued

Figure 9: An online discussion using Moodle forums

© 2013 British Educational Research Association

740

British Journal of Educational Technology

Vol 45 No 4 2014

Table 3: Means, standard deviations, t-test significance, and effect size for CCTST and HCTSR Pretest Scale Analysis Inference Evaluation Induction Deduction CCTST Overall HCTSR Overall

Posttest

M

SD

M

SD

p

d

4.36 6.49 6.37 6.82 8.75 17.23 1.95

1.33 1.89 2.14 2.18 2.38 3.94 0.80

4.89 7.11 7.96 7.73 10.40 19.96 2.66

1.10 1.67 1.88 1.69 2.42 3.37 0.79

0.00* 0.00* 0.01* 0.00* 0.00* 0.00* 0.00*

0.43 0.35 0.80 0.47 0.69 0.74 0.89

*p < 0.05. Table 4: Means and standard deviations for overall TOEIC reading scores Pretest Level Basic Intermediate Advanced

Posttest

n

M

SD

M

SD

5 39 39

124.40 243.33 365.64

21.91 35.52 38.78

264.60 333.59 405.71

42.70 49.57 39.74

Table 5: Summary of group (2) × occasion (2) mixed design ANCOVA for reading scores (excluding Basic group) SV Covariate Between-subjects Group (A) Error (S/A) Within-subjects Occasion (B) Group × occasion (A × B) Occasion × error (B × S/A)

SS

df

5368.16 566176.63 373805.36 192371.27 102431.71 22481.99 21934.01 58015.71

1 76 1 75 77 1 1 75

MS 5368.16

F 2.09

p

Partial η2

0.15

0.03

373805.36 2564.95

145.74*

0.00

0.66

22481.99 21934.01 7734.54

29.06* 28.36*

0.00 0.00

0.28 0.27

*p < 0.05.

To evaluate the interaction effect, a 2 × 2 mixed design analysis of variance (ANOVA), with proficiency level as the independent variable, indicated that pretest mean scores (F = 210.94, p = 0.00) and posttest means scores (F = 50.26, p = 0.00) were significantly different. Given the significant main effect for occasion, ANOVA analysis was performed to demonstrate that reading TOEIC posttest mean scores were significantly higher than pretest mean scores for the Advanced group (F = 41.62, p = 0.00) and Intermediate group, (F = 187.80, p = 0.00). These results indicate that the instructional treatment was successful in improving students’ English reading proficiency. These results can be compared to a nonadaptive methodology for infusing CT into EFL instruction, in which we found an insignificant difference between the pretest and posttest for scores on the General English Proficiency Test (p = 0.71). In order to better evaluate the effectiveness of proficiency level grouping based upon TOEIC guidelines, we evaluated the differences between basic and “low-intermediate”; “lowintermediate” and “high-intermediate”; “high-intermediate” and “low-advanced”; and “high© 2013 British Educational Research Association

Online adaptive CT-infused English literacy

741

Table 6: Means and standard deviations for overall TOEIC writing scores Pretest Level Basic Intermediate Advanced

Posttest

n

M

SD

M

SD

15 34 34

62.67 103.82 157.94

21.70 12.12 13.43

128.00 140.00 167.06

32.17 26.20 19.47

Table 7: Summary of group (2) × occasion (2) mixed design ANCOVA for writing scores (excluding Basic group) SV Covariate Between-subjects Group (A) Error (S/A) Within-subjects Occasion (B) Group × occasion (A × B) Occasion × error (B × S/A)

SS

df

215.45 85141.56 55280.54 29861.02 25374.24 3278.88 6208.89 15886.47

1 66 1 65 67 1 1 65

MS 215.45

F 0.47

p

Partial η2

0.50

0.01

55280.54 459.40

120.33*

0.00

0.65

3278.88 6208.89 244.41

13.42* 25.40*

0.00 0.00

0.17 0.28

*p < 0.05.

advanced” and “low-advanced” learners. Due to the small sample sizes resulting from the selection of learners with high or low pretest scores from each group, gain scores were calculated and evaluated using t-tests and Mann–Whitney U-tests. In terms of reading, only one significant difference was found between “low-advanced” (n = 12) and “high-advanced” learners (n = 11), t = 3.10, p = 0.01; U = 20.00, p = 0.00. “High-advanced” learners, unsurprisingly, noted less improvement (12.50 points) than “low-advanced” learners (60.21 points), likely due to a ceiling effect resulting from “high-advanced” learners scoring near the maximum on the TOEIC pretest. English literacy: writing Mean scores and standard deviations for the three proficiency groups are reported in Table 6. Substantial improvements in writing proficiency appear to have resulted from the treatment. Excluding the basic group (due to low sample size), English writing proficiency was analysed using a 2 × 2 mixed design ANCOVA (Table 7). By including students’ additional writing hours as a covariate, the results indicate a significant main effect for occasion and proficiency level, as well as a significant interaction effect for occasion × group. The effect sizes for the main effects and interaction effect were all large. To evaluate the interaction effect, a 2 × 2 mixed design ANOVA with proficiency level as independent variable indicated significant differences between pretest (F = 304.01, p = 0.00) and posttest mean scores (F = 23.37, p = 0.00). Given the significant main effect for occasion, ANOVA was performed to determine that TOEIC writing posttest mean scores were significantly higher than the pretest mean scores for the Advanced group (F = 7.03, p = 0.01) and Intermediate group (F = 79.36, p = 0.00). These results indicate that the instructional treatment effectively improved students’ English writing proficiency. As with reading scores, differences between high- and low-scoring learners from different proficiency groups were evaluated. In terms of writing, significant differences were found between basic (n = 15) and “low-intermediate” learners (n = 13), t = 3.12, p = 0.00; U = 38.00, p = 0.00 as well as “high-intermediate” (n = 12) and “low-advanced” learners (n = 11), t = 2.47, © 2013 British Educational Research Association

742

British Journal of Educational Technology

Vol 45 No 4 2014

p = 0.03; U = 35.00, p = 0.04. Basic group learners showed a higher level of improvement (65.33 points) than “low-intermediate” learners (29.31 points). Both improvements were significant, but it seems as though the provision of a basic group was a beneficial decision, although only 15 students were classified as having a basic writing ability. Interestingly, there was no significant difference between “low-intermediate” and “high-intermediate” learners in terms of reading or writing, suggesting that the range of scores used to classify this group was effective in informing the course design for that level of proficiency. Nevertheless, a significantly greater improvement in writing scores was noted for “high-intermediate” writers (35.00 points) than “low-advanced” writers (17.27). Given the fact that “high-advanced” writers did not improve over the course of the intervention, we also suggest that this effect was due to advanced level learners approaching a maximum TOEIC writing score (as an increase of over 17 points brought them to within 30 points of a maximum score of 200). Interaction analysis of online discussions As shown in Table 8, a total of 226 and 236 messages were posted in the second online discussion (SQ) and fourth online discussion (Debate). The total number of units of analysis for each activity was slightly larger than the total number of messages as some messages involved two phases of interaction. For the SQ activity, 52, 20 and 28% of the analysis units were coded as Phases I–III respectively. There was only one example of Phase V interaction and none for Phase IV. In the Debate activity, 30, 29, 28, 2 and 11% of the units were coded in the respective phases. A chi-square test, χ2 (4, n = 497) = 45.63, p = 0.00, ϕ = 0.30, shows that during the Debate activity, the discussion moved towards higher phases (see Figure 10). Based on data from the three groups in online discussions, English proficiency appears to influence the social construction of knowledge, as measured by the IAM. The results of chi-square analysis, χ2 (4, n = 239) = 11.08, p = 0.03, ϕ = 0.22, show that during the SQ activity, students with higher levels of English proficiency tended to post messages in higher phases of interaction. A chi-square test, χ2 (8, n = 257) = 15.43, p = 0.05, ϕ = 0.17, shows a similar tendency for the debate activity. Discussion CTS CT-infused online adaptive English instruction clearly had a positive effect on students’ acquisition of CTS. This success might stem from both direct CTS instruction and from CT-infused adaptive language activities. Significant improvement in CCTST scores, a measure of discipline-neutral reasoning skills, illustrates the effectiveness of direct CTS instruction, whereas significant improvement in HCTSR, a measure of observable CTS in student writing, suggests that the CT-infused approach to language learning successfully instilled the CTS necessary to write critically. The success of direct CTS instruction has been noted in previous research (Marin & Halpern, 2011). In our case, we find that each subscale of the CCTST (analysis, inference, evaluation, induction and deduction) improved over the course of the intervention. As direct instruction and CTS-targeted practice activities were provided for each of these skills individually, students not only had a clear idea of what these skills involved but also were able to put them into practice. Furthermore, we utilised structured asynchronous online discussions with SQ in order to enhance students’ CTS, an approach effective in previous studies (Yang, Newby & Bill, 2005). Before online discussions, the instructor introduced and modelled the six categories of SQ. After direct instruction, students worked collaboratively to discuss how to apply SQ. After posting on the asynchronous online discussion board, the instructor offered feedback. The success of our © 2013 British Educational Research Association

Intermediate (n = 34)

Advanced (n = 34)

Total (n = 83)

Total number 31 97 98 226 of messages Phase I 24 (70.59%) 55 (53.92%) 45 (43.14%) 124 (51.67%) Phase II 7 (20.59%) 19 (18.63%) 21 (19.61%) 47 (19.58%) Phase III 3 (8.82%) 27 (26.47%) 38 (37.25%) 68 (28.33%) Phase IV 0 (0.00%) 0 (0.00%) 0 (0.00%) 0 (0.00%) Phase V 0 (0.00%) 1 (0.98%) 0 (0.00%) 1 (0.42%) Total 34 (100.00%) 102 (100.00%) 104 (100.00%) 240 (100.00%)

Basic (n = 15)

Socratic Questioning

34 (35.78%) 31 (25.41%) 78 (30.35%) 25 (26.32%) 32 (26.23%) 75 (29.18%) 22 (23.16%) 45 (36.89%) 72 (28.02%) 2 (2.11%) 2 (1.64%) 4 (1.56%) 12 (12.63%) 13 (9.84%) 28 (10.89%) 95 (100.00%) 123 (100.00%) 257 (100.00%)

236

Total (n = 83)

13 (33.33%) 18 (46.15%) 5 (12.83%) 0 (0.00%) 3 (7.69%) 39 (100.00%)

111

Advanced (n = 34)

88

Intermediate (n = 34)

37

Basic (n = 15)

Debate

Table 8: Results of synchronous online discussions by interaction categories for students of three proficiency levels for two types of online collaborative activities

Online adaptive CT-infused English literacy 743

© 2013 British Educational Research Association

744

British Journal of Educational Technology

Vol 45 No 4 2014

Figure 10: Percentage of online discussion messages by interaction category

approach confirms the effectiveness of SQ through generic questioning prompts (King, 1995) and the positive effects of scaffolding CT by offering questions and directing feedback. CT-integrated reading and writing activities were also responsible for improvement. For example, the CTS topic of Reading 3 was logical syllogism, a type of deductive reasoning. For this activity, students needed to comprehend the entire article, deconstruct arguments, make inferences and draw conclusions about the content, leading to the development of CTS. Writing was also a process for generating ideas, structuring those ideas, and drafting and revising a paper. Process writing provided the opportunity for students to think through an issue and state their ideas clearly, leading to improved CT (Marzano & Pickering, 1997). English literacy CT-infused online adaptive English instruction had a positive and significant impact on students’ English literacy. The use of an adaptive online learning system was integral to the development of English reading and writing proficiency. Because the courses were presented on the Moodle system, students were allowed to learn at their own pace. In terms of adaptive instruction, possible reasons for improvement lie in adaptive grouping, adaptive learning materials and adaptive individualised feedback. With accurate measurement of students’ English proficiency, online adaptive grouping helped students learn at a level suited to their individual abilities, with the provision of different learning materials matching their individual learning needs, reducing cognitive overload. Moreover, adaptive individualised feedback enabled students to avoid errors and acquire correct patterns, resulting in better understanding of the content and the ability to respond correctly in the future (Jaehnig & Miller, 2007). Both instructor/peer feedback and automatic, system-generated feedback were provided. In the writing unit, individual face-to-face feedback was provided and teaching assistants crossed out errors and provided the correct form or structure (Ferris & Roberts, 2001), allowing students to correct their errors, with subsequent writing becoming more accurate (Chandler, 2003). As with previous research showing learning gains from informative feedback (Hwa, Fook, Atan, Majid & Luan, 2007), the online adaptive system was programmed to provide multiplechoice review questions for each unit, with simultaneous feedback from the system. CT-integrated asynchronous online discussions were useful in increasing students’ engagement and enhancing their understanding of issues (Persell, 2004). Students used supplementary learning resources to conduct research before discussing their findings online, thereby enhancing active engagement and extension of knowledge, thinking and perspective (Reis et al, 2008). For © 2013 British Educational Research Association

Online adaptive CT-infused English literacy

745

example, the asynchronous online discussion topic of writing 4 was a debate. Students considered both advantages and disadvantages, before composing a concluding statement, thereby deepening their understanding of issues while increasing their use of English. Interaction analysis of online discussions Considering differences in English proficiency, both Basic and Advanced groups posted more messages for the Debate activity compared wth the SQ activity. All three groups posted longer, clearer messages with supporting evidence in arguing their positions in the Debate activity, as compared with the SQ activity. Increased higher phase interaction for the Debate activity may be due to purposive scaffolding and the instructional design. Students grew familiar with online posting and comfortable with engaging in a critical or “argumentative” manner through practicing SQ in the first online discussion. Despite the increased interactions in Phases II and III during the Debate activity, the negotiation of meaning of specific terms (Phase IIIA) appeared to be less demanding compared with other sub-scales, possibly because the Debate topic did not involve the twisted meanings of specific terms. Indeed, the extended processing time and the easy access to consult other resources in text-based CMC might also reduce the demands of negotiating or clarifying the meaning of unfamiliar terms. It is worth noting that although interactions of Phases IV and V increased in the Debate activity as well, the results in this study appeared to partially corroborate those of Gunawardena et al (1997) who found the debate format made it difficult for participants to move towards the last two phases, particularly the phase of synthesis (Phase IV). Participants in their study reflected that debating was different from discussing and thus they found it difficult to propose synthesis. Although the number of Phase IV interactions increased in the Debate activity, as compared with the SQ activity, it was the least observed of the five phases for both activities. Unsurprisingly, students with higher levels of proficiency performed better in online discussions, as the activities involved English comprehension and expression. Because all groups posted messages demonstrating a range of interaction and tended towards higher phase interaction and longer, more complicated messages in the Debate activity, we suggest that both Debate and SQ activities are appropriate for learners of all levels and that CT-related interaction through online discussions is a suitable approach for ESL/EFL. Conclusions We found that students receiving CT-infused online adaptive English instruction demonstrated greater improvement in CTS, as students learned CTS concepts through both direct instruction and practice during reading and writing activities. Additional reasons for this improvement include the facilitation of the adaptive grouping system, adaptive learning materials and adaptive individualised feedback. Based on the analysis of online discussions, students appeared to benefit from opportunities for social interaction with peers to improve their CTS and English abilities. As many studies suggest that cultural attributes of Confucian societies might explain why Asian learners of English tend to be silent in EFL classrooms, Cheng (2000) addressed misinterpretations of Confucian teachings and found that differences in teaching methods and English language inadequacies were more likely to be the causes for the phenomenon. Along this line of argument, the present study also found that the oral production among Taiwanese students could be successfully encouraged by designing and modelling activities such as debating and through the use of CMC, as suggested by authors such as Kern (1995). Differences between Debate and SQ activities were also found and may result from either maturation or instructional design elements, including scaffolding. Because online discussions might improve communicative competence better than other kinds of online writing activities, the relative advantages of SQ and Debate activities can be further evaluated in future research. © 2013 British Educational Research Association

746

British Journal of Educational Technology

Vol 45 No 4 2014

No comparison group was included in the design of this study due to limitations on available courses and reluctance by the university to offer a nonadaptive version of the course. Therefore, future research would benefit from a comparison of adaptive and nonadaptive conditions, as well as the evaluation of a fully individualised approach to CT-infused instruction. Future research could also integrate elements of item-response theory that could allow learners to progress from one proficiency level to another as the course progresses. This may require greater automation and continual assessment, which introduces further complexity to the design, but may provide a potential alternative for adaptive online instruction to that which is presented in this paper. As this study has successfully developed a CT-infused online adaptive English instruction system, our findings have implications both for instructional designers and classroom teachers. Instructional designers can develop CT instructional models and adaptive online courses for a variety of academic courses. Instructors can enhance learners’ understanding of CT concepts and the factors that influence students through CT-infused instruction. Acknowledgements The funding for this research was provided by the National Science Council of Taiwan, under grants NSC 96-2520-S-006-003-MY3 and NSC 99-2628-S-006-001-MY3. References Abrami, P. C., Bernard, R. M., Borokhovski, E., Wade, A., Tamin, R. & Surkes, M. (2008). Instructional interventions affecting critical thinking skills and dispositions: a stage one meta-analysis. Review of Educational Research, 78, 4, 1102–1134. Alagozlu, N. (2007). Critical thinking and voice in EFL writing. Asian EFL Journal, 9, 3, 118–136. Aukerman, M. (2006). Who’s afraid of the big “bad answer”? Educational Leadership, 64, 2, 37–41. Bailin, S., Case, R., Coombs, J. R. & Daniels, L. B. (1999). Conceptualizing critical thinking. Journal of Curriculum Studies, 31, 3, 285–302. Brandl, K. (2005). Are you ready to “Moodle”? Language Learning & Technology, 9, 2, 16–23. Chandler, J. (2003). The efficacy of various kinds of error feedback for improvement in the accuracy and fluency of L2 student writing. Journal of Second Language Writing, 12, 3, 267–296. Cheng, X. (2000). Asian students’ reticence revisited. System, 28, 435–446. Cohen, J. (1988). Statistical power analysis for the behavioral sciences (2nd ed.). Hillsdale, NJ: L. Erlbaum Associates. Darhower, M. (2002). Interactional features of synchronous computer-mediated communication in the intermediate L2 classroom: a sociocultural case study. CALICO, 19, 2, 249–276. De Wever, D., Schellens, T., Valcke, M. & Van Keer, H. (2006). Content analysis schemes to analyze transcripts of online asynchronous discussion groups: a review. Computer and Education, 46, 1, 6–28. Educational Testing Service (2009). Examinee handbook: speaking and writing. Princeton, NJ: Educational Testing Service. Ellis, R. (1999). Learning a second language through interaction. Philadelphia: John Benjamins. Ennis, R. H. (1987). A taxonomy of critical thinking dispositions and abilities. In J. B. Baron & R. J. Sternberg (Eds), Teaching thinking skills: theory and practice (pp. 9–26). New York: Freeman. Facione, P. A. (1990). Critical thinking: a statement of expert consensus for purposes of educational assessment and instruction. Millbrae, CA: The California Academic Press. Facione, P. A. (2007). Critical thinking: what it is and why it counts. Millbrae, CA: California Academic Press. Facione, P. A. & Facione, N. C. (1994). Holistic critical thinking scoring rubric. Millbrae, CA: The California Academic Press. Fahim, M., Bagherkazemi, M. & Alemi, M. (2010). The relationship between test takers’ critical thinking ability and their performance on the reading section of TOEFL. Journal of Language Teaching and Research, 1, 6, 830–837. Ferris, D. & Roberts, B. (2001). Error feedback in L2 writing classes: how explicit does it need to be? Journal of Second Language Writing, 10, 3, 161–184. Gunawardena, C., Lowe, C. & Anderson, T. (1997). Analysis of a global online debate and the development of an interaction analysis model for examining social construction of knowledge in computer conferencing. Journal of Technical Writing and Communication, 17, 4, 397–431. Hernandez, A., Kaplan, M. A. & Schwartz, R. (2006). For the sake of argument. Educational Leadership, 64, 2, 48–52. © 2013 British Educational Research Association

Online adaptive CT-infused English literacy

747

Hwa, Y. L., Fook, F. S., Atan, H., Majid, O. & Luan, W. S. (2007). A study on different types of feedback in a language multimedia courseware. International Journal of Instructional Technology & Distance Education, 4, 4, 23–34. Jaehnig, W. & Miller, M. L. (2007). Feedback types in programmed instruction: a systematic review. The Psychological Record, 57, 4, 219–232. Kern, R. G. (1995). Restructuring classroom interaction with networked computers: effects on quantity and characteristics of language production. The Modern Language Journal, 79, 4, 457–476. King, A. (1995). Designing the instructional process to enhance critical thinking across the curriculum: inquiring minds really do want to know: using questioning to teach critical thinking. Teaching of Psychology, 22, 1, 13–17. Kovalik, D. L. & Kovalik, L. M. (2007). Language simulations: the blending space for writing and critical thinking. Simulation and Gaming, 38, 3, 310–322. Lantolf, J. P. & Thorne, S. L. (2006). Sociocultural theory and second language learning. In B. van Patten & J. Williams (Eds), Explaining second language acquisition (pp. 197–221). Mahwah, NJ: Erlbaum. Liaw, M. L. (2007). Content-based reading and writing for critical thinking skills in an EFL context. English Teaching and Learning, 31, 2, 45–87. Limongelli, C., Sciarrone, F. & Vaste, G. (2011). Personalized e-learning in Moodle: the Moodle_LS system. Journal of E-Learning and Knowledge Society, 7, 1, 49–58. Marin, L. M. & Halpern, D. F. (2011). Pedagogy for developing critical thinking in adolescents: explicit instruction produces greatest gains. Thinking Skills and Creativity, 6, 1, 1–13. Marzano, R. J. & Pickering, D. J. (1997). Dimensions of learning. Alexandria, VA: Association for Supervision and Curriculum Development. Neenan, M. (2009). Using socratic questioning in coaching. Journal of Rational Emotive & Cognitive Behavior Therapy, 27, 249–264. Nunan, D. (2003). The impact of English as a global language on educational policies and practices in the Asia-Pacific region. TESOL Quarterly, 37, 4, 589–613. Oertig, M. (2010). Debriefing in Moodle: written feedback on trust and knowledge sharing in a social dilemma game. Simulation & Gaming, 41, 3, 374–389. Ohta, A. (1995). Applying socio-cultural theory to an analysis of learner discourse: learner-learner collaborative interaction in the zone of proximal development. Issues in Applied Linguistics, 6, 2, 93–122. Persell, C. H. (2004). Using focused web-based discussion to enhance student engagement and deep understanding. Teaching Sociology, 32, 1, 61–78. Reis, S. M., Eckert, R. D., McCoach, D. B., Jacobs, J. K. & Coyne, M. (2008). Using enrichment reading practices to increase reading fluency, comprehension, and attitudes. Journal of Educational Research, 101, 5, 299–314. Simpson, E. & Courtney, M. (2008). Implementation and evaluation of critical thinking strategies to enhance critical thinking skills in Middle Eastern nurses. International Journal of Nursing Practice, 14, 6, 449–454. Swain, M. & Lapkin, S. (2002). Talking it through: two French immersion learners’ response to reformulation. International Journal of Educational Research, 37, 3–4, 285–304. Tarone, E. (2010). Social context and cognition in SLA: a variationist perspective. In R. Batstone (Ed.), Sociocognitive perspectives on language use and language learning (pp. 54–72). Oxford: Oxford University Press. Vygotsky, L. S. (1962). Thought and language. Cambridge, MA: MIT Press. Wang, M. & Lindvall, C. M. (1984). Individual differences and school learning environments. Review of Research in Education, 11, 1, 161–225. Wang, M. C. & Walberg, H. J. (1983). Adaptive instruction and classroom time. American Educational Research Journal, 20, 4, 601–626. Watson-Gegeo, K. A. & Nielsen, S. (2003). Language socialization in SLA. In M. H. Long & C. J. Doughty (Eds), The handbook of second language acquisition (pp. 155–177). Oxford, UK: Blackwell. Wu, W. S. (2008). The application of Moodle on an EFL collegiate writing environment. Journal of Education and Foreign Languages and Literature, 7, 45–56. Yang, Y.-T. C., Newby, T. J. & Bill, R. L. (2005). Using Socratic questioning to promote critical thinking skills through asynchronous discussion forums in distance learning environments. American Journal of Distance Education, 19, 3, 163–181. Zohar, A. (1994). Teaching a thinking strategy: transfer across domains and self learning versus class-like setting. Applied Cognitive Psychology, 8, 6, 549–563.

© 2013 British Educational Research Association