1 Research Context

The empirical case discussed in this book was drawn from a longitudinal study, in which GroupScribbles (GS) was used in some classes of a Future School in Singapore (Looi et al. 2009, 2010a, b; Wen et al. 2010, 2011, 2012). Future schools in Singapore are schools involved in the FutureSchools@Singapore project initiated by the Ministry of Education, which aims at harnessing ICT for engaged learning and keeping the education system and programs in Singapore relevant to preparing students for the future. Future schools are expected to develop new ICT solutions to support innovative teaching approaches and to consolidate these innovations and experiences and extend them to the wider education community.

As a Future School, the school in this study leverages ICT to better engage students. Following the ICT-enabled one-to-one learning approach—it refers to providing every teacher and student with a portable laptop, notebook, or tablet PC for continuous use both in the classroom and at home (Bocconi et al. 2012)—each student is equipped with a laptop (Apple MacBook), which offers a rich variety of resources and learning platforms. Teachers in all subjects are required to maximize the use of various computer technologies so as to optimize class teaching and learning. Because of the frequent use of computer technologies in school, teachers and students are technologically savvy. The purpose of the GS project was not just to introduce the GS technology itself but more to instill the socio-constructivist pedagogy enabled by GS into the classroom.

Within this research endeavor on collaborative learning in GS classrooms, I positioned my work on studying collaborative language learning in a networked classroom and focused on analyzing multimedia-based social interactions by which learning took place. Singapore is known as a multiethnic and multilingual country, and its specific language environment requires and provides a valuable support for L2 learning research. In Singapore, an “English-knowing” bilingualism policy dictates that English is the de facto national language, which is used not only in official workplaces but also as the medium of instruction in the school system (Pakir 1991). Chinese, Malay, and Tamil languages are defined in specific Singaporean terms as “Mother Tongue Language (MTLs)” that are mainly used for communicating and maintaining the cultural heritage. The Ministry of Education in Singapore defines MTL by student ethnicity rather than by the language used at home or the first language learned (Tan 2006). Chinese/Mandarin language is the MTL for all ethnic Chinese. Learning the Chinese language is a key feature of the school system (MOE 2004) as the ethnic Chinese population (2.7 million) accounts for 75% of the total population of Singapore.

Since the 1980s, there has been a gradual but significant shift from using Chinese predominantly in informal domains to using English in the Chinese community. A survey administered by the MOE (2004) showed that the number of primary 1 Chinese student who spoke predominantly English at home had risen from 36% in 1994 to 50% in 2004, and this trend was expected to continue. It has been widely reported in public media and academic studies that the Chinese language proficiency of the new generation of Chinese Singaporeans is declining, and the recession was most obvious in reading and writing. Besides language competency, students are observed lacking motivation for studying the Chinese language as well (Silver 2005; Wee 2003). Facing this problematic situation, the MOE has made a number of changes to the Chinese language policy in the past decade. The most recent recommendation on the MTL reform was made in the review report of the MTL at the beginning of 2011. The review committee particularly emphasizes the importance of interactivity, collaboration, and critical thinking in language learning.

This case study was conducted based on the assumption that both the teacher and students had been sufficiently enculturated into the GS-based networked classroom environment. Before data collection, technical training during which the teacher and students learned how to use GS was conducted in July 2010. The teacher and all the students had developed a good understanding of the features of GS and competence in using GS technology (more examples can be found in Chap. 4 of this book). Yet, though the teacher knew how to work with GS, she still had a problem integrating GS into the concrete activity design and enactment. It is believed that “if a teacher could not understand the essence of design behind activity design, s/he could only emulate a similar process of activity, but the efficacy of design might not be fully unlocked” (Wen et al. 2012, p. 115). Hence, alongside technical trainings, a series of professional development sessions (1 h per week, about 5 weeks) were carried out to ensure that the teacher had a good understanding of GS-based language learning design before the actual enactment of GS-based lessons. In the professional development sessions, the Rapid Collaborative Knowledge Improvement (RCKI) concept and its related nine principles were introduced (see Sect. 4.1 for more details about the principles).

In addition to technical training and teaching professional development, activities using paper scribbles (real Post-it notes) were implemented in the class to acquaint the teacher and students with GS activities in the first stage of enculturation. This phase was critical to GS enculturation as the relevant protocols and social etiquettes were introduced and applied (Ng et al. 2008). Once students were familiar with the technical function of GS and the teacher had a rough knowledge about how to utilize GS affordances to design class activities, a series of GS Chinese lessons were designed and implemented. The focus at this stage was on improving students’ collaborative learning awareness (e.g., to be aware of the ideas, knowledge, and activities of the others) in the Chinese language class and empowering the design and enactment of GS-based Chinese lessons by the teacher. Designing and enacting GS-based lessons by integrating the collaborative tool with pedagogy and content knowledge is always a big challenge for teachers. During that period of time, GS lessons were still codesigned by the teacher and me. Several cycles of GS activity design, implementation, evaluation, and redesign with the guidance of the principles were carried out. From February 2011 to May 2011, GS was routinely used in Chinese language reading comprehension lessons. A total of seven GS Chinese lessons around the theme of reading comprehension was conducted. After conducting a lesson, the teacher and I reflected on the activity design together to identify both the “good” and the “bad” and to see how to make improvements in the following lessons. Through this process, we both developed a better understanding of the GS affordances, and it helped us improve the pedagogical innovation, consequently and iteratively.

From August 2011 to October 2011, altogether four GS lessons (60 min per lesson) on composition were designed and implemented to achieve learning objectives prescribed in the school’s Chinese language syllabus. All the collaborative writing tasks were codesigned by the teacher and me.

2 Participants

The subjects of this study were from a class of the future school in secondary level 2 (grade 8). Their age ranged between 14 and 16. There were five Higher Chinese classes and four Normal Chinese classes at secondary grade 2 in the school. In most secondary schools in Singapore, students are channeled into a Higher Chinese or Normal Chinese class based on their language proficiency. Compared with the Normal Chinese class, the Higher Chinese class requires a high level of linguistic proficiency and cultural knowledge (MOE 2004). The class (N = 19, 6 female students and 13 male students) involved in this study belonged to a Higher Chinese class. At the beginning of the GS project, 19 students of the class were separated into 5 groups in terms of language proficiency. They formed a comparative high-ability group, a medium-ability group, a comparative low-ability group, as well as two mixed-ability groups, respectively. In order to build and sustain group culture, the group compositions remained unchanged from the stage of technical training until the end of the implementation of this study. Before this study, they had been familiar with the technical functions of GS and even the culture of a GS classroom.

Table 6.1 shows the general information of each group (all the names are pseudonyms). The student language proficiency level was evaluated by the school final year examination scores of Chinese language. The examination was administered before GS-based collaborative writing lessons. The diagnostic test assessed a broad spectrum of language skills—including writing, reading, and oral speaking—and the test scores could reflect students’ general language proficiency. The total score of the test was 170. As shown in Table 6.1, students in the class varied greatly in Chinese language proficiency (M = 118.29, SD = 16.28). The students’ mean score was 118. Based on this, all the students were classified into levels: High (H, >118 + 5), Medium (M, 118 − 5 < 118 < 118 + 5), and Low (L, < 118 − 5).

Table 6.1 Participants and group composition
Full size table

As indicated in Table 6.1, Group 2 (HHHM) and Group 5 (HHMM) belonged to high language proficiency groups. Students in these two groups, on average, had higher test scores than those in other groups. Group 4 was a low language proficiency group, consisting of only one student with medium language proficiency and three students with low language proficiency. According to the average score, both Group 1 and Group 3 belong to medium language proficiency group. Whereas, all three students’ scores in Group 3 were around the average, in Group 1 students of high-, medium-, and low language proficiency were mixed.

The Chinese language teacher of the class was Chin (pseudonym). She was in her early 40 s. She had studied the Chinese language abroad (4 years in Taiwan and 2 years in Peking University), and has experiences in teaching and school management at the primary and secondary school levels for almost 20 years. She once worked in the Ministry of Education in Singapore, as a curriculum planning officer and national chief trainer for primary school education. With such training and working experience, she was quite creative and effective in designing learning activities. Furthermore, compared with other local Chinese language teachers, especially those without an overseas educational background, she was more aware of the necessity and importance for Ethnic Chinese students to master the Chinese language. She believed that the Chinese language plays an essential role in creating cultural identity, preserving cultural heritage as well as cultivating students’ thinking abilities. She was willing to try new teaching approaches to arouse students’ interests in Chinese language learning. She held that every student has potential and what teachers need to do is to assist students to realize their potentials.

3 Classroom Setting

The seating arrangement in the GS classroom was not typical of most secondary schools in Singapore. Usually, single desks and chairs were arranged in rows facing a whiteboard. In the GS class, the students sat in groups of three or four at their tables, looking toward the teacher, who was standing at the front of the class. Each student in the classroom had a personal computer with a GS client software installed. In this way, students in the same group were able to do both face-to-face and GS-based interactions. Besides this, an interactive whiteboard was set up in front of the classroom to help teachers visualize and monitor every single group’s GS-based interaction process. The GS classroom layout can be seen in Fig. 6.1.

Fig. 6.1
figure 1

GS classroom arrangement

Full size image

4 Task Design

The GS-based collaborative writing lesson reported in this study was the last GS lesson enacted. The primary goal of this study was not to investigate how to foster a community culture for knowledge improvement. Instead, this study was focused on examining how L2 learners learn together in a networked classroom environment where face-to-face and online interactional modes with distinguished characteristics are combined. Therefore, the last GS lesson was selected on the assumption that the teacher and students had developed familiarity with GS-based collaborative activities.

Stahl (2005) stated that “in order to observe effective collaboration in an authentic educational setting … we must bring together groups of students who will work together well, both by getting along with and understanding each other and by contributing a healthy mix of different skills” (p. 87). Following this thought, the data collected in this study was from a case where both the teacher and students were already familiar with relevant GS class protocols, and they were used to group work and were satisfied with the lesson. The teacher mentioned in the post-lesson interview that “today’s lesson is completed smoothly as we planned … I felt much better in this lesson than in the first few GS lessons … I have already started to like collaborative learning.” According to students’ individual feedbacks on the lesson, it was known that all the students in the class were positive about the lesson. Some students expressed:

I like the lesson. I like to write together in a group, because we can discuss with each other and provide ideas and suggestions for each other. My own idea might not be the best, but we can keep improving these ideas in group work.

In today’s lesson, we worked well … I prefer to write collaboratively than individually. Students own different ideas toward the same topic, thus everyone will be involved in intensive discussion on how to write an article. During this process, we can learn from each other, and in this way, our abilities of critical thinking and collaborative learning get opportunities to be improved.

In the case of collaborative writing practice of a planning activity, students were required to make decisions on the ideas they intended to express and then formulate the language structure to express these ideas as they produced a text together. Students not only generated, clustered, and ordered ideas, they also considered both hierarchical and structural relations among the ideas, to make sure the small group’s outline was internally consistent. In this kind of collaborative writing task, writers plan, produce text, and revise together, and their knowledge of writing is made explicit in ways that are neither possible nor traceable when individuals compose alone (Milian 2005; Ndlovu and Geva 2008). According to Chai (2006), writing performance is highly relevant to the planning activity, no matter what the language proficiency the learners are. Students can also benefit from articulating their ideas as they organize the task, plan the content, and air their viewpoints about the audience, purpose, and form of their text. In this way, they jointly deepen understandings about linguistic knowledge as well as writing content and strategies.

The main learning objective of the lesson was to help students understand that an argumentative essay can be conceptualized and composed from exploring the contributing factors of a phenomenon, followed by finding its impacts and providing solutions if needed. The lesson task included five main phases (Table 6.2).

Table 6.2 Main phases of the collaborative planning task
Full size table

In the lesson, students were encouraged to generate their group ideas via collecting individual wisdom within the group and borrowing ideas from other groups. Respecting and encouraging cognitive diversity, the activity began with the creation and presentation of different ideas. In the subsequent phases, the synergy of ideas was sought. The final phase of idea convergence and consensus seeking thus could lead to knowledge convergence (Ficher and Mandl 2005) and advancement. The task was designed with an expectation of more mutual coordinating activities in the dialogue, which would result in a more consistent shared knowledge structure and in a better mutual solution.

The topic for the writing was “ ” (Is plastic surgery ethically right?). A visual organizer (Fig. 6.2) was uploaded as the background of the GS group board in order to provide tangible scaffoldings for students to follow the teacher’s instruction and to allow them to pay attention to the three elements (cause, consequence, and solution) necessary in writing an argumentative essay.

Fig. 6.2
figure 2

A graphic organizer for the planning task

Full size image

5 Data Collection

I included all five small groups’ data in this study, considering different groups still evolve alternative approaches to appropriating the technology even though in the same classroom setting (Dwyer and Suthers 2006; Larusson and Alterman 2007; Overdijk and van Diggelen 2008). The aim of this type of study was not to make a statistical comparison of the results but to check certain conceptually established dimensions in various contexts (Yin 2003). In this study, I sought to know if there was evidence that any productive group understanding development took place as a result of the specific medium transition I observed. I also wanted to explain why partition in the same classroom has apparently led to different learning effects among students in different groups. Instead of analyzing a case longitudinally, I chose to do a case study by focusing on what happened holistically at some point in time, so as to provide a more complete and truthful analysis on which situation a representational tool can be used for facilitating productive interactions in an authentic context.

The main data sources for this study were the video data of face-to-face and GS-based interactions within groups, between groups, and between the teacher and students, as well as the teacher’s instruction. Besides this, I gathered additional information to provide the analysis with more contextual information in terms of my classroom observation notes, the teacher’s interview, and students’ self-reports after the lesson.

Eisner (1998) stated, “We need to listen to what people have to say about their activities, their feelings” (p. 183). Interview is one of the most important sources of case study information and can take one of several forms, including open-ended, focused, and structured (Merriam 1998). In this study, generally, the teacher’s interview questions were about the teacher’s own reflection about the lesson and issues about which I was curious as an observer in class. I pursued reflections raised by the teacher herself by providing guiding questions for example, “How do you feel about the GS lesson we just had?” (for the complete set of questions, see Appendix A). Additionally, in this future school, the students were encouraged to reflect on their learning experiences and share their individual reflection through their class blog. While doing a post-lesson reflection was not compulsory after every single lesson, most of the students had been used to doing it. In this study, every student of the class was required to provide their individual feedback to open-ended guiding questions such as, “How do you feel about using GS in your Chinese writing class?”, “What did your group members do that helped you learn?” (for the complete set of questions, see Appendix B). These questions arose for which I needed answers to confirm my observation conclusions and to understand what was going on in the small groups. Additionally, I also collected the grades given to each group’s GS-based group inscriptions by the teachers, which could partially indicate the task performance of every single group.

While the teacher was carrying out the lesson, I sat in the class, observing teacher and student actions and taking field notes. One video camera was set at the back of the classroom to record the classroom session. Five other video cameras were set behind each individual group to capture their group work. The screen-capturing software iShowU was installed on every student’s MacBook to record all the actions of individual students on the computers, as well as their verbal talks and facial expressions. Taking use of the video data, I had opportunities to review classroom actions and to observe or isolate individual/group parts of what was going on. The digital video technology is able to facilitate data collection and empirical investigation of real-time learning processes (Park and Kinginger 2010). In this study, the use of iShowU provided an explicit indicator of learners’ cognitive processes, and presented an in-depth, moment-by-moment analysis of a collaborative learning process.

After that, the video data captured by video cameras and iShowU during the whole learning process were transcribed verbatim. I first watched the video captured by iShowU for each individual student and transcribed the face-to-face and GS-based interactions within the group. Another researcher (a colleague of mine) helped to check the accuracy of the transcription by comparing the classroom/group video and iShowU data with the transcript. The emphasis of the transcription was on the accuracy of the discussion content and sequence of turns. Speaker intonation or other discourse properties were not specifically examined. Transcripts were parsed into turns. In transcribing, there were occasions when the speaker was interrupted while speaking. Even when the content was not fully expressed, the turn was considered completed. Backchannel response, such as “yes” and “uhm,” were also considered as turns. In the lesson, students were encouraged to communicate in Chinese, and thus their verbal conversations and postings were mainly in Chinese. English translations were added in square brackets. There were sporadic instances of conversations in English. They were transcribed verbatim, including all the typing and grammatical errors if any.

Talk is the most important means of interaction to ensure that a student’s engagement in a series of activities contributes to their understanding development in classroom education (Mercer 2008). Not only the talk between students, but also some occasional instances of teacher–student talks that occurred within small groups were also included in the transcripts. On the top of that, inscription is a valuable means of interaction in a networked classroom as well. Prior work in learning sciences has emphasized the role of learner-constructed inscriptions for meaning making practices (Kozma and Russell 2005; Medina and Suthers 2012). In the transcript, students’ talks (verbal discourses) and GS-based interactions (mainly focused on inscriptions) were blended together and presented in a chronological order (Table 6.3).

Table 6.3 Example of interaction transcript of a group
Full size table

6 Data Analysis

In qualitative research, data analysis is a complicated process of systematically searching and arranging the field notes, interview transcripts, and other materials that researchers accumulated to increase their own understanding of the data and to present what they have discoursed to others (Bogdan and Biklen 1982). Analysis involves working with data, organizing them, breaking them into meaningful units, synthesizing them, searching for patterns, discovering what is important and what is to be learned, and deciding what you would like to tell others (Bogdan and Biklen 1982). As Stake (1995) stated, the function of the qualitative researcher during data gathering is clearly to maintain vigorous interpretation; collection and analysis of data often occur simultaneously in qualitative research ( Merriam 1998). In the present study, in accordance with its objectives, the data collection and analysis aim to obtain information on seven specific aspects:

  1. (1)

    The distribution of different functions of interactions (the knowledge or information that is handled in interactions) in different media.

  2. (2)

    The quality of group output at different phases of the task.

  3. (3)

    The various patterns of medium transition observed as students completed the task in groups.

  4. (4)

    The trajectories of group understanding development at different phases of the task, and their interplay with the patterns of medium transition.

  5. (5)

    The various types of plenary and group-level instruction/assistance offered by the teacher.

  6. (6)

    The teacher’s perception about students’ learning and her interpretation of her own classroom behaviors.

  7. (7)

    The students’ perception about their group working experiences, task design, and learning gains.

To this end, a five-stage analysis procedure was followed (Fig. 6.3). Stages 1 and 3 of data analysis were mainly focused on analyzing the process data (transcript). At stage 2, the analysis revolved around group output (all the group inscriptions on the group board). Stages 4 and 5 accentuated the teacher’s interview and the students’ reflection, respectively. Whenever necessary, observation notes were used in all the stages of data analysis for the purposes of clarification and confirmation.

Fig. 6.3
figure 3

The five-stage analysis procedure

Full size image

At Stage 1 of data analysis, statistical analysis of interactions was conducted to provide an overall picture of the distribution of interactions in different functions and media as small groups completing the given task. In other words, to analyze learning processes in a networked classroom environment, I first focused on the medium and functions of interactions.

Approximately 500 min (19 students × 25 min per student) of iShowU data were transcribed. In the transcript of each small group, each student’s verbal discourses and GS-based interactions were presented in a chronological order. Following the multilevel analytical approach introduced in Chap. 5 and adopting the unit of event, I first classified all the transcribed events (face-to-face-mediated or GS-mediated) into the five categories related to functions performed to complete the task. In Table 6.4, the categories are presented with description and examples.

Table 6.4 Categories used to identify the function of interaction in small groups
Full size table

Since this study aims to investigate how L2 learners in different group settings appropriate GS for productive collaborative interaction, assessing group learning outcomes (reflected by the quality of group final product) can provide information to help analyze and assess group interaction process. Hence, at Stage 2 of data analysis, a rubric, in accordance with the task requirements, was developed to evaluate group final product (output) at each phase of the task so as to examine whether different groups successfully completed their group task at each phase and assess their performance.

The rubric was built on the Complexity, Accuracy, and Fluency (CAF) framework (Skehan 1998; Ellis 2003) which has been widely used in applied linguistics research to measure learners’ writing and speaking performances. The rubric was further modified after consultation with the Chinese language teacher, Chin. We replaced “Fluency” with “Structure” by considering the activity design. In this planning task for argumentative essay writing, students were required to contribute ideas about the topic and to organize and synthesize their group ideas. A completed essay was not expected. Such activity design aimed to provide opportunities for every student to share and improve their ideas. The main point of the activity was to create a diversity of ideas and achieve accuracy of expression on the matter, not necessarily to obtain fluency in expression.

The rubric (shown in Table 6.5) developed focused on assessing three aspects: (1) “Syntax and lexis complexity” that is, “the extent to which learners produce elaborated language” (Ellis and Barkhuizen 2005, p. 139); (2) “Content richness and accuracy” that is, the extent to which learners provide diversified and innovative arguments/ideas and generate reasons/evidences to support these arguments, and the extent to which learners produce correct and appropriate language to deliver the contents; and finally, (3) “Structure appropriateness” that is, the extent to which the contents delivered are well organized. The group output at the end of each phase of the task was rated on a four-point scale, from 1 to 4 points in units of 1, from these three dimensions, respectively.

Table 6.5 Rubric for coding quality of group output
Full size table

A contextualized qualitative analysis allows for a fuller utilization of the available information (Erickson 1996; Barron 2003) to make interpretations of the quantitative results about the interaction and group learning outcomes. At Stage 3 of data analysis, the transcripts (including both verbal conversations and inscriptions) of each group were separately marked off by the content of social interaction as indicated by its function in accomplishing the collaborative writing planning task at different phases. The analysis at this stage was always guided by the two research questions. In order to summarize the patterns of medium transition that emerged, I referred to those excerpts in which inscriptions were generated and further utilized. Meanwhile, the microanalysis of interaction at this stage also concentrated on investigating in what way group understanding development relates to the medium transition and, during the process, what role the inscriptional devices can play. Intrinsic and extrinsic factors, such as group language proficiency, task design, and teacher’s activity, were both taken into consideration. Section 3.3.2 of the literature review showed that the teacher, as a facilitator or moderator, is still crucial to the appropriation of collaborative technologies and the effects of students’ collaborative learning in a networked collaborative learning classroom. In this study, the qualitative microanalysis of interaction not only paid close attention to interactions among small-group students but also concentrated on the teacher’s interactions with students at both class level and group level. As the main body of data analysis in this study, the ultimate purpose of this stage of data analysis was to extract the principles for designing and enacting collaborative activities in L2 learning classrooms productively.

At the last two stages, the teacher’s interview data and the student’s reflection were analyzed, respectively. This stage of data analysis served to support the validity of the research findings as an interpretation of the transcript. I first reviewed the answer from the teacher and all the students. Then I reread the answer with the intent of looking for themes in the answers to each of the questions. Third, I revisited these themes to decide which research questions they addressed and organized the themes according to my research questions. Finally, based on this part of data analysis, I developed narrative statements in the form of a specific or general description that address the research questions.

To sum up, in this study, I adopted the inductive process for data analysis to identify the key functions of interaction that emerged through the use of the detailed reading of raw data from the transcripts, the field notes, the teacher’s interview, and the students’ reflection. I immersed, summarized, and condensed the intensive raw data, noted the primary themes and categories that emerged, and linked the research objectives and summary findings derived from the raw data to produce reliable and valid findings. In addition to data source triangulation, one of my colleagues helped to check the transcripts, classroom observation notes, and data coding for the validation of this case study.