2.1 Online intercultural exchange

Nowadays, more and more people endorse Kramsch’s (Reference Kramsch1993) idea that teaching culture has to be an interpersonal process with meaning emerging only through social interaction. Such a proposition is well in line with a social constructivist approach in which learning is a learner-generated, social, and collaborative process (Vygotsky, Reference Vygotsky1978). Following this trend, online intercultural exchange projects are gaining increased popularity in second language education in recent years. These projects have evolved from mainly asynchronous communication to multimodal environments that involve both synchronous and asynchronous communication (Guth & Helm, Reference Guth and Helm2010; Guth & Helm, Reference Guth and Helm2012; Lamy & Hampel, Reference Lamy and Hampel2007).

While asynchronous communication is regarded as a better tool by some researchers for deeper learning because of its flexibility of access and time for reflection (Abrams, Reference Abrams2003; AbuSeileek & Qatawneh, Reference AbuSeileek and Qatawneh2013; Anderson, Reference Anderson1996; Collison, Elbaum, Haavind & Tinker, Reference Collison, Elbaum, Haavind and Tinker2000; Green, Reference Green1998; Im & Lee, Reference Im and Lee2004), other researchers have suggested that its lack of spontaneous feedback may prevent people from moving to a more profound analysis of a topic (Pilkington, Reference Pilkington2004), and that the community-nurturing characteristic of synchronous communication is helpful in encouraging deeper reflection of issues (Salmon, Reference Salmon2000; Schwier & Dykes, Reference Schwier and Dykes2004).

So far, the major proportion of data in the literature was collected from asynchronous communication such as forum postings and emails, and most of the models used today have not yet been able to provide an efficient and systematic way for researchers to observe and analyze learner interaction in synchronous online exchange characterized by spontaneity and fluidity (Wang, Reference Wang2005). Furthermore, despite the popularity and potential of online intercultural exchange projects, researchers caution against oversimplification and assumptions about their effectiveness and productiveness for language and culture learning. In his proposal of a social-realist approach to online foreign language education, O’Dowd (Reference O’Dowd2007) called for an extensive systematic movement and change in order to get the best out of new technologies for students’ language learning. To achieve this goal, educators need to be well informed on findings from empirical studies focusing on diverse aspects in online communication. For instance, the question of whether synchronous online exchanges would end up with superficial conversations with little educational value can only be answered by investigating the actual exchange process itself. In such an investigation, the use of self-generated questions is no doubt an important factor to explore.

2.2 Questioning for productive discussion

Cognitive process is one of the major aspects examined in recent research on promoting online discussions (Gao, Wang & Sun, Reference Gao, Wang and Sun2009), in which questioning is a topic seen repeatedly. Many researchers believed that the level of thinking that occurs was influenced by the level of questions asked (Chin & Langsford, Reference Chin and Langsford2004; Cox, Carr & Hall Reference Cox, Carr and Hall2004; Elder & Paul,Reference Elder and Paul1998; King, Reference King1990; Qatipi, Reference Qatipi2011; Savage, Reference Savage1998; Taba, Reference Taba1966). Gao, Zhang and Franklin (Reference Gao, Zhang and Franklin2013) further elaborated on this idea by suggesting that information was more likely to be understood and retained from a discussion when individual learners were actively engaged in cognitive efforts such as questioning.

Furthermore, some studies found that online discussions very often failed to support a high level of knowledge construction (Kanuka & Anderson, Reference Kanuka and Anderson1998) or truly conversational modes of learning (Thomas, Reference Thomas2002). Efforts have been taken to find out ways to better understand the online communication process and to foster and support higher-order thinking of the participants. Using questions was one of the major strategies found to be essential for increasing interaction and language output as well as critical thinking within an online communication system (Lai, Reference Lai2006). Different models and approaches have been put forward for creating a productive and meaningful online exchange, and the use of questions is often one of the main components (Garrison, Anderson & Archer, Reference Garrison, Anderson and Archer2000, Reference Garrison, Anderson and Archer2001; Lipman, Reference Lipman1991; Newman, Webb & Cochrane, Reference Newman, Webb and Cochrane1997).

For instance, Garrison and his colleagues (2000) proposed an analytical framework known as “Community of Inquiry.” Their approach was later used by McLoughlin and Mynard (Reference McLoughlin and Mynard2009) in their small-scale study into the efficacy of online discussion forums in facilitating higher-order thinking. Evidence of higher-order thinking processes was indeed found from participants’ online postings when the correct conditions were present. In order to better understand the dynamics and productiveness in online interaction, it would be worthwhile to also look for evidence of higher-order thinking in synchronous communications. This served as the starting point of the present study.

In addition to the Community of Inquiry model (Garrison et al., Reference Garrison, Anderson and Archer2000, Reference Garrison, Anderson and Archer2001), effort was also taken by Gao, Wang, and Sun (Reference Gao, Wang and Sun2009) to provide a more systematic and comprehensive framework for understanding how learning occurs through online discussion. In the productive online discussion model they have developed, facilitating thinking and discussions by raising questions was listed among the most desired learner actions.

Empirical studies were also found investigating the effect of questioning and instructional devices such as question prompts in directing student thinking through productive discussion in an online or offline learning environment (Choi, Land & Turgeon, Reference Choi, Land and Turgeon2005; Ge, Chen & Davis, Reference Ge, Chen and Davis2005; Yang, Newby & Bill, Reference Yang, Newby and Bill2005) as well as how question types influenced the quantity and quality of undergraduate students’ online discussions (Bradley, Thom, Hayes & Hay, Reference Bradley, Thom, Hayes and Hay2008). These studies indicated that students’ critical thinking skills could be fostered and demonstrated through structured online exchange. Additionally, the use of questions was often found in these studies to be an important factor that instructors could work on to promote a more productive and effective exchange between language learners. However, as some studies found that when students were asked to generate questions on their own, factual rather than thought-provoking questions were generally posed (King, Reference King1990), the present study sought to verify such a claim and to find out whether and how language learners would actively participate in this small-scale project by using self-generated questions.

In addition, to date, little research was found investigating the use of questions in synchronous online intercultural exchange between language-learning dyads, let alone projects involving Chinese English learners and U.S. Chinese learners. Thus, in order to better understand the characteristics of online learning and the potential of synchronous communication for substantive learning in language and culture education for learners across the globe, there is a practical need to study what these two groups of learners would be actually doing in the collaborative online space.

3.1 Participants

The participants were six English major students enrolled in a General English class at a university in Southern China and six students enrolled in a third-year Chinese Language and Culture class at a public university in the U.S., ranging in age from 18 to 22. The students from the U.S. were native English speakers, and the students from China were native Chinese speakers.

A pre-survey was conducted to elicit information about the participants’ foreign language and culture learning background and their expectations towards the project. The results indicated that the Chinese participants had all studied English as a subject in school for twelve to thirteen years, and none of them had been to the U.S. Compared to the rather homogeneous language-learning experiences of their counterparts in China, the U.S. participants had a more diverse background in Chinese language and culture learning. The number of years they had spent studying Chinese ranged from three to seven years. Three of them had studied Chinese for four years. While two participants had never been to China before, the other four had all visited China.

3.2 Procedures and data

The desired outcomes for this exchange project were greater cross-cultural understanding and language development. The project itself lasted for six weeks, and it consisted of two main steps. Participants first used an online collaboration tool (Google-forms) for questionnaires (see Appendix) composed of word associations, sentence completion, and situation questions (similar to the initial stage in the Cultura model (García & Crapotta, Reference García and Crapotta2007)). They then had synchronous discussions within dyads using instant chatting tools (Wechat or QQ).

After completing the pre-survey, the participants were asked to answer the questionnaires in their native language. Afterwards, the participants were able to read, compare, and reflect on the answers given by all the participants, which had been complied into a spreadsheet. In order to provide a richer input for the online discussion, the questionnaire was also sent to other students from both sides who were interested in answering the questionnaires but not able to participate in the subsequent exchange. A total of 31 responses were collected, 14 of them from U.S. students, and the rest from Chinese students.

In the second stage, dyads were randomly arranged with one Chinese student and one American student in each dyad. No additional criteria were used to match the participants. It is believed that, in a natural intercultural exchange environment, people are most often brought together by common interests instead of other considerations such as class credits or level of language proficiency. After all, no matter which language, English or Chinese in this case, was used in the communication, each dyad would always have a native speaker in that language. Participants then chatted with their partners about a particular word association, sentence or situation they had chosen by using instant online chatting tools. They were required to use text chat only. In order to make the project beneficial to language learners on both sides, each dyad was asked to finish discussion of four topics in four weeks’ time, two in Chinese and two in English. Upon completion of the four topics, a post-survey was given to collect participants’ evaluation and suggestions for the project. The data collected from the post-surveys would be analyzed in a separate study.

Due to different reasons such as incomplete data because of technical issues, analyses were conducted on 20 chat logs. English was the main language in twelve chat logs, and Chinese was used in the other eight. Code-switching is a common phenomenon in all the chat logs. The chat logs were analyzed in two ways. First, descriptive statistics such as counts and frequencies of questions used were calculated in light of the research questions. Second, a qualitative examination was conducted on a chat discussion for a clearer picture of how questions were actually used to engage participants in complex cognitive processes.

3.3 Question types and coding scheme

Bloom and his co-researchers (1956) outlined six hierarchical levels of thinking within the cognitive domain: knowledge, comprehension, application, analysis, synthesis, and evaluation. Despite some controversies (Anderson & Krathwohl, Reference Anderson and Krathwohl2001), the first three levels are often viewed as constituting lower-order thinking and the last three as constituting higher-order thinking (Notar, Wilson & Montgomery, Reference Notar, Wilson and Montgomery2005; Schrire, Reference Schrire2006).

A higher-order thinking question is defined as a productive question that requires significantly more complex thinking that can stimulate mental activities and take students forward in their thinking to construct understanding (Applegate, M., Quinn & Applegate. A., Reference Applegate, Quinn and Applegate2002; Chin & Langsford, Reference Chin and Langsford2004). The classification scheme of questions (see Table 1) for the current study was thus created by combining Graesser, Rus and Cai’s (Reference Graesser, Rus and Cai2008) question classification schemes and Bloom’s (1956; Krathwohl, Reference Krathwohl2002) taxonomy. English translation is provided hereafter for all the Chinese examples, indicated by a slash (/). All the examples and excerpts are presented as they were in the original chat logs.

Table 1

Question classification scheme and examples from current study

Numbers 1–9 are used to represent and code the nine question types (see Table 1). Given that it would be much harder to decide and speculate on the intention of the questioner, and the current focus was to see if and how questions were used to engage participants from both sides, both question stems and responses solicited were thus used for question classification. In addition, a question can sometimes be a hybrid of two or more categories. In this case, if a question is a combination of categories under the same level of thinking such as analysis and synthesis, it would only be coded once; if it is a hybrid under different levels of thinking, such as comprehension and synthesis, it would be coded twice, one under each category. The six dyads are denoted by A to F in this paper.