Exploring Types,Levels,and Processes Of Intercultural Communication In A MOOC

The MOOC(Massive Open Online Course),as a rising phenomenon of the digital age,is a platform for intercultural communication involving large groups of individuals across all the divisions of gender,age,and more importantly,culture.This research highlights a MOOC on“Intercultural Communication”(Run 3,Fall,2016),a five-week course presented by Shanghai International Studies University(SISU),as a case to explore the types of communication dynamics(network analysis)starting with a grounded theory approach and exploring the relevance of network analysis and machine learning tools.The MOOC case chosen attracted8225 enrollments,with 4635 learners starting the course,from which criteria were set to examine those “active” in the course(N=733)by making three or more comments.This dissertation provides a brief review of MOOCs’ definition and background,theoretical foundations for the analysis of such big online databases,and how it may facilitate intercultural competence as a learning goal and foster the engagement of speech acts as online comments.Data mining tools followed step by step and detailed grounded analysis of the extensive data pool to help uncover the main demographic and typological information participants voluntarily provided throughout the course.Further extractions from the embedded data helped reveal a viable set of participant categories relevant to this research: country of residence,being an inhabitant or a sojourner,original country if a sojourner,gender,and being a learner or a mentor.The various ways learners disclosed these identity components are examined and discussed,along with the efficacy of various data mining methods.Posting comments is the main measurable activity of course participants.Commenting behavior is evaluated for each of the participant categories.To address the participants’ cultural diversity,country(of residence)affiliations at the micro-level were considered along five macro geo-cultural regions: 1.Europe,North America,and the Pacific;2.Greater Asia;3.Latin America and the Caribbean;4.The Middle East and North Africa,and 5.Sub-Saharan Africa.“Following” and “Replying” are determined to be two primary directional communication behaviors(from participant A to B)and analyzed using Node XL Pro software to determine networks and interactive communities in the course.Examining this multiplex network of communication helped categorize the types of participants.Mentors were first analyzed as a specific group of participants volunteering and selected to willingly involve in communication to encourage other participants to communicate.Then the converse group of noncommunicative participants was examined.For the other “active learners,” three centrality measures were extracted(Betweenness centrality,Eigenvector centrality,and Page Rank in the following and replying network)to categorize three groups initially.(1)Low-communicative participants(n=117)ranked in the lower third on all three measures of centrality,and(2)Highcommunicative participants(n=103)ranked in the high third on all three measures.All others were categorized as medium-communicative participants(n=299).Among the comments made by each group,five significant topics were extracted for all of the above groups using the LDAvis topic modeling method to examine aspects of engagement and competence.Further calculation of centrality measures showed differences across countries and geocultural regions and revealed how participants within a region could be divided into types of social learners.Among 88 “communities” developed in the course,61 stayed in small communities(with an average of 2.8 participants in each),24 were part of medium-sized communities(with an average of 6 participants in each),and three large communities were formed(with an average of 15.3 participants in each).Exploring participants’ intercultural diversity within communities reveals that they tend to communicate with people with different cultural backgrounds,meeting the course’s stated goals.The four key research questions were answered,providing findings on(1)participant types and the essential demographics in this type of MOOC(showing that even identity-protected hidden demographics can be reliably elicited with software tools),(2)two types of communication dynamics(replying and following),(3)five ways that levels of depth(“communities”)are formed based on varying types of online networking patterns(based on the degree of active engagement: low,middle and high,in addition to the highly involved mentors and contrasting non-communicative participants),and(4)ways that level of activity and types of topics facilitate intercultural competence(in sequences similar to Deardorff’s process model).The study also shows how selected social networks and machine learning software can help identify,elicit,or analyze MOOC big data(to determine missing demographics,types of dynamics,communication levels,and processes that show competence).Recommendations are provided on how these methods and analysis techniques could be applied to other MOOC research or improvements for this or similar courses.