In this section, we try to reexamine the connection between our empirical ap- proach and the learning theory on which this work is based. We will first dis- cuss how social development theory backs up the learning on MOOCchat and GLIDE.
7.2.1
Revisiting MOOCchat and GLIDE: Focusing on Social
Development Theory
Social development theory by Vygotsky [127] influenced this work as the fun- damental framework of how learning occurs and how to facilitate it. The key takeaways of the theory around which this work built were 1) “guided partici- pation”, also known as scaffolding, that helps learning in the zone of proximal development and 2) the two crucial roles of language, sign systems, and tools
that mediate interaction with more knowledgeable others and students inter- nalize learning materials, and 3) indirect interaction with the broader society through language as a socio-cultural product that better better contextualizes learning. This section reflects on the two empirical approaches we made re- garding these three aspects.
Relying on the theoretical implications of the social development theory, the MOOCchat was an attempt to introduce student interaction into extremely decontextualized education settings. From the perspectives of social develop- ment theory, MOOCs open more knowledgeable others and interaction with them. On the contrary, the current work developed MOOCchat approach real- time matching feature to form ad-hoc groups of students, which gives them the chance to interact with the more knowledgeable others.
The way the students in a small group on MOOCchat interacted each other implements the ideas of interaction as the source of learning and language as the mediator for learning. The specific element intended for learning on MOOCchat is where the students exchange their ideas in the discussion and have a chance to change their answers to finally submit. During the discussion, the real-time chat induces them to reason their initial answers, persuade others, identify er- rors from others’ references, and negotiate on the final answers, which in turn help the active process of internalization where each student uses one’s own language for this whole procedure, rather than passively copying others’ refer- ences. Thus, we argue that MOOCchat enables student interaction for learning and also structures their interaction for effective internalization through active processing of language.
First, the design of GLIDE is rooted in guided participation. The feature branch workflow is generally considered as a difficult topic as reported in the inter- views with the teachers. The tasks required to perform a project using the fea- ture branch workflow are out of the students’ reach, considering their actual de- velopmental level and potential developmental level. Therefore, to pull those tasks down to the area within the zone of proximal development, selected tasks among the hard ones had to be scaffolded; using command line interface, hav- ing many options per command, and understanding the abstract model of Git workflows. To achieve this, GLIDE was designed as a web-based application with the graphical user interface instead of command line, dedicated to sup- port one simple target workflow (the feature branch workflow) while sacrificing the high levels of flexibility with the commands and options, and tightly cou- pled with the lecture to help building the abstract model of the workflow before working on software development. This is where the participating teacher’s advice was critical in finding the appropriate levels of difficulty of the tasks, to include them in the common zone of proximal development for the students in the class.
The second way that learning in GLIDE approach reflected social develop- ment theory is that the lesson plans for the lecture placed a significant focus on being able to explain the feature branch workflow to others and discuss it. The lesson plans had the dedicated time for peer discussion where the students ask, answer, and explain what they understand in their own language. This reflects the mediation role of language during the interaction with the more knowledgeable others in the theory and the findings from the qualitative anal- ysis suggested that it might have worked as the source of learning. The ideas exchanged are also supposed to be internalized in language form, which is the
key step of learning according to the theory.
Lastly, GLIDE approach also put significant focus on giving the students the exposure to the standard terminology and widely accepted norms (e.g., not writing code off directly on the master branch and keeping the commit tree as simple as possible) as implicit sign systems and language from the culture of collaborative programming in industry practice. Social development theory ex- plains the value of this teaching practice that the learners can indirectly interact with the broader society through the mediating language. By teaching with this concentration, we argue that learning on GLIDE can be better contextualized in the society of collaborative programming, which potentially reduces the gap between learning context and the practice in industry or academia.
7.2.2
Theoretical Contributions of Exploratory Study on In-
structional Designs
The theoretical contribution of the interview study includes a discussion of how instructional designs from cognitive load theory and cognitive apprenticeship have been contextualized in web programming tasks. We also illustrated how those theoretical constructs interplay with the task types that students perform and collaborative approaches. Another theoretical contribution is that this work proposed the instructional design and collaboration task matrix as an analytical framework that other researchers can use to understand and explore the inter- action between these factors.