Deep Learning Design & Context

My research path this week has led me to discover a literature review that does an excellent job in investigating and illustrating the concepts and developments around effective design that enables Deep Learning (DL) within Technology Enhanced Learning (TEL) . The authors aim to provide some guidelines for designers of TEL.  I hereby summarise and comment upon the contents of the article and also discuss additional content that I have further explored as a result of the themes raised.

To begin with, to highlight the methods and approaches being discussed, I illustrate how the authors assert that design for DL is mainly concerned with the task at hand – the required/desired achievement and the activities and resources needed to join these together.

Screen Shot 2015-12-20 at 15.58.58

(Boyle & Ravenscroft)

The article begins by stating that one of the key issues in TEL today is that technology is developing at such a rate that e-learning courses are being led by the technological abilities available rather than fundamental DL objectives ‘The technology is alluring; this can distract from deep design in a surface rush to exploit the affordances of the new technology, We need a basis for design, and a conceptual unit of organisation, that are applicable across constant technological change.‘ (Boyle & Ravenscroft).


The authors go on to state that in order to offer a stable basis for general design principles to be developed, TEL needs a focus point around which all other concerns evolve. They go on to propose that this focus point should be ‘context’ as this is a theme that has been discussed as having importance in literature regarding TEL but that has not yet been clearly defined or articulated.

The Oxford English Dictionary defines ‘Context’ as: ‘The circumstances that form the setting for an event, statement, or idea, and in terms of which it can be fully understood’.  This definition is incredibly broad by definition and needs further explanation. A neat way to further describe context in terms of how it affects human interactions and communication can be found in ‘The Six Kinds of Context‘ by Lee McGaan, Ph.D, Monmouth College, Illinois, USA: 

Types of communication context:
Cultural, Inner, Physical, Relational, Situational and Symbolic.

All of these contexts play a part in TEL. I have matched context terms to associated and relevant elements of TEL as I see them below:

Pre-conceived approaches to learning material based on existing rules and patterns learned from our cultures. Colour is most likely the most important variable in TEL to consider under cultural context.
Emotions experienced before approaching learning material that could affect data retention and knowledge transfer (see HEIDIG, S. et al.).
Actual objects that can be handled during the learning activity to enhance learning points. Also environmental conditions such as weather and location.
The student should consider the teacher or teaching program to be knowledgable and trust him/her/it to impart accurate knowledge.
The situation in which a student considers him/herself to be in and the association expected responses to each (lecture, quiz, conversation, game, problem-solving etc.)
Messages which are received before or after a learning event that influence how one understands the learning event in part or in its entirety.

So my learning points obtained from reviewing these context definitions are: Thinking pragmatically, an eLearning course that could justifiably refer to itself as TEL with DL would perhaps be one that incorporated or utilised various contextual data. i.e:

  1. It manipulates emotions pre-learning
  2. It is visually customisable to allow for regional preferences and language – perhaps even customisable on an individual basis.
  3. It offers clear identification of the teaching delivery method  – being varied and diverse throughout the course.
  4. 3D visualisations would be the next best thing to handling actual objects or seeing scientific experiments in action etc. This would be something to experiment with and carry out further research on.
  5. eLearning that is presented in the environment that is being taught on could be fruitful – i.e. – learn about a roman ruin via mobile e-Learning whilst actually standing in the ruins themselves.
  6. Additionally schema could be assessed through a pre-course quiz. (See my separate blog)

There are of course numerous other considerations in deciding on the included elements of an eLearning course that would facilitate DL (such as layout, multimedia inclusion, device sophistication, learning complexity etc.). Here we are simply discussing context.

On a related note, see my previous blog post ‘The Effect of Color‘ to view a video which discusses human relationships with colour in both ‘Cultural’ and ‘Inner’ contexts.

A subject also discussed by Boyle & Ravenscroft is that of Ubiquitous Computing (UC). This is a relatively new term that can be summarised as:

… a concept in software engineering and computer science where computing is made to appear anytime and everywhere. In contrast to desktop computing, ubiquitous computing can occur using any device, in any location, and in any format. A user interacts with the computer, which can exist in many different forms, including laptop computers, tablets and terminals in everyday objects such as a fridge or a pair of glasses. The underlying technologies to support ubiquitous computing include Internet, advanced middlewareoperating systemmobile code, sensors, microprocessors, new I/O and user interfaces, networks, mobile protocols, location and positioning and new materials (

In exploring this field further and relating it to DL; context aware technologies can incorporate environmental data such as (for example) location, weather, physical conditions plus the personal location (actual) of peers. This data could enrich learners experiences, for example: enable visits to real-world sites for undertaking learning in situ. Additionally learners could meet up with peers who are in the vicinity on an ad-hoc basis in order to discuss a problem and brainstorm. It has to be said at this point that, although useful, “this (UC) doesn’t amount to understanding and modelling of context” (Boyle & Ravenscroft) so we need to really consider how we can enable truly contextualised deep learning using UC.

The word ‘context’ originates from the latin phrase ‘to weave together’ (Contextus, from con- ‘together’ plus texere ‘to weave’) (Oxford Dictionary Online). So to achieve true DL we need to weave real-time UC data into authentic learning experiences. It has to be considered at this point that real-time data are neither stable nor standalone elements so the learning platform really needs to be adaptive to live context. Learning experiences, mobile or otherwise are usually built upon fixed frameworks which are laid out in advance and dependant upon stable conditions. The authors ask if they should be more flexible?

“A major theme within TEL is the critique of traditional ‘dis-embedded’ formal context and the argument that we need to replace these with the design of more authentic contexts for learning”. (Boyle & Ravenscroft)

This leads me to think about whether eLearning should emulate the classroom context (artificial classrooms) or encourage learning aligned to real-life situations, known as authentic learning, where we create contexts that are meaningful and engaging (rather than a didactic approach) where we learn within and around a context where the skills will be used – such as in apprenticeships. Here we see a parallel with e-textbook learning research in the ‘paper-metaphor’ conundrum. (See separate post)

Constructivist Learning (CL) is considered an example of authentic learning. In CL, the users are considered as both active and re-active constructors of the learning situation. This approach allows users to be creative v.s. a traditional more behaviourist approach (rigid with pre-defined steps).

CL can be summarised as facilitating the below qualities in learners:

  • Critical discussion and reasoning
  • Exploratory dialogue
  • Creative thinking.

This approach will allow learners to coherently develop context as they progress along a learning path. But as the authors highlight, the ability to develop active real-world context will depend on the contextul depth and quality of the initial TEL program:

“The degree to which participants have freedom to weave contextual coherence from below is shaped by the strength or weakness of the perceived contextual structure” (Boyle & Ravenscroft)

I would comment here though that TEL designers need to be careful that ‘authentic’ learning programs are neither simplistic nor too commercial. ‘Real-world’ contexts can be limited by commercial, political or socio-economic agendas. Also, is this approach suitable for all ages of learners?

Another important part of the discussion in context-centric learning is the assertion made by the authors that learners need both a shared generic understanding of context in addition to high-quality, local-contingent learning. A shared generic understanding of context would give learners  a starting point upon which to build real-world live contexts.

In exploring this subject further, the authors discuss how historical context gives a starting point whereas real-world live contexts need to be generated by the participants. A metaphor is given in that games are conveyors of context: the game rules and play method are pre-determined – but the game itself is co-ordinated by the players over an amount of time. (See my separate post on role-play and games as facilitators of Deep Learning)

Another question arises here surrounding the perception of context – can we achieve a balance between the context perceived by the user and that perceived by the designer? How remote is the designer of TEL?:

“Is the designer an actor in the context, like a teacher who both designs and runs an interactive seminar? Alternatively, is the designer a pre-constructor of the learning context, but not a participant?” (Boyle & Ravenscroft)

Traditionally TEL designers are actors but visual design has suffered in that it plays second fiddle to pedagogic concerns. (although I aim to propose in my research that visual design plays a key part in pedagogy). I would propose that the designer as pre-constructor is a healthier construct – the finished TEL product has to be more complete, explicit in its aims and shareable, thus open to criticism.

“Design requires not just a construction of the overall learning context, but detailed concern with the tasks, the activities of learners, and the means of knowledge representation used”. (Boyle & Ravenscroft)


In conclusion, the authors propose these essential elements for DL TEL:

  • Setting engaging and authentic tasks
  • Appropriate and effective use of representational media
  • Learner control, with appropriate scaffolding support
  • Productive feedback on errors.

The authors finalise their article by asking how can a designer of TEL find the knowledge and insights to accurately inform the design of these features and thus facilitate DL. They quote that Psychology has the answers… (See my separate post regarding this subject).


BOYLE, T. and RAVENSCROFT, A. (2012). Context and Deep Learning Design. Computers & Education (59). p.1224-1233.

HEIDIG, S., MULLER, J. and RECHELT, M. (2014) Emotional design in multimedia learning: Differentiation on relevant design features and their effects on emotions and learning. Computers in Human Behavior. 44. p.81-95.
(Accessed: 12.12.15)
(Accessed: 12.12.15)
(Accessed: 13.12.15)
(Accessed: 20.12.15)


One thought on “Deep Learning Design & Context

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s