Tuesday, 27 August 2013

FIT6021 Workshop 7: Design Science

Part One
Design Science was the theme of Workshop 7, and it was the workshop I was most looking forward to this semester. While I didn't explore it in any great depth at the time, I actually used aspects of design science in my Honours thesis in 2011, where I created an interactive fantasy eBook prototype and gave it to users to test and see how much they liked reading it compared to a non-interactive eBook. Design science research involves the researcher creating an artifact that will (ideally) be useful to some particular section of society. They will then test this artifact and evaluate it on how successfully it achieved its purpose. A design science research artifact can take one of four forms, as described by Hevner, et al (2004):
  • Constructs: A conceptual object created to describe and represent some aspect of the real world, such as classes and subclasses of things (eg. educational institutions and primary schools/universities), components of things (eg. a teacher at an educational institution), properties of things (eg. academic standing of a university), states of things, events that occur (eg. university has funding cuts and loses a department) and processes that things undergo. Essentially it is the symbols and language used to define and communicate problems and solutions. 
  • Models: A conceptual object made up of constructs and links between these constructs as a way to represent or abstract a real world situation or entity. An example would be a model for a database showing the three main levels (internal, conceptual and external).
  • Methods: A set of actions used to achieve a certain outcome or solve a problem. It may take the form of a mathematical algorithm or a general textual description for the process, or any combination of the two.
  • Instantiations: An actual system (hardware or software) that shows how constructs, models and methods can be implemented for its intended purpose. It can almost be seen as the 'end result' of the other three forms, and it allows researches to perform a more concrete assessment of its effectiveness and suitability for real world situations, how people use it and any benefits or limitations to its use.
Unlike Hevner et al., Gregor and Jones view Design Science as resulting in two key things; a generic hardware/software system or method, and a theory to explain the system's structure, behaviour and effect on its environment (2007). Weber (2013) proposes an alternative method of evaluating design-science research; by looking at the problem specification, contribution to knowledge and solution derivation method as a process, and at the construct, model, method and/or instantiation as a product.

I think Design Science is a useful and relevant research method in the field of Information Technology and Systems, however one drawback is that it can quickly become out of date. The rate at which technology advances continues to increase, which means that prototypes of systems developed as a result of design science research may become obsolete almost as soon as they are created, or the research itself may become irrelevant if the technology it is based on gets phased out in favour of something better. It is also important that the 'problem' being addressed by the research is actually a problem and that solving it will help a large number of the community, rather than just satisfying curiosity. Another thing to consider is that the design process is iterative and largely relies on trial-and-error; this means that the researcher must be careful if they are under any time constraints, as the longer it takes them to get their artifact right, the shorter lifespan it will likely have.

Part Two
I think Design Science will have a very strong role in my research. I will be looking at how an interactive narrative - coupled with the motion capture technology of the XBox Kinect - can be used to teach sign language and give feedback to the user (there are many programs etc that teach sign language already but few, if any, that actually give the user feedback based on their gestures). At some point I will need to have at least some sort of basic prototype (artifact, in this case an instantiation) to test on actual users. As well as finding out whether such an approach is effective in teaching sign language (or at least more effective than more common educational software), I hope that the program/system resulting from my research can be further developed into a tool for children to learn sign language, making it a practical contribution to society as well as a theoretical contribution to research.

Design Science Online Resources
Here are the online resources I found for anyone who wants to know more about Design Science:

  • "Design Science Research in Information Systems" by Ting-Peng Liang. Provides a brief overview of Design Science and discusses a few examples. LINK
  • "ICT-Design Science Research" by Raimo Halinen. Another slightly more detailed overview of Design Science and how it can be applied and evaluated (a couple of the slides are not in English, but the majority of it is). LINK
  • "Design Research Citations" by Unknown Author. Though this page does not directly contain information about Design Science itself, it does provide an extensive list of references/citations about the Design Science method and various sub-categories (eg. general references, philosophical references, methodology). LINK
Rex approves of Design Science:

No comments:

Post a Comment