Much has been developed, researched and written about the power of high fidelity simulations (especially in defense and healthcare) and their ability to provide far more effective training outcomes and better measurability of performance. And I will include Serious Games as well in the same context. I think it is perhaps a good time to raise a few questions.
Firstly, are simulations (and/or serious games) more suited for assessing performance than traditional paper/pencil or online tests? If yes, does this apply uniformly across all subjects/domains or are these particularly suited for certain types of assessments?
Secondly, what are the essential attributes of such simulation or game based assessments? What are the criteria upon which a simulation or game may be said to reliably, accurately and efficiently assess a student’s performance?
Thirdly, what are the more accessible ways in which such simulations or games can be developed? Learning implements like Multiple Choice or Drag and Drop questions are fairly quick and easy to design and develop, and there are a host of tools around that make the development process fairly rapid. But assessments based on simulations and games may take up too much time and the effort to develop them rises exponentially as the number of variables increase.
Fourthly, what kind of features in a simulation are required in order for the simulation to be more effective as compared to traditional alternatives? Are there techniques, like perhaps adaptive testing, that can be applied to simulation based assessments?
Fifthly, what evidence can be reliably obtained to show that simulations can indeed assess performance reliably?
I may be missing other questions, but the intent is to try to understand how simulation based assessments can be brought into the mainstream education, if indeed they can be proven a reliable and accessible alternative to traditional techniques. It will bring the fun back into taking exams for millions of schoolchildren. That itself should be motivation enough for us to research the space!
Viplav Baxi's Meanderings 
I wonder how simulation can be used for assessments, but I think situations like laboratory practicals or Practical conditions experiments can be used to suffice the requirement and could also help institutions which dont have availability of resources for lets say a Lab.
I believe the basic purpose and essence of using Simulations lies around reproducing reality and enhancing it, and for that to be achieved in any scenario like Academic assessment few thing should be noted:
1.Simulation should be made with a focus to inculcate curiosity and increased Interaction. ( curiosity being essential otherwise the whole experience of learning is ruined.
2. A rich simulation must provide the user/student a feeling of Indulgence or to experience the virtual realm.etc.
And Thanks alot for bringing up this questions, I am also working upon a personal project for Laboratory Simulation and I would add love to add in few points later. ( I have exams tomorrow :D)
Thanks Abhishek. Look forward to learning from you!
There’s another aspect – that to simulate something you must understand its essential conditions completely. That is, a simulation is a world where the simulation creator knows the rules and defines the interactions in the microworld they invent. Interesting problems worth solving often do not come with such a neat set of known rules.
Even if designers addressed your other points (which all make sense), there is a limited set of things that simulations could assess.
Thanks for your response.
There is, of course, the challenge of fidelity. Simulations have been effectively been used in a lot of areas, but these “microworlds” are more or less finitely defined. They could be adaptive or “intelligent” stretching the boundaries of what is finitely defined. The question is whether there are finitely interesting problems, and my guess is that there are enough there across the board to make an impact – from K12 to corporate uses.One of the ways the microworlds can be extended is to allow judgment and creativity in ways that can ultimately be assessible in terms of outcomes. We have also generally thought of simulation outcomes being assessible by the simulator itself – however, we should allow manual interpretation of results from actions in a simulator.