Comments on: Scenarios and Conceptual Clarity https://blog.learnlets.com/2015/12/scenarios-and-conceptual-clarity/ Clark Quinn's learnings about learning Tue, 31 Oct 2023 13:31:13 +0000 hourly 1 By: When Should You Avoid Branching Scenarios? - Experiencing eLearning https://blog.learnlets.com/2015/12/scenarios-and-conceptual-clarity/#comment-1418536 Tue, 31 Oct 2023 13:31:13 +0000 http://blog.learnlets.com/?p=4695#comment-1418536 […] Linear scenarios (Clark Quinn) […]

]]>
By: Clark https://blog.learnlets.com/2015/12/scenarios-and-conceptual-clarity/#comment-821926 Fri, 11 Dec 2015 14:22:31 +0000 http://blog.learnlets.com/?p=4695#comment-821926 In reply to arun.

Arun, yes, in situations where you do have social learning possibilities, having some ambiguity and discussion is a great opportunity. Thanks for sharing.

]]>
By: arun https://blog.learnlets.com/2015/12/scenarios-and-conceptual-clarity/#comment-821900 Thu, 10 Dec 2015 21:57:07 +0000 http://blog.learnlets.com/?p=4695#comment-821900 Thanks for the post Clark. I’ve been using what we call ‘mini-sims’ essentially your linear simulations, then more nuanced branched scenarios (your contingent scenarios) as part of our blended learning solutions. We’ve been trying to develop more efficient ways to create them looking at BranchTracker for the mini-sims for example.
For the more complex branched scenarios we currently use twine to plan and then create them using articulate or storyline but I find both so limited in terms of using variables for really interesting/personalised options. I appreciate the mention of SimWriter which I hadnt heard of and will check out. The other thing we’re really enjoying playing with is posing scenarios via simple elearning and then debating out responses in discussion forums which is great in terms of seeding community discussion.

]]>
By: Clark https://blog.learnlets.com/2015/12/scenarios-and-conceptual-clarity/#comment-821899 Thu, 10 Dec 2015 20:41:52 +0000 http://blog.learnlets.com/?p=4695#comment-821899 In reply to Chris Riesbeck.

Chris, always welcome your insightful feedback; so please don’t hesitate! You’re absolutely right that the market has changed since I first made these distinctions, and there are indeed some solutions that fit in-between branching scenarios and full simulations. NexLearn’s SimWriter tool allows variables to be passed on from earlier branches to later to influence outcomes, and Avalanche ST has some local processing available at particular locations, very much like your ‘scenes’. Actually, when I talk about designing serious games (e.g. in Engaging Learning), I very much focus on only the necessary underlying engine to create and sequence the desired decisions. Yes, you can learn from a full simulation, but I talk about creating first a scenario that requires the right decisions, and then tuning the experience to get an engaging series of important decisions (leveraging Sid Maier’s “a game is a series of interesting decisions” for my evil purposes). Your approach nicely combines a good design approach with a practical implementation plan. Thanks for sharing!

]]>
By: Chris Riesbeck https://blog.learnlets.com/2015/12/scenarios-and-conceptual-clarity/#comment-821896 Thu, 10 Dec 2015 19:51:11 +0000 http://blog.learnlets.com/?p=4695#comment-821896 A nice overview of the different types of simulation, Clark. As always, if I pop up, it’s to disagree somewhere. This time, it’s to argue that there’s something between contingent scenarios and full simulation, that I think is often better than real simulation. I called it outcome-driven simulation. I made a sketch for comparison, along the lines of yours: http://cs.northwestern.edu/~riesbeck/dynamic-scenes.png

The key idea is that you first author the meaningfully different situations, or scenes, that can occur in your domain. Each scene has its own coherent set of issues and possible actions. Then you author rules to select the most appropriate scene to go to, based on the actions taken. Structurally, you have a graph, richer than your linear scenarios, but not an unbounded tree like your contingent scenarios.

There’s room for a fair amount of complexity here. Rules can accommodate multiple action choices per scene and scene variables to hold details such as money spent so far. These provide a history and some individualization to the simulation path, but are not meant to encode a true simulation model.

In a true simulation, lots of things can result from certain choices, not all of them that interesting to explore. With the outcome-driven approach, you can maximize the exploration of interesting situations and challenges.

A few slides on the motivation here: http://cs.northwestern.edu/~riesbeck/outcome-driven-simulation.pdf

A commercial application: https://echristensenblog.wordpress.com/technology-solutions/dynamic-scene-adapter/

]]>