I just read ‘SimSketch: Multiagent Simulations Based on Learner-Created Sketches for Early Science Education’ by Lars Bollen and Wouter R. van Joolingen (http://ieeexplore.ieee.org/xpl/articleDetails.jsp?reload=true&arnumber=6471713)
I also build a few models and looked at the sample models. Software is available at http://modeldrawing.eu/our-software/simsketch/
For those familiar with our Behaviour Composer (modelling4all.org) I see three major differences:
+ Simpler interface (their target users are 8 to 12 year olds)
+ Sketches used (instead of pre-defined shapes as in NetLogo) — could/should be extended to work with both sketches and pre-built shapes. Together with more drawing tools simulations can be created that look very nice and, when desired, ‘informal’. This use of sketches was perhaps inspired by the article in Science ‘Drawing to Learn in Science’. Related (but significantly different are drawing programs with physics. e.g. http://crayonphysics.com/, http://www.algodoo.com/, and http://www2.futurelab.org.uk/resources/documents/project_flyers/moovl.pdf. This idea was pioneered by Randy Davis at MIT. These, unlike SimSketch, do not have the notion assembling behaviours for model elements.
– Their ‘micro-behaviours’ are opaque and not user definable. Also the set of available micro-behaviours is well-suited for young users but more capable and customisable ones are needed to create a wide range of models.
It is interesting to consider what SimSketch could become if the behaviours could be defined and shared in a user-friendly manner by ordinary users.