(This was composed in response to a colleague looking for guidance on how to develop simulated software training. If any of you can use it in your practice, feel free...any questions, let me know.)
I've been at the "simulating software without harm to the actual application" for a number of years now. We happen to work with an Electronic Health Record (EHR) and, not only out of fear of altering a patient record but, also, fear of HIPAA ramifications, do we have to create this 'imaginary reality'. We do it in a very streamlined development process, using Adobe Captivate. We create CBT and, for some of the more limited users of the application, CBT-alone training is all they require to get up and running (as their access might not be as impactful as others). For said other users, though, we offer a hybrid approach to our training, of both a synchronous ILT/CBT session for new users.
This hybrid approach takes the best of both of worlds and mashes them together. The CBT delivers the more foundational/basic elements and, once the users have completed the CBT elements, we regroups as a class, discuss, and delve into deeper topics. We've found this approach to be beneficial for a number of reasons. First off, it allows for individualized questioning on the basic elements, rather than one question stopping the whole class. For example, if it was a standard ILT experience, and there were fifteen students, one student raising their hand puts everyone else's learning on hold till that learner's question was answered. By using the CBT to get the learners' feet wet, questions can be addressed individually (which is more comfortable, I've found for adult medical professionals).
Additionally, by using CBT straight away, it puts MIH ("Mouse In Hand") and gets learners cued in straight away to the fact that this is a computer application. Too often, those of us with the knowledge fall in love WITH our knowledge and try to train it via standard lecture. By allowing the users to jump right in, they start making connections right away, where they might not have before. Then, when the group 'regroups', they already feel experienced in the topic at hand and can begin to discuss more finite details of the application. And, really, when it's all said and done, when you're dealing with an adult audience, you're looking for pragmatic, you're looking for practical - let's not hide it, let's start clicking! (Side note - The adult audience point is also why we don't go with 'strictly CBT', as they still need that 'traditional' experience to feel comfortable).
Finally, there's the time factor. By alleviating the 'group questioning' referenced above, and the typically 'tangential' nature a lecture based class would take, we've been able to cut typical class time in HALF (which, for anyone well-spoken on the art of the ROI, you know that's pretty huge). By doing so, not only have we saved human resources, time, and, therefore, money, we've been able to double, and in some cases triple, the number of class offerings we as a training team can make. The ROI here touches on so many levels.
Long story short, as I tell a lot of folks, this solution works and works HUGE for us. There's no such thing as a 'universal solution; to training needs (let alone application simulation training needs). I can say with utmost certainty that in a similar organization, looking to tackle a similar task, such an approach would be worth, at least, taking a long look at. The money, time, and personnel saved are priceless, but, then again, so is an effective training product that allows learners to get in, get out, and get using.