Educational Simulations Overview

It’s been awhile since the last post. I’ve switched jobs at  Harvard Business School, moving from the educational technology group that builds educational technologies for HBS faculty and students on campus to Harvard Business School Publishing’s Higher Education Group. I’ll be helping launch new product initiatives, one of which is developing a product line of business simulations for distribution to the academic market.

Having had the privilege of working on the teams on campus that developed the Venture Capital Game and the Airline Pricing Game simulations as well as on the teams that enhanced and supported the UpTick financial markets and Beer Game supply chain simulations, I had some familiarity with team-based products in this genre. But what people think simulations are and how they are used differs widely depending on where you look and whom you ask, so this post serves as an introductory assessment of learning simulations in large part by recapping a great book on the subject: Clark Aldrich’s Learning By Doing: A Comprehensive Guide to Simulations, Computer Games, and Pedagogy in e-Learning and Other Educational Experiences (Wiley, 2005). This book helped me develop a framework for understanding and assessing simulations and I think it could do the same for others interested in this field.

I use the term “recapping” very loosely as this post doesn’t do his book justice despite my plagiarism of many of Aldrich’s key concepts. I focus here only on some of the higher-level categories and elements that define different aspects and types of simulations, and in particular I focus on those that matter most to me given my work with business simulations and given my personal interests and thoughts on the subject. The nice thing about the book is that each reader should be able to similarly draw out aspects most relevant for themselves. Clark Aldrich was the lead designer of the Virtual Leader business simulation (at the time he worked for SimuLearn and formerly for Gartner Group as well). He actually has authored another book called Simulations and the Future of Learning: An Innovative (and Perhaps Revolutionary) Approach to e-Learning (Pfeiffer, 2003). That book apparently details his team’s challenges as they built Virtual Leader. I mention it in case anyone might want to start with his earlier text.

Simulations Defined

The book begins by giving some rough definitions of simulations. I found this definition which I felt was a good short one:

Simulation involves building a dynamic model of a process or system, then performing what-if analysis to see how changes would effect the actual process. By mimicking its operation you can understand the system better, explore alternative strategies, optimize performance, and train personnel – all at a fraction of the cost and time it would take to experiment with the real system.

The book quotes Jane Boston from Lucas Learning explaining that simulations work best when used for:

• Understanding big ideas & concepts
• Depicting time & scale
• Situations where it is important to give people practice in decision making (because in real life it is critical or dangerous)
• Taking us to a time/place we are unable or unlikely to experience directly

HBS Professor Dorothy Leonard, in her book Deep Smarts: How to Cultivate and Transfer Enduring Business Wisdom, points out that information cannot become knowledge until it connects with something we already know. We have to constantly build receptors for such information based on our experiences. Simulations can be a highly efficient means of creating receptors, allowing people to make better sense of later real experiences.

Genres of Simulations

Aldrich outlines some of the major genres of simulations found in most corporate and educational products:

Interactive Spreadsheets

This is what Aldrich calls the “business process” simulations where users (often in groups or teams facilitated by a simulation expert) allocate resources along competing categories at successive turn-based intervals. It is this aspect of simulations that draws largely from Systems Dynamics theory – models are created consisting of causal maps of variables, feedback loops are identified, and then stocks (accumulations) and flows of variables are mapped. Simulating the model then consists of calculating the values of the stock variables by progressing the system through time. MIT’s original Beer Game is the most famous example of this, since it was developed then enhanced by systems thinking founder and member, Jay Forrestor and Robert Sternam, respectively. Software products like Forio’s Broadcast and iSee Systems’ iThink or Stella products allow modeling using systems elements. While some software vendors and consultants believe that this itself defines simulations, it is in fact only one genre.

Branching Stories

Branching stories involve users making multiple-choice decisions along an ongoing sequence of events, where those decisions impact the story or feedback. These types of decision trees allow for very specific location-based feedback. They’re also good at modeling conversations between people. I’ve seen many types of these simulations, from simple Skillsoft-type sims, to more complex performance simulations produced by business process firms like Accenture, to complex video-based simulations like SIMmersion that work with companies to create decision trees so customized and detailed that they are attempting to mimic a type of artificial-intelligence experience for the user. Many of the “soft-skills” simulations utilize this genre heavily since it lends itself to a more qualitative approach (vs. the quantitative systems theory approach).

Game-Based Models

The goal here is “making learning fun”. There is a lot of research that supports motivation as a key element of learning. At the very least the “fun” element increases time-on-task which, for quality products, should increase learning. There is usually a time-based, competitive element to these. Vendors include firms like Games2Train.

Virtual Labs/Products

Here users interact visually with accurate representations of actual products or environments, unencumbered by the restrictions of reality. Aldrich distinguishes virtual labs from products in that the former forsakes some of the fidelity of the latter in order to focus on a situation where the product is used. This genre is used heavily for training and marketing purposes. Medical simulations like those produced by SimQuest usually fall heavily into this category.

Aldrich details other genres as well, including role-plays, flight simulators, marketing mini-games, etc.

Picking the Right Genre

Aldrich then provides several helpful matrices for understanding the benefits and limitations of various simulation genres. For instance, basic pedagogical models from lectures to story-based learning can be mapped against linear-to-dynamic skills. You may require a linear approach if your goal is to provide learning as part of a highly-established process that leads to certification or legal requirements. You may want to create a facilitated simulation because it allows more flexibility but is obviously more costly to deliver and support. I won’t go into details here because we’re not ready to outline our design plans for our product line, but this section is worth reading for those still in the formative stages of determining how to create the greatest impact using simulations.

New Types of Scalable Content

Aldrich then goes on to detail new types of dynamic (vs. linear) content categories. He defines systems-based content as that which exposes users to complex relationships. This is the Systems Thinking approach described earlier. He then describes cyclical-based content as the combination of interface features that combine to impact the user environment and create outcomes. These are usually dealt with at the interface level and add a level of realism to the simulation that impacts the transferability of skills.

Genres like role-plays are “all cyclical all the time” – there is no systems element present at all. Flight simulators teach cyclical content through muscle memory. Again, Aldrich has helpful diagrams that help you start to frame various types of content and understand which genres of simulation are best at delivering each. The ultimate message is that with computer simulations we now have the capability to cover dynamic and linear content simultaneously – and together these form the equation for the fidelity of a simulation (its accuracy or loyalty to the original process it replicates).

Elements of Educational Experiences

Aldrich lists 3 essential elements that can be used to create successful educational experiences: simulation elements, game elements, and pedagogical elements.

Simulation Elements

Simulation elements enable discovery, experimentation, role-modeling, practice, and active construction of systems, cyclical, and linear content. They enable transferability of learning from the educational environment to reality by providing context alignment – a learning situation similar to a performance situation that will be encountered later. (Or as Dorothy Leonard puts it, they allow creation of receptors that can later capture information and turn it into knowledge.)  Simulations should be used when transferring skills is difficult or when users need to better understand the systems or cyclical nature of the content.

Game Elements

These provide familiar and entertaining interactions that drive up the time spent by the user on/within the educational experience. They do not directly support the learning objective(s) but are rather, as Aldrich puts it, the “spoonful of sugar that helps the medicine go down.” They can be used creatively to reward achievement within a larger learning context. Much research continues to be done examining the power of computer and video games, how they empower users as the key to success, how they instill the user with the understanding that mistakes are necessary on the path to success, etc. Leveraging the benefits of that genre (without having the budget to replicate it) is something the educational arena continues to struggle with.

Aldrich spends ample time later in the book detailing game genres: strategy games, first-person shooter games, management games, role-playing games, massively-multiplayer online role-playing games (MMORPGs), military games, etc.

Pedagogical Elements

These are the background material, case studies, user interface descriptions, etc, that present, annotate, and support the content. A powerful educational experience focuses on the learning objectives and frames these other elements around it to ensure maximum impact. And the pedagogy is the most critical element – you can’t decide what to model (simulation elements) or what to reward (game elements) until you define the learning objectives and craft the pedagogical elements.

Crafting the Solution

Again, Aldrich’s diagrams and tables help frame your perspective so you can help decide when simulations are beneficial and if so, which type or which elements will help you most. For instance, he details common challenges and then identifies the best simulation elements to use (use branching stories to reduce employee turnover, use interactive spreadsheets to increase understanding of complex systems, etc).

He also provides insights through extensive interviews into how successful simulation vendors approach design and development. Defining key issues, understanding the scope of what you are trying to achieve, determining how success will be measured – these are all critical elements of designing an appropriate solution. Aldrich quotes James Hadley of JHT, Inc.:

The goal of instructional simulations is to stimulate the creation of mental models within the learning by having them discover rules and principles through experimentation. Designers should constantly be asking themselves, “How do I help the learner discover this principle and then verify that they know it?”

Aldrich then gives his own insights into how users proceed through simulations:

• First they internalize the goals of the program, the models used, the time frame involved, background, etc.
• Then they experiment with the interface and interact with isolated, simple systems.
• Next they engage with the full simulation.
• Finally they engage in an un-chaperoned fashion.

Based on this understanding, the author has thoughts on the sequence that the designer should use when creating simulations:

• Design the simulation
• Add feedback
• Game it up
• Finalize the story
• Add relevant introductory material
• Add relevant background material

He goes into much more detail about the process, much of which will be familiar to those of you who have built software. But his points are valid that simulations are a special kind of software product that, when built correctly, employ various attributes and elements to create an immersive and effective learning environment.

I recommend the book to anyone interested in simulations, business or otherwise. You can keep up with Clark Aldrich's latest thinking on his Learning Circuits Blog which is hosted by the American Society for Training and Development.