The conference gave participants a wealth of information, assistance and a chance to meet and exchange ideas. Presenters included Systems Thinking leaders such as Mark Paich, Andy Ford, Lees Stuntz, and John Morecroft. Round table discussion topics were suggested by participants and ranged from world dynamics and climate policy, to community building and social policy, to the economic crisis.
The isee systems lab was open for business throughout the conference. STELLA and iThink developers and trainers were available to answer end-user questions, provide instruction in new features, and gather ideas for future versions.
A wide variety of fun activities – door prizes at keynote sessions, cocktails and hors d’oeuvres during poster presentations, and a cruise on Lake Champlain – helped cement connections that participants began to make during sessions and round tables.
Eureka!
Mark Paich delivers
opening keynote At their best, models and simulations provide “ah ha” moments for Systems Thinkers; real insight into the relationships between variables in a system. Those moments of insight fuel further exploration and even greater understanding and significantly improve the quality of decisions.
In his keynote address “Eureka – Maximizing Insights from Dynamic Models”, Mark Paich, Principal at Lexidyne, LLC, suggested that “most models leave money on the table.” That’s not the model’s fault. People just tend to spend much more time creating models then they do testing and conducting policy (or decision) analysis and finding deeper insights.
Mark suggested that real insights are found when Systems Thinkers use their models and simulations to look for synergy, timing effects, and robustness. Synergy occurs when the sum outcome of two alternative policies is greater than either policy alone. Timing effects indicate that the order of policy implementation is important; the outcome is better if policy A is implemented before policy B. Robustness tests policies in relationship to various uncertainties. If you find that policies hold up no matter what uncertainties are faced, then the model is robust.
Mark explained synergy, timing effects, and robustness using a model he created for a commodity manufacturer that was trying to implement process improvements. The manufacturer was never able to fully implement improvements and needed to understand what was getting in the way of needed change.
Mark started by engaging the managers in a discussion of process improvements that would have a positive impact – fairly mundane but presumably helpful changes like increased training on machine use, getting new machines, changing the repair schedule, concentrating on machine maintenance, etc. He used iThink to run sensitivity analyses on many assumptions about particular improvements, their combinations, and their implementation order.
The data generated by those analyses was exported to an Excel spreadsheet and then to SPSS. Scatter diagrams created with SPSS showed which improvements had the most positive impact and allowed Lexidyne to spot outlier improvements which could offer insight. An exploration of outliers considered implementing combinations of improvements and time sequencing effects. Last, they checked for robustness to make sure that various uncertainties wouldn’t get in the way of successful policy implementation.
Simply creating a model was not enough to give the manufacturer the insights it needed to begin improving processes. Only by analyzing simulation-generated data did Lexidyne find the insights the manufacturer needed.
How to Engage Newcomers in Systems Thinking
Teaching people a new way of thinking can be difficult, whether those people are business executives or grade school students. As is true with so many things, it’s easier for newcomers to catch onto Systems Thinking when they start off with exercises that are fun, familiar, and entertaining.
John Morecroft, Senior Fellow in Management Science and Operations at London Business School, uses Romeo and Juliet and their love to introduce business clients to Systems Thinking. At Making Connections, John described an engagement at Brasken in Brazil where he taught managers to explore the implications of entering the international chemical manufacturing market using Systems Thinking.
John Morecroft discusses
Romeo & Juliet model
Rather than learning about causal loops, stocks, and flows in the context of high stakes chemical manufacturing, Brasken managers modeled Romeo and Juliet, a familiar couple with an entertaining relationship. Client teams explored the question “Does the love Juliet has for Romeo and the love Romeo has for Juliet increase and/or decrease at the same rate” by creating a model and simulation using Systems Thinking basics.
Brasken managers immediately noted that no two of their models were the same and simulations produced a variety of results; love was flat, love was always rising, love was always falling. They discussed possible leverage points including angry families or another woman.
The exercise made it clear that Systems Thinking is a great way to understand and easily test leverage points outside of real-life pilot testing. It also illustrated that Systems Thinking is useful even in the absence of hard data. The lessons they learned in the entertaining world of fictional love modeling were then applied to the real-world challenge of marketing and distribution of chemical products.
Greg Orpen, principal of the Innovation Academy Charter School, uses games and reading to “get kids hooked” on Systems Thinking. Eighth grade math students play an epidemic game that simulates the spread of infection. Students move about the classroom exchanging pieces of paper that note their status – infected or susceptible. The epidemic spreads; slowly at first and then with alarming speed.
The rapid rate of infection is impressive and intrigued students graph data they collected during the game. This allows a meaningful discussion of S-shaped curves, causal loops (people are susceptible, get infected, infect other susceptible people, etc. etc.), stocks (numbers of infected and susceptible people), and related flows. With all that under their belts, students are ready and eager to move on to modeling and simulations that explore infection rates in different scenarios.
Once students, business people, scientists, and others have applied Systems Thinking to familiar, fun, memorable relationships and changes, they are ready to dig into their own issues, problems, and potential leverage points.
Collaboration Through Modeling
Thorough investigations of complex systems often require the collaboration of professionals or integration of data from more than one discipline. Collaboration though, is often difficult. Professionals bring their particular points of view, methodologies, and datasets and getting everything to work together introduces hurdles into already complicated issues.
Richard Palmer, Chair of the Environmental Studies Department at the University of Massachusetts Amherst, has been studying water resources since the early 1990s. His Shared Vision Modeling, developed with Bill Werick, has been applied to water resources projects and conflicts around the world.
Richard described a project that focused on drought problems in the Potomac River basin. The project involved stakeholders – specifically, people who cared about Civil War battle grounds that would be impacted by flooding or drought – and engineers. The two groups came to the project with different information, concerns, and styles of thinking.
Richard pointed out that “people in conflict are not typically early adopters of new ways of thinking.” Instead, they lean on their experience and areas of expertise. He described engineers as professionals who like intricacy, complexity, and details. The stakeholders were more interested in the protection of historic/sacred ground than the technical nature of any proposed water project. STELLA paved the way for a collaborative examination of solutions and outcomes that included the concerns of both groups.
Andy Ford shares
collaboration stories
Andy Ford, author of Modeling the Environment (the second edition will be published in 2009) and Professor of Environmental Science at Washington State University discussed a student project completed for his course in interdisciplinary modeling. His student, a fluvial geomorphologist interested in river restoration, created a model to simulate the impact of river restoration on salmon population and harvesting potential.
Of course, his model required a thorough consideration of the salmons’ river to sea migration and spawning process which falls squarely in the area of fish biology, not geomorphology. A model of salmon maturation, death/survival, spawning, and population change, also considered human investment in river restoration and accompanying natural restoration. It showed that, as the student expected, salmon counts increase as rivers improve.
What the student didn’t expect to find is the very fast payback in salmon population once a river is restored. He was impressed by the resiliency of fish and the rate at which the population bounced back. This “ah ha” moment was only possible when the fish life cycle model was combined with river restoration data. As the student himself pointed out, without joining information from the two disciplines, he never could have made the argument that river restoration is essential to quick gains in fish harvesting.
Modeling Soft Variables
You know how much a car costs. You know how many cars are made in an hour. You know how many sales people are currently employed by dealerships. But how does the market feel about those cars? How can you describe consumer buying behavior?
Soft variables – attitudes, opinions, feelings, behaviors – are at the heart of many Systems Thinking models. Kevin Austin, founder of Enzyme International a consulting firm based in Australia, has applied a key lesson learned as a medical researcher specializing in pain management to business management. That lesson – Ask the patient.
Doctors and nurses understand that by asking the patient if he is in pain and whether the medication dose is effective, they get the information they need to adjust the dosage. Kevin is helping businesses increase revenue by asking employees about things that cause them the greatest job “pain” and would most likely lead to their resignation.
Kevin and his Enzyme associates quantify soft variables (feelings of stress, disappointment over career advancement, deflation due to lack of recognition, etc.) by asking employees to vote on their biggest sources of pain. He collects similar data on their sources of job pleasure (chances to attend training classes, recognition, etc.) and perception of management effectiveness. All of that data is graphed in relation to actual attrition rate and then linked to financial performance. Staff attrition models and simulations project profit gains over time as various soft variables are addressed. Kevin pointed out that, often, improving on just one variable increases positive perception across the board effectively reducing attrition rates.
CAPT Brett Pierson
connects at breakfast
Equally intriguing was Captain Brett Pierson’s (Joint Staff, Warfighting Analysis Division, Pentagon) system dynamics model of the counterinsurgency doctrine created by General David Petraeus and social scientists. The model examines leverage points in waging and winning irregular, unconventional wars like those currently being fought in Iraq and Afghanistan.
Measuring the effectiveness of irregular warfare strategies and tactics requires measuring behavior, attitude, and relationships between Americans, the people they are fighting, and the people they are helping. Capt. Pierson noted that the Pentagon is most comfortable with hard data that is expressed in units. But what exactly is, for example, a “unit of motivation”? People need to “feel safe enough to go to work” but how is the feeling of safety noted? How is satisfaction with electrical services noted?
Capt. Pierson’s model notes the movement of people from one behavior stock to another as a means of using and understanding soft variables. More people going to work indicates that more people feel safe enough to go to work. So going to work moves a person from the unsafe to the safe stock. The main model combines five sub-models that consider security, governance, services, economic development, and host nation security forces to provide a thorough examination of leverage points (e.g. adding surge soldiers to security detail, increasing access to electricity) that could, in combination, result in victory.
Many other Making Connections presentations and round table discussions spent time on soft variables reinforcing their importance to many scientific and social questions and ability of Systems Thinking models to contain all variables – even those that are hard to quantify – in a system.
Keynote Presentations DVD
Making Connections Keynote Presentations are available on DVD for $100.
You’ll view 4 sessions that outline a wide range of Systems Thinking applications and techniques, and help you get the most out of your iThink and STELLA models.
|
 |
|
|
