Packer, David W. with William Glass-Husain, "Designing Interactive Multi-User Learning Laboratories", 1997 August 19-1997 August 22

Online content

Fullscreen
Designing Interactive Multi-User Learning Laboratories

David W. Packer, GKA Incorporated (dpacker@gka.com)
William Glass-Husain, Powersim Corporation (wilgl@powersim.com)

Introduction
Experiential activities have long been recognized as a powerful way to demonstrate the

significance of a systemic view to learning. Simulations in which participants become
part of the feedback structure create an environment where people can encounter many of
the challenges and frustrations of life in a complex system. The “Beer Game,” “Fish
Banks,” and “People Express,” have become effective, popular ways of enhancing
learning, through active participation in a system less mysterious than the real world but
with enough of its characteristics to provide a rich experience. Such participation
typically involves both playing a simulation and debriefing the experience to expose the
role of underlying structure. The whole activity is often called a Learning Laboratory.

This paper deals with design and development of multi-user learning laboratories,
consisting of a simulator (system dynamics model) that accommodates multiple decision-
makers, an interface that creates a “realistic experience,” and workshop facilitation that
aids the debriefing activity.

This paper is largely experiential as opposed to theoretical. It stems from a number of
projects we have done in this area, largely at GKA Incorporated, and also from our
growing recognition of a set of recurring fundamental design issues. In the paper we
attempt to organize and articulate our thoughts about the design of learning laboratories
and the factors that lead to learning effectiveness.

A Real Example
In the Spring of 1996, GKA worked with a U.S.-based manufacturing firm that, although

over a century old, had in the last decade undergone significant expansion. A new CEO
had recently been hired who had a strong vision of turning the company from a
manufacturing to a global market driven firm. To communicate that vision to his
company he invited his top managers from each division to participate in a weeklong

conference focusing on company strategy and performance.

The conference began with an introduction to a custom-designed simulation of the firm’s

industry and its major global markets. The hundred managers were soon split into ten

cross-divisional teams and sent into separate rooms that held nothing but a flip chart and
a computer with a tiny monitor. Each team represented a competing company and each
meal period represented a year. During the week the teams pored over computer
printouts giving current information on sales, customer satisfaction, product quality, and
finances. They eagerly crowded around posted bulletins of newspapers with light-hearted
articles on company successes, failures, and industry trends. The teams created their own
decision-making processes, filling the walls with posters showing data and business plans
that they studied before they typed in their decisions on investments in capacity, staffing,
and marketing. And each day at 8 am., 1 p.m, and 8 p.m. they clicked the “Send
Decisions” button on their computer in order to have their decisions counted before the

year advanced.

Although the simulation was a key part of the learning lab experience, it did not exist by
itself. The conference included a number of presentations given by corporate
management on company issues. GKA also ran workshops on systems thinking issues,
including the beer game and some work with causal loops. Most importantly, the week
ended with a series of debriefing activities that involved teams making presentations on
company strategies as well as some causal loop analysis of important competitive issues.
The grand finale included a banquet and prizes.

Design Trade-offs
t In creating a learning lab such as this one, there are many lnterf
| decisions that must be made. These decisions include both | Sophistication

bigger ones (what type of issues should the simulation , rr be tal == jienty
address) and smaller ones (what type of computer 2D ic Complexity
equipment to use). Three choices that drive many of these

design decisions are the levels of detail complexity, dynamic complexity, and simulator

80

|
|
e

:
:
i
H
i
1
4
i

interface sophistication. These can be summarized on a three dimensional axis as shown.
These choices apply to all simulations and learning labs, but are particularly significant in

a multi-user simulation.

A high degree of detail complexity is characterized by the presence of a significant
amount of dis-aggregation and a large amount of detail. High detail complexity
significantly adds to the challenges of managing a project as it adds to the time spent on
design, development, data collection and testing while developing a learning lab. A very
important factor is the likelihood of pressure from clients to include large amounts of
detail under the often mistaken assumptions that more detail automatically leads to more

realism and more realism leads to more learning.

On a different axis, a simulation or learning lab with high dynamic complexity contains a
model with a robust underlying structure with a significant amount of feedback. These
models are usually of high quality and are based on extensive research and testing with
regard to similar companies and industries. There is a wide range of possible behaviors
that may emerge from the simulation. Consequently how to design the workshop to
enable participants to experience and learn about those behaviors is an important
consideration of using high levels of dynamic complexity.

The level of interface sophistication involves the visible aspect of the simulation. It
includes the type and manner of input by the user and also of the presentation of
simulation results. Increased sophistication in the interface can lead to a higher degree of
ease of use (for example with on-line help) as well as make the simulator more appealing
(for example, with multi-media and other technological glitz). Note that developing
elaborate interfaces requires different skill sets than traditional system dynamics efforts.
A strong focus on interface sophistication can significantly add to the cost, and may in
fact get in the way of the learning if the interface becomes too video game-like or
overwhelming. Our experience has been that there are significant diminishing returns on
investment in the interface sophistication. A simulator with a compelling story and a
number of interesting dynamics can often add to the engagement of the participants as

81

|
|
{
|
:

much as an elaborate multi-media simulator interface, yet provides a much greater

learning experience.

There are many different trade-offs between these design drivers, Note that the learning
lab that is “Ideal for Systemic Learning” has just enough detail complexity to make the
simulation tangible, just enough interface sophistication to make it easy to use,
interesting, and fun, and has a significant amount of investment in dynamic complexity

(helping to create those dynamic “ah-ha’s”).

Design Process
The process of designing a learning lab is an iterative one in which each step generates

the information needed for the next step but also feeds back and helps to refine the

previous one.

Typically, the design proceeds through these five steps:

1. Articulating the objectives of the learning laboratory
Identifying the systemic and dynamic issues
Developing a model
Designing a simulator

~ &

Designing the learning laboratory workshop

Articulating the Objectives

The first step in designing a learning laboratory is to articulate the objectives for the
participants. The ultimate design of the simulator and workshop will vary considerably
depending on the purpose of the learning laboratory. Some possible objectives are to:

1. Learn how to operate in the business. The simulation is then used to develop
familiarity with terms, processes, decision factors, and information sources used by
employees in the real organization.

2. Learn dynamic lessons such as the impact of feedback, delays, and other systemic

issues on the business.

82

3. Provide a forum for development and testing of new strategies on all or part of the
business.
4. Help employees gain holistic view of the business system by acting as “President for a

Day.”

These objectives have considerable interplay with the level of detail in the simulation, the
sophistication of the interface, the duration of the experience, and the content and manner

of other workshop experiences.

Identifying the Issues
Producing a useful, relevant learning lab requires that the simulation and workshop

address on the core concerns of the client organization. In particular, issues of a systemic
or dynamic nature that involve long time delays, feedback, and connections across
multiple parts of the business environment are particularly well suited to be examined in
learning labs. Conceptualization techniques such as FASTBreak™ can be used to assist
a cross-functional group of managers in the client organization in the process of
generating important issues, identifying areas in which to focus, and coming up with
variables that can then be used to develop causal loop diagrams and system dynamics

models.

Multi-user learning laboratories are particularly effective at highlighting issues involving
competitive strategy. While single-user microworlds such as People Express
Management Flight Simulator have long included a competitive sector, interactions in
these simulators between participants’ decisions and cornpetitor reactions occur on a
highly abstract level. On the other hand, multi-user simulations offer highly visible, often

emotionally strong lessons of the interplay between company, competitor, and market.

A recent example involved a multi-user simulation of the fast-food restaurant industry. In
the third year, team “Charlie Chicken” made a bid to gain market share by offering a
steep discount to customers with great fanfare (expressed both verbally and by choosing
from a list of pre-set promotions available in the simulation). Over the next two years,

each of the other four teams also offered similar discounts. During that time “Charlie

Chicken” not only lost the market share it had gained, but all the teams involved suffered
significant drops in profit. At the mid-day debriefing we discussed what had happened’
and brought out “Escalation” systems archetype. Seeing this dynamic play out back and
forth across a classroom during the course of a few hours not only dramatized the
relevance of this competitive issue but also provided a context in which we could teach

the more general issues that were involved.

Developing the Model
The models that underlie the simulations used in our learning laboratories are developed

with the system dynamics methodology using one of the standard SD modeling software
packages. Developing a dynamic model of the business environment that is separate
from the simulation interface makes the model more visible and hence easier to create,
modify and test.

Typically we begin the model by combining generic system dynamic structures (such as
inventory, capacity, and market stock-flow sectors) and customize them to match the
objectives and issues identified previously. We then work with a small group in the client
organization who have a broad understanding of their business and market to further
refine and validate these assumptions. Often, the development of the model and the
design of the simulator interface will overlap, allowing for the progressive stepwise

refinement of the overall simulation.

Designing the Simulator
In designing a simulator, it is helpful to think of it as being composed of four modules:

e User Interface: An attractive, convenient software that can be used to enter decisions
and see simulation results. May include multiple windows, notebooks, on-line help,
pop-up messages, and multi-media displays.

© Numerical Information: The set of current/historical data representing the state of the
business system.

« Textual Information: On-line help, pop-up messages and other textual descriptions

that provide context for the simulation.

System Dynamics Model: Contains the set of mathematical relationships that make up
the business, competitive, and market system. Calculates the new simulation results

each time period, based on user decisions and the old results.

Elaborate simulators may use separate tools in developing each of these modules. For
example, Powersim may be used to build the model; Visual Basic may used to develop
the user interface; Microsoft Access may be used to develop databases for the numerical
and textual information. (When distributing such a simulation the end user does not
actually need all these tools; run-time versions for each of these are available free and can

be bundled with the simulation software.)

Simpler simulations may combine these modules together. For example, many single-
user simulations have the numerical and textual information integrated into the model and
user interface. In addition, for ease of distribution, the system dynamics model may be
translated from its native format into mathematical equations that are embedded directly
in the user interface software. Taking the opposite approach, most of the current modeling
software programs contain some interface developing capability that can be used to make

prototype simulators or even simple interfaces for final versions.

Creating a Workshop that Enhances Learning
The workshop encompasses both experiencing the simulator and stepping back to reflect

on the experience and its meaning. The role of facilitation is, we believe, a critical factor
in learning. Facilitation should, for example, encourage (even demand) theory building
and testing that provides difection for experimenting with coherent strategies, assessing
outcomes, and discussing perceptions jointly among participants. It moves the experience
away from a cut-and-try video game approach with dubious learning potential toward one

that develops understanding at the systemic level.

On a final note, we believe that to create a successful learning laboratory, designers must
help participants to move their mental models from Winning to Learning. This includes
encouraging far-out approaches that test the limits of the situation creating an exciting

atmosphere where players want to make mistakes in order to truly learn!

85
References

Diehl, Emst. 1990. Participatory Simulation Software for Managers: The Design:

Philosophy behind MicroWorlds Creator. European Journal of Operations Research
59 (1): 203-209.

Kim, Daniel H. 1992. Learning Laboratories: Practicing Between Performances. Systems
Thinker 8 (3).

Kreutzer, W. Brian. Building Learning Laboratories to Create More Effective-Distributed
Decision Making. 1995. In Learning Organizations, ed. Sarita Chawla and John
Renesch, 216-227. Portland, OR: Productivity Press.

Kreutzer, David. FASTBreak™: A Facilitation Approach to Systems Thinking
Breakthroughs. 1995. In Learning Organizations, ed. Sarita Chawla and John
Renesch, 228-241. Portland, OR: Productivity Press.

Meadows, Dennis. Fish Banks, LTD Game Kit. Available from Laboratory for
Interactive Learning, Institute for Policy and Social Science Research, Hood House—
UNH, Durham, NH 03824.

Senge, Peter and William Isaacs. Overcoming Limits to Learning in Computer-Based
Learning Environments. 1990. System Dynamics Group Working Paper D-4185.
Sloan School of Management, MIT, Cambridge, MA 02139

Simons, Ken. 1993. New Technologies in Simulation Games. System Dynamics Review
9 (2): 135-152.

Sterman, John. 1992. The Beer Distribution Game: An Annotated Bibliography
Covering its History and Use in Education and Research. System Dynamics Group
Working Paper D-4281-1. Sloan School of Management, MIT, Cambridge, MA
02139

People Express Management Flight Simulator™. 1990. Available from MicroWorlds,
125 CambridgePark Drive, Cambridge, MA 02139. Software.

Powersim®, 1995. Available from Powersim Corporation, 1175 Herndon Parkway,
Suite 600, Herndon, VA, 20170. Software.

86

sia

Metadata

Resource Type:
Document
Rights:
Date Uploaded:
December 18, 2019

Using these materials

Access:
The archives are open to the public and anyone is welcome to visit and view the collections.
Collection restrictions:
Access to this collection is unrestricted unless otherwide denoted.
Collection terms of access:
https://creativecommons.org/licenses/by/4.0/

Access options

Ask an Archivist

Ask a question or schedule an individualized meeting to discuss archival materials and potential research needs.

Schedule a Visit

Archival materials can be viewed in-person in our reading room. We recommend making an appointment to ensure materials are available when you arrive.