Business Simulation Duration Issues
The drive to pack the most learning in to the time available means that a key requirement for choosing the right business simulation means that duration is a major constraint.
One of, perhaps the major, constraint when choosing and using a business simulation is the amount of time required for it to deliver good learning. Here, I discuss the problem and ways to shorten duration during design and use while packing the most learning into the time available - ensuring efficient and effective learning. As illustrated by the icon above, where, instead of trying to get a quart in a pint pot often we are trying to get a gallon in a shot glass!
The diagram to the left list the key design and use solutions. To get more information on an issue double click on it.
But before looking at the solutions, lets look at the problem.
A professionally designed business simulation will have a desired duration and (as shown below). If it is run in less time then it will be too complex and not provide effective learning and if given too much time the learning is participants time ill be wasted (and bored).
Too Complex = Ineffective
|Desired Duration = Effective & Efficient
Every business simulation has a duration that makes effective and efficient use of learner time and will be central to simulation design, choice and use. A time that ensures that participants think deeply about their decisions and fully analyse the results.
My empirical research (in conjunction with Imperial) showed that the amount of time required for a business simulation is very highly correlated with complexity (the number of decisions that learners make each period) . This research involved a range of business simulations designed for company training use, each of which had been used many times. The research explored the relationship between complexity & duration, how this affected realism, cognitive process and cognitive pressure and how this impacted andragogic effectiveness (adult learning).
Design & Acquistion Cost
A second issue with duration is how it impacts design time and hence design and acquisition cost. The diagram to the right shows for a broad range of my business simulations how how the model size correlated with model size. My research  showed that design time correlated with model size. My and other research also showed there was a link between business simulation duration and design time.
The first step to minimising duration while delivering good learning is a design approach that eliminates waste and ensures learning. The main ways to minimise duration during design are a learning focus, simplification and stylisation, the learning journey, cognitive processing, meta composition and the choice of delivery mode.
Learning Focus - matching to learning needs
The use of a business simulation can be seen as two overlapping sets - A+B (the learning provided by the simulation) and B+C (the learning required). Where B is relevant learning. The duration of the business simulation depends on the size of A+B. To minimise duration the size of area A must be small compared with the size of area B - the simulation model, its decisions and results must be just those that are relevant.
In other words, duration is minimised by basing design on learning needs rather than on merely replicating reality - a design process that involves appropriate simplification and stylisation, designing a "learning journey", cognitive processing design and meta composition.
Beyond matching to learning needs (leaving out irrelevancies) the simulation model, decisions and results need to be simplified and stylised just as a comic strip is a stylised and simplification of the real world. (In fact I see a business simulation as the mathematical equivalent of the comic strip !)
The Learning Journey - how the simulated experience evolves
A business simulation is not a static experience, the experience evolves as the simulation progresses - the Learning Journey. Why is this important? And, how do you design the learning journey?
The diagram to the right shows how workload (cognitive load) evolves during the simulation. The basic curve shows what happens if there is not a learning journey - initially work is hign but then tails off as the simulation progresses. Above this, there is economic calibration, ramped compexity and tutor interventions. These keep learners busy and increase the amount of learning done (cognition) - increasing simulation effectivness and efficiency without increasing the duration.
This means that as the simulation progresses the economic challenge evolves, changes and becomes more difficult. So, for example, participants may take over a company that is cash rich, unprofitable but with an opportunty for growth. As the simulation progresses and as sales grow profitability increases but cash outflows mean that liquidity becomes a major problem . Another example is my Teamskill Production Management simulation - initially during the a dip in seasonal sales here participants can learn how to manage the factory so that late, when seasonal sales peak they must schedule well and operate at peak effciciency. These economic patterns mean that the simulation is"easier" at the start (when participants are learning about their business) but then becomes "difficult" later forcing deeper thought and learning.
This means that as the simulation progresses issues (topics) are introduced. This is illustrated below for the Training Challenge Simulation . Here only Time Decisions are made every period, decisions about Time Restriction are made most periods but other decisions about Capital Investment, Fee Changing, etc are spread through the business simulation and are only made once. As illustrated below this limits total cognitive load (thinking time) and shortens duration (without reducing learning). In this example, progressively introducing topic reduced duration by at least 50%.
Not only does phasing decisions and results shorten duration but it improves effective learning and engagement. As illustrated above, introducing new issues (such as Capital Investment) means that there is a focus on and concentration on the issue at that point during the simulation. This allows the trainer to concentrate on the relevant issues.
Where the simulation is used as a Course Theme the decisions and results can be introduced to embed the last topic and introduce the next. In terms of engagement (as illustrated by this quote from Schneider Electric) "The continuous introduction of new ideas kept everyone interested, and the competitive nature of the simulation encouraged the sales mentality to try to win. Throughout the training, there were never problems with people checking email, voicemail and so on. Most would voluntarily work through lunch on their (virtual) business."
Ambiguity determines the amount of thinking (cognitive processing) and its level must be determined by the amount and depth of thinking about individual decisions and results .
For learning to take place participants must unravel the links between decisions and results. If the links are too ambiguous then the learners cannot make sense of their decisions and the impact. Equally, if the links are too unambiguous they are trivial and participants do not thnik deeply enough and the simulation is a trivial waste of time.
To optimise duration and learning, principle decisions and results should be to the right of the ambiguity band above with secondary decisions to the left as this ensures that participants spend appropriate time thinking about the decision or result.
Granularity determines the range of choices for decisions and the amount of detail shown in results. Decision granularity is determined by the amount of choice. For example a granular decision might be promotional expenditure where one has freedom to choose any amount, a mid-granularity decision where you can select from a list of options and a low granularity decision is where you have two options (such as create website (yes/no)).
For results (as shown below) granularity determines balance of processing by the simulation software with the amount of processing need by the participants and through this the amount of time spent by the participants processing the results.
To the left of the granularity spectrum, participants are provided with raw data that they need to analyse and interpret and although this takes time, it necessitates significant thought (cognitive processing) and so elicits learning. As one moves to the right duration is reduced as the need for processing is lessened but this is at the expense of cognition.
To optimise duration and learning, granularity like ambiguity should be decided based on the need for and importance of cognitive processing.
Here we are looking at the way decisions, the models and results link together and the complexity of these links. Meta composition impacts cognitive processing and if not done appropriately will make it hard (or impossible) to understand the way decisions impact results and this adds to duration (and may lead to confusion and no learning). Equally, if the link between decisions and results is too obvious, cognitive processing will be trivial and lead to ineffective learning.
The example above defines the key causal links for a simple simulation but even so it will be difficult to separate ot the causal links
My business simulations either where each team operate separately and where the participants enter their decisions directly into their own computer or involve several teams interacting in the same market places (where the tutor submits decisions to the simulation model).
Direct Use Simulations are where teams enter decisions asynchronously and are, generally, the simplest simulations. my experience running such simulations many hundred times showed that there are two characteristic patterns:
Consequentially both types of participants spend the same amount of time overall. If the decisions are results are shared between teams this tend to speed the slower teams.
Tutor Mediated Simulations are where teams interact (compete) in the same market places and are, generally, the most complex simulations.. Here decisions are made synchronously and these is a short delay while decisions are entered, results printed and returned to teams. Besides taking time to enter decisions, the decision schedule must accomodate slower teams.
Overall, Direct Use Simulations have shorter durations than Tutor Mediated Simulations.
Learning to use the simulation software, entering decisions, result display and (if necessary) recovering from problems is time wasted. The relative importance of these depends on whether the simulation is used by the participants (Direct Use Simulation) or only by the trainer (Tutor Mediated Simulations). With Direct Use Simulations participants will have no prior knowledge of the software and so it must be easy to use. The nature and manner of use of Direct Use Simulations means that decision entry, and recovery are less of an issue. With Tutor Mediated Simulations as the only person using the software is the trainer and he or she can become familiar before hand learning to use is not a major concern. But decision entry, result display and problem recovery can be as, typically, this type of simulation involves all teams in direct competition. All these factors must be addressed to minimise duration during use.
Crucial to minimising duration and ensuring learning is the role of the tutor and how he or she manages the learning process. Comparing business simulation use with and without tutor support suggests that (for similar learning) durations without a tutor can be three times longer.
However, there is the need to provide information ot support the tutor's learning management the simulation needs to provide special reports and reflection triggers.
The special tutor's report (to the right) compares three team's efficiency. None of the teams are wasting money on promotion or idle capacity but differ in terms of sales forecasting and sales losses with A-team over stimulating demand significantly. The tutor can use this information to discuss forecasting and scheduling with A-team.
A second way to support learning management is to produce comments about problems, strengths and weaknesses that can either be fed directly back to the participants or fed back via the trainer.
Learning management ensures that the simulation is tailored to participants' needs and, by constraining, the mistakes they make ensures effective and efficient learning - packing the most learning in the time available.
Here I'm discussing the tutor responding to participant initiated questions from participants. Commonly, to reduce the time take to analyse results the amount of information provided is limited. For example, results may just show unit sales and closing inventory but not detail how these were calculated but occasionly (rarely) participants will ask how these were calculated. Here time (and tutoring stress) can be saved where a report is available that details the calculation.
(Facilitation support reports also play a major role during simulation design as they ensure the validity of calculations).
Participants get very involved and this encourages them to work through lunch, coffee breaks and, if a residential course, late into the night. For example, on an Assessment Centre, coffee breaks were carefully scheduled and I remember well when a group was reminded that they should be taking the break, one assessee stated emphatically "Coffee Breaks are for WIMPS". Thus coffee, meal brakes etc. are free time that you can time table into simulation without robbing from scheduled time. (It is worth mentioning that if you are running a business simulation on a residential course over night, you do not let the participants know your room number - otherwise you will get phone calls in the middle of the night with questions - some thing I learned the hard way!)
Neither the participants or the tutor should have to waste time learning how to use or using the simulation software. This means that they should be taken through an intuitively clear process (illustrated below),
get very involved and (as illustrated above) work through lunch, coffee breaks and, if a residential course, late into the night. For example, on an Assessment Centre, coffee breaks were carefully scheduled and I remember well when a group
Let face it - things can go wrong a team or the tutor might enter a wrong decision, the printer may jam or run out of ink or or the computer loses power at an inopportune time. Not only does it take time to recover from errors but during this time participants are becoming more and more agitated and disaffected. This means that error recover should be seamless and quick.
It is standard data processing practive to "journalise" data - to take save copies of data to the hard disc regularly. As illustrated above, at the start of the siulation process, the data saved the previous period is loaded. After simulation, the current period's data is saved. This means that a copy of each period's data (and where approriate, the period's decisions) are saved to disc and is available if it is necessary to revert to an earlier period.
Wrong Decisions: Occasionally a wrong decision is entered because the tutor misreads a decision form or because participants misunderstand decisions. Where this happens it is necessary to revert to the earlier period, correct the error and re-simulate. When data is journalised this takes a minute or so. Where it is not journalised it may be necessary to restart the simulation, re-enter all the decisions and re-simulate all the periods.
Printer Problems: In my experience, the printer is the greatest problem - it jams or run out of ink at the worst time. As a consequence, as show above, data is saved before the results are printed. This means that if there is a printer problem there is no need to re-enter decisions and re-simulate. Rather, all that is necessary is to deal with the printer and print.
Power Supply Problems: Happily, these are very rare. On one occasion a client came to me because the simulation they had had to have the computer on for the whole of the course - other wise all the data was lost right back to the beginning. Unfortunately, despite notices, the cleaning crew switched of the computer regularly. Proper journalisation protects against this and even if power is lost while a period is being simulated it is easy and quick to revert back to the start of the period, reload decisions and resimulate.
Decision Entry is both a major bottleneck and source of errors - problems that are addressed by using prepopulated templates and decision checking.
Pre-populated Templates (such as the one on the right) not only define what decisions can be made but also save entry time as only changes need to be entered (the entries in red).
Decision Checking involves parsing decisions and after all decisions are entered screening the decisions to flag illegal and unusual decisions.
 Jeremy J. S. B. Hall & Benita M. Cox (1994) Complexity is it Really that Simple Developments in Business Simulations and Experiential Exercises Volume 21 - download full paper here
 Jeremy J. S. B. Hall (2006) Computer Business Simulation Design: Novelty and Complexity Issues Developments in Business Simulations and Experiential Exercises Volume 33
 Jeremy J. S. B. (2008) Corporate Cartooning: The Art of Computerised Business Simulation Design, Developments in Business Simulations and Experiential Exercises Volume 35
 Jeremy J. S. B. Hall (2012) Designing the Training Challenge Developments in Business Simulations and Experiential Exercises Volume 39 - download the full paper here
Most recent update: 02/01/15
Hall Marketing, Studio 11, Colman's Wharf, 45 Morris Road, London E14 6PA, ENGLAND
Phone +44 (0)20 7537 2982 E-mail firstname.lastname@example.org