ACADEMIC INTERVIEW

Professor Vedran Lelas, Georgia State University

Bio: Professor Vedran Lelas, currently a Visiting Assistant Professor at Georgia State University, teaches generalized modeling techniques and operations management courses at the undergraduate level. Prior to this position, he was an Assistant Professor in Residence at the University of Connecticut, Stamford, and worked as a risk constants and analyst for banking, government, and manufacturing companies. Professor Lelas earned a Ph.D. in Management Science and Information Systems from The University of Texas at Austin, and a BS in Mathematics from the University of Zagreb, Croatia.

Q: What classes are you teaching with Crystal Ball?

Lelas: I am currently teaching a course titled Generalized Modeling Techniques with Applications. This semester, I teach four sections of the course for the approximate total of 160 students.

Q: How are you teaching the software in your classes?

Lelas: Simulation is introduced using a simple NPV model and Excel functions first. Then I expand that there is an easier way to accomplish the same thing with a simulation add-in. I talk a little bit about add-ins in general. Some of the students already have a 140-day Student Edition of Crystal Ball if they purchased optional text book (Moore-Waterford: Decision Modeling with MS Excel, Prentice Hall). A few students use seven-day evaluation versions of the software.

Next, using a NPV example, I demonstrate how to set up and run the simulation. I discuss only four buttons on the CB toolbar:

1. Define Assumption, which is used to set up the input distributions.
2. Define Forecast, which is used to designate the output cell.
3. Run Preferences, which is used to set the number of simulation trials.
4. Run/Reset Simulation buttons, which is used to run the simulation.

The example used is a simple profit or NPV model, where there are number of input parameters (prices, costs, etc..) modeled with different statistical distributions, the profit or NPV is the forecast, and we run it for 500 or 1000 trials. The resulting profit or NPV distribution is then examined using standard statistical techniques (mean, st dev, st error, confidence interval) and question such as what is the probability that the profit or NPV will be negative is answered.

Q: How does Crystal Ball help your students?

Lelas: Hopefully, it helps them better understand how uncertainty can be modeled and utilized in business applications.

Q: Why do you like the functionality of Crystal Ball?

Lelas: Its fairly easy to use, has sufficient number of different input distributions, and all you need to describe a simulation output (stats and frequency chart)