| Useful Techniques Example Models |
|
|
|
|
Model: How to Build a Frequency Chart
From: Decisioneering
Detail: Looking for a way to visualize how a frequency
chart is built? This worksheet contains macros designed to aid your
understanding of Crystal Ball frequency charts.When you click the
single step button, you see one iteration of a simulation. Crystal
Ball randomly selects values for Income and Expense, Excel recalculates
Profit, and the the Excel histogram (a frequency chart) is updated.
This example is limited to 100 steps. |
 Download
For:
Crystal Ball
Level:
Simple |
|
|
Model: An Executive Summary
From: Paula Jennings, Managing Director, Decision Insight
> More about Decision
Insight, Ltd.
Detail: A model may be used by a number of stakeholders
ranging from those
who are interested in the detail to those who want just a snapshot.
Through my experience of working with this range of needs, I have
found that the inclusion of an executive summary is invaluable.Tips
when developing an executive summary:
- Link all summary information back to the working model
- Include only key performance measures of interest
- Highlight key financial and operational assumptions on which
results are based
- Identify key inputs and assumptions used in sensitivity analysis
- Use graphics, helping to identify trends over time
- Position the sheet near the front of the model; if the working
model is large, consider an automated extract process to generate
a single page into a separate file
- Protect the sheet from alteration, using Excel's protection
feature, since
the summary is only for viewing
|
Download
For:
Crystal Ball
Level:
Simple |
|
|
Model: Auditable Excel Spreadsheet Tips
From: George Laking, Clinical Research Fellow, Cancer Research
UK PET Oncology Group, george.laking@csc.mrc.ac.uk
Detail: Many authors of large models may feel a little
"lost" when returning to
their Excel spreadsheet after a period of time (e.g., weeks to months)
away. The author suggests a number of tips that help alleviate confusion.
Updated 1/21/03. |
Download
For:
Crystal Ball
Level:
Simple |
|
|
Model: Creating a Bimodal Distribution
From: Decisioneering
Detail: This model describes how to create a bimodal distribution
by combining multiple distributions. |
Download
For:
Crystal Ball
Level:
Simple |
|
|
Model: Creating a Switch Cell for Ranges
of Assumption Values,
Part 1
From: Decisioneering
Detail: This model is related to a
CB Tip of the same name located in our technical support section. |
Download
For:
Crystal Ball
Level:
Simple |
|
|
Model: Creating a Switch Cell for Ranges
of Assumption Values,
Part 2
From: Decisioneering
Detail: This model is related to a
CB Tip of the same name located in our technical support section. |
Download
For:
Crystal Ball
Level:
Simple |
Model: Using Dummy Variables in a Regression Model to Determine if Seasonality Exists in Order to Select the Correct Time-Series Forecasting Technique
From: Mark E. Zwolak, CFM
Detail: When using CB Predictor, one must select whether the time-series forecast to be used should include seasonality or not. Rather than anecdotally guessing, is there a quantitative technique to ascertain whether or not there is seasonality in the data set so that the appropriate forecasting technique can be utilized?
If a seasonal time-series forecasting technique is utilized for data that is not seasonal, one might get an incorrect forecast. This example presents results in an additional 1% degree of error in aggregate, however the monthly error rate is much higher. Using historical data, one can assign dummy variables to each of the months over several years and then perform a multi-variable regression to ascertain the p-value
of each of the months to determine statistical significance and hence, seasonality. |
Download
For:
CB Predictor
Level:
Moderate |
|
