## Using Parametric Estimating

That'll be \$465 per metric ton.

How about \$125 per network drop?

These are all examples of parameters that can be integrated into a parametric estimate. Parametric estimating uses a mathematical model based on known parameters to predict the cost of a project. The parameters in the model can vary based on the type of work being completed and can be measured by cost per cubic yard, cost per unit, and so on. A complex parameter can be cost per unit, with adjustment factors based on the conditions of the project. The adjustment factors may have several modifying factors, depending on additional conditions.

There are two types of parametric estimating:

• Regression analysis This is a statistical approach that predicts what future values may be based on historical values. Regression analysis creates quantitative predictions based on variables within one value to predict variables in another. This form of estimating relies solely on pure statistical math to reveal relationships between variables and to predict future values.

• Learning curve This approach is simple: The cost per unit decreases the more units workers complete, because workers learn as they complete the required work (see Figure 7-2). The more an individual completes an activity, the easier it is to complete. The estimate is considered parametric, since the formula is based on repetitive activities, such as wiring telephone jacks, painting hotel rooms, or other activities that are completed over and over within a project.

EXAM TIP Don't worry too much about regression analysis for the exam. Learning curve is the topic you're more likely to have questions on.