Question

A cost estimating relationship (CER) describes the cost as a function of design variables. What are the basic steps in developing a CER?

Answer 1

The Cost Estimating relationship (CER) is not a quantitative framework. It is technique which uses appropriate techniques to effectively help quantify a relationship between cost or price and the independent variables.
There is a 6 step process which is used to develop CER

Step 1:
It is necessary that the dependent variable be defined properly so that the data assimilation and gathering can be done in an easier way. The dependent variable is the price or cost function which is to be predicted. Its choice depends on the purpose of the cost function. It may also be called the response variable.

Step 2:
These may be referred to as the independent variable. It may also be said to be independent, or the explanatory variable. The change in this independent variable i.e. cost driver would cause a change in the total cost of the activity. A few examples of cost drivers are viz. Machine Hours, Number of units, No. of orders, No of production runs, Direct labour hours.

This cost driver should make some economic sense and should be calculated very accurately as it would have a direct impact on the cost. The following factors should be kept in mind while selecting the same.
1. The variables selected should be such that they may be quantitatively measurable.
2. Make sure that the variable has some historical data availability without which it is futile to analyse.
3. For a choice of developing CER, the performance characteristics choice would always be preferred above the physical characteristics as performance characteristics are usually known before design characteristics

Step 3:
This steps involves collecting the data. This is a time consuming process. It is imperative that all the data be checked and that all observations are relevant and free of unusual costs.
A lot of historical data and sufficient past observations would help derive an acceptable cost function. This should not be adjusted to shows us any circumstancial change that may happen. e.g. changes in equipment type, If there are any changes in price levels due to inflation etc. The time period to meausre the cost driver and the cost should ideally be the same.

Step 4:
This step now involves plotting the scatter diagram. This will give and establish a relationship between the cost driver and the dependent variable. It will also give us a indication visually as to whether a linear function of the cost can approximate cost behavior. Outliers would definitely be highlighted in the scatter diagram.

Step 5:
After exploring the scatter diagrams it would be finally convenient for us to focus and select the best cost driver which would help in predicting the dependent variable. This cost driver would best help in predicting the values of the dependent variable. A higher correlation of the independent variable with the dependent variable helps us in the selection of the best predictive tool. It is important to note that the values of the independent variable be available to make proper and timely estimates. The various methods to estimate cost function include i.e. High Low Method, Engineering method, Regression Analysis, Time Series Method, Account Analysis, Simulation Analysis

Step 6:
To test whether the function is reliable we would have to do the following three main tests. These include: Logical Relationship Tests; Goodness of fit tests; Specification Tets (Tests of the assumptions of the model)