Yes. Let’s do regression analysis today. This post comes up consequent of a humble request by one of the blog readers. And since I have a thing for maths, i thought why not? Hey Stan! Your post is finally up. Straight on to basics.
Regression analysis is a quantitative forecasting technique that is used to develop an equation that will estimate costs given a certain number of unit production. For starters, determining how cost will change with output or other measurable factors of activity is of vital importance in decision making, planning and control. The challenge for most organisations is to determine how costs will be in future because costs behave variably in different seasons. You know they pose a challenge to predict them.
A regression equation allows one to mark the relationship between variables that are dependent and those that are independent. In most cases cost is the dependent variable and cost drivers the independent variables. The regression equation is written like this:
y = a + bx
Once the equation is identified, you can use the equation to predict costs for the future. This equation enables managers develop budget plans since they can forecast future costs with some degree of certainty.
The cost equation will most of the times be derived from past data. Past data could involve looking at company’s records, interviewing managers or conducting special studies. The disadvantage of this, as you will agree, is that it assumes past trends will be replicated in future. This is not always the case. The following steps are followed in estimating a cost function using regression analysis:
- 1. Choose the dependent variable – This will depend on the cost function being estimated. Remember that the dependent variable is the cost to be predicted and managed.
- Identify the independent variable – This is the factor to predict the dependent variable. An independent variable should be measurable and have an economically plausible relationship with the dependent variable. Economic plausability means that the relationship describing how changes in the cost driver leads to changes being considered is based on a physical relationship, a contract or knowledge of operations and makes economic sense to the manager concerned. Of importance to note, is that all the individual items of cost included in the dependent variable should have the same independent variable (cost driver).
- Collect data on the dependent and independent variables – These data could be time series or cross-sectional data.
- Plot the data – The plot provides insight into the relevant range of the cost function and reveals whether the relationship is approximately linear.
- Estimate the cost function – This has over time appeared to be the most difficult stage of regression analysis. In class assignments this is where the work begins. All the above steps are usually already done for us.
- Interpret the equation.
That was the basic introduction to regression analysis. In Let’s do Regression Analysis Part 2 I will show you how to carry out step 5 and 6. I will also explain what correlation coefficient and coefficient of variation mean and what their implications are.
Keep it here for more insights.