 You don't have to be a financial wizard to predict future costs, but you do need to know a company's cost equation. There are two primary methods used to predict costs at various levels of volume. They are the high-low method and regression analysis method. This short video will focus on regression analysis. Regression analysis is a statistical calculation to find the line that best fits a data set. The line is the cost equation. It uses all the data points, unlike the high-low method, which only uses two data points. And it is more accurate than either visually looking at a scatter plot or calculating a cost equation using the high-low method. So all of the data points on a scatter plot would be used. This is an example of an output report you would get when you perform regression analysis in Excel. By the way, this is not from the same data set used in the high-low example. Anyway, you can see some of the highlighted items on the output report. Let's start with the two at the bottom. The intercept is the fixed cost, and the x-variable one is the variable cost. We use the coefficient amounts for the cost equation. Since we know the variable cost per unit and the fixed cost, we can write the cost equation, which in this case is y equals 7.849766x plus 14,538.5 cents. The final number we need in regression analysis is the r-squared value, which is a predictor of the goodness of fit or how well the line fits the data points. r-squared values closest to 1 means the cost equation is good at predicting costs at various levels of volume. r-squared values close to 0 means the cost equation does a poor job at helping us predict costs. r-squared is an important number and makes regression analysis significantly better at predicting costs than the high-low method, which provides no such confidence factor. Regression analysis is most commonly performed in Excel. This table is the output table from Excel. If you would like to see how to use the Excel add-on feature to perform regression analysis, there are literally thousands of YouTube videos that can help with that. My purpose is to show you how the output data can be used to determine a cost equation.