way to measure cost behavior (or develop a cost-volume formula) according to what costs should be, not by what costs have been. It entails a systematic review of materials, labor, support services, and facilities needed for product and services. Engineers use time and motion studies and similar engineering methods to estimate what costs should be from engineers' specifications of the inputs required to manufacture a unit of output or to perform a particular service.

statistical procedure for estimating the average relationship between the dependent variable (sales, for example) and one or more independent variables (price and advertising, for example). It is a popularly used method for estimating the cost-volume formula(y = a + bx). simple regression involves one independent variable, e.g., direct labor-hours or machine-hours alone, whereas multiple regression involves two or more independent variables. Assuming a linear relationship, the simple regression model indicates that the relationship is y = a + bx, where a, and b are unknown constants, called regression coefficients. The multiple regression model is y = a₀ + a₁x₁ + a₂x₂ + ... + a_kx_k, where a's are coefficients and x's represent the number of independent variables.

In estimating the cost-volume formula, regression analysis attempts to find a line of best fit. To find the line of best fit, a technique called the least-squares method is widely used.

algebraic procedure used to separate a semivariable cost or mixed cost into the fixed and the variable components. The high-low method, as the name indicates, uses two extreme data points to determine the values of a (the fixed cost portion) and b (the variable rate) in the cost-volume formula y = a + bx. The extreme data points are the highest and lowest x - y pairs.