Decision Analysis
Armstead M. Deas
BA520 Week#2, May 9, 2015
Dr. Jackson

Abstract
A variable is defined as a value that can take different values depending on particular circumstances. As such a variable can take on many values as and when the situation dictates. Subsequently a variable is not static; its altering nature can help the estimator determine the possible values of a function for different variable values (Render, Stair, Hanna and Hale, 2009). For example, assume that the total amount of gas used is a function of distance travelled and time taken. The total gas cost can be determined for many distances travelled i.e. 10, 50, 100 miles.
A simple linear regression is a function that has a constant and variable portion. The variable portion and total value (Y), of the function are affected by different values of a variable (X). One important characteristic of a linear regression is the fact that when drawn the results of the function must always follow the straight line. A scatter diagram is a graphic representation of the relationship between two variables. In a scatter diagram one variable will be on the horizontal axis while the next variable exists on the vertical axis (Jobson, 1991). As an example, assume that the cost, of a car wash parking lot, is influenced by the number of cars washed. Also assume that a flat payment must be made each period. This then forms the basis for a linear regression model as defined below:
Total Car wash cost = Constant + variable portion = $500 + $20 (Number of cars)
Variables
Constant portion = $500 = the amounts that must be paid for the lot
Variable portion = $20 = the direct cost of each car washed
Number of car = the number of cars that is used to estimate the total parking cost.
Multiple regression analysis is an estimation model that uses several variables as a means to determine the value of a function. In comparison to the simple regression model which has only uses one variable portion; a multiple regression may have as many variables