In this Discussion you will use one of the Discussion Board data sets to identify two variables that are correlated and then create the best prediction equation for those variables. Finally, you will use the prediction equation you created to make a prediction for one variable (the Y variable) using a value for the second variable (the X variable).
Correlated Variables: Use one of the Discussion Board data sets to identify two variables that are correlated. Use SPSS to document the correlation. For example, you may think that height and weight are correlated. From the Female Health data set there is a correlation of 0.364 between height and weight.
Best Prediction Equation: Use the Linear Regression procedure to generate the best prediction equation (regression equation) using one variable as the dependent or predicted variable (Y) and the other as the independent variable or predicting variable (X). To identify the dependent variable and independent variable think about which variable impacts the other variable. For example, would weight help determine our height or would height help determine our weight? In this example, the height would be the independent variable. In other words, you will use someone’s height to predict their weight using the equation: Weight (Y) = -169.839 + [5.001 *Height (X)]
Making a Prediction: Use your best prediction equation to predict a value for the Y variable using a hypothetical value for the X variable. For example, what is the predicted weight of a female who is 63 inches tall? Important: you must remember and use the same measurement of the data as was used in the data set! The predicted weight of a female who is 63 inches tall is:
Y = -169.839 + 5.001*63
Y = -169.839 + 315.063
Y = 145.224
State the results of your prediction: For example, the predicted weight of a female who is 63 inches tall is 145.224 pounds. What is the result of your prediction?
Optional: Find another variable in your data set that correlates with your dependent variable and add it as another independent variable to your regression equation to create a multiple regression equation. Post your multiple regression equation to the Discussion Board.