STEP-BY-STEP TECHNOLOGY GUIDE: Multiple Regression

TI-83/84

The TI-83/84 calculators do not perform the calculations for multiple regression.

EXCEL

  • Step 1 Load the response variable y into column A, and the predictor variables into columns B, C, and so on.
  • Step 2 Click on Data, then Data Analysis, then Regression, and click OK.
  • Step 3 For Input Y Range, select the cells in column A. For Input X Range, select the cells in columns B, C, etc.
  • Step 4 If you want to verify the regression assumptions, select Residual Plots and Normal Probability Plots.
  • Step 5 Click OK.

Note: Excel does not construct prediction intervals and confidence intervals for multiple regression.

MINITAB

  • Step 1 Load the response variable y into C1 and the predictor variables into C2, C3, and so on.
  • Step 2 Click on Stat > Regression > Regression > Fit Regression Model….
  • Step 3 Select C1 as your Responses, and select C2, C3, etc. as your Continuous predictors.
  • Step 4 If you want to verify the regression assumptions, click the button labeled Graphs…, select Four in One, and click OK.
  • Step 5 Click OK.

Note: See the Guide on page 741 for instructions on how to construct confidence intervals and prediction intervals for multiple regression.

SPSS

  • Step 1 Load the response variable y into VAR00001 and the predictor variables into VAR00002, VAR00003, and so on.
  • Step 2 Select Analyze > Regression > Linear….
  • Step 3 Move the y variable to Dependent, and the x variables to Independent(s).
  • Step 4 If you would like to verify the regression assumptions, click the button labeled Plots…, select Normal Probability Plot, and click Continue. Click OK.

Note: See the Guide on page 741 for instructions on how to construct confidence intervals and prediction intervals for multiple regression.

JMP

  • Step 1 Load the response variable y into Column 1 and the predictor variables into Column 2, Column 3, and so on.
  • Step 2 Select Analyze > Fit Model.
  • Step 3 Move the y variable to Y. Select the x variables and click Add under Construct Model Effects. Click Run.
  • Step 4 If you would like to verify the regression assumptions, observe the Residual by Predicted Plot under Effect Tests.

Note: See the Guide on page 741 for instructions on how to construct confidence intervals and prediction intervals for multiple regression.

CRUNCHIT!

  • Step 1 Enter data into the spreadsheet.
  • Step 2 Select Statistics, highlight Regression, and select Multiple Linear.
  • Step 3 Select the Dependent Variable from the drop-down menu, and check the Independent Variables. Click Calculate.

Note: CrunchIt! does not construct prediction intervals and confidence intervals for multiple regression.