STEP-BY-STEP TECHNOLOGY GUIDE: Confidence Intervals and Prediction Intervals

741

MINITAB

  • Step 1 Enter the x data into column C1 and the y data into column C2.
  • Step 2 Select Stat > Regression > Regression > Fit Regression Model….
  • Step 3 For Responses, select Messages. For Continuous predictors, select Age. Click OK.
  • Step 4 Select Stat > Regression > Regression > Predict….
  • Step 5 Enter 20 in the first element of the Age column. Click OK. The results are shown in Figure 18.

SPSS

  • Step 1 Enter the x and y data into the first two columns.
  • Step 2 Select Analyze > Regression > Linear….
  • Step 3 For Dependent, select the y variable. For Independent(s), select the x variable
  • Step 4 Click Save…, and select Mean and Individual under Prediction Intervals. Click Continue, then OK.
  • Step 5 Return to the Data window and the Data View tab. The new variables LMCI_1 and UMCI_1 represent confidence intervals for the corresponding x value, and the new variables LICI_1 and UICI_1 represent prediction intervals for the corresponding x value.
  • Step 6 To make confidence or prediction intervals for x values not in the dataset, add the value (for example, 21) in the first empty element of the x column, and repeat Steps 2 through 4. Look for the variables LMCI_2 and UMCI_2, and LICI_2 and UICI_2.

JMP

  • Step 1 Enter the x and y data into the first two columns. Rename the columns Age and Messages.
  • Step 2 Select Analyze > Fit Y by X. For Y, Response, select Messages. For X, Factor, select Age. Click OK.
  • Step 3 Click the red triangle beside “Bivariate Fit of Messages by Age” and select Fit Line.
  • Step 4 Click the red triangle beside “Linear Fit” and select Mean Confidence Limit Formula and then Indiv Confidence Limit Formula. Confidence intervals and prediction intervals for all existing Age values are supplied.
  • Step 5 To make confidence or prediction intervals for Age values not in the dataset, add the value (for example, 21) in the first empty element of the Age column, and hit ENTER. The intervals will automatically appear.

CRUNCHIT!

  • Step 1 Enter the x and y data into the first two columns.
  • Step 2 Select Statistics, highlight Regression, and select Simple Linear.
  • Step 3 For Dependent Variable, select Var2. For Independent Variable, select Var1. For Predict (optional), enter the value (for example 21). Click Calculate.