# Regression in Minitab

## Stat >> Regression >> Fitted Line Plot ...

By default, after you specify the predictor and response, this command produces a plot which comes up in a new window and contains:

• a plot of the estimated regression equation on a scatter plot of the original data
• the estimated intercept (b0) and estimated slope (b1)
• S = the square root of the mean square error
• R-sq = estimated R-square

By default, it also produces the following in the session window:

• the estimated regression equation with estimated intercept (b0) and estimated slope (b1)
• S = the square root of the mean square error
• R-sq = estimated R-square
• Analysis of variance table for assessing whether there is a significant linear association
• if they exist, an indication of any data points that yield extreme residuals or that have undue influence on the estimated regression line

You can ask Minitab to do other things, too, within the command:

• On the main pop-up window, you can specifiy the type of regression model. Choices include: linear, quadratic, cubic. Default is linear.
• Under Options ..., you can ask Minitab to not analyze the X data, but rather to analyze the log10 of the X data. To do so click on the logten of X box.
• Under Options ..., you can ask Minitab to not analyze the Y data, but rather to analyze the log10 of the Y data. To do so click on the logten of Y box.
• Under Options ..., you can ask Minitab to display confidence bands on the fitted line plot. To do so, click on the display confidence bands box, and specify the desired confidence level in the confidence level box (default is 95.0%).
• Under Options ..., you can ask Minitab to display prediction bands on the fitted line plot. To do so, click on the display prediction bands box, and specify the desired prediction level in the confidence level box (default is 95.0%).
• Under Options ..., you can ask Minitab to put a title on your fitted line plot. Type the desired title in the title box.
• Under Storage ..., you can ask Minitab to store the estimated residuals (RESI), the estimated (fitted) y values (FITS), and the estimated coefficients (COEF) in the data worksheet. To do so, click on the Residuals box, the Fits box, and the Coefficients box, respectively.

## Stat >> Regression >> Regression ...

By default, after you specify the predictor(s) and response, this command produces the following in the session window:

• the estimated regression equation
• the estimated intercept (b0) in the row labeled Constant and column labeled Coef
• the estimated slopes (b1, b2, ...) in the rows labeled by the X variable names and column labeled Coef
• the standard deviation of the estimated intercept, s(b0), in the row labeled Constant and the column labeled SE Coef
• the standard deviation of the estimated slopes, s(b1), s(b1), ..., in the rows labeled by the X variable names and the column labeled SE Coef
• the t-statistic and P-value for testing H0: Beta0 = 0 in the row labeled Constant and columns labeled T and P, respectively
• the t-statistic and P-value for testing H0: Beta1 = 0, H0: Beta2 = 0, ..., in the rows labeled by the X variable names and the columns labeled T and P, respectively
• S = the square root of the mean square error
• R-sq = estimated R-square