Macros for ANOVA & Regression

1. Added Variable Plots
2. ANOM for 2 Level, 2 Factor Design
3. Box-Cox Transformation for Regression and Response Surface Models
4. Box-Tidwell Procedure
5. Calibration or Inverse Regression
6. Coefficient of Multiple Correlation
7. Confidence Interval for a Correlation Coefficient
8. Confidence Intervals for Regression Coefficients
9. Creating Indicator Variables -- Part I
10. Creating Indicator Variables -- Part II
11. Example of Using Jackknife Techniques in Regression
12. Fitted Line Plot Through the Origin
13. Hildreth - Lu Procedure
14. Multiple Case Cook's Distance
15. Multiple Case Influence Analysis
16. One-way ANOVA Confidence Intervals
17. PRESS Statistic for Regression data using a Power Transformation
18. Regression by a Grouping Variable
19. Ridge Trace Plot
20. ROC (Receiver Operating Characteristic) Curve
21. Rolling Regression
22. Simplex Method for Linear Programming Problems
23. Sorted One-Way ANOVA
24. Sorted Simple Regressions
25. Standardized Regression Coefficients
26. Tukey's 1 Degree of Freedom Test of Nonadditivity

Added Variable Plots
This macro creates added variable plots in a linear regression analysis. These plots are also known as partial regression plots. Written by Mike Delozier Back to Top

• Minitab 15 and 16
• Code
  Sample Data 1
  Documentation
  

ANOM for 2 Level, 2 Factor Design
This macro creates an ANOM chart for a 2 factor, 2 level factorial design. The interaction between the 2 factors is displayed on the same scale as the main effects. The default decision limits are calculated at alpha = .05. Written by Cathy Akritas Back to Top

• Minitab 15 and 16
• Code
  Documentation
  

Box-Cox Transformation for Regression and Response Surface Models
This macro determines the likelihood estimate of the Box-Cox power transformation parameter in regression and response surface modeling applications. A plot of the log-likelihood function over a range of parameter values is displayed showing the likelihood estimate and an approximate 95% confidence interval for the parameter. Also displayed is a plot of the values of the PRESS statistic transformed back to the original response scale over the 95% confidence interval. Optionally, the user may choose to specify the range of parameter values in the plot of PRESS, display an index plot due to Cook and Wang (1983) showing the influence of individual cases on the likelihood estimate, and store all computed results.

Some of this functionality has been added to Minitab 16 as part of the analysis at Stat > Regression > General Regression. Written by Steve Orlich and Mike Delozier Back to Top

• Minitab 15 and 16
• Code
  Sample Data 1
  Sample Data 2
  Documentation
  

Box-Tidwell Procedure
This macro executes the Box-Tidwell procedure to determine appropriate predictor variable power transformations for a regression model linear in the transformed predictors. It is important to note that this procedure can be numerically unstable resulting in error conditions for some data sets. Written by Mike Delozier Back to Top

• Minitab 15 and 16
• Code
  Sample Data 1
  Sample Data 2
  Documentation
  

Calibration or Inverse Regression
Obtains the point and interval estimate for a new value of X, the independent variable in a simple regression equation, given a new determination of Y, the dependent variable. This is referred to as the statistical calibration, or inverse regression. Written by Janice Derr Back to Top

• Minitab 15 and 16
• Code
  Sample Data 1
  Documentation
  

Coefficient of Multiple Correlation
This macro stores the coefficient of multiple correlation for each column regressed on the others. Written by Andy Haines Back to Top

• Minitab 15 and 16
• Code
  

Confidence Interval for a Correlation Coefficient
This macro calculates a confidence bound or interval for a correlation coefficient. Written by Eli Walters Back to Top

• Minitab 15 and 16
• Code
  

Confidence Intervals for Regression Coefficients
This macro calculates confidence intervals for coefficients from a regression model. This functionality is new in Minitab 16, and can be found at Stat > Regression > General Regression. Written by Veronica Bubb Back to Top

• Minitab 15
• Code
  

Creating Indicator Variables -- Part I
This macro calculates the number of columns needed for storage when creating indicator variables from multiple categorical variables. The result from this macro is used in the follow-up macro "Creating Indicator Variables -- Part II" which creates the indicator variables. Written by Daniel Griffith Back to Top

• Minitab 15 and 16
• Code
  

Creating Indicator Variables -- Part II
This macro creates indicator variables for multiple categorical variables and stores the indicator variables in the current worksheet. Written by Daniel Griffith Back to Top

• Minitab 15 and 16
• Code
  

Example of Using Jackknife Techniques in Regression
This macro is a simple example of using jackknife techniques to obtain the MSEs and coefficients from regression. For each iteration, a row of data is excluded, a regression analysis is performed on this reduced set of data, the MSE and coefficients are stored. The process is repeated for each row of data. All of the MSEs and coefficients for each iteration are displayed in a table in the Session window. Written by Cathy Akritas Back to Top

• Minitab 15 and 16
• Code
  

Fitted Line Plot Through the Origin
This macro creates a fitted line plot that goes through the origin. Written by Cathy Akritas Back to Top

• Minitab 15 and 16
• Code
  

Hildreth - Lu Procedure
The Hildreth - Lu procedure corrects for serial correlation (autocorrelation) in regression type data. Written by Cathy Akritas Back to Top

• Minitab 15 and 16
• Code
  

Multiple Case Cook's Distance
This macro computes the multiple case extension of Cook's single case distance measure. Depending on the data set size, the distance measure can be computed for all case pairs and triplets. In addition, the distance measure can be computed for user selected subsets of up to ten cases. Graphics produced include a plot of Cook's distance for single cases against case number, an influential case pairs ID plot, and fixed-pair effect plots which display the effect, or change in Cook's distance, due to adding a third case to a fixed pair of cases. Like functionality is available for models with no constant term. Written by Mike Delozier and Steve Orlich Back to Top

• Minitab 15 and 16
• Code
  Sample Data 1
  Documentation
  

Multiple Case Influence Analysis
This macro employs a novel backward elimination approach in searching for influential multiple case subsets in linear regression using Cook's multiple case distance measure. Written by Mike Delozier and Steve Orlich Back to Top

• Minitab 15 and 16
• Code
  Sample Data 1
  Documentation
  

One-way ANOVA Confidence Intervals
This macro calculates and displays the endpoints of the confidence intervals that are given by Stat > ANOVA > One-Way, Stat > ANOVA > One-Way (Unstacked) and Graph > Interval Plot. Written by Cathy Akritas Back to Top

• Minitab 15 and 16
• Code
  

PRESS Statistic for Regression data using a Power Transformation
This macro computes the model fits, residuals, deleted fits, deleted (PRESS) residuals, and the PRESS statistic in the original units of the response when a power transformation of the response is applied in a linear regression. Written by Mike Delozier Back to Top

• Minitab 15 and 16
• Code
  Sample Data 1
  Sample Data 2
  Documentation
  

Regression by a Grouping Variable
This macro runs a regression analysis of Y on X.1 - X.n for each level of a grouping variable. Written by Cathy Akritas Back to Top

• Minitab 15 and 16
• Code
  

Ridge Trace Plot
Produces a ridge trace plot and a plot of RSS VS K for data in 'y' and predictors formed into the M1 matrix. M1 and 'y' should be in "correlation form". These plots are useful data analytic tools for ridge regression. Written by Berton Gunter Back to Top

• Minitab 15 and 16
• Code
  Sample Data 1
  

ROC (Receiver Operating Characteristic) Curve
This macro performs three functions as a follow up to a binary logistic (BLR) regression analysis:

1. Generates a classification table
2. Generates an ROC (Receiver Operating Characteristic) curve
3. Given the event probabilities, stores an event probability for each row, not just the first instance of a unique set of predictor values Written by Mindy Tomlinson Back to Top

• Minitab 15 and 16
• Code
  Documentation
  

Rolling Regression
This macro performs a rolling regression, which allows you to check for changes in the regression coefficients over time. The first iteration runs a regression on rows 1 through k. The second iteration runs a regression on the rows 2 through k+1. The third iteration runs a regression on rows 3 through k+2 and so on.

There is an optional subcommand that allows you to use rows 1 through k for the first iteration and rows 1 through k+1 for the second iteration and rows 1 through k+2 for the third iteration and so on. The data used in the regression is anchored at row 1. Written by Annie Molhoek and Daniel Griffith Back to Top

• Minitab 15 and 16
• Code
  

Simplex Method for Linear Programming Problems
This macro finds the optimal solution of a linear program, using the Revised Form of the Simplex. Linear programming (LP) deals with an objective function with only linear terms, and assumes only linear constraints exist. Written by Eduardo Santiago Back to Top

• Code
  

Sorted One-Way ANOVA
This macro performs One-Way ANOVA for a single response versus each specified factor, one factor at a time. For each factor, the R-squared and p-value are calculated. These values are then sorted from the smallest to the largest p-value and displayed in a table. Written by Andy Haines Back to Top

• Minitab 15 and 16
• Code
  

Sorted Simple Regressions
This macro performs simple regression for a single response versus each specified predictor, one predictor at a time. For each predictor, the R-squared and p-value are calculated. These values are then sorted from the smallest to the largest p-value and displayed in a table. Written by Andy Haines Back to Top

• Minitab 15 and 16
• Code
  

Standardized Regression Coefficients
This macro calculates the standardized coefficients for simple and multiple regression analysis. Written by Cathy Akritas Back to Top

• Minitab 15 and 16
• Code
  

Tukey's 1 Degree of Freedom Test of Nonadditivity
This macro performs Tukey's 1 degree of freedom test of nonadditivity, which is a test for an interaction in a two-factor experiment with a single replicate. Written by Eli Walters Back to Top

• Minitab 15 and 16
• Code
  

Disclaimer:

Minitab Inc. provides the Macro Library as a convenience only. Minitab neither endorses, supports, nor verifies the accuracy of any content, information, or functionality of any macro found in the Macro Library. Minitab specifically disclaims any and all responsibility or liability arising from or related to any reliance upon, use, or incorporation of any content, information, or macro found in the Macro Library.