SPSS On-Line Training Workshop

 

Link to Table of Contents

In this Tutorial: 

linear regression

binary logistic
regression

multinomial
logistic
regression

Regression Menu Regression includes linear regression and logistic regression.

camera.gif (1166 bytes) MOVIE: Linear Regression camera.gif (1166 bytes)

camera.gif (1166 bytes) MOVIE: Stepwise Regressioncamera.gif (1166 bytes) 

camera.gif (1166 bytes) MOVIE: Logistic Regression camera.gif (1166 bytes)

Back to Statistical Procedures

 

In this on-line workshop, you will find many movie clips. Each movie clip will demonstrate some specific usage of SPSS.

Linear regression:  This helps us analyze the variables to set up a model of the data in the form of an equation.

bullet

Method: This lets you select how independent variables are entered into the analysis.  The "Enter" method enters all variables at the same time.  The other methods involve some sort of step-wise regression.

bullet

Statistics: To check for collinearity problems, we need to make sure that Estimates and Model fit are selected.  (These are usually selected by default.)  R-squared change is needed for variable selection methods. If you want to check for collinearity problems, you can select "Collinearity diagnostics".  You can make other selections, if needed.  

bullet

Plots:  You have the option of plotting the residuals.

bullet

Save:  You can save some of the results of the analysis, either to the data editor or to a new file.

bullet

Options:  If you are doing stepwise regression, you can change the criteria for entry and removal under this submenu.

click here to watch Linear Regression

click here to watch Stepwise Regression

Binary Logistic and Multinomial Logistic Regression:   This is used to determine factors that affect the presence or absence of a characteristic when the dependent variable has two levels (binary logistic) or more (multinomial logistic).

bullet

Method:  This lets you select how independent variables are entered into the analysis.  The "Enter" method enters all variables at the same time.  The other methods involve some sort of step-wise regression.

bullet

Select:  This button is used if you want to limit your analysis. 

bullet

Categorical:  Here is where you  identify categorical variables and specify how you want this data compared.

bullet

Save:  This allow you to save output as new variables in the data editor window.

bullet

Options:  If you are doing stepwise regression, this is where you can set your entry and removal criteria.   Another option that can be checked here is the "CI for exp(B):".    This gives you the odds ratio and is helpful in interpretation of parameter estimates.

click here to watch Logistic Regression

 

 

Navigation for Home,  Tutorials and Contact Us

©1999 Dr Carl Lee, Dr Felix Famoye, Central Michigan  University.  Joyce Sharp, student assistant.  All rights reserved.