Multiple Regression


This is the same as regression but accounts for multiply independant variables.
The values can be seen below:
 


 

Data input is similar to linear regression each observation is entered into a row in the data table and each variable is
represented by a column.

1. Enter the data  into 3 columns.  Named appriopately.

2. Click on analysis the n regression, and then linear.  As shown below in figure 1

Figure 1.

3.  Then move the dependent variable into the the box labelled dependent.  Followed by moving the
     independent variables into the box marked independent as in Fig 2.

Figure 2

4.  There are anumber of option available under the save item see Figure 3.  These include predicted values, residuals,
     leverage values to detect outliers and confidence limits.  These will then be attached to the datset as new columns.

Figure 3

5.  Also available under statistics is collinearity diagnostics to check for multi collinearity as it is assumed in mulitple
     regression that there is no multi collinearity among independent variables see Figure 4.

Figure 4


 

6.  Clicking OK on linear regression will perform the regression analysis this produces a regression of the dependent variable
     vs the independent variable.  Output will consised of the variables entered or removed, model summary including R and R square,
    Anova statistics and coefficients.  Additional output may include collinearity diagnostics.  For an example see Figure 5.
 

Figure 5


 
 
 

Return to SPSS homepage

Return to Linear Rgeression