I am conducting regression analysis in which there are high correlations among the predictors. I would like to employ the following measures of relative predictor importance: Kruskal's Relative Importance Algorithm; Shapley Value Regression; Penalty and Reward Analysis. Are these techniques available in SPSS?
Resolving the problem
The importance measures that you listed - Kruskal's relative importance, Shapley Value Regression, and Penalty & Reward - are not available in SPSS. A request has been filed with SPSS Development that they be added to a future release of SPSS.
There is an extension command, STATS_RELIMP, available from developerWorks (https://www.ibm.com/developerworks/mydeveloperworks/files/app?lang=en#/person/270002VCWN/file/447acd1f-35ac-43cd-b6d0-39a9f1381a9f) that provides a dialog box interface to the R relaimpo package, which includes several measures of relative importance, including Shapley values, but not Kruskal's measure or Penalty & Reward analysis.
Rate this page:
Copyright and trademark information
IBM, the IBM logo and ibm.com are trademarks of International Business Machines Corp., registered in many jurisdictions worldwide. Other product and service names might be trademarks of IBM or other companies. A current list of IBM trademarks is available on the Web at "Copyright and trademark information" at www.ibm.com/legal/copytrade.shtml.