IBM Support

Missing data imputations - lack of pooled information in output

Troubleshooting


Problem

I have run several analyses in SPSS Statistics with a multiply-imputed data set, i.e. a data set that was created by the Multiple Imputation procedure in SPSS Statistics, which was supplied by a colleague.. The original data, with split file variable Imputation_ equal to 0, is followed by five imputed data sets. When I run Cox Regression, for example, pooled results should be available for the Regression coefficients and for the covariate means. (See the SPSS Help topic "Analyzing Multiple Imputation Data" for a list of pooled statistics for each procedure.) However, Cox Regression on imputed data is printing model coefficients for the original data and for each of the created (imputed) data sets. The output should also provide a final model line that displays the Pooled, or combined, model representing a summary of all the imputation models. This combined model uses a complex algorithm to compute the necessary SEs, adjusting for the imputations. Unfortunately, my output does not provide a Pooled model line. Since this is imputation data and split file variable, Imputation_, is spelled correctly, why is the pooling not performed?

[{"Product":{"code":"SSLVMB","label":"IBM SPSS Statistics"},"Business Unit":{"code":"BU048","label":"IBM Software"},"Component":"Not Applicable","Platform":[{"code":"PF025","label":"Platform Independent"}],"Version":"Not Applicable","Edition":"","Line of Business":{"code":"LOB76","label":"Data Platform"}}]

Log InLog in to view more of this document

This document has the abstract of a technical article that is available to authorized users once you have logged on. Please use Log in button above to access the full document. After log in, if you do not have the right authorization for this document, there will be instructions on what to do next.

Historical Number

91986

Document Information

Modified date:
16 June 2018

UID

swg21488850