IBM Support

CROSSTABS: Problems with Fisher Exact test, percentage of exp. counts < 5 with integer mode, empty rows or columns

Troubleshooting


Problem

I'm running SPSS CROSSTABS with tables that sometimes involve empty rows and/or columns, and am running into the following problems: 1) In one case CROSSTABS indicates that the Fisher test cannot be computed because there is an empty row or column in the table. However, another command also involving an empty row or column works without complaint. 2) When computing the percentage of cells with expected counts < 5, the total number of cells in the table, including the null rows or columns are used in the numerator, but only the cells from the non-null rows or columns are used in the denominator. This results in notes like 225% of cells have expected counts < 5. 3) In one case, the probability for the Fisher test is given as > 1. Are these known problems?

[{"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

67880

Document Information

Modified date:
16 April 2020

UID

swg21478437

Page Feedback