Skip to Main Content
IBM Data and AI Ideas Portal for Customers

Shape the future of IBM!

We invite you to shape the future of IBM, including product roadmaps, by submitting ideas that matter to you the most. Here's how it works:

Post your ideas

Post ideas and requests to enhance a product or service. Take a look at ideas others have posted and upvote them if they matter to you,

  1. Post an idea

  2. Upvote ideas that matter most to you

  3. Get feedback from the IBM team to refine your idea

Help IBM prioritize your ideas and requests

The IBM team may need your help to refine the ideas so they may ask for more information or feedback. The product management team will then decide if they can begin working on your idea. If they can start during the next development cycle, they will put the idea on the priority list. Each team at IBM works on a different schedule, where some ideas can be implemented right away, others may be placed on a different schedule.

Receive notification on the decision

Some ideas can be implemented at IBM, while others may not fit within the development plans for the product. In either case, the team will let you know as soon as possible. In some cases, we may be able to find alternatives for ideas which cannot be implemented in a reasonable time.

Additional Information

To view our roadmaps:

Reminder: This is not the place to submit defects or support needs, please use normal support channel for these cases

IBM Employees:

The correct URL for entering your ideas is:

Status Future consideration
Workspace SPSS Statistics
Created by Guest
Created on Apr 6, 2021

Add column proportions z-test for post-hoc testing in the Complex survey module

I would like to request for a future SPSS release to include column proportions z-test for post-hoc testing in the Complex survey module.

Surveys are an often used resource to collect informative data from a large number of people. However, analysis of survey data is often complex. Luckily, SPSS has developed a Custom Survey module to help with this type of analyses. However, when running a chi-square test in this module, there is no solution to run post-hoc tests on significant chi-square tests (this applies to analyses that are greater than 2x2). However, this solution is available for non-survey data or non-complex survey data under the Crosstabs feature or in Custom Tables. However, these modules do not take into account the complex survey design and so post-hoc results are not as accurate.

There were several options provided to me from the SPSS statistics team:

  1. Use Custom Tables with effective base weighting (using the final sample weight variable used in Complex Samples as the effective base weighting variable on the Options tab in Custom Tables) and specifying column proportions tests.

    1. However, the results from this should more closely track those from Complex Samples than using CROSSTABS, but they will only be approximations to the results from a full Complex Samples treatment

  2. Within the formal Complex Samples module, there are some options, but none are exactly what's desired.

    1. One is to specify table percentages in the Complex Samples Crosstabs Statistics dialog and check the box for Adjusted residuals. These adjusted residuals provide approximate Z values for testing a cell vs. the average of the row (or column) rather than against another column in that row.

    2. Another option is to use the Complex Samples Logistic Regression procedure, which will allow you to specify contrasts of columns against each other using the Odds Ratios option, but unless the row variable in the original table has only two rows, a recoded row variable where one value represents the original row and a second value combines all the other original rows would have to be used to match the proportions comparisons from column proportions tests , and this would have to be done for each row (using this recoded binary variable as the dependent variable). Then, within the analysis for each row, the Odds Ratios specifications would let you compare each level of the original column variable against a chosen reference level, so to get all pairwise, multiple runs would be required, using a different reference category for each.

    3. Finally, the Odds Ratios output produces estimates of odds ratios and confidence intervals, not p values.

While I appreciate the thoughtfulness that went into these recommendations, it would be far easier (and helpful!) to be able to do post-hoc testing directly in the Complex Survey module itself, rather than the work-arounds.

Needed by Date Apr 20, 2021