SPSS Statistics 32 freezes during CSV export on macOS 27? Stop repeating the export on your only project file.
IBM has recorded that SPSS Statistics 32 Classic on macOS 27 may hang when exporting CSV. Work from a separate project copy, test the export there, and do not treat this specific issue as proof that analysis, ordinary saving, or every export function is broken. Avoid moving a critical study to this setup until the actual deliverable has been checked.
This guide is for graduate students, doctoral researchers, and research staff who need a dependable data handoff. It also helps university lab support staff document the failure, review a test plan, and set boundaries for any remote reproduction.
Last updated November 2, 2026; verified against IBM’s known-issue record and IBM licensing documentation. The record was updated on October 12, 2026, and was still open as of the latest review. IBM’s issue record describes the possible CSV export hang; it does not establish that every installation will reproduce it or promise a fix date.
[ SECTION_01 ] What IBM has confirmed—and what remains unproven
The documented problem is specific: IBM’s support record says SPSS Statistics 32 Classic has been tested running on macOS 27, while CSV export may hang. The record identifies that export behavior as a known issue. It does not say that all SPSS operations fail, that all Mac users will see the same symptom, or that a particular setting is the confirmed cause.
Keep those distinctions in the incident notes. A report that says “SPSS is unusable” is less helpful than one that identifies the exact operation that stopped responding. IBM’s SPSS Statistics 32 documentation is a reference for the product, but the known-issue record is the source to check for this specific compatibility issue and its current status.
Before changing the research environment, capture the observations that could confirm or rule out a match:
- The exact SPSS product and build shown in the application’s version information.
- The macOS version displayed by the system.
- The export format selected, including whether the failing operation is specifically CSV export.
- The actions immediately before the application became unresponsive.
- What “stuck” means in this case: a progress indicator that does not change, an unresponsive window, or an export that appears to finish but leaves an unusable file.
- Whether the project still opens, analysis commands still run, and ordinary project saving still completes.
These are facts to record, not diagnostic conclusions. For example, a frozen application window during export does not by itself prove that the source data is damaged. Likewise, a file appearing in the destination folder does not prove that it contains a complete, usable export.
[ SECTION_02 ] Preserve a reproducible copy before testing
The safest first action is to preserve the original project and make a separate copy for reproduction. Do not repeatedly force-quit the application while it is working on the sole copy of an active study. If a project is stored on a shared drive or synchronized folder, follow the lab’s file-handling procedure before duplicating or moving it.
Use a copy that contains only what is needed to reproduce the export behavior. If the project includes personal, clinical, or otherwise restricted information, do not upload it to an unapproved service or send it in an ordinary support request. A sanitized test project can help determine whether the behavior is repeatable without exposing research participants or confidential records.
In your incident notes, keep the input and output locations distinct. Record which project copy was used and where the attempted export was written. Preserve any partial output separately rather than replacing a known-good result. If the export attempt changes the project, stop and return to an untouched copy instead of treating the altered file as a reliable baseline.
Use this evidence sequence:
- Record the starting state. Note the application version, operating system version, project copy, and intended file format.
- Run one controlled attempt. Use the copy, not the only original. Record the visible behavior and the point at which the operation stops progressing.
- Preserve the outcome. Keep the error message, partial file, or other evidence where lab policy permits. Do not include sensitive case data in screenshots or support attachments.
- Stop if the source copy becomes uncertain. If you cannot tell whether the project was saved or altered, do not continue testing on that copy.
- Escalate with a minimal reproduction. Share only the information and files approved by the institution and the software license terms.
This creates a useful support record without claiming a root cause. It also gives a lab administrator enough detail to compare the observed behavior with IBM’s documented issue.
[ SECTION_03 ] Does the CSV problem affect analysis and ordinary saving?
Not automatically. The documented symptom concerns CSV export. To determine whether your own project has a broader problem, test each operation separately on a working copy and record the result. Do not infer that analysis or ordinary saving is affected simply because a CSV export hangs.
Keep the checks independent:
- Opening the project: Can the copy be opened again after the application is closed and restarted?
- Running an analysis: Does a representative analysis complete on that copy? Record the command or menu operation and any output or error.
- Saving the project: Does the normal project save operation complete, and can the saved copy be reopened?
- Exporting CSV: Does the same copy hang specifically at the CSV step?
- Other output formats: If the research workflow needs another format, test it separately. A successful alternative export does not establish that CSV is fixed.
IBM’s data-file-type and saving guidance distinguishes data file types and saving approaches. Use it to identify the intended output type; do not use it as evidence that the macOS 27 CSV issue has been resolved. The SPSS Statistics 32 documentation can also help you check the product’s supported operations and workflow details.
If analysis and ordinary saving work but CSV alone hangs, document that separation. It narrows the issue to the observed export path without proving why it occurs. If other operations fail too, record each one as a separate symptom and include it in the support request rather than expanding IBM’s CSV finding into a general failure claim.
[ SECTION_04 ] Verify the exported file before treating it as deliverable
A generated file is not necessarily a valid deliverable. Check it independently before sending it to a collaborator, importing it into another tool, or using it as the basis for a reported result.
Begin with the research question and the receiving workflow. Confirm that the file opens in an approved tool and that its columns correspond to the variables expected by the next analysis step. Compare the variable names and data rows against the project or a trusted earlier output. Check representative records at the beginning, middle, and end of the file, while avoiding unnecessary exposure of sensitive values.
Then review the details that can quietly change interpretation:
- Rows and variables: Confirm that expected cases and fields are present. Investigate unexplained omissions rather than assuming the export is complete.
- Missing values: Check whether missing data remain distinguishable from valid values and match the project’s established interpretation. IBM explains how missing values are handled in SPSS Statistics 32.
- Text and separators: Check whether text appears intact and whether the receiving tool reads the intended field separators and text options. IBM’s CSV text-options documentation describes relevant export options.
- Downstream import: If the file is intended for a specific approved analysis pipeline, import a test copy there and confirm that fields are parsed as expected.
- Reproducibility: Keep the validated output associated with the project copy and the recorded export settings. Do not overwrite a known-good output until the new one passes the lab’s checks.
These checks do not certify that the dataset is scientifically correct. They establish whether the exported file is sufficiently complete and readable for the intended handoff. If the file is incomplete, has altered text, or handles missing values unexpectedly, mark it as unverified and return to the project copy or escalate the issue.
[ SECTION_05 ] Choose an action based on the cost of a failed handoff
The next step depends on whether the CSV is essential now, whether the project is reproducible, and whether an approved alternative environment is available. This comparison is a decision aid, not a claim that any option fixes IBM’s known issue.
| Option | Evidence required | Risk for an active study | Fit rating |
|---|---|---|---|
| Continue with a validated existing workflow | A known-good project copy and an output that passes the lab’s checks | Lower if the current workflow is already accepted and unchanged | Strong when a verified path exists |
| Test the CSV export on a separate copy | A controlled reproduction record and a review of the resulting file | Moderate; stop if the copy or output becomes uncertain | Strong for diagnosis, not proof of a fix |
| Pause a planned environment migration | A note that CSV is a required deliverable and has not passed acceptance testing | Lower than changing an active workflow without validation | Strong when the study depends on the export |
| Request institutional or IBM support | Product and OS versions, precise steps, and a sanitized reproduction if permitted | Depends on the data shared; follow institutional rules | Strong when the issue blocks required work |
For a study in progress, keep the last verified workflow in place until a replacement passes the same acceptance checks. This does not mean macOS 27 is unusable for every SPSS task. It means a required CSV handoff should not be moved into an unverified setup based only on the fact that the project opens or an analysis runs.
Check IBM’s issue record again before scheduling a migration or repeating a validation. As of the review date noted above, the record remained open; IBM’s record does not provide a guaranteed repair date. Re-test after IBM changes the issue status, publishes relevant release notes, or provides a version update that could affect the export path. Treat that as a reason to validate again, not as evidence that the export is already fixed.
When filing a support request, include the reproducible steps and the separate results for opening, analysis, ordinary saving, and CSV export. Keep claims limited to what you observed. If the institution’s support process requires a license check, use IBM’s installation and licensing guidance and have the license administrator confirm the applicable terms.
[ SECTION_06 ] Can a remote Mac help if you do not have a local one?
A remote Mac can be a candidate for an isolated reproduction, but it cannot be assumed to remove the documented SPSS issue. Reproduction is meaningful only if the environment, product version, project copy, and export procedure are recorded and the test is permitted by the institution. A different result on a remote machine would be an observation about that test environment, not proof of a general fix.
Before using a remote machine, confirm these boundaries:
- License permission: Ask the license administrator whether the institution’s license permits the planned remote access and usage pattern. IBM provides separate guidance for authorized-user licenses and concurrent-license administration. The applicable terms depend on the institution’s license; do not assume remote access or multiple sessions are allowed.
- Data approval: Confirm that the project data may be used on the chosen remote environment. Use a sanitized copy if approval for real research data is absent.
- Purpose: Define the test narrowly: reproduce the CSV export behavior and inspect the output, not migrate an active study by default.
- Evidence: Record the environment and procedure so the result can be compared with the local test.
- Stop condition: If access terms, data approval, or project integrity are unclear, do not upload or continue the reproduction.
If the lab needs a broader review before changing its setup, use the NOVAKVM remote Mac service overview as a starting point, then confirm the actual environment and usage terms before making a decision. A remote Mac is appropriate only when the license and research-data rules permit it.
For this specific problem, the current workaround—continuing on an existing lab machine or postponing the export test—can have real drawbacks: the required Mac may be unavailable, the team may lack an isolated test environment, and an unverified local setup can delay a deliverable. Renting a Mac through NOVAKVM may be a more convenient way to obtain a separate environment for a short, controlled reproduction, rather than buying hardware solely for a test. It still cannot guarantee that SPSS will export successfully, and it is not suitable for restricted data unless the institution approves that use. If the school’s license and data rules allow the test, review the NOVAKVM Mac options and use only a sanitized project copy.