Table of contents
At a glance
How do you validate Minitab for a GMP-relevant analysis?
Validate the defined analysis workflow using the actual Minitab version, data transfer, method, configuration and intended use of the result. Check meaning and completeness of data, calculations, scripts and retained records. Assess what the supplier Validation Kit covers. A common calculation engine does not establish that desktop and web functions are identical or that every statistical method suits your process.
- A mathematically correct metric may be based on incomplete data or an inappropriate method.
- Minitab Statistical Software, Web-based application, add-on modules, and associated products are subject to a specifically defined scope.
- Project, worksheet, and export retain different information.
- Supplier certificates and internal inspections are combined based on their actual coverage.
1. Define the quality decision prior to software implementation
Start with a specific question: Should Minitab support exploratory process development, perform periodic trend analysis, or provide data to support a defined quality decision? Identify input data, expected insight, and subsequent action. Installing the software alone does not indicate which functions are relevant to your regulated process.
This guide addresses computerized systems within human medicinal product GMP activities. The current EudraLex volume references Annex 11 and Chapter 4. Annex 11 is a GMP guideline; it establishes a validation depth based on a justified risk assessment. Clinical trials, medical devices, and other regulatory areas require separate consideration. The term "statistical software" does not imply the general application of all GxP requirements.
A useful scope description, for example, includes the acceptance and incorporation of released laboratory results, predefined analysis, and the report for technical review. It also defines boundaries: study planning, external data sources, or additional predictive models may lie outside this approved workflow. This clearly indicates when a new application must be evaluated initially.
2. Clearly distinguish desktop, web applications, and add-on products
The core product is Minitab Statistical Software. Additional components include Minitab Solution Center, Connect, Workspace, Real-Time SPC, and optional modules. These names do not imply a predefined, common delivery package. Document which application employees open, which modules they use, and where data or projects are stored. Integrated data preparation is also part of the described transition.
For inventory assessment, you require the specific version record and the corresponding release notes. For web usage, additionally, the provided functional status, the browser path, and the associated storage location must be traceable. A desktop version number does not describe every cloud component.
The public kit description refers to a common computing core. This does not imply full functional equivalence. For example, the web documentation restricts certain macro and file operation commands. When transitioning from desktop to web environments, verify the actual workflow used, rather than relying solely on an existing results comparison.
3. Explicitly verify erroneous values and data transfers
Minitab flags missing numerical and date/time values with an asterisk. This aids in detection but does not establish a defined meaning for the source data. For example, a system may return −99 to indicate a missing measurement. Without appropriate recoding, this becomes a regular numeric value. As a result, a mathematically correct analysis may yield a technically incorrect conclusion.
Prior to import, specify which columns contain measurement values, units, sample identifiers, and status. Describe allowed substitute codes and the handling of incomplete datasets. Verify not only the total number of imported rows: duplicate identifiers, swapped samples, or mixed units may go undetected even when the row count appears correct. These checks serve as a support mechanism for the specific data pathway.
The decision on whether to proceed with calculations in the presence of missing values shall be based on the established procedure and intended use. Verify the behavior of the actually implemented function and the specified data basis. Successful execution of an analysis does not automatically imply that the intended sample was complete or that the report may be deemed suitable for use.
4. Method and settings to be documented in a testable evaluation agreement
An approved analysis pathway requires more than just the name of a menu command. It specifies data type, selection criteria, method, key parameters, and interpretation. For example, Minitab explains for a Normal Capability Analysis that data must be continuous, subgrouping must be rational, process stability must be established, and appropriate distributional assumptions must be made. A calculated capability index does not retrospectively validate these prerequisites.
The following internal work matrix combines professional and technical specifications. Complete it for your specific routine process. The role listed is a required contribution, not a universally mandated signature assignment. The same professionally qualified individual may undertake multiple responsibilities, provided your organizational rules and required independence are maintained.
| Component | Definition within the process | Validatable evidence | Contribution |
|---|---|---|---|
| Underlying data | Source, unit, identifier and outlier rule | Expected import with full reconciliation of identifiers | Laboratory or process owner |
| Method | Procedure, Assumptions, and Permitted Data Volume | Rationale selection and appropriate reference cases | Statistical Responsible Party |
| Execution | Client, Version, Module and Script Status | Configuration plus repeatable workflow | System owner and IT |
| Result | Key figure, data volume and interpretation limit | Comparative evaluation and documented technical review | Decision Responsible Party |
| Record | Project, expenditures and required context | Readable, recoverable analysis status | Quality Assurance and IT Department |
5. Example: Same Mean Value with Different Completeness
The following example is a custom set-up, not a minitab test. An example report expects five clearly identified measurements of the same unit. The full values are 98, 99, 100, 101 and 102. In the source system, −99 is explicitly missing. For the intended reporting use, completeness is required in this example.
Run variants with defined initial state and separate test data. The expectation is derived from independent calculation; the actual Minitab output, function used, and handling of missing values will be captured during review. No native locking function is assumed. Unauthorized report usage may be technically and organizationally prevented depending on the reviewed design.
| Test state | Independent calculation | Expected process outcome | Proof to be recorded |
|---|---|---|---|
| All five measurements available | Total 500; N = 5; Mean value 100 | Completeness requirement met; additional method validation remains required | Identifiers, input values, settings, and results |
| Third measurement value is missing | Four values yield 400 / 4 = 100 | Average value, but incomplete; report not yet intended for the defined purpose. | Incorrect value display, flagged N and decision |
| −99 is used as a measurement value without verification | 301 / 5 = 60.2 | Import rule violated; root cause investigate and re-evaluate under control. | Unaltered test source, identified error and correction reference |
| Missing measurement correctly supplemented | Again 500 / 5 = 100 | Provide evidence of a new, complete dataset basis and updated analysis status. | Source of the addition, version, and re-evaluation |
| Only numbers exported without analysis context | The numerical mean can remain unchanged at 100. | The scope of required evidence remains open. | Comparison of Project, Export, and Storage Purpose |
6. Treat macros as a separate execution pathway
Execution files contain Minitab commands; macros can automate repetitive tasks or extend functionality. Even a workflow saved from the history requires a defined input, a controlled state, and verifiable outcomes. Macro-free operation is not equivalent to evidence of low risk: a mistaken manual selection can still compromise the same quality decision.
The web application supports executives and macros with limitations. According to the vendor, nested macro calls, opening of data files within a macro, and certain write operations are not supported. ODBC and PYSC are listed among unavailable web commands. Therefore, a workflow that functions on the desktop cannot be unreviewedly considered equivalent to a web-based process.
Check the response when a file is missing, an incorrect column is present, or execution is interrupted. Particularly important is whether an old report could be mistaken for a new one using updated input data. A restart must begin from a unique and consistent state. Demonstrate for your workflow how errors are detectable and which data is utilized upon restart.
7. Project, worksheet, and export to be retained according to their intended use
A Minitab project contains, according to documentation, worksheets, outputs, session logs, and dialog settings. A separately saved worksheet has a different scope. When exporting to external worksheet formats, stored constants and matrices may be lost. Text outputs inherit the displayed time format; undisplayed time components are not recorded.
Therefore, determine which data and relationships are required for subsequent evaluation. A readable results document may accompany this but does not automatically suffice. Based on a complete evaluation status, verify that the data foundation, selection, method, and decision remain traceable and logically justified. Session history serves as an information source; its designation does not establish an immutable audit trail.
Minitab describes autosave to OneDrive or Google Drive for web projects. Local files are not automatically saved. Neither autosave nor a BAK file replaces a justified backup and retention policy. Assign responsibility and define access, recovery and retrieval after version changes. Chapter 4 and Annex 11 provide a regulatory framework without prescribing a universal file format.
8. Assess the Validation Kit and updates following their coverage
Minitab describes a Validation Kit comprising quality documentation, templates, and automated scripts along with expected outcomes. This public description serves as a useful starting point. The specific version, environment, and functionalities covered by a particular package must be verified prior to its implementation.
Align supplier invoices with your requirements. A verified reference invoice can serve as a basis for a calculation justification. It has not yet been determined whether your source data is correctly matched, whether the method is appropriate, or whether the export preserves all necessary information. Gaps will be intentionally addressed; existing, relevant documentation does not need to be simply copied without reason.
The release notes also provide specific reasons for a review of impacted components. At the documented status of 22.5.2, for example, release and file transfers are distinguished; known issues include missing expiry markings when weight columns are modified. Verify whether the used analysis pathway is affected. Test scope, transition solution, and usage decision follow this assessment, not a blanket, full revalidation of every update.
9. Start with a complete routine case in the SME
First, collect an actual, routinely occurring case that is genuinely required: source, example data, valid work instruction, Minitab project, output document, and responsible decision-maker. The department specifies the intended use of the result. A statistically qualified responsible party evaluates the method and assumptions. IT confirms version, access, storage, and technical dependencies; QA integrates the evidence into the quality management system.
Subsequently, gaps are prioritized and testing criteria are defined prior to execution. Actual results and deviations are documented, and a clear, justifiable decision is made regarding the extent of usage. This workflow serves as a planning aid and does not prescribe a fixed number of separate documents nor a general QA signature for each program run.
The effort primarily stems from diverse analytical methods, data sources, individual macros, manual transfers, and unclear historical data evaluation practices. A robust planning process emerges only when these differences are clearly visible and the relevant subject-matter experts are actively involved.
10. Checklist before using an evaluation pathway
Use this custom checklist for a specific process. An open point initially represents a finding whose significance and handling require determination. A generic confirmation such as "Minitab validated" does not make the remaining limitations visible.
- Intended use, GMP relevance, and explicitly excluded applications are specified.
- Client, Version, Add-on Modules, Scripts, and Data Storage are clearly traceable.
- Data origin, identifiers, units, alternate codes, and expected completeness are defined.
- Method, assumptions, and relevant parameters were technically justified.
- Normal, incomplete and incorrect inputs have certain predetermined reactions.
- Supplier documentation was evaluated regarding version and required functionalities.
- Project and release retain the required data and decision context.
- Access, modifications, backup, and restoration are appropriately defined and verified.
- Variations, intended use, and responsibilities are clearly justified and traceable.
- Updates and new applications trigger a suitable assessment of the released process.
Project experience and collaboration
Verify the Minitab analysis against reference cases
A Minitab analysis should provide a traceable answer to the underlying question. With the process owners, we examine the input data, analysis parameters, project version and conclusion. For recurring analyses, reference cases and a focused comparison of evidence may help. A new analytical workflow first needs requirements and responsibilities. Together, we establish what must be retained for review and later reconstruction.
Frequently asked questions about Minitab
Must every Minitab installation within the company be validated under GMP?
No, actual usage determines the scope. For example, exploratory analyses should be distinguished from a predefined, quality-relevant routine evaluation. For the human medicinal product GMP process under consideration, the intended use, data pathway, and result application must be assessed collectively. The same computer system may support multiple application purposes; therefore, neither the release of all functions nor the requirement to individually evaluate every conceivable analysis follows.
Has Minitab been validated for our process using the Validation Kit?
No. The public package description provides useful manufacturer documentation and reference scripts, but does not include a review of your specific application. Obtain the appropriate package and align its actual scope with your requirements. Data import, method, script modifications, and storage may require additional validation evidence.
Can we use an audited desktop workflow unchanged in the web application?
Not without scrutiny of differences. A common computing core does not imply identical file, macro, and integration functionalities. The web-based documentation does not list unsupported commands or macro variants. Compare data transfer, execution, error handling, and storage of your workflow. A matching average under normal conditions does not address, for example, a command that is no longer functional.
Why is a plausible mean value insufficient as a test result?
Since different databases may yield the same mean value, as illustrated in this article, both five complete measurements and four existing values result in a mean of 100. Only the first approach satisfies the example-based completeness requirement. Therefore, in addition to the calculation, verify identifiers, data count, missing values, and subject-matter applicability. The allowable treatment of missing values must be determined specifically for the particular analysis.
How do we handle −99 or other surrogate codes for missing measurements?
Clarify their significance within the source system and establish a controlled transfer procedure. Minitab explicitly states that a non-reencoded substitute code is treated as a numeric value. Maintain traceability to the original source and verify the conversion. Later removal of outliers without a justifiable rule does not constitute a valid data transfer.
Do we need to use at least 100 test values for validation?
Not based on the Minitab help for Normal Capability Analysis. The guidance provided there pertains to a robust estimation of process capability using appropriate data, not to the general scope of a software test. For validation, select meaningful normal, boundary, and error cases based on risk. For the actual statistical evaluation, additionally determine the professionally required data volume and the suitability of the method.
Is a PDF or an Excel file acceptable as evidence of evaluation?
A worksheet export does not provide the same information as a complete Minitab project including output, traceability, and settings. Verify the evaluation evidence required and confirm that the selected format retains all necessary data. Data loss may be significant for constants, matrices, or time-based components. A readable PDF report may be included in the file record, but it cannot serve as the sole and complete validation evidence.
Is the Minitab history a complete audit trail?
A command history alone does not constitute a complete GxP audit trail. The described session workflow supports the traceability of command sequences. Whether relevant data modifications—including identity, timestamp, reason, and appropriate protection mechanisms—remain detectable must be demonstrated in your specific environment. Establish the necessary controls based on usage and risk assessment.
How many batches must we retest after a Minitab update?
First, assess the change and its impact points with your approved workflow. Relevant considerations may include new calculation functions, altered user interfaces, file compatibility, or known output issues. These lead to justified comparison and regression testing, as well as potential additional actions. A fixed test volume cannot be derived from the version number; likewise, no update may be accepted automatically without additional validation evidence.
What documents and personnel do we need for project launch?
Start with a real routine workflow, representative data, the employed method, project files, and existing supplier documentation. Designate a subject-matter expert, statistical expertise, and points of contact for IT and Quality. Together, define data relevance, acceptance criteria, and gaps. The number of procedures, scripts, and interfaces significantly drives effort far more than the mere number of installed Minitab licenses.
Sources and review date
As of 6 October 2026. Supplier reference: Minitab Statistical Software 22.5.2 and rolling desktop/web help; specific installation and modules to be confirmed separately. The Validation Kit was categorized solely based on its public description and not reviewed or executed. Regulatory focus: Human medicinal product-GMP, Annex 11, and Chapter 4. Tables and figures are internal work aids and do not constitute product testing.
- Minitab Statistical Software
- Minitab Software Updates — 22.5.2
- Confidence in Every Calculation: How Minitab Supports FDA-Regulated Validation
- Request the Software Validation Kit
- Execute a script in the web application
- How Minitab denotes missing data
- Data considerations for Normal Capability Analysis
- Save and export Minitab projects and worksheets
- European Commission: EudraLex Volume 4
- EU GMP Annex 11: Computerised Systems, January 2011 revision
- EU GMP Chapter 4: Documentation, January 2011
Sources checked: .
Ensure traceability of the Minitab analysis pathway
We support you in defining the relevant application, evaluating supplier documentation, and jointly developing specific data, method, and output reviews with the respective department, QA, and IT.
- Computer system validation →
Bring your GxP systems into validated operation, from laboratory software, MES and eQMS to ERP. We define the scope, plan risk-based tests using GAMP 5 and hand the evidence to QA for release.

