RESEARCH SUPPORT
Excel data analysis support:
what the ToolPak can and cannot do
A workbook is reviewed for the tests the add-in can run reliably and the point at which the design needs a statistical package, then transferred with variable types, missing values and codes intact and verified against the original file.
No Commitment Required
Strictly Confidential
In-House PhD-Level Statisticians
Fixed Project Fee
Same-Day Response
What Can the Add-In Do for Research Data?
Excel data analysis support is help judging whether a workbook can carry your analysis and, where it cannot, moving the data to a statistical package without altering a value. It is research-analysis support, not Microsoft product support or help installing the add-in. Whether Excel is a statistical software has a practical answer for research: it is a spreadsheet, and the Analysis ToolPak adds a fixed menu of procedures, among them descriptive statistics, t-tests, single- and two-factor ANOVA, correlation and regression. It offers no models for binary outcomes, no mixed models and limited diagnostics. Judging which analyses can remain in the workbook falls to the in-house PhD-level statisticians of OnlineSPSS, a specialist statistical consultancy for academic, clinical and professional research.
WHEN TO ENGAGE
Analyses That Have Outgrown a Workbook
A workbook can serve a project well until the design, the data or a reviewer asks for something its tools cannot supply.
Tests Beyond the Menu The research question needs a method the add-in does not offer, such as a model for a binary outcome, repeated measurements or a rank-based comparison.
Results to Be Reproduced A supervisor or journal expects the analysis to be rerun from the data, and static tables, pasted values and edited cells cannot show how a number was produced.
A Workbook to Move Years of data sit in tabs with merged cells, mixed formats and codes typed as text, and the file must become a clean dataset before any package can read it.
A Result in Doubt A t-test or regression run in the add-in has produced a result the researcher distrusts, and the question is whether the setup, the data or the method is at fault.
ADD-IN AND BEYOND
What the Add-In Covers and What Follows It
The add-in is judged against the design, so some analyses are checked and kept while others are moved to a package that records them.
Tests the Add-In Runs
Descriptive statistics, t-tests, correlation, single- and two-factor ANOVA and regression run in Excel are checked for setup, with the t-test version matched to the design.
Missing-Value Rules
The regression tool expects a complete block of numeric cells, so blanks and text entries are resolved and recorded before any procedure is run on the data.
Transfer to a Package
The transfer to a package defines variable types, missing-value codes and a codebook of labels once, with dates, leading-zero identifiers and numbers stored as text converted deliberately.
Pivot Summaries
Pivot tables give counts, sums and means by group, a sound way to inspect a dataset, and they are labeled as descriptive since they carry no tests or intervals.
Reproducing the Figures
Before an analysis leaves the workbook, its key figures are reproduced in the new program, so the move can be audited and any change in a result is explained.
The Wider Analysis
Models, diagnostics and reporting for the whole project sit within Quantitative Data Analysis, with the workbook kept as the source file and every change to it recorded.
SPREADSHEET PITFALLS
Risks in Analyzing Research Data in a Workbook
A spreadsheet changes data quietly and records little, which is why problems appear in the tables long after their cause.
Regression Inputs That Break The regression tool expects predictors in adjacent columns, takes at most 16 of them and stops on a blank or text cell, so a larger model must be rearranged or moved to a package.
Ranges That Miss Rows A range that stops one row short, or hidden rows that are still counted, changes a result with no message to warn the researcher, and the output looks entirely normal.
No Welch ANOVA, No Post Hoc Tests The ANOVA tools return an F test without a correction for unequal variances and without post hoc comparisons, so a significant result cannot say which groups differ.
Two-Factor Layouts Must Be Balanced Two-factor ANOVA with replication needs the same number of observations in every cell, so the unequal group sizes of real studies call for restructuring or another tool.
Significance Without an Estimate The add-in's t-test output gives a p-value but no confidence interval for the difference, so reports rest on significance alone unless the interval is computed separately.
Pivot Means Hiding Gaps A pivot table averages only the cells that hold values, so a mean can rest on far fewer records than the table suggests unless the count is shown beside it.
DELIVERABLES
A Checked Dataset and a Record of the Move
Together they show that the numbers survived the change of program and that no value in the data was altered on the way.
Workbook Audit Note A review of the workbook listing merged cells, mixed formats, hidden rows, text-stored numbers and formulas that touch the data, with what was changed and why.
Clean Analysis Dataset One row per case and one column per variable, with consistent types and a defined missing-value code for every variable.
Codebook and Labels Variable names, labels, response scales and value codes in one document, so that the dataset can be read by a collaborator without the original workbook.
Add-In Results Reproduced Each ToolPak result recomputed in the new package, with the differences, if any, explained, and a confidence interval added to every comparison reported.
Descriptive Summary Tables Counts, means and spread by group, with the number of non-missing values beside every mean, replacing the pivot tables that were used for inspection.
Transfer Checklist The checks run on the converted file, such as row counts, ranges and a comparison of summaries with the original, recorded as a short list you can keep.
ILLUSTRATIVE EXAMPLE
A Method Comparison Begun in a Spreadsheet
An illustration of how a result obtained with the add-in is rechecked and extended when the research question asks for more.
ILLUSTRATIVE EXAMPLE · CLINICAL CHEMISTRY ASSAY RESEARCH · UNITED STATES
A doctoral researcher in laboratory medicine compared two assay methods on the same patient samples, with the results kept in an Excel workbook and a paired comparison of means already run through the ToolPak.
That comparison of means answered whether the methods differed on average but not whether they agreed sample by sample, which is the question in a method comparison. The workbook also stored sample identifiers with leading zeros as numbers, so some rows had been matched to the wrong partner. The dataset was cleaned and moved to a statistical package, where the mean difference with its interval and the limits of agreement were estimated, and the paired comparison was kept as one part of the report. The conclusion was limited to agreement within the concentration range that had been sampled.
Outcome: The researcher received the workbook audit note, the clean dataset and codebook, the mean difference with its 95% confidence interval, the agreement analysis and the transfer checklist.
This example does not describe a real client or project. We never share details of actual client work, in line with the confidentiality we promise every researcher. This illustrative example is included to show the type, scope, and standard of work involved in this kind of analysis.
SOFTWARE
Destinations for Analyses That Leave the Workbook
The package is chosen by the analysis the design needs and by the tools your team or field already uses, and each of them can read a spreadsheet file.
HOW WE ASSESS YOUR PROJECT
We Review Your Actual Research
Your assessment is based on a review of your actual research materials and analytical requirements, not a generic project category.
You Submit Your Project Details
Share whatever is available at the current stage, such as the research aim, protocol or proposal, dataset, variable list, manuscript or statistical requirements.
We Review Your Research and Define the Scope
Your statistically trained Research Coordinator reviews the materials to understand the study design, research questions, data structure, analytical requirements and expected deliverables. If clarification is needed, additional documents or information can be requested before the assessment is finalized.
Your Analysis Is Completed and Delivered
Once the scope and fixed project fee are agreed, the in-house statistician completes the work and the Research Coordinator remains the communication point through delivery, with the analysis, deliverables and timeline as agreed.
RESEARCH RESPONSIBILITY
A New Program, the Same Owner of the Conclusions
A change of program changes how an analysis is run and recorded, and it never changes whose study it is or who answers for its conclusions.
Agreed Statistical Support OnlineSPSS provides the agreed methodological, analytical and statistical reporting support, defined in a scope that is confirmed before any work begins.
No Intellectual Contribution OnlineSPSS does not claim any intellectual contribution to your research. Authorship of the thesis or paper that reports the analysis remains yours.
Your Research Decisions The research questions, the meaning of each variable, the interpretation of the results and the conclusions remain yours, with overall responsibility.
Confidential Handling Research data and project details are handled confidentially, project information is not published, and an NDA is available where required.
FREQUENTLY ASKED QUESTIONS
Moving Research Data Out of a Workbook
The add-in's reach, whether Excel counts as statistical software, pivot tables in a paper and fragile identifiers decide how much of a workbook can be trusted as it stands.
What can the Excel Analysis ToolPak do for research data?
The Excel Analysis ToolPak runs descriptive statistics, t-tests, correlation, single- and two-factor ANOVA and regression, and it returns static tables of values. It is adequate for simple comparisons on clean data. Models for binary outcomes, for more than two measurements per person and for clustered data, post hoc comparisons and detailed diagnostics need a statistical package.
Is Excel accepted as statistical software for a thesis or paper?
Excel is a spreadsheet with statistical functions and an add-in, not a statistical package. Descriptive statistics, and a simple t-test or regression on clean data, can be reported when the procedure, version and settings are named. Reviewers often expect confidence intervals, diagnostics and a rerunnable record, which the ToolPak does not supply, and some designs cannot be analyzed in it. The design decides, not the program.
Can pivot tables produce the tables for a paper?
For descriptive counts and means, yes, provided the number of non-missing values is shown and the base of every percentage is stated. Pivot tables carry no tests or confidence intervals, so a comparison between groups is analyzed separately and reported with its estimate and interval.
How are dates and identifiers protected during a conversion?
Identifiers are treated as text so that leading zeros survive, and dates are converted from the serial numbers that a spreadsheet stores into proper date values with the format checked. After conversion, row counts, ranges and a few spot-checked cases are compared with the original workbook, and the checks are recorded.
How do you protect my research data and confidentiality?
Your data and project details are handled confidentially, and project information is not published. Researcher names and institutions are not used publicly without permission. An NDA is available where required and can be arranged before sensitive materials are shared. Read more about confidentiality and data security.
What happens when I request an assessment, and how long does it take?
You share your research question, design and relevant materials, such as the workbook, a list of variables and any add-in output, and a statistically trained Research Coordinator reviews the request and responds the same day. The analysis, deliverables, timeline and one fixed project fee are then confirmed before any work begins, so the timeline follows from the agreed scope.
Still have a question? Find more answers in our full FAQ.
NEXT STEP
Keep What the Workbook
Does Well, Move the Rest
A project rarely has to abandon its spreadsheet entirely: descriptive work can stay, while the tests and models that need records, diagnostics or intervals move to a program that keeps them.
No Commitment Required
Strictly Confidential
Fixed Project Fee
In-House PhD-Level Statisticians
Same-Day Response
Serving Researchers Since 2014