SPSS
A single-user commercial statistics licence.
$99/moto several thousand $/year
Depends on region, modules, and licence type; IBM lists no single public rate.
Range shown for comparison only
SPSS alternative for clinical research
The SPSS alternative that writes up the result, not just runs it.
You already know the routine in SPSS: pick the test, run it, then rebuild the output into a table and write the paragraph by hand. StatsPlease does the whole job: import your existing .sav file, and it checks the assumptions, chooses and computes the right test, builds your baseline-characteristics table, and drafts Methods and Results in journal style, ready to check and paste. No licence, no seat to renew, no syntax window.
Every statistic is computed by our deterministic engine.
Built by a surgeon who knew research statistics shouldn’t require an SPSS licence to get right. Every test is chosen and computed by a fixed engine, the same way an analyst would do it by hand. The AI never computes a number.
Free during open access. No credit card.
Switching from SPSS
What you get instead of an SPSS licence
StatsPlease replaces the day-to-day reasons researchers reach for SPSS: choosing the right test, running it correctly, and writing it up in journal style.
- The right test chosen and computed for your data, with normality and variance checked automatically
- A baseline-characteristics (Table 1) built for you, not assembled by hand
- Methods and Results drafted in journal style, ready to check and paste into a manuscript
- Runs in your browser: no installation, no licence key, no syntax to learn
- Import your existing SPSS (.sav), REDCap, CSV, or Excel files directly
- Every result reproducible: run it again and get the identical numbers
Cost
What an SPSS licence costs, next to StatsPlease
IBM does not publish a single global SPSS price: subscriptions range from roughly $99/month to several thousand dollars a year depending on region and modules. StatsPlease is free for every feature during open access.
StatsPlease
For one researcher, or a whole lab.
Free nowduring open access
No credit card.
Start freeNo credit card
See it work
A finished result, from a real dataset
This is real output from the public Vanderbilt diabetes study, the same starter dataset every new account gets. The numbers are computed from the data; the wording is drafted for you to check and edit.
Results · Vanderbilt diabetes study (n = 390)
Glycemic control in the Vanderbilt diabetes cohort
Stabilized serum glucose was strongly correlated with glycated hemoglobin (HbA1c), r = 0.75, P < .001: as glucose rose, so did HbA1c.
| Measure | HbA1c < 7% | HbA1c ≥ 7% | P |
|---|---|---|---|
| n | 330 | 60 | |
| Age, y | 44.7 (16.1) | 58.4 (13.1) | <.001 |
| Glucose, mg/dL | 91.6 (26.9) | 194.2 (77.4) | <.001 |
Mean (SD). Mann-Whitney U; groups split at the clinical HbA1c cutoff of 7.0%.
And a null result, reported the same way
HbA1c did not differ between men and women (median 4.90 vs 4.79%; P = .225). Not every comparison is significant, and StatsPlease says so.
How it compares
How StatsPlease compares to SPSS
The tool you already know, and the two other places researchers land when they’re deciding what replaces it.
| Capability | SPSS | R | ChatGPT & other LLMs | StatsPlease |
|---|---|---|---|---|
| Numbers you can independently check | ✓ | ✓ | ✗ | ✓ |
| Correct test selected automatically | ✗ | ✗ | ⚠ | ✓ |
| Drafts Methods & Results for you | ✗ | ✗ | ✓ | ✓ |
| No licence, syntax, or stats knowledge needed | ✗ | ✗ | ✓ | ✓ |
| Publication-ready output | ✗ | ✗ | ⚠ | ✓ |
| Reproducible in SPSS | ✓ | ✓ | ✗ | ✓ |
ChatGPT and other LLMs can produce statistics that read plausibly but cannot be independently checked or reproduced.
AI chat tools that write fresh code for every question can give different answers to the same data on different days. StatsPlease runs the same audited engine every time, so a rerun reproduces every value exactly.
Free during open access. No credit card.
And GraphPad Prism
Where StatsPlease fits next to GraphPad Prism
GraphPad Prism is the standard on the lab bench: dose–response and other nonlinear curve fitting, Kaplan–Meier survival curves, and the highly customisable, publication-quality figures biomedical and pharmacology journals expect. If your analysis centres on those graphs, Prism is very good at them, and StatsPlease does not try to replace them.
- Reach for Prism when the deliverable is the figure: curve fitting, survival curves, and finely styled publication graphs from bench experiments.
- Reach for StatsPlease when the deliverable is the write-up: the right test chosen and computed for your clinical design, a baseline-characteristics table (Table 1), and Methods and Results drafted in journal (AMA) style, ready to check and paste.
In Prism you still decide which test fits your design and type up the Methods and Results yourself. StatsPlease reads your columns, checks normality and variance, picks the test, and writes the paragraph around it. Both compute honest numbers you can reproduce. The difference is guided test selection and a finished manuscript section, not the graphs.
GraphPad Prism is a per-seat annual subscription (student, personal, and lab or group plans, priced by region); it lists no single global rate. StatsPlease is free for every feature during open access.
Reproducibility
Don’t trust us. Recompute us.
Each result below is a real analysis StatsPlease ran on a published clinical dataset. The test, the numbers, and the data are all here. Download any one, run it in SPSS, and check every value. Some results are strong and some are small, exactly as the data is.
Pearson correlation
Vanderbilt diabetes study · n = 390
Stabilized glucose and HbA1c were strongly correlated; r = 0.75; P < .001.
Both measures continuous and approximately linear, so Pearson's correlation.
Download the data (CSV)Mann-Whitney U
Heart-failure cohort · n = 299
Serum creatinine was higher in patients who died than in survivors (median 1.30 vs 1.00 mg/dL); U = 14190; P < .001; r = 0.46.
Creatinine was non-normal with unequal variances, so Mann-Whitney U.
Download the data (CSV)Mann-Whitney U
Melanoma survival study · n = 205
Tumour thickness was greater in ulcerated melanomas (median 3.54 vs 1.29 mm); U = 8520; P < .001; r = 0.65.
Thickness was non-normal in both groups, so Mann-Whitney U.
Download the data (CSV)Questions
Common questions about switching from SPSS
Is StatsPlease a full replacement for SPSS?
For the everyday analyses of clinical and epidemiological research, yes: group comparisons, correlations, regression, and categorical tests, in parametric and nonparametric forms, plus the write-up. It is not built for survival analysis, longitudinal mixed models, complex survey designs, or Bayesian methods; for those, a biostatistician remains the right choice.
Do I need to learn SPSS syntax, R, or any code?
No. Upload a CSV or Excel file and StatsPlease reads your columns, checks assumptions, and chooses the test for you. There is no syntax window and nothing to install.
Can I bring in a file from SPSS?
Yes. StatsPlease imports SPSS (.sav) files directly, alongside REDCap, CSV, and Excel, so switching over does not mean re-entering your data.
Are the numbers computed or invented?
The numbers are computed. Every test statistic, p-value, effect size, and confidence interval is calculated directly from your data through a fixed, transparent decision tree, the same way an analyst would compute them by hand. They are not generated by ChatGPT or another large language model, so they reproduce exactly and hold up in review. The Methods and Results wording around those numbers is drafted for you in journal style, then yours to check and edit.
Can I check the results against SPSS myself?
Yes. Every test statistic, p-value, effect size, and confidence interval is computed directly from your data through a fixed, transparent decision tree, so a rerun in SPSS should land on the same numbers. The validation page has worked examples on real datasets, with the data available to download and check yourself.
Why not just use a free tool like Jamovi or JASP?
Jamovi and JASP are genuinely good, free tools, and they compute the same tests correctly: if you already know which test your data call for, they will run it and the numbers will be right. StatsPlease is built for the two steps on either side of that number. First, guidance: it screens each group with Shapiro-Wilk and Levene’s and routes to a parametric, Welch, or rank-based test based on what it finds, so choosing between a t-test and Mann-Whitney is not left to you, and it explains, in plain language, why it chose what it did. Second, the write-up: a baseline-characteristics table (Table 1), a Methods paragraph, and a Results sentence for every test with its effect size and 95% CI, in the P = .04, “1.2 to 3.4” style journals expect. A free GUI hands you a correct results table; StatsPlease hands you the test decision, made and explained, plus the manuscript paragraphs you would otherwise write by hand.
Is it really free?
Yes. StatsPlease is in open access and every feature is currently free, with no credit card required.
Free during open access. No credit card.