Published  |  Last updated

Quick Answer: The best statistical software for social science research depends on your budget, your methods and what your field expects. R and Python are free; SPSS, Stata and SAS are paid products. In a review of 6,469 PubMed articles, SPSS was named most often (52.1%), ahead of SAS (12.9%) and Stata (12.6%).1 That review covers health sciences up to 2017, so treat it as a guide to reporting habits, not to what is best.

Which Statistical Software Do Researchers Actually Report?

Masuadi and colleagues searched PubMed for articles naming one of 14 statistical packages, and after exclusions analysed 6,469 articles from 1997, 2007 and 2017. The paper is open access under CC BY 3.0, so we recomputed the shares from the reported counts and redrew the figures in R.1 SPSS was named in 3,368 articles (52.1%), SAS in 833 (12.9%) and Stata in 815 (12.6%). The percentages add to more than 100% because some articles named more than one package.

Bar chart of statistical software named in 6,469 PubMed articles from 1997, 2007 and 2017: SPSS 52.1% (3,368 articles), SAS 12.9% (833), Stata 12.6% (815)
Share of 6,469 PubMed articles naming each package, 1997, 2007 and 2017. Percentages recomputed in R from the reported counts.

SPSS led in every period: 27.9% of articles in 1997, 59.7% in 2007 and 51.3% in 2017. R rose to third place in 2017 at 11.4%.1 The leading package also depended on the study design.

Bar chart of the most-used statistical package by study design: SPSS 61.1% in observational studies, SPSS 65.3% in experimental studies, Review Manager 43.7% in systematic reviews and meta-analyses, R 60.2% in Review articles/Research support
Leading package for each study design, as reported in the authors’ Table 1. Redrawn in R. The authors’ label for the last group is Review articles/Research support (n = 218); the chart shortens it to Review articles.

What this shows: which packages authors named in health-sciences abstracts and full texts in three sample years, and that the choice differs by design. What it does not show: social-science practice, anything after 2017, which package gives better results, or how many analysts use each package. The authors also note that they searched only PubMed and excluded articles without accessible text.1

The Five Packages Compared

We picked these five because they cover the main routes a social scientist takes: free code-based, free scripting, and three commercial products. Details, prices and version numbers below are as of 30 October 2026, read from each vendor or project page.

R

R is free software under the GNU General Public License, and its site lists R 4.6.2 as the current release.2,3 It suits researchers who want every step in a script that others can rerun, and it led among articles labelled Review articles/Research support in the PubMed review (60.2%, n = 218).1

SPSS Statistics

IBM sells SPSS Statistics as subscription editions: a base edition listed from US$99 per authorized user, plus add-on modules from US$79 covering regression, advanced statistics, custom tables, complex samples and missing values. Perpetual, campus and student licences are listed without prices on IBM’s page.4 It was the most reported package in the PubMed review.1

Stata

Stata 19 is the version StataCorp describes. Its new features include high-dimensional fixed effects, conditional average treatment effects, Bayesian methods, panel-data VAR models and new graphics.5 Stata was third in the PubMed review.1

SAS

SAS/STAT includes more than 100 prewritten procedures covering analysis of variance, regression, Bayesian inference, exact techniques for small data sets and methods for missing values.6 SAS OnDemand for Academics is free for teaching and learning, with a browser-based SAS Studio and up to 5 GB of uploaded data.7

Python

Python releases are open source, and python.org lists Python 3.15.2 dated 25 October 2026.8 For statistics, the statsmodels package estimates models and runs tests, accepts R-style formulas and is released under a modified BSD licence.9 It fits researchers who also clean, scrape or automate data in the same language.

Side-by-Side Comparison

Cost model as listed by each vendor or project on 30 October 2026; prices and versions change, so check the linked pages.

PackageCost model
RFree, GNU General Public License
SPSS StatisticsSubscription, base from US$99 per authorized user; add-on modules from US$79
StataCommercial
SASCommercial; free SAS OnDemand for Academics for learning
PythonFree, open source

Other Statistical Software You May Be Asked to Use

Beyond the five packages above, a supervisor or lab may ask for one of these. Excel: Microsoft says you can get free access to Excel using Microsoft 365 for the web, and it also lists paid plans.10 MATLAB: MathWorks says its Statistics and Machine Learning Toolbox supports t-tests, nonparametric tests, distribution tests and ANOVA, including repeated measures ANOVA.11 Minitab describes itself as statistical software for quality and process improvement, listing t tests, chi-square tests, control charts and design of experiments.12 JMP lists histograms, regression, distribution fitting and design of experiments.13 It offers a free 30-day trial.14

Free Point-and-Click Options: JASP and jamovi

If you want a free program without writing code, two projects are worth testing. JASP describes itself as open-source statistics software that is free and supports classical and Bayesian analyses.15 jamovi is a free and open statistical spreadsheet built on R, with a desktop version and a cloud version that has a free Guest plan, and it shows the R syntax behind each analysis.16

How to Choose Statistical Software for Social Science

Work through these five questions in order.

  1. 1What do your supervisor and target journal use? Ask for a recent paper in your field and see which package its methods section names. Sharing a package makes help and review easier.
  2. 2What can you license? Check whether your university provides SPSS, Stata or SAS. If not, R, Python, JASP and jamovi cost nothing.
  3. 3What methods do you need? List them first, for example multilevel models, panel data, survey weights or Bayesian analysis, then check the vendor page for each.
  4. 4Do you need a rerunnable script? R, Python, Stata and SAS are built around code. Save SPSS syntax if you use SPSS.
  5. 5Can you test it on your data? Run one small analysis in your top two choices before committing.

What to Write in Your Methods Section

Name the package, its version and any add-on packages or modules, and state the test or model by name. Choosing the test comes first: our free selector below helps with that, and the sample size calculator helps you plan before data collection.

Need the Analysis Run or Checked? Consult Editverse

Editverse subject-matter experts work to COPE, ICMJE and GPP 2022 guidance, with institutional and country-specific ethics followed.

Editverse logo

Editverse

SPSS, R and Stata Analysis

Chosen your software but not sure the analysis is right?

You send: your dataset, codebook and research questions.

You get: the analysis run in SPSS, R or Stata, with assumption checks, tables and a methods paragraph for your manuscript.

When: You know what you want to test and need it run or checked.

SPSS/R/STATA Statistical Analysis · $400

Get my analysis run →

Editverse logo

Editverse

Manuscript Writing

Results ready and the paper still unwritten?

You send: your data, results and target journal.

You get: a manuscript written and edited by PhD experts, with statistical review, journal selection and submission support.

When: The analysis is done and the paper is not.

Manuscript Writing Services · $1,500

Write my paper →

Editverse logo

Editverse

Journal Selection

Not sure which journal fits your findings?

You send: your abstract and the kind of journal you are aiming for.

You get: a shortlist of suitable journals matched to your topic and design.

When: You have results and need the right place to send them.

Scientific Journal Selection Services · $115

Find my journal →

Free Research Tools from Editverse

Frequently Asked Questions

What is the best statistical software for social science research?

There is no single best package. It depends on your budget, your methods and what your supervisor, co-authors or target journal use. R, Python, JASP and jamovi are free, while SPSS, Stata and SAS are commercial products with SAS offering a free academic learning environment. In one review of 6,469 PubMed articles, SPSS was the most reported package.

Is SPSS still the most used statistical software?

In a bibliometric review of 6,469 PubMed articles from 1997, 2007 and 2017, SPSS was named in 52.1%, and in 51.3% of the 2017 articles. That review covers health sciences and ends in 2017, so it does not show current use in social science.

Is R really free?

Yes. The R Project states that R is free software under the GNU General Public License, and its site lists R 4.6.2 as the current release. Free to use does not mean free of effort: you will write code or use a point-and-click front end.

Can I use SAS for free?

For learning and teaching, yes. SAS OnDemand for Academics is a free, cloud-based environment with browser-based SAS Studio, and SAS says even people outside a college or university can get free Studio access for learning.

How much does SPSS cost?

As of 30 October 2026, IBM lists a base subscription from US$99 per authorized user and add-on modules from US$79, excluding taxes, with prices varying by country. The subscription renews at the then-current price after the first year. Check IBM’s page for your country and for student or campus options.

Which software is used for systematic reviews and meta-analyses?

In the PubMed review, Review Manager was the most used package for systematic reviews and meta-analyses (43.7% of those articles), followed by Stata (38.3%). SPSS was used in 2.7% of them.

Is Python good for social science statistics?

Python can do it. The statsmodels package estimates statistical models and runs tests, and it accepts R-style formulas, under a modified BSD licence. It suits researchers who also want to scrape, clean or automate data work in the same language.

Does the software I choose change my results?

This post’s sources do not compare results between packages, so we make no claim. Name the package and version in your methods section, keep your analysis script or syntax, and run the same model in a second package if a result matters.

References

  1. Masuadi E, Mohamud M, Almutairi M, Alsunaidi A, Alswayed AK, Aldhafeeri OF. Trends in the usage of statistical software and their associated study designs in health sciences research: a bibliometric analysis. Cureus. 2021;13(1):e12639. doi:10.7759/cureus.12639. Open access, CC BY 3.0; percentages recomputed and figures drawn by Editverse in R. pmc.ncbi.nlm.nih.gov/articles/PMC7872865/
  2. The R Project for Statistical Computing. Home page and news (R 4.6.2 listed as released 29 October 2026; accessed 30 October 2026). www.r-project.org/
  3. The R Project for Statistical Computing. What is R? (free software under the GNU General Public License; accessed 30 October 2026). www.r-project.org/about.html
  4. IBM. IBM SPSS Statistics pricing and editions (accessed 30 October 2026). www.ibm.com/products/spss-statistics/pricing
  5. StataCorp. What is new in Stata 19 (accessed 30 October 2026). www.stata.com/new-in-stata/
  6. SAS Institute. SAS/STAT software (accessed 30 October 2026). www.sas.com/en_us/software/stat.html
  7. SAS Institute. SAS OnDemand for Academics (accessed 30 October 2026). www.sas.com/en_us/software/on-demand-for-academics.html
  8. Python Software Foundation. Download Python (Python 3.15.2 listed, dated 25 October 2026; accessed 30 October 2026). www.python.org/downloads/
  9. statsmodels developers. statsmodels documentation, version 0.15.0 (accessed 30 October 2026). www.statsmodels.org/stable/index.html
  10. Microsoft. Use Microsoft 365 apps for free on the web (page offers free access to Word, Excel and PowerPoint using Microsoft 365 for the web, and links to paid plans; accessed 30 October 2026). www.microsoft.com/en-us/microsoft-365/free-office-online-for-the-web
  11. MathWorks. Statistics and Machine Learning Toolbox (hypothesis tests and ANOVA section; read in a browser, accessed 30 October 2026). www.mathworks.com/products/statistics.html
  12. Minitab. Minitab Statistical Software (accessed 30 October 2026). www.minitab.com/en-us/products/minitab/
  13. JMP Statistical Discovery. JMP data analysis software, with a Try JMP free button (accessed 30 October 2026). www.jmp.com/en/software/data-analysis-software
  14. JMP Statistical Discovery. Download JMP free trial (free 30-day trial request form; accessed 30 October 2026). www.jmp.com/en/download-jmp-free-trial
  15. JASP Team. JASP: free and user-friendly statistical software (accessed 30 October 2026). jasp-stats.org/
  16. The jamovi project. jamovi: a free and open statistical spreadsheet (accessed 30 October 2026). www.jamovi.org/

Editverse Research Team | Updated October 2026