| Package | Cost to a South African student | How you drive it | Best for | Main drawback |
|---|---|---|---|---|
| IBM SPSS Statistics | Usually free through your university’s site licence; a paid GradPack exists if not | Menus, with a syntax file behind them | The default in most SA faculties; what your supervisor can help with | Access ends when your registration does |
| R (with RStudio) | Free, open source | Code | Anything, including methods no menu offers; fully reproducible | Steepest learning curve of the four |
| jamovi | Free, open source; desktop “free forever” per the project | Menus, on an R engine, showing the R syntax as you click | A student with no site licence who wants SPSS-style workflow at no cost | Smaller module ecosystem than SPSS or R |
| JASP | Free, open source (AGPL v3), supported by the University of Amsterdam | Menus, with results updating live | Anyone needing Bayesian alongside classical analysis | Fewer data-management features; less common in SA supervision |
Before you compare anything: check what you already have
The most expensive mistake a South African postgraduate makes here is buying a licence for software their university already provides. Terms differ sharply between institutions and change from year to year, so treat the summary below as a prompt to check your own institution’s software page rather than as the final word.
- UCT states that it pays for a number of packages under site licence, and that students may use them “for as long as you are a student” and must remove them on leaving. Its published statistical list includes SPSS, Statistica and JMP Student Edition, with MATLAB for numerical computing and NVivo for qualitative work, and the software is described as installable on your own computer as well as a university machine.
- Wits runs an SPSS licence through ICT, obtained by completing a declaration form. Its postgraduate software list also covers SPSS Amos, JMP, Mathematica, MATLAB, R, SAS, Stata and Statistica for registered students. Stata is handled faculty by faculty — free through named contacts in Health Sciences and in Commerce, Law and Management, but Wits publishes a R750 fee for a Stata licence for all other faculties, which is the single most concrete rand figure in this whole comparison.
- Stellenbosch distributes institutionally licensed software to registered students through its Software Hub, with SPSS, Stata, Statistica, Mathematica, MATLAB, ATLAS.ti and EndNote among the packages named.
- Unisa makes SPSS available to all staff and postgraduate students on request to its College of Graduate Studies statistics support address, and holds a SAS site licence extending to undergraduates. Note one real constraint: Unisa describes its MATLAB licence as available on campus or at a regional centre for students, rather than at home.
- UP asks registered students to request SPSS, Stata, Atlas.ti and Qualtrics through the service portal, and has R, SPSS and Atlas.ti installed on Research Commons workstations.
Two implications matter more than the individual entries. First, if you are at any of these universities, your realistic default is SPSS at no personal cost. Second, that access is tied to registration — which is a genuine argument for learning something you will still be able to open after you graduate.
The ranked shortlist
1. IBM SPSS Statistics — the default recommendation, if your university provides it
Who it suits: the large majority of South African master’s students doing questionnaire-based research in education, nursing, psychology, social work or commerce. It is what your department teaches, what your supervisor can troubleshoot, and what the statistical consultation service will expect to see. That support ecosystem is worth more to a part-time student than any feature list, and our step-by-step guide to analysing questionnaire data in SPSS walks the full workflow.
Where it falls short: your access disappears when your registration does, which is a real problem if examiners request further analysis after you have graduated. It is also expensive without an institutional licence — IBM sells a student GradPack in Base, Standard and Premium editions, but only through authorised resellers, and IBM does not publish a price itself, so the figure depends on which vendor and which country you buy in. Finally, menu-driven work is only reproducible if you actually use the Paste button to build a syntax file, and most students do not.
2. jamovi — the best free replacement for SPSS
Who it suits: any student without a site licence, and anyone who wants their analysis to remain openable for the rest of their life. jamovi is free and open source, describes its desktop version as “free forever”, and looks and behaves enough like SPSS that a student who has used one can find their way around the other in an afternoon. It runs on the R engine and shows you the underlying R syntax for each analysis as you click, which makes it an unusually gentle on-ramp if you ever want to move to code.
Where it falls short: the module library, while substantial, is smaller than what SPSS and R offer, so a specialised technique may not be available. It is also less familiar to South African supervisors, so you may be the person in the room who knows the software best — manageable, but worth agreeing with your supervisor before you commit.
3. JASP — the one to choose if you need Bayesian analysis
Who it suits: students whose work calls for Bayesian methods alongside conventional ones. JASP is open source under an AGPL v3 licence and supported by the University of Amsterdam, and it offers standard procedures in both classical and Bayesian form — which no other package on this list does as accessibly. Results update live as you change options, which makes it genuinely good for learning what a test is doing.
Where it falls short: data cleaning and variable management are thinner than in SPSS or jamovi, so many users prepare data elsewhere and bring it in. If your analysis is entirely conventional, JASP’s main advantage does not apply to you, and jamovi is the easier like-for-like swap.
4. R with RStudio — the long-term investment
Who it suits: doctoral candidates, anyone intending an academic or analytics career, and anyone whose analysis is not in a standard menu. R is free under the GNU General Public License; RStudio’s open-source edition is free under AGPL v3 and remains the standard environment for working in R. Because everything is a script, your entire analysis is reproducible by definition — an examiner query two years later is answered by re-running a file.
Where it falls short: the learning curve is real, and a master’s dissertation with a hard submission date is a costly place to meet it. If you are writing around a full-time job, the honest arithmetic is that weeks spent learning R are weeks not spent writing chapters — and our guide to writing a dissertation while working full-time explains why those weeks are the scarcest resource you have.
The recommendation, stated plainly
Use the package your university licenses — for most South African postgraduates that is SPSS, at no personal cost — and if you have no licence, use jamovi rather than paying for anything. The differences between these four packages will not change your findings. A t-test returns the same answer in all of them. What will change your findings is choosing the wrong test, ignoring a violated assumption, or running out of time; and the package that best protects you against those is the one your supervisor and your faculty’s statistician can help you with at short notice.
The one case for going against that default is if you expect to keep doing quantitative work after this degree. Then learning R during a doctorate — not during a master’s with a deadline — pays for itself many times over.
Two things the package will not do for you
It will not choose your test. Every one of these programs will happily run an analysis that is wrong for your design, and none of them will warn you; our guide to choosing a statistical test is the decision that has to happen first. And it will not write your chapter. Turning output into an argument, justifying your choices in the methodology, and keeping every citation straight is the part that takes the longest and carries the marks — which is what Tesify is built to hold, with your chapter structure, drafts and reference list in one project while the analysis stays yours.
FAQ
Is SPSS free for South African students?
At most large South African universities it is provided under an institutional site licence at no personal cost to registered students, though the request process differs — a declaration form at Wits, a service-portal request at UP, an email to statistics support at Unisa. Check your own institution’s software page, because terms change annually.
What happens to my SPSS licence when I graduate?
Site licences are tied to registration. UCT, for example, states plainly that you must remove the software once you are no longer a student. If you may need to re-run analyses after graduating, keep your data in an open format and consider learning a free package.
Is jamovi as good as SPSS?
For the analyses in a typical master’s dissertation, yes — it covers the standard tests, produces publication-style tables, and is free. SPSS has a wider ecosystem of specialised procedures and far more local support, which is the real difference for a student.
Can I switch packages halfway through my analysis?
Yes, provided your data is in a portable format such as CSV and you re-run everything rather than mixing outputs from two programs. Report which software produced your final results, with its version.
Do I have to declare which software I used?
Yes — name the package and version in your methodology chapter and cite it in your reference list. Get the version from the program’s own About screen rather than from a website.
Is R worth learning for a master’s dissertation?
Usually not, if you have a deadline and a licensed alternative. It is worth learning during a doctorate, or if your analysis genuinely needs methods that menu-driven software does not offer.
What about Stata and SAS?
Both are available at several South African universities — Wits and Stellenbosch list Stata, Unisa holds a SAS site licence, and UP provides Stata on request. Stata is common in economics and public health; if your department uses it, use it. Note that Wits publishes a R750 Stata fee for faculties outside its two supported ones.
Does the software affect whether my results are trusted?
No. What is examined is whether the test suited the design, whether assumptions were checked, and whether the reporting is complete. A correct analysis in a free package beats a flawed one in an expensive package every time.
Can I run these on a low-spec laptop or limited data?
All four run comfortably on ordinary student hardware, and all are desktop applications, so the analysis itself uses no data once installed. Download the installer on campus or workplace Wi-Fi rather than over mobile data.