Three packages dominate UK undergraduate social science dissertations. All three will run every test a typical project needs. The differences that actually matter are what happens to your access after graduation, how long you will spend learning before you produce a single result, and whether anyone can reproduce what you did.
| IBM SPSS Statistics | jamovi | R | |
|---|---|---|---|
| Cost to you | Commercial; normally via a university site licence | Free | Free |
| Licence | Proprietary | Open source (AGPL3; jmv analyses GPL2+) |
Open source (GPL-2 | GPL-3) |
| Interface | Point and click, with optional syntax | Point and click, spreadsheet-style | Code only (or via an editor such as RStudio) |
| Time to first result | Under an hour | Under an hour | Days to weeks |
| Access after you graduate | Usually lost with the site licence | Keeps working | Keeps working |
| Reproducibility | Good if you save syntax; poor if you only click | Analyses re-run automatically when data changes | Excellent — the script is the record |
| Taught in UK departments | Most widely | Growing | Common in economics, increasingly elsewhere |
| Best for | Following departmental teaching materials exactly | Standard undergraduate analyses on any budget | Anyone continuing to postgraduate or data work |
The shortlist, ranked
1. jamovi — the best default for most undergraduates
jamovi is free and open source, released under the AGPL3 with its analysis package jmv under GPL2+, and is built on top of the R statistical language. In practice it looks and behaves like a modern SPSS: a data spreadsheet, a menu of analyses, output that appears in a panel beside your data.
Who it suits: anyone running the standard undergraduate set — t tests, ANOVA, correlation, regression, chi-square and the non-parametric equivalents — who wants results today and does not want to lose access to their own analysis after submission.
Where it falls short: the analysis library is narrower than SPSS’s or R’s. Highly specialised procedures may not have a module, and community modules vary in maturity. If your supervisor has given you SPSS click-paths for a specific technique, translating them costs a little time.
A genuine advantage worth knowing: jamovi re-runs your analyses automatically when the underlying data changes. If you discover a coding error in participant 43 at 11pm, you fix the cell and every table updates. In SPSS, a click-only workflow means redoing the analysis by hand and hoping you remember every option you set.
2. SPSS — pick it when your department has built the course around it
IBM SPSS Statistics, currently at version 32, is the package most UK psychology, sociology and education departments teach. It is commercial software; students normally reach it through an institutional site licence rather than buying it, and those licences are restricted to university members and to non-commercial use.
Who it suits: students whose module materials, lab handouts and supervisor feedback are all written in SPSS. Fighting that is a poor use of your remaining weeks. If every worked example you have been given shows an SPSS dialog box, use SPSS.
Where it falls short: two real risks. First, access is tied to your enrolment — students routinely discover over the summer, or after graduating, that they can no longer open their own analysis. Second, a click-only workflow leaves no record of what you did, which makes correcting an error late in the project genuinely painful.
Mitigation: use the Paste button. Every dialog in SPSS can write its syntax to a file instead of running silently. Build a syntax file as you go and you get reproducibility roughly equivalent to R, plus a document you can include in an appendix.
3. R — the long game, if you have the time
R is a free software environment for statistical computing maintained by the R Core Team, with copyright held by the R Foundation for Statistical Computing in Vienna. The current release is 4.6.1, published in June 2026.
Who it suits: students heading for postgraduate study, research assistantships or any data-facing career; students in economics and quantitative geography where R is already the departmental norm; and anyone whose analysis needs something the point-and-click packages do not offer.
Where it falls short: the learning curve is real and it arrives at exactly the wrong moment. Starting R from scratch six weeks before submission is a way to spend your remaining time debugging rather than analysing. Learn it before your final year, or alongside a package you already know.
The compromise most people miss: because jamovi runs on R, you can work in jamovi’s interface and switch on syntax mode to see the equivalent R code for every analysis. That is a genuinely gentle on-ramp — you get your results now and learn the language from your own project.
The recommendation
Use jamovi unless your department has given you SPSS materials, in which case use SPSS and press Paste on every dialog. Learn R deliberately, in advance, rather than under deadline pressure.
The reason to lead with jamovi is not ideological. It is that a free, open-source package removes the single most common statistics emergency in UK undergraduate dissertations — losing access to your analysis at the point you most need to revisit it.
What none of them will do for you
No package chooses your test. All three will happily run an independent-samples t test on repeated-measures data and return a confident-looking p value. The decision about which test your design calls for is made before you open the software, from your hypothesis type, your levels of measurement and your design — our guide to choosing the right statistical test works through that in order.
Equally, none of them will tell you whether you recruited enough people. That is a power calculation you should have run before collecting data, and if you are writing it up retrospectively, the honest route is a sensitivity analysis — covered in our guide to sample size for an undergraduate dissertation.
And whichever package produces your tables, your methodology chapter still has to name it, name the version, and justify the analysis strategy before the results appear. That is a writing job, not a software one, and it is set out in our guide to writing the methodology chapter.
How to cite the software you used
Name the package and the version in your methods chapter — “analyses were conducted in jamovi 2.7” — and include a formal citation if your handbook asks for one. jamovi asks to be cited as The jamovi project (2026). jamovi (Version 2.7) [Computer Software], retrieved from its website. R has a standard citation available from within the software itself. For SPSS, cite IBM and the version number.
Markers notice when a version number is missing, because it is the difference between a reproducible method and a vague one.
Once your analysis is done, the results chapter still has to explain what the numbers mean. If that is where you are stuck, you can draft it in Tesify from your own output — your data, your interpretation, 100% written by you.
Frequently asked questions
Is jamovi accepted by UK universities for dissertations?
Yes. It is legitimate, citable statistical software and produces standard output. Your department cares that the analysis is correct and properly reported, not which menu you clicked. If in doubt, tell your supervisor which package you are using before you start.
Will I lose access to SPSS after I graduate?
Usually, yes — university site licences are restricted to current members of the institution. If you expect to revisit your analysis after graduating, either work in a free package or export your data and syntax before your account closes.
Can I switch packages halfway through my analysis?
You can, and it is sometimes the right call, but re-run everything in the new package rather than mixing outputs. Two packages can produce slightly different values for the same test because of differing defaults, and an unexplained inconsistency in your results chapter invites questions.
Do I need to learn R to get a good mark?
No. Marks come from a defensible design, a correct analysis and a clear interpretation. R is a career investment, not a marking criterion.
What about Excel?
Fine for data entry and simple descriptives, and genuinely useful for cleaning. It is a poor choice for inferential statistics: the tests available are limited, the output is not in a reportable format, and it does not record what you did. Enter data in Excel by all means, then analyse elsewhere.
Is JASP worth considering as well?
Yes, if you want Bayesian analyses alongside conventional ones. JASP is free, open source and supported by the University of Amsterdam, and it sits in much the same niche as jamovi. If your project has no Bayesian component, either is a sound choice — pick the one your supervisor can help with.
Does my university’s site licence let me install SPSS on my own laptop?
Usually yes, for the duration of your enrolment and for non-commercial academic use only. Check your IT services pages for the exact terms, since they vary by institution and are enforced.
Which package should I use if my data are qualitative?
None of these three. SPSS, jamovi and R are built for numerical analysis. Interview transcripts and open-text responses are coded in qualitative analysis software, or perfectly respectably by hand with a spreadsheet and a clear coding frame. Using a statistics package to count word frequencies in interview data is not thematic analysis, and markers will say so.
My supervisor uses Stata. Should I switch to that?
Only if your department teaches it and can support you. Stata is common in economics, public health and some social policy departments, and where it is the local norm the same logic applies as with SPSS: use what your supervisor can help you debug. What matters is that you can explain and reproduce your analysis, not which package produced the table.
Do I need to include my raw output in an appendix?
Check your handbook, as practice varies. Where it is required, include the syntax or script rather than screenshots of every dialog box — it is shorter, it is searchable, and it demonstrates that your analysis was planned rather than clicked together.