The best free alternatives to SPSS for thesis work are JASP and jamovi (point-and-click, APA-style output), PSPP (an SPSS look-alike that reads .sav files) and R (the most powerful, with a learning curve); all of them compute the same statistics as SPSS, so a t-test, Cronbach's alpha or regression gives identical numbers. What changes is the interface and the range of analyses. Your examiner cares that the analysis is correct and reported against cited thresholds, such as alpha ≥ .70 (Nunnally, 1978) or Cohen's (1988) effect size benchmarks, not which program produced it.
The main options
| Tool | Best for | Strengths | Limitations |
|---|---|---|---|
| JASP | Students who want SPSS-like menus | APA tables, effect sizes by default, Bayesian tests, SEM module | Fewer data-management features |
| jamovi | Teaching and coursework | Spreadsheet view, add-on modules, shows equivalent R code | Some advanced models need add-ons |
| PSPP | Users with SPSS habits and .sav files | Familiar layout and syntax | No SEM, limited graphs |
| R | Anything advanced | Every method, including lavaan for SEM (Rosseel, 2012) and psych for EFA | Needs coding |
| Browser tools | Quick, specific analyses without installation | Works on any device, APA output, data stay on your computer | One analysis per tool |
What you need for a typical survey thesis
Most quantitative dissertations need five things: data cleaning (including reverse coding), descriptive statistics, reliability (Cronbach's alpha), factor analysis (EFA, sometimes CFA) and a hypothesis test (correlation, regression, t-test, ANOVA, or SEM). JASP and jamovi cover all of these except PLS-SEM. R covers everything. PSPP covers the basics and EFA but not SEM.
Worked example: the same t-test, four programs
A student compares satisfaction scores between two groups of 50 participants. Group A has M = 3.92 (SD = 0.71), group B M = 3.58 (SD = 0.68). In SPSS, JASP, jamovi and R, Student's t-test gives the same result: t(98) = 2.45, p = .016. JASP and jamovi report Cohen's d by default, here d = 0.49; you can check it from d ≈ 2t / √df = 4.90 / 9.90 = 0.49. By Cohen's (1988) benchmarks of 0.2, 0.5 and 0.8, this is a medium effect, just below 0.5.
Group A reported higher satisfaction (M = 3.92, SD = 0.71) than Group B (M = 3.58, SD = 0.68), t(98) = 2.45, p = .016, d = 0.49, a medium effect (Cohen, 1988).
Note that d keeps its leading zero in APA 7, because effect sizes like d can exceed 1, while p does not (American Psychological Association, 2020). Small differences can appear in default settings, for example whether Welch's correction is applied, so always state which version of a test you ran.
Practical tips
- Keep a clean master file in CSV, which every program opens. Never overwrite raw data.
- Record every step (syntax in R, the analysis tree in jamovi or JASP) so you can rerun it when your supervisor asks for a change.
- Check defaults: missing-data handling, one- or two-tailed tests, rotation method in EFA.
- Cite the software and version in your method chapter. Each program provides a recommended citation in its "About" page or documentation.
Reliability and factor analysis in the free tools
Cronbach's alpha is under "Reliability" in JASP and jamovi and under Analyze, Scale in PSPP; in R it is psych::alpha(). All four report the same coefficient and item-total statistics as SPSS, so you can apply the same .70 threshold (Nunnally, 1978). For EFA, JASP and jamovi report KMO and Bartlett's test, and their default extraction and rotation settings differ from SPSS, so state the method and rotation you chose rather than relying on defaults.
Does free software affect your marks?
No reputable examiner requires SPSS. What they look for is a justified analysis, correctly reported statistics and thresholds with sources. If your programme provides SPSS and your supervisor uses it, learning it is a sensible choice; but you will not lose marks by using R or JASP, and you keep access after you graduate.
Run it free
Without installing anything, the free survey data analysis tool runs descriptives, alpha, EFA, correlation and regression in the browser and exports APA tables to Word. See all tools in thesis order, and the comparison with R.
FAQ
Can JASP or jamovi open SPSS files?
Yes, both open .sav files, as does PSPP.
Is PSPP the same as SPSS?
It imitates the interface and syntax for core procedures, but lacks many advanced modules.
Which free tool runs PLS-SEM?
The R package seminr, or a browser-based PLS-SEM tool. JASP and jamovi focus on CB-SEM via lavaan.
Will my results match a classmate who used SPSS?
Yes, with the same data and settings. Differences point to different settings or different data cleaning.
References
- American Psychological Association. (2020). Publication manual of the American Psychological Association (7th ed.). https://doi.org/10.1037/0000165-000
- Cohen, J. (1988). Statistical power analysis for the behavioral sciences (2nd ed.). Lawrence Erlbaum Associates.
- Nunnally, J. C. (1978). Psychometric theory (2nd ed.). McGraw-Hill.
- Rosseel, Y. (2012). lavaan: An R package for structural equation modeling. Journal of Statistical Software, 48(2), 1–36. https://doi.org/10.18637/jss.v048.i02