Free and (Mostly) Open Source Data Analysis Software for Academic
Research
Dr. Chinchu C.
Abstract
Data analysis is a crucial task in knowledge creation in social sciences. Free resources
for data analysis provide researchers with greater freedom and make the research process
more accessible and democratic. This article lists some free software which can perform basic
and advanced statistical data analysis tasks. Some software which can perform other tasks
such as text mining are also introduced. Ease of use and functionality are the major criteria
for selecting these software packages.
Keywords: Free Software, Data Analysis, Social Science, Research
Free and (Mostly) Open Source Data Analysis Software for Academic
Research
Dr. Chinchu C.
Introduction
Research is a knowledge creation process and hence crucial to human progress.
However there have been attempts to transform and treat knowledge as a commodity and use
it for economic gain. This is generally viewed as an obstruction to the growth of research and
to the propagation of knowledge. Research in social sciences generally requires the analysis
of data in various forms. Use of software tools is an imperative in this situation. The range of
software needed range from spreadsheets and statistical software to big data analysis and
qualitative data analysis tools. Nonetheless the proprietary software tools available for use in
social science research are priced so high that it is almost impossible for individuals or even
small institutions to buy them. This may also lead to instances of piracy which is again a
punishable offence and an unethical act, detrimental to the quality of research as well. It is in
this context that the use of free and open-source (FOSS) software tools becomes a necessity
and a political act for keeping knowledge free of restrictions. In a broader sense, use and
promotion of free software is an act of responsibility towards the future generations. This
paper presents an introduction to a number of free and/or open source software packages
which can be used for various purposes, mainly by social science researchers.
The Concept of Knowledge Freedom
The idea that knowledge, in its various forms should be freely available to all, is not
new. Foucault (1980) has maintained that knowledge is inextricably linked to power. Thus
knowledge can be viewed as a form of power. In the academia too, knowledge acts as a form
of soft power which can determine the prospects of one's career growth. From such a
position, it becomes imperative that anyone who works for equality and egalitarianism should
also work for making knowledge accessible to all.
The free software movement is based on the philosophy that considers software as a
form of knowledge. The movement for software freedom was initiated because of the
rampant commercialization in the field of software. Proprietary software, also known as
licensed software, imposes a number of restrictions on the user, even after she/he has bought
it. This typically includes restrictions on copying, sharing, modifying etc. The restraint on
modifying the source code of software packages is also a hindrance to the learning process
inherent in the process. Thus it can be argued that even after one buys a software, the
ownership is not entirely transferred to the buyer.
Free and Open Source Software (FOSS) are different from proprietary software in that
they give freedom to the user to run, change or modify, share, and copy a software. The term
‘free’ denotes freedom, not free-of-cost.
Software Licenses
Software packages are distributed under a variety of different licenses, each with a
different view on the rights of the end-user of the product. The software are usually referred
to with the type of license that they belong to. The license types that are relevant to research
tools fall under any of the categories listed below.
Proprietary: The absolute ownership of the software rests with the provider, not with the
customer. Usually the customer is given various rights to use the software, under an end-user
license agreement (EULA). There will be restrictions on the copying, sharing, modifying or
any such operations on the product. Some software need monthly or annual licensing with
payments to the provider.
Open Source: Open source software give the underlying programming code of the software
to the user. The user may modify and even re distribute the modified version, under the terms
of the license agreement. A number of license formats are used to distinguish between
different agreements between the provider and the user.
Free Software: Free software may either be provided free of cost, or at a price. The term
‘free’ refers to liberty of usage, not price. Even though all free software are open-source, all
open-source software may not be classified as free software, according to the philosophical
issues existing between the two movements.
Apart from these, there exists other forms of licensing such as freewares, which are
packages provided free of cost, but without the source code, and freemium packages, where a
basic version of the software is provided freely and more advanced versions have to be
purchased.
Statistical analysis software
A list of free or open-source software packages, mainly used for quantitative
statistical analyses, is given here. Even though most of them provide support for advanced
procedures like text mining, the primary usage is taken as criteria to include them under the
category of statistical software. The list is not exhaustive, and is curated mostly based on
individual preference and experience of the author.
R: Even though touted as a free alternative to proprietary statistical analysis packages
including SPSS, R is not a software package in itself. Rather, R is a programming language
and an environment for statistical analyses (Ihaka & Gentleman, 1996). It is a dynamic
environment, offering many facets for development (Chambers, 2009). R Being a
programming language, R offers a steep learning curve for those without a basic training in
programming. A number of Graphical User Interfaces (GUI) that use R in the background are
available offering reliable and fast data analysis capabilities. This include R Commander,
RKWard, Deducer, Rattle, Jamovi etc.
R Commander: R Commander, developed by John Fox, is a GUI for the R programming
language. It offers one of the most powerful and user-friendly interfaces to use R for
statistical data analysis (Downie, 2016). R Commander, along with its plug-ins like the EZR
and FactoMineR provide a number of functionalities to perform a range of advanced
statistical analyses. R Commander has to be installed as a package from within R, and is not
available as an executable installation file.
Deducer: Deducer is another GUI built as a potential replacement for proprietary data
analysis software (Fellows, 2012). Unlike R Commander, Deducer, along with the JGR (Java
GUI for R) console of R, can be downloaded and installed as a separate package. This makes
installation easier. Once installed, Deducer offers an analysis experience that is much similar
to SPSS, with its intuitive and familiar menu structure. Descriptive and inferential statistics
are available as menu items similar to the popular proprietary packages.
PSPP: PSPP is a GNU project for statistical data analysis (GNU Project, 2015). It is
sponsored by the Free Software Foundation. For users familiar with SPSS, PSPP offers an
interface which is almost a clone of the SPSS menu system. It has a data/variable view
window and menus labeled Data, Transform, Analyze, etc. all of which are reminiscent of
SPSS. Unlike SPSS, PSPP can be downloaded and used free of cost, and offers a good
portion of capabilities provided by the proprietary package. PSPP supports over one billion
cases and variables, and is inter-operable with multiple software like LibreOffice and
OpenOffice.Org. The data files and syntax are compatible with SPSS as well. Elaborate
documentation is available and there is an active online global community that supports the
users.
Openstat: Openstat is a free package for statistical analysis developed by William Miller
(Miller, 2013). Initially designed as an instructional package, it provides a large number of
statistical operations to users, with detailed documentation. Openstat is designed for the
Windows platform only, and may need compatibility software to run in Linux or Mac
platforms.
RapidMiner: RapidMiner is a software platform, first developed by Ralf Klinkenberg, Ingo
Mierswa, and Simon Fischer in 2001. It was initially named YALE (Yet Another Learning
Environment) and later changed to RapidMiner in 2013. It is a freemium software, which
means a low-key edition is available as free software and the full version is proprietary.
RapidMiner is available for Windows, Linux and Mac platforms.
SOFA: SOFA (Statistics Open For All) is one of the most user-friendly and intuitive
statistical packages currently available. The software, developed by Paton-Simpson &
Associates Ltd, can be run on Windows, Linux and Mac platforms (SOFA, 2017). A good
collection of descriptive and inferential statistical tests, as well as visualizations are provided.
The software adopts an instinctive approach where the user is guided through the various
steps to be followed in the analysis, starting from choosing the procedures based on the type
of data and results expected.
JASP: JASP is one of the most sought-after free and open-source alternatives to the
proprietary packages. JASP was created by pooling resources from various Universities and
research funds. In addition to the frequentist procedures offered by traditional software, JASP
provides Bayesian inferential tests as well. Available in Windows, Linux, and Mac platforms,
the tables and plots produced by JASP are in APA style, which helps in preparing
manuscripts. The addition of Bayesian statistics gives an edge to JASP in terms of scientific
rigour. (The JASP Team, 2015).
Jamovi: As of 2021, Jamovi is probably the frontrunner in terms of ease of use and
functionality among the free and open source packages which aim to provide a simple and
powerful alternative to proprietary data analysis software such as SPSS and Stata (The
jamovi project, 2020). Jamovi is an R based GUI which is easy to install and use, with most
functions needed for an academic research project in social sciences bundled with the default
installation. Besides, there is the option of adding more specialized modules using the jamovi
library. These modules add a plethora of additional capabilities to jamovi. With continuous
updates being released, what also make jamovi special is its ability to dynamically modify
analysis results when the underlying data is modified. Cross-platform compatibility is also
ensured, with the ability to import data in formats such as csv, sav (used by SPSS), dta (used
by Stata) etc. Detailed documentation and third party tutorials are also available for jamovi.
In India, Jamovi has become part of many workshops and training programmes in research
methodology, where SPSS used to be the default choice.
MicrOsiris: Microsiris was developed at the University of Michigan. It is based on an earlier
package, OSIRIS IV, developed at the same University. It is a data management and and
statistical analysis tool. Speed and lightness are the distinct features of Microsiris. Microsiris
offers a good user experience, with a clean interface and features like provision to deal with
missing data. It can also import data from various platforms, including SPSS and SAS.
PAST: PAST (PAleontological STatistics) (Hammer et al., 2001) is another free software
package which has recently been recognized as a viable alternative to proprietary software
across social sciences. Originally developed in 2001 for use in Paleontology research, it has
later been adapted to be used in various other fields. PAST offers a rich menu of statistical
procedures to researchers, with some features like simple phylogenetic analysis exclusive for
the paleontology and ecology fields. Features like Google Maps integration are unique to this
package. Only Windows platform is supported at present.
Gretl: Gretl is an open-source and multi-platform software package designed mainly for
econometrics. It is considered as a free software alternative to the popular econometrics
software, EViews, which is a proprietary product. Launched in 2000, gretl is an acronym for
Gnu Regression, Econometrics and Time-series Library. A large range of menus are available
for procedures like time series analysis. Apart from English, gretl supports more than 15
languages, including French, Spanish and Chinese.
Bluesky Statistics: Bluesky became an open source project in 2018, and is being developed
by a community ever since. This is yet another R based GUI which aims to provide a point-
and-click data analysis experience to social science researchers and students. It provides a
familiar user interface for users who have used SPSS or similar software, and some elements,
such as the ribbon below the menu bar resemble popular office packages. The desktop
version of Bluesky is available for free download and use. Like modules in jamovi,
extensions are available for Bluesky. There is a 'variable and data view' functionality similar
to SPSS, and a number of basic and advanced data analysis options too. Books featuring
Bluesky for advanced statistical applications have been published (Lamprianou, 2019).
SalStat: SalStat is another free data analysis software written in the Python language. Ease of
use is a guiding principle for its development, and the developers have designed it
particularly for researchers in social sciences, agriculture and biological sciences, business
analytics, market research etc. Some feature like scraping data from web pages is also
available in SalStat, making it more than just a data analysis tool. A number of descriptive
analyses are available, and decent amount of inferential statistics are provided, including
linear regression. SalStat does not support importing the sav file format used in SPSS. No
new updates have been released since 2018.
Ministat Application: Ministat is an Android application which tries to provide a complete
data analysis experience in a smartphone, and it does this job reasonable well. Ministat
provides data visualizations, t-tests, ANOVA, correlation tests, regression analyses, and much
more. It also provides some useful e-books which can be accessed only if one agrees to watch
an advertisement. A premium version of the application without the advertisements is also
available for those willing to pay. Though not much popular, with only 10k+ downloads as of
December 2021, Ministat provides a good experience of data analysis with a smartphone and
is worth a try.
Software for other research purposes
The software packages in this list are either data mining packages or CAQDAS
(Computer Aided Qualitative Data Analysis Software). Most of them are able to perform
descriptive and inferential statistical operations as well.
Tanagra: Tanagra is a free data mining platform, intended for use in research and academics.
It was developed as an experimental platform for students and researchers (Rakotomalala,
2005). It provides support for machine learning along with exploratory data analysis. The
source code is provided for those interested in working on similar projects. Even though
written as instructional software, Tanagra is widely used in actual studies as well. It has also
been recommended for small-scale industrial usage (Enright & Klippenstein, 2004).
Weka: Named after a flightless bird found only in the New Zealand, Weka (Waikato
Environment for Knowledge Analysis) is a product of the University of Waikato, New
Zealand. It is a multi-platform machine learning software written in Java. Weka is used for
educational and research purposes, along with industry applications (Frank et al., 2009). The
simplicity and user-friendliness of user interfaces is an important feature of Weka which
makes it popular among researchers.
Rattle: Rattle (the R Analytical Tool To Learn Easily) is a GUI package for R, primarily
intended for the purpose of data mining. Rattle is used both as an instructional tool for data
mining in universities and also by consultants and data scientists worldwide (Williams,
2009). Interactions with the interface are also recorded as scripts in the underlying R console,
so that these scripts can be used independently of the GUI. Text mining capability has also
been added to rattle in its later versions.
Orange: Orange is an open-source program for data mining, data visualization, and machine
learning. Python scripting is used in the back-end. It provides an interactive visualization and
analysis platform (Demšar et al., 2013). Operations are done through interactive workflows
combined of widgets, which can be either custom-defined or predefined. Widgets for
Bioinformatics, Time Series Analysis, Natural Language Processing and text mining etc. also
available as add-ons in Orange, along with customary data mining capabilities.
KH Coder: KH Coder is a light-weight program which is developed specifically for
quantitative content analysis and text mining. It was developed by Koichi Higuchi for
personal research purpose and later offered as an open-source software under the GNU GPL
license. Various text-mining procedures can be performed with KH Coder. An elaborate
documentation is also available online.
Conclusion
Usage of software in the various stages of academic research is on the rise. Hence it is
imperative that researchers pay more attention to the various issues and options in the area of
research software, irrespective of their area of academic expertise. Early career researchers
and students should have a keen interest in the various happenings in this area to stay updated
in their academic pursuit. There is no claim that the list of software presented here is
exhaustive or complete. It is hoped that this introduction to the software various packages
available to the researchers in social science will inspire some of them to explore the hitherto
unexplored areas, since research is also identified as a process of defying the beaten path to
discover hidden knowledge.
References
Chambers, J. M. (2009). Facets of R. The R Journal, 1(1), 5–8.
Demšar, J., Curk, T., Erjavec, A., Gorup, Č., Hočevar, T., Milutinovič, M., Možina, M.,
Polajnar, M., Toplak, M., & Starič, A. (2013). Orange: Data mining toolbox in
Python. The Journal of Machine Learning Research, 14(1), 2349–2353.
Downie, T. (2016). Using the R Commander: A Point-and-Click Interface for R. Journal of
Statistical Software, 75(Book Review 3). https://doi.org/10.18637/jss.v075.b03
Enright, J., & Klippenstein, J. (2004). Tanagra: An Evaluation. University of Alberta.
https://webdocs.cs.ualberta.ca/~zaiane/courses/cmput695-04/work/A2-
reports/tanagra.pdf
Fellows, I. (2012). Deducer: A data analysis GUI for R. Journal of Statistical Software,
49(8), 1–15.
Foucault, M. (1980). Power/knowledge: Selected interviews and other writings, 1972-1977.
Pantheon.
Frank, E., Hall, M., Holmes, G., Kirkby, R., Pfahringer, B., Witten, I. H., & Trigg, L. (2009).
Weka-a machine learning workbench for data mining. In Data mining and knowledge
discovery handbook (pp. 1269–1277). Springer.
GNU Project. (2015). GNU PSPP for GNU/Linux (0.8.5) [Computer software]. Free Software
Foundation. https://www.gnu.org/software/pspp/
Hammer, Ø., Harper, D. A. T., & Ryan, P. D. (2001). PAST: PALEONTOLOGICAL
STATISTICS SOFTWARE PACKAGE FOR EDUCATION AND DATA
ANALYSIS. Palaeontologia Electronica, 4(1), 1–9.
Ihaka, R., & Gentleman, R. (1996). R: A Language for Data Analysis and Graphics. Journal
of Computational and Graphical Statistics, 5, 299–314.
Lamprianou, I. (2019). Applying the Rasch Model in Social Sciences Using R (1st edition).
Routledge.
Miller, W. (2013). Statistics and Measurement Concepts with OpenStat. Springer-Verlag.
//www.springer.com/in/book/9781461457428
Rakotomalala, R. (2005). TANAGRA: a free software for research and academic purposes.
Proceedings of EGC’2005, 2, 697–702.
SOFA. (2017). SOFA Statistics (1.4.6) [Computer software]. Paton-Simpson & Associates
Ltd. https://www.sofastatistics.com
The jamovi project. (2020). Jamovi (1.2) [Computer software]. https://www.jamovi.org
The JASP Team. (2015). Software to Sharpen Your Stats. APS Observer, 28(3).
https://www.psychologicalscience.org/observer/bayes-or-bust-with-new-softwares
Williams, G. J. (2009). Rattle: A Data Mining GUI for R. The R Journal, 1(2), 45–55.
About the author: Dr. Chinchu C. is a Consulting Psychologist, Podcaster, Writer, Educator, and
Research Consultant from Kerala, India. His short bio is available at
https://www.mhinnovation.net/profile/dr-chinchu-c
Sources of Funding: No funding agencies have provided financial support or funding for writing this
article.
