This flowchart walks you through some important considerations in deciding whether or not to learn R. Take a look, and check below for a bit more detail. If you do decide to learn R, here's our rundown of how to get started.
If you're using SPSS, Stata, or SAS, the only reason not to switch to R is that you have to learn something new. R can do just about anything those packages can do, but it's free. R has superior visualization and reporting capabilities. Plus, it has a fantastically welcoming and ever-growing community.
The great thing about using a spreadsheet is that it's easy to see what's going on. Everybody can read a spreadsheet—which makes them a great option for collaborating with people who don't know how to use more advanced tools.
The bad thing about spreadsheets is that it's hard to see why things are going on. Typos and bugs in formulas can be incredibly hard to find, which means you may end up doing things manually; putting you in a situation where you have to remember what you did and do it again with new data.
Replacing or supplementing your spreadsheet use with R has advantages for both reproducibility and automation. In R, you can script operations on data: importing, cleaning, analyzing, graphing, and reporting. This gives you an audit trail focused on the process of what is happening, which is easier to debug and adapt to new purposes. Even better, you can re-run scripts on new or updated data with a single command.
Spreadsheets are the foundational tool for working with data. This means that 1) you'll always use spreadsheets in some instances, especially when sharing data with other spreadsheet users, and 2) R isn't necessarily the best next step. If most of your needs are graphs and simple statistics (like counts, percentages, and rates), a business intelligence tool might be a better choice. This is especially true if you want to share interactive graphs with lots of people in your organization.
Tableau, Power BI, Qlik, Quicksight, or other business intelligence tools are the best way to share interactive dashboards and graphics with colleagues. However, R can add to their capabilities with more advanced and more flexible analytic and statistical options, and with more sophisticated data cleaning and processing.
If data cleaning is your main need, you may prefer a dedicated data processing tool like Alteryx or TriFacta that provides a similar graphical user interface to business intelligence tools instead of R. These tools won't be free like R, but they will be easy to pick up.
Python and R are pretty similar in their capabilities for working with data. Python is best-in-class for machine learning (R can do it too, it's just a little harder). R has better visualization and reporting capabilities (Python can do it too, it's just a little harder). R is very popular in the social sciences, and with its wonderful reporting capabilities (primarily through
knitr packages). We think that R is generally the better recommendation for people working in the social sector, but you can certainly be very successful with either.
If you're ready to learn R, please visit this site to download the appropriate packages. Once you've downloaded it, see our article on Getting Started with R for concrete next steps, including some of our favorite resources for learning R.