Referencing software in publications

Often, researchers will ask questions about how to reference the computing resources that you use in your research publications. As one of our vendors just replied to a user as to how they would like their software cited, I thought that it would be good to give some examples. Note that the standards used in the journal or other location may have rigid styles that you will have to adhere to, and the examples will have to be modified to fit within those formats. The Library has assistance with how you might satisfy those requirements.

A question that some ask is “why would you do that?” The primary answer is the same as for citing other papers – you are acknowledging the source from which you are getting support. There is a secondary answer – you will want to include this information to help others reproduce your results. This may require a bit more work than simply giving a location where the resource can be found. The version of the software and perhaps the type of computing device used can be important information that needs to be referenced. For example, the software package that you’ve used may have a “quirk” that only is found in particular versions of the software. For example, a reviewer of your work is trying to replicate your results with a different version of the software, but that is producing different results as the reviewer’s version has a bug that was fixed in your version.

Keith’s post on REDCap is timely in many ways. For my purposes, it permits me to use as an example how the REDCap people want their software cited. (See the tab marked “Citing REDCap”.) The R Project actually tells you how to reference it when it starts – run the command “citation()” to see the format to be used. (I would highly recommend adding the version you’re using to the information listed there.)

Other vendors or products that provide guidance for citations are: Mathematica, Maple, SPSS,  and Stata.

ESRI, the makers of ArcGIS have a specification on how to reference their data, but not their program. In cases like this, the standard information to include would be the “Author” (i.e. the company making the software), the “Date” (of manufacture), the “Title” (name of the program), “Edition” (version number) and “Publisher” (company including location). So for ArcGIS 10.2.2, the information would be (re-arranged to fit your citation style):

ESRI, 2014. ArcGIS for Desktop, version 10.2.2. Environmental Systems Research Institute, Redlands, CA, USA. (

Another instance of a vendor unspecified format is MATLAB. Someone that had a somewhat frequent need to cite MATLAB wrote a blog article on the way to do it, so that he could easily find it again. (I used to do things like that, but the blog I used is now gone. Now, I make a voyage of re-discovery when I have to recall these kinds of things. In the future, perhaps you’ll see my notes to myself here. And perhaps you just did. 😉

Leave a Reply