Do you think in terms of longevity when you choose a software that can collect, process, and analyse data? We all have examples of applications that we know well and are comfortable with, but we often forget that data files need to be managed for longer than a few years; we need to have a long-term perspective in mind. Digital evolution is fast. In a matter of a decade, or a few decades, it may be difficult to find a program that can open old files. Regardless of whether the files belong to a continuous longitudinal study or a valuable research material collected twenty-five years ago, it is important to think ahead when you decide which software to use.
The most important consideration is that there should a good chance that the files you save at the end of a research project can open again in the future. Using open software with open, well-documented file formats gives you the best odds that the files will be readable, even if the program you use now can no longer be used then – or perhaps doesn’t even exist anymore.
Many of the most common software applications are proprietary, meaning that there is an owner (normally a company) who limits how the program can be used and who has access to it. This doesn’t have to be a problem, provided that you can export data from the program in a file format that is a good candidate for long-term use and storage. Then you can export the data files to that file format at the end of the project. However, some data sources (machines, scanners, drones, etc.) are equipped with analysis software that can only save data in its own, undocumented format. If you need to use any of these tools, please be aware that there is a risk that the data becomes unaccessible and unreadable over time, and to use their software only for data that have to be processed that way.
If you develop your own analysis software, bear in mind that the program should have an option to save data in a format that can be read without your software (for example an XML-based format). If this isn’t possible, it should at least be possible to export data to a simple file format, even if it means that some information gets lost. Even if you add the code to the documentation, or make it available in another way, in a few decades’ time it may still be difficult to run the program. (If you use open source code, dedicated people may possibly keep it alive for longer, which decreases the risk that data become unreadable.)
Think about the future when you choose software for managing research data.
Does the software you use have functionality for documenting the project data? From a research data management perspective, you should consider which options the software has for documenting and describing metadata. To document your work and write metadata while you process and analyse data comes naturally to some people, but seems almost impossible to others. It is, however, necessary in order to make the data understandable and reusable for future users. If you can document the process in the same tool as you use for something else, it creates less resistance and makes documentation easier. If you choose between several software applications, pick one with good functions for documentation, instead of one without documentation capabilities.
Remember to explore your options for making documentation easier. Your local research data support unit may be able to assist you with how various analysis programs can be used to document data.
Does your software force you to store data in a non-secure and possibly illegal way? Even though you have made sure that it allows access to the material for all who need to work with it, there are other things to consider. Some software programs don’t store data locally, but in their own storage locations or in cloud locations. Do you need to investigate whether the offered storage solution is legal? How are data backed-up? Is that something you have to order especially, or take care of yourself? Is it even possible to back-up the data? In a worst case scenario, there are user conditions that allow the software manufacturer free access to, or even ownership over, stored data – or the right to at any time erase their servers and everything on them.
You also need to confirm that the storage offered complies with the legislation for the data you process or analyse. Are personal data stored in a way that is allowed according to the General Data Protection Regulation? Are the servers secure enough to meet the requirements of the information classification level of your material?
Avoid using programs that force you to take risks with data and the information they contain. Consult with your local IT Services if you are unsure of what to do.