Software aware or software driven?

Friese is right, of course: The researcher must be in charge, not the software. A point that is made by her time and again (compare Friese 2019 and Friese 2016) and also shared by Marres (2018) who calls in his book “Digital Sociology” for “device aware” instead of “device driven” research.
Indeed, research must be software aware, not software driven! However, the argument that I was trying to make in my blog post was not about the ideal (software aware) but about my frequent observation (software driven). Especially as a software trainer I more often than not see many researchers being driven by the technology they use.
Using the “Digital” Label

Digitised methods and the “neutrality” of data
When it comes to considering the appropriateness of research method Rogers (2013) makes an interesting distinction: He differentiates digital native methods and digitised methods. While digital native methods were created in an already digital environment, digitised methods were “translated” into a digital software, yet are based on analogue forerunners. Although this distinction is not undisputed (see Jarke 2018), I think especially the transformation process is worth thinking about.
Whether we are aware of it or not: The digital transformation is changing our research and, I argue, even the way we think. What software tools we know and what data we have available will influence the way we design our research methods and even our research questions.
When it comes to the data side of the analysis, not even here we can hope for a method neutral or “objective” data base. As Kitchin (2014) points out: Data is never neutral. It is always framed by the technological, economic, ethnic, temporal, local and philosophical context in which it was created. A good example for this is a public database were people can report damages on the road. Due to the varying degree of response by the population in poor and rich areas, the rather wealthy districts are overrepresented. Relying on this data for road improvement would even strengthen the social inequality that created it.
Similar bias can influence our research: Neither the tools nor the data are ever objective, unbiased or methodologically neutral. The (pseudo) objectivity and strongly suggestive power of digital data and tools must not lure us into shutting down the most essential skill of any humanist or social scientist: contextualising our data and reflecting on our tools and methods. After all, the best tool and the best method are always those that allows us to answer our research question in the most appropriate and valid way.
Really nice website and blog posts, Andreas! I found this very useful doing my first steps of software-based QDA. Thanks.
Thank you, Andreas, as always I agree with much of your statement. Software must not determine the researcher and definitely not be used as an argument to raise research funds. There is only one aspect that is important to me which is missing out a little: As humanists we know terms like formation of insights, progress of knowledge and so on. We see ourselves as researchers on a path to a broader understanding. Shouldn’t digital humanists, who really mean “digital” seriously, also participate more in the development of tools and software? Historical-critical methods have also developed over the years. So why not reflect on tools (as methods), improve their usability and contribute to technical (not just intellectual) progress (of existing tools and software)? — In this case, the research questions will be equally digital and humanistic.
You are right, of course, Stefan! If the software is part of the research question, this changes the entire situation. Yes, as a Digital Humanist one is not limited to reflecting on tools but can also change and improve them. However, I thinks this is limited to disciplines that have a strong affiliation towards IT and might not apply for a regular social scientist or humanist.