The interviews with DARIAH-FI participants in Autumn (by Anna Sendra Toset from WP5) brought up some interesting results, which we wanted to test by conducting a short follow-up survey. In total 20 people from DARIAH-FI took part in the survey, which is not fully representative, yet gives some indications. Of those 20 people surveyed, 9 associated themselves with “data science, computer science, language technologies, or other technical discipline”, and 9 with “humanities, social sciences, arts, or other non-technical discipline”, thus providing equal share of answers for technical and non-technical professions.
DARIAH-FI project goal and main target audience
The information on DARIAH-FI homepage states that it is a “network for data-intensive social science and humanities research”, however when asked about their vision of what are the specific goals DARIAH-FI, there was a variety in interviews’ answers.
One approach could be characterized by focus on strengthening existing research teams and enhancing the existing research – support the research that is already ongoing, support the connections between the teams in the project.
On the other hand, there were views that DARIAH-FI should primarily focus on people outside the project – the results should be directly usable by a wider community of researchers with computational skills or even without such skills.
To test opinion of other DARIAH-FI project participants, the question about the target audience was formulated as a multiple-choice question (see Figure 1).
The main target group seems to be the larger community of SSH researchers, but the patterns across the answers again are manifold. The patterns are equally different for technical and non-technical respondents.
One key take-away from this is that work packages and the tools being developed and their target audiences are very different, and perhaps different target groups should be targeted at different parts of DARIAH-FI, without trying to enforce a common goal and a fixed audience.
Documentation and User Support
Documentation is a crucial part of every digital tool or interface – it serves the purpose to document the tool and its use, both as a reference for the research team working with this tool daily and for outer use. Documentation as a description of functions, API calls and installation of the tool is useful when communicating between data scientists and software developers. In order for the tool to be accessible for non-technical users, tutorials and user support are often needed.
When working in combined research teams, the technically more complex tasks, like running models, as well as installation and setting up the tool and troubleshooting errors is done by computer scientists or code developers. It would be unrealistic to think that an outside humanities researcher will solve technical problems, which an in-house researcher delegates to technically educated colleagues. Therefore, the survey question about documentation was formulated with explicitly giving multiple options (Figure 2).
Survey results show that both technical and non-technical ranked tutorials and documentation equally high as “very important” or “important”, while user support got significantly lower scores.
Apart from recommending every digital tool of DARIAH-FI to have a documentation, tutorials and users support, it is worth noticing that project partners are different. There are “service-oriented” institutions like National Library and CSC and more “research-oriented” ones like teams in different universities (see Figure 3).
The difference is crucial, as “service-oriented” institutions serve large number of people, they have dedicated staff (bibliographers, consultants) who work with users and clients; they have manuals and instructions and documentation to facilitate support of users; tools are beginner to mid-level, usually, (CSC being an exception for they offer services from easy to complex).
Research institutions focus on scientific excellency; teams are rather closed; content and tools might be unintelligible to non-experts of the field; technical support targeted at research team’s hardest problems.
Here service oriented DARIAH-FI institutions, particularly CSC, can serve as a model for research-oriented institutions on how to implement the best practices of documentation, tutorials and users support.
Cooperation inside DARIAH-FI
Several areas of cooperation were identified, and to better measure the general attitude a multiple-choice question was structured as in Figure 4.
From the options in Figure 4, sharing good practices gets the highest score for both technical and non-technical respondents. Bringing together researchers from different projects seems to be more important for the non-technical people. DARIAH-FI website gets the lowest score for both technical and non-technical respondents, perhaps reflecting the previous experience of how hard it is to maintain a usable, up-to-date portal from data gathered from different sources.
Conclusions
The above answers from the survey further demonstrate that the ways to achieve the fundamental project goal of DARIAH-FI in 2022-2023 – to facilitate data-intensive social science and humanities research – are diverse across researchers and work packages. Some research is directed at skilled researchers and is promoting novel research methodologies, which will benefit the wider research community only indirectly. Other work packages are aimed at creating ready-to-use tools for not yet experienced researchers in data-intensive social sciences and humanities, and in these cases thorough documentation and tutorials are recommended.
By Harri Haralds Matulis