facility repairs, equipment main- tenance and calibration, off-campus Internet access and personal memberships, also influence data engagement activities (41); severe time pressures (42); cost of wi-fi at home; internet access, esp. pay as you go (42); money to pay access fees for papers.
They do not discuss their data practices in depth but note that they spent 5 months in the field over the course of 2 years. In the first footnote they note the number of interviews in the two countries and the length of time of their research trips (2-3 weeks each). It sounds like they asked participants to keep a “data journal” but the raw journal entries are not discussed or linked to.
I noted the word “data” being used in different ways throughout the paper. My understanding of “data” is research data” which the analyst seems to be using initially. However, then, esp. during the ethnographic aspects, the language of “data” becomes used for internet data bundles. This slippage is not analyzed or discussed.
The analysts note: “purchasing software and hardware was usually the responsibility of the individual instead of the institution. This raises an important concern: the inability to regularly update research hardware and software places researchers in a position in which they are unable to effectively make use of online resources.” It was unclear how / why they note this as unique or different from anywhere else in the world. Are there locations where computers and software/hardware are the responsibility of the institution? Perhaps this is a disciplinary difference since all graduate students in my department (anthro) are responsible to buy and own their own hardware and software. If someone cannot afford a computer, the university, so far as I know, does not do anything about it. Is this perhaps not the case in Europe? This lack of clarity on my part might also be partly because it is unclear who the “who” is in this paper. Students? Staff? Employees?
Underlying assumption that tackling these challenges to usage of data will help to “generate more innovative and productive solutions for the publics and public health settings within its reach”
“Professional membership organizations, networking initiatives, data sharing sites, and Web 2.0 tools offer scientists important means of communicating with peers, gaining access to data resources, and disseminating their own data outside of traditional publication routes” (41)
They note that “data engagement channels and access to information are severely curtailed due to micro-economic concerns.” (43)
They look at the “data engagement activities” of scientists in (bio)chemistry laboratories in Kenya and South Africa but it is unclear what exactly they mean when they say “data” (or “data engagement” for that matter”).The analysis looks very broadly at use of professonal networking sites (as places where researchers build social networks to ask for data or research outputs?).
Specifically, they “selected sites in countries with a robust history of scientific research and that represent major contributors to Africa’s scientific output. We also chose to focus on university laboratories, which, while having engaged in foreign collaborations and received foreign grants, were not part of large research networks. This selection allowed us to come to grips with how the ideals of OS are taken up within the context of a specifically African public institution, illuminating perspectives and practices that might otherwise be obscured by the cultures of a transnational research.” (40)
The authors draw on growing critiques of data and platforms more generally to highlight making data open: 1) is labour intensive, costly and dependent upon infrastructural capacities ; 2) relies on socio-cultural values and political biases are embedded into the very aesthetics of platform design; 3) requires further examination of the variety of information and communication technology (ICT) systems, national infrastructures and research environments necessary to generate, process, disseminate and re-use data.” (40)
“To conceptualize the role and reach of data engagement among African scientists” the authors adapted Amartya Sen’s Capabilities Approach (CA) for human development. (41) They looked at how “an individual transforms those resources into assets and the contexts that constrain that capacity (Sen, 1999, p. 109). Goods and services, in short, are not valuable in and of themselves – they must be ‘converted’ into utilities that a person can use to advance their goals and from which they can ultimately derive some form of value.” (41)
Echoing the discussions raised in the collaboration essay, funding emerges as an important aspect that structures what STS work is done: “chief among these necessary conditions is basic financing. Issues of funding were discussed by every single interview participant and, as in other parts of the world, determined the scope and scale of research activities, which scientific questions were pursued, and how. Rather than simply acknowledging that ‘money is tight’, however, a number of interviewees made a noteworthy distinction between the research activities supported by grant money, and the shortfall not covered by the grants” (41)
This is an important point: “the dichotomies that underpin the very notion of the ‘divide: – e.g. online/offline and access/no access – do not tally with the partial and uneven ways in which Internet resources circulate.” (45)
The authors are worried that the Open Access and Data movement in Africa isn’t taking into account the labor required to “open” the data. They note that mainstream (Western rhetoric) about Open Science is out of sync with the physical, social and regulatory research environments in which scientists operate in Kenya and South Africa. They question the assumption that access = use. This has also been debunked in ICTD (see Donner’s “After Access”) and also increasingly coming to the fore in Open Access discussions as well.
“Rendering science ‘open’ involves contextual- izing, cleaning and curating data so it can be searched and utilized – a process which is labour intensive, costly and dependent upon infrastructural capacities that tend to be obscured by policies seeking to maximize the availability of information online (Leonelli, 2010, 2013).”
Bezuidenhout et al. are largely analyzing why Kenyan and South Africans are not able to / don’t use Open Access and Open Data resources more. However, they are not able to study usage (because it isn’t being used). My project differs in this regard because I am interested in the sociality of the sharing and the questions that arise when we begin to share. Because of my methodology (building organizational archives for the various organizations and using those as probes for discussions), I hope to be able to move beyond what is not (narratives of lack of / deficit) to discussions that take as a starting point that it is possible. I think such an analysis will be able to avoid a well-worn narrative of deficit that is often common in studies of Africa.
Analysts underlying assumption seems to be that research data and work *should* be connected (for better science?) but because of these barriers (named in paper), they are not being connected or used as much as they should or could be. (“These costs, however small in comparison to research budgets, inhibited many of the research participants from engaging with these different platforms – thus missing many opportunities to profile and connect their research to that of others” (42).
Bezuidenhout et al. are strongest on their techno and macro levels of analysis but their nano and micro levels are wanting as it is hard to get a sense of who they were engaging (students? employees? faculty? administrators?) and also what exactly was involved in the “data engagement” that was being discussed. The concept of “data” was used in different valances (by them as well as the interlocutors) and that was not noted or analyzed. I think their use of “data” is largely to mean research outputs? But there is a conflation between open access and open data which are not the same thing. Authors also are conflating Open Data with Open Science which are also not the same thing. I would argue that this is more the “Open Data” movement rather than the maxim of OS (“OS understands democratizing science as increasing the amount of data available” p. 45).
There is also a missing aspect which is the contribution BACK into the online data platforms. Access appears to be a one-sided “use” of outputs of data but the paper doesn’t discuss the importance of circulation (with creation by / from African scholars as well).
There seems to be an underlying assumption that data sharing should be designed to be at a global scale: “clearly what is necessary is a global, coordinated, inter-disciplinary and multi-focused discussion on how to pull these diverse aspects together into a coherent approach.” (45) My own opinion is that data sharing can be valuable even within a very micro-research community, esp. if there is the ability to gain access to it or if it is archived and managed in a way that has the potential for it to be available to others in the future. I do not think data sharing needs to be automatically at a global scale. In fact, that is what is currently being done and it is problematically centralized by private Western players so I don’t think that is the solution.
I also question the assumption that because there are more people on the continent using their mobile phone that data sharing capabilities need to be built for the mobile phone (“The dominance of cellular phones on the African continent – in comparison to other ICTs – and their increasingly effective use in a wide range of financial and health-related applications should provide a focal point for OS initiatives on the continent – too many websites critical for data engagement are highly cumbersome for use on mobile devices”, pg 45). I think the question is *who* the data is for. Is it for lab based researchers? Are lab based researchers all using laptops at work? What functionalities are available on the laptop vs the mobile phone? These are the very “user-specific” contextual factors that should be considered. Who/when/where/why are people using their phones and who/when/where/why are African scientists using data sharing platforms?