Page:The battle for open.pdf/177

 The global, uncontrollable nature of such events puts the relationship between the individual academic and their institution under a new type of strain. Similarly, for academics who work in potentially sensitive subject areas, such as Middle-​­Eastern politics, climate change or evolutionary psychology, then pressure to be open and establish an online identity may subject them to particular groups with strong interests.

A further issue to consider with relation to openness is that of cost. Individuals often overestimate the time it takes to engage with tools such as blogs and social networking. While establishing an online identity does take some time, there is a period of investment, which has benefits once an identity has been established. Online networks can act as effective information filters, respondents to specific queries, research groupings for formal projects and dissemination routes, making it a ­time-​­saving practice. However, the cost of other aspects of openness may be underestimated. One example is that of open data. It may seem fairly trivial to release data for a particular ­project – whether this is through the project’s own website, attached to a relevant publication or in a central repository. For many projects, in the hard sciences especially, this is the ­case – ­publicly sharing data from a collection of geology samples for instance. But as soon as human subjects are involved, data sharing becomes more complex. While it is easy to anonymise data, it turns out that deanonymisation is also not as difficult as one might imagine. In order to make any data that deals with people open, whether it is surveys, data records or interviews, researchers either need their consent to make it available as it is (a video interview for example), or they need to anonymise it. This involves removing identifiers such as name or student ID number. However, other pieces of data which are required for the data to be useful for researchers are also sufficient to allow for