Page:The battle for open.pdf/178

 reidentification. In the US a person’s date of birth, gender and zip code has been found to be unique for between 61% (Golle 2006) and 87% (Sweeney 2000) of the population. So to release this data requires considerable effort to make it truly anonymisable, and in order to do so, the reduction in the data quality may make releasing it worthless. Ohm (2009) concludes, ‘Data can be either useful or perfectly anonymous but never both.’

These examples are used to illustrate that openness brings with it its own set of problems. One reaction to these types of challenges is often to withdraw, but that is to hand control over to others and for education and academics to be removed from the society in which they exist. Establishing the type of credible online identity discussed in the previous chapter is one element of this, but it will also require understanding and support from the institutions who have a relationship with those individuals.

Conclusions

As well as these issues, previous chapters of this book have raised other problems with the open approach, including:


 * The Gold route for open access leading to unequal publishing opportunities
 * Forcing students to adopt open behaviours that they may be uncomfortable with
 * The low completion rates of MOOCs
 * A route that permits increased commercialisation of education
 * The ­long-​­term sustainability of OER projects

Each of the issues raised in this overview arises because of the open nature of the practice, and in addition there will be other