Number Crunch

First published in Research Fortnightly HE (February)

Number crunch

Big data affords universities both big opportunities and big responsibilities.

Information is the oxygen of the modern age, the former American president Ronald Regan declared in 1989. The ability of modern technology to harness prodigious volumes of information, including information about the behaviour of individuals, shows how prescient his comment was. As the Higher Education Commission showed in its report published on 26 January, the enormous databases that universities hold can identify trends and insights that lead to more effective methodologies and enhance the student experience. Data that would hitherto have remained dormant in the bowels of university archives can now create profiles of individual students. And personal data, such as students’ access to facilities and learning resources, can be analysed to enable universities to deploy targeted student support mechanisms. Noble purposes indeed.

With great power comes great responsibility, particularly when data are used to make decisions about individuals. Big data potentially represent a form of significantly increased scrutiny of individuals. Responsibility must be exercised, therefore, not only in relation to how the data are applied, but also in relation to how they are collected.

First any collection or use of data from which individuals can be identified must comply with the 1988 Data Protection Act. This oft-maligned legislation contains very valuable protections for individuals, but it also seeks to balance those protections with the need for institutions, such as universities, to collect and use personal data for legitimate purposes. The legislation is not a bah humbug to the march of technology and its ability to inform decisions. It does, however, require that the collection and use of personal data is fair, lawful and proportionate.

Use of personal data must also be justified by complying with one of the prescribed conditions set out in Schedule 2 to the act. One such condition is contractual necessity. Clearly universities cannot discharge their contractual duties to students without engaging in some form of monitoring, profiling and intervening. But the prevailing model of university education has focussed on monitoring performance rather than engagement, output rather than input. The student contract has therefore been performed, some would argue, quite well without recourse to input data. Where therefore is the contractual necessity?

The most obvious condition justifying the use of such personal data is the informed and freely given consent of the students themselves. If students are partners in the teaching and learning project, then obtaining their consent to far greater scrutiny is a reasonable expectation of that partnership. Obtaining informed consent would also fulfil the fairness requirement of the legislation, which relates to the method by which the data have been obtained, including whether individuals have been misled about the purpose for which their data are being collected.

Proportionality requires that personal data used in creating individual student profiles be relevant to the general purposes for which it has been obtained―for example, an assessment of engagement would need to aim to achieve improved performance. If the data do not contribute in some material way to those purposes, then continuing to retain the information is likely to be construed as excessive and a breach of the data protection act.

Data can, of course, be anonymised, in which case the act does not apply. Trends in student behaviour can therefore be analysed not by reference to individuals, but to various categories of individual, such as colour or nationality. Data can reveal differences in approaches to learning and success rates in different methods of assessment. Understood properly, the information can benefit those who may otherwise be disadvantaged by current practices.

But interpreting data is a highly specialised skill and spurious causal inferences can easily be drawn if left to the uninitiated. Where there are differences in treatment of individual students based on flawed inferences relating to protected characteristics, such as colour or national origins, institutions may be vulnerable to complaints of discrimination or of failure to comply with the duty to ensure equality of opportunity.

Big data are precious assets to any institution, but their value lies not in the raw material but in the consequences of its use for individuals. Only fair and transparent collection that does not interfere with individuals’ reasonable expectations of privacy, together with competent and accurate analysis, will ensure that the ability to harness data will result in real benefits for all.


Geraldine Swanton
Senior Associate Solicitor
Education Team
DD: 0121 214 1455
International DD: +44 870 763 1385


Comments - No comments posted yet.

Leave a Comment


Please add 7 and 8 and type the answer here: