Freedom of Information in the UK and its Implications for Research in the Higher Education Sector

Freedom of information legislation came into effect in the UK in 2005. All universities that receive block grants from the Higher Education Funding Councils in England, Wales, Scotland and Northern Ireland are subject to the legislation. Recent cases where universities have received requests for data and other information generated by researchers, working in areas such as climate, have given rise to controversy and widespread concern in the research community. This paper examines some of those concerns, relating to responsibilities for the ownership and holding of information, for data and records management, and for the handling of requests under the legislation. It also considers the implications relating to personal data, and to information that may affect the commercial interests of universities operating in a competitive environment, or the interests of the many other organisations which may be involved in research partnerships with universities; and it outlines concerns about the possible impact on quality assurance, peer review, and scholarly discourse. Finally, the paper emphasises the need for support and training for researchers so that they become more aware of the legislation and its implications, and how to deal with requests when they arise. International Journal of Digital Curation (2012), 7(1), 57–71. http://dx.doi.org/10.2218/ijdc.v7i1.214 The International Journal of Digital Curation is an international journal committed to scholarly excellence and dedicated to the advancement of digital curation across a wide range of sectors. The IJDC is published by UKOLN at the University of Bath and is a publication of the Digital Curation Centre. ISSN: 1746-8256. URL: http://www.ijdc.net/ 58 Freedom of Information doi:10.2218/ijdc.v7i1.214 Background: The Legislation Over eighty countries across the world now have laws that provide for access to information held by the state. In the UK, this legislation comprises the Freedom of Information (FoI) Act 2000, the related but distinct FoI (Scotland) Act 2003 and the Environmental Information Regulations (EIR) 2004. They came into force on 1 January 2005, and for the purposes of this paper, they are referred to collectively as FoI legislation. The aim of the legislation is to make public bodies more open and accountable, and like similar legislation in other countries, it starts from a presumption in favour of disclosure. It is important also to note that access to information under the legislation is motive-blind: anyone, anywhere in the world, can request information, and their reasons or motives for so doing are irrelevant to a decision on whether or not to meet the request. Moreover, once someone has received information under the legislation, they are free to pass it on to anyone else: the information is effectively in the public domain. A significant feature of the UK legislation is that it explicitly defines universities – as well as publicly-funded research institutes – as ‘public authorities’ so that they and the staff they employ are therefore subject to its requirements. This means that anyone can submit a request for any data or information generated by researchers during the course of their work at a university or research institute. The university or institute then has to provide the information requested unless there are good reasons (under ‘exemptions’ or ‘exceptions’ prescribed in the legislation) why they need not do so. These requirements apply, unlike those in other countries such as the US, regardless of the source of the funding for the research. Although UK universities – with the exception of the independently-financed University of Buckingham – depend on public funds for a significant part of the income they receive for both teaching and research, public policy has also laid great stress on their autonomous status. Hence they are unlike most of the other bodies defined as ‘public authorities’ under the legislation. Even before the Government’s decision to implement a system of much higher student fees, universities were becoming decreasingly reliant on public funds. Less than half the funding for major research-intensive universities comes from public sources. Moreover, unlike most ‘public authorities’, universities operate in a highly-competitive environment, and with their own strong commercial interests. Most of them are also charities, which means that they have a duty under the Charities Act 2006 to preserve and exploit their assets, as well as to disseminate knowledge, including the results of research. All these considerations can give rise to tensions and difficulties in complying with the legislation. 1 Freedom of Information (FoI) Act 2000: http://www.legislation.gov.uk/ukpga/2000/36/contents 2 FoI (Scotland) Act 2003: http://www.legislation.gov.uk/asp/2002/13/contents 3 Environmental Information Regulations (EIR) 2004: http://www.legislation.gov.uk/uksi/2004/3391/contents/made. These Regulations are based upon the requirements of a European Union Directive. 4 A recent study suggests that UK universities enjoy greater freedom from state control than those in any other European country. See Estermann, Nokkala & Steinel (2011). The International Journal of Digital Curation Volume 7, Issue 1 | 2012 doi:10.2218/ijdc.v7i1.214 Michael Jubb 59 It is notable that many of the issues and concerns now being raised in the UK relating to access to research data under the legislation were raised more than thirty years ago in the US, following decisions under the FoI Act passed there in 1966. Thus in 1977, the National Commission for the Protection of Human Subjects of Biomedical and Behavioral Research was asked to “investigate and study the implications of public disclosure of information contained in research protocols, hypotheses and designs submitted to the Secretary of Health, Education, and Welfare in connection with applications or proposals for grants, fellowships or contracts under the Public Health Service Act” following a court decision which held such information generally to be disclosable under the FoI Act. Its report noted that the decision had “caused concern to many members of the research community, who take the position that an investigator’s ideas and methodology are his or her ‘stock-intrade’ and thus deserving of protection from disclosure.” It made a number of recommendations to mitigate those concerns (National Commission for the Protection of Human Subjects of Biomedical and Behavioral Research, 1977). Two decades later, the Shelby Amendment to the US Act gave rise to further controversy. The amendment required Federal awarding agencies to ensure that all data produced under an award would be made available to the public through procedures established under the Act. Leading universities and bodies, such as the National Academies, lobbied against the amendment and recommended significant modifications to the draft language. While not disagreeing with the basic objective of making scientific data publicly available, they argued that doing so through the FoI Act would be costly and burdensome, that public release of data prior to publication in scholarly journals would seriously short-circuit the scientific research process, that it would have a chilling effect on university-industry collaborations, and that it could lead to special interest groups harassing researchers. All these issues, as we shall see, are now being raised in the UK. Consultations and Guidance Until recently, very few researchers in the UK have been aware that the information and data they gather, create and analyse in the course of their research are subject to the provisions of the FoI legislation. However, recent and well-publicised cases surrounding the disclosure of data and other information about climate change at the University of East Anglia and Queen’s University Belfast have raised some concerns about the implications of the legislation. The Independent Climate Change E-mail Review headed by Muir Russell (Russell et al., 2010) and a follow-up inquiry and report from the House of Commons Science and Technology Committee (2011a) highlighted the “confusion and unease” in the research community, and the need for guidance to the higher education (HE) sector. The Committee recommended that the ICO should provide such guidance as a matter of urgency; and the guidance was issued in September 2011 (ICO, 2011a). 5 See, for example, the 15 July 1999 statement of the President of the President of the National Academy of Sciences to the sub-committee on Government Management, Information and Technology, Committee on Government Reform, US House of Representatives (http://www7.nationalacademies.org/ocga/testimony/foia.asp); and the Memorandum of 6 December 1999 to Harvard University Deans, Department Chairs, Faculty and Research Staff (FOIAmemo11-2999-1.doc). The International Journal of Digital Curation Volume 7, Issue 1 | 2012 60 Freedom of Information doi:10.2218/ijdc.v7i1.214 Following the Muir Russell Report, three related initiatives were put in place. First, the Information Commissioner’s Office (ICO) held a roundtable meeting with representatives of the HE sector (ICO, 2010). As a result of that meeting, an ICO-HE Sector Panel has been established to provide a forum for discussion and to address issues of concern, along with a sub-panel with a specific focus on research (ICO, 2011b). Second, a Q&A briefing paper on FoI and research data was published by JISC (2010). Third, a series of three workshops in London, Manchester and Glasgow was organised by the Research Information Network (RIN) in concert with the ICO, the Scottish Information Commissioner, JISC Legal, the National Archives, the Information and Records Management Society, and the UK Data Archive. The purpose of the workshops was twofold:  To raise awareness of the FoI legislation and its implications; and  To gather evidence from researchers and others in the HE sector about thei

The International Journal of Digital Curation is an international journal committed to scholarly excellence and dedicated to the advancement of digital curation across a wide range of sectors.The IJDC is published by UKOLN at the University of Bath and is a publication of the Digital Curation Centre.ISSN: 1746-8256.URL: http://www.ijdc.net/

Background: The Legislation
Over eighty countries across the world now have laws that provide for access to information held by the state.In the UK, this legislation comprises the Freedom of Information (FoI) Act 20001 , the related but distinct FoI (Scotland) Act 2003 2 and the Environmental Information Regulations (EIR) 2004 3 .They came into force on 1 January 2005, and for the purposes of this paper, they are referred to collectively as FoI legislation.
The aim of the legislation is to make public bodies more open and accountable, and like similar legislation in other countries, it starts from a presumption in favour of disclosure.It is important also to note that access to information under the legislation is motive-blind: anyone, anywhere in the world, can request information, and their reasons or motives for so doing are irrelevant to a decision on whether or not to meet the request.Moreover, once someone has received information under the legislation, they are free to pass it on to anyone else: the information is effectively in the public domain.
A significant feature of the UK legislation is that it explicitly defines universitiesas well as publicly-funded research institutes -as 'public authorities' so that they and the staff they employ are therefore subject to its requirements.This means that anyone can submit a request for any data or information generated by researchers during the course of their work at a university or research institute.The university or institute then has to provide the information requested unless there are good reasons (under 'exemptions' or 'exceptions' prescribed in the legislation) why they need not do so.These requirements apply, unlike those in other countries such as the US, regardless of the source of the funding for the research.
Although UK universities -with the exception of the independently-financed University of Buckingham -depend on public funds for a significant part of the income they receive for both teaching and research, public policy has also laid great stress on their autonomous status. 4 Hence they are unlike most of the other bodies defined as 'public authorities' under the legislation.Even before the Government's decision to implement a system of much higher student fees, universities were becoming decreasingly reliant on public funds.Less than half the funding for major research-intensive universities comes from public sources.Moreover, unlike most 'public authorities', universities operate in a highly-competitive environment, and with their own strong commercial interests.Most of them are also charities, which means that they have a duty under the Charities Act 2006 to preserve and exploit their assets, as well as to disseminate knowledge, including the results of research.All these considerations can give rise to tensions and difficulties in complying with the legislation.
It is notable that many of the issues and concerns now being raised in the UK relating to access to research data under the legislation were raised more than thirty years ago in the US, following decisions under the FoI Act passed there in 1966.Thus in 1977, the National Commission for the Protection of Human Subjects of Biomedical and Behavioral Research was asked to "investigate and study the implications of public disclosure of information contained in research protocols, hypotheses and designs submitted to the Secretary of Health, Education, and Welfare in connection with applications or proposals for grants, fellowships or contracts under the Public Health Service Act" following a court decision which held such information generally to be disclosable under the FoI Act.Its report noted that the decision had "caused concern to many members of the research community, who take the position that an investigator's ideas and methodology are his or her 'stock-intrade' and thus deserving of protection from disclosure."It made a number of recommendations to mitigate those concerns (National Commission for the Protection of Human Subjects of Biomedical and Behavioral Research, 1977).Two decades later, the Shelby Amendment to the US Act gave rise to further controversy.The amendment required Federal awarding agencies to ensure that all data produced under an award would be made available to the public through procedures established under the Act.Leading universities and bodies, such as the National Academies, lobbied against the amendment and recommended significant modifications to the draft language.While not disagreeing with the basic objective of making scientific data publicly available, they argued that doing so through the FoI Act would be costly and burdensome, that public release of data prior to publication in scholarly journals would seriously short-circuit the scientific research process, that it would have a chilling effect on university-industry collaborations, and that it could lead to special interest groups harassing researchers 5 .All these issues, as we shall see, are now being raised in the UK.

Consultations and Guidance
Until recently, very few researchers in the UK have been aware that the information and data they gather, create and analyse in the course of their research are subject to the provisions of the FoI legislation.However, recent and well-publicised cases surrounding the disclosure of data and other information about climate change at the University of East Anglia and Queen's University Belfast have raised some concerns about the implications of the legislation.The Independent Climate Change E-mail Review headed by Muir Russell (Russell et al., 2010) and a follow-up inquiry and report from the House of Commons Science and Technology Committee (2011a) highlighted the "confusion and unease" in the research community, and the need for guidance to the higher education (HE) sector.The Committee recommended that the ICO should provide such guidance as a matter of urgency; and the guidance was issued in September 2011 (ICO, 2011a). 5See, for example, the 15 July 1999 statement of the President of the President of the National Academy of Sciences to the sub-committee on Government Management, Information and Technology, Committee on Government Reform, US House of Representatives (http://www7.nationalacademies.org/ocga/testimony/foia.asp); and the Memorandum of 6 December 1999 to Harvard University Deans, Department Chairs, Faculty and Research Staff (FOIAmemo11-29-99-1.doc).Volume 7, Issue 1 | 2012 Following the Muir Russell Report, three related initiatives were put in place.First, the Information Commissioner's Office (ICO) held a roundtable meeting with representatives of the HE sector (ICO, 2010).As a result of that meeting, an ICO-HE Sector Panel has been established to provide a forum for discussion and to address issues of concern, along with a sub-panel with a specific focus on research (ICO,  2011b).

The International Journal of Digital Curation
Second, a Q&A briefing paper on FoI and research data was published by JISC (2010).
Third, a series of three workshops in London, Manchester and Glasgow was organised by the Research Information Network (RIN) in concert with the ICO, the Scottish Information Commissioner, JISC Legal, the National Archives, the Information and Records Management Society, and the UK Data Archive.The purpose of the workshops was twofold:  To raise awareness of the FoI legislation and its implications; and  To gather evidence from researchers and others in the HE sector about their practices and concerns, and the support they would like to have in their approach to FoI.
All three initiatives informed the guidance drafted by the ICO and published in September 2011.This paper is based, in the main, on the results of the workshops.It also takes account of discussions at the sector panel, recent cases and decision notices issued by the ICO, and the guidance the ICO has now issued.

The Workshops
The workshops were held at the University of Manchester (22 March), University College London (1 April) and the University of Strathclyde (12 April).About 80 people in total took part.Half were researchers; the other half included research managers, compliance and governance officers, records and information managers, and policy staff.
Each workshop had two parts.First, there was a series of presentations on the nature of the legislation (from the ICO or the Scottish Information Commissioner's Office), and on the challenges that it can pose to researchers, as well as the opportunities it can provide for those who wish to gain access to information (from researchers and other experts).Following initial discussion relating to the presentations, there was then a series of facilitated group discussions about participants' experiences and concerns relating to FoI and its implications for them and their work.Much of the discussion focused on the data that researchers gather and create, but there was some discussion too of other kinds of information generated in the course of research, including emails, logs and lab books, working papers and drafts.
What follows is structured around the key points that arose in the discussions at the three workshops, focusing on the perceptions and misperceptions, the questions and concerns raised by researchers and those who support their work, operating in an HE context.

The International Journal of Digital Curation
Volume 7, Issue 1 | 2012

Key Issues
FoI requests relating to research University managers and administrators have, for the most part, well-established procedures for dealing with requests for information under FoI legislation.They receive a regular flow of requests, but the number relating to research is small, and most are to do with the funding and management of research rather than the processes undertaken by researchers themselves or the outputs they produce.However, when such requests are received they may give rise to complex issues, so that even a small number can have a significant impact on management time within the university.Moreover, there are concerns that the numbers might grow, and about the disproportionate impact that concerted information-seeking campaigns might have across significant parts of the HE sector.Universities have been reluctant to refuse disclosure under the provisions for dealing with vexatious or repeated requests covered in Section 14 of the FoI Act.6 Research cultures: Openness, sharing, and control Researchers in the HE sector typically share information with a range of colleagues, partners and peers in the course of their work.It is a standard part of the culture, operating through largely informal processes and procedures which researchers understand, and which give them a sense of control: they decide what to share, with whom, when, and how. 7In the last few years there have been moves towards more openness, driven in part by funders and research policy-makers8 .They see data and other outputs of research as valuable resources which should -alongside the formal reporting of results in journal articles, conference proceedings and so onbe shared with the wider research community and others.However, it is notable that none of the UK Research Councils make reference to the requirements of FoI legislation in the guidance they offer to researchers on managing and sharing data.Moreover, it is recognised that cultures vary across disciplines and subject areas.While researchers in genomics make their data freely available as soon as it is produced, those in many other disciplines are more restrictive in their attitudes towards what they will make available and to whom.Many are disconcerted by the loss of control they perceive in dealing with an FoI request for information.

Procedures: Formality vs informality
In order to meet their obligations under the legislation, universities have established procedures to log requests; to determine whether the information should be made available or whether one of the exemptions allowed under the legislation applies; and to ensure that appropriate action is taken. 9The formalities involved in such procedures, and the language used, make many researchers feel uncomfortable.doi:10.2218/ijdc.v7i1.214They may perceive FoI requests and the procedures for handling them as impersonal, bureaucratic, or even confrontational.Some university FoI officers acknowledged at the workshops that the language of the legislation itself, and the associated codes of practice, can engender an adversarial approach.
Hence researchers may be less positive in their response to a formal FoI request than to a more informal approach, face-to-face or over the phone.Some FoI officers recommend establishing a dialogue between requester and researcher, not least to help in establishing exactly what information is being sought, and how easily it can be provided, particularly when requests are often vague or imprecise.Others are more cautious and stress the need for formality in following procedures.All recognise that establishing dialogue may be difficult or impossible when responding to requests that are intentionally broad and imprecise (sometimes called 'fishing expeditions') or those that are expressed in hostile language.Similarly, the FoI officers stress that even where informal dialogue is established, it is essential to keep them informed about what is happening, and to record the nature and results of the discussions.They thus underline the importance of guidance for researchers on how to deal with FoI requests.

Ownership and holding of information
The ownership of information is a matter of much confusion.Researchers often have a strong feeling of 'ownership' -in a metaphorical sense -of data and other information relating to or resulting from their research, particularly before they have formally published results in a journal article or other format.Where ownership or copyright lies in a legal sense may be a complex matter, depending on the detail of relationships between researchers, their employing university, funders or sponsors of research projects, and others who have supplied information.The answer may vary in different circumstances, and between different individuals: research students, for instance, may be in a different position from members of academic staff.Moreover, complex issues can arise when members of academic staff conduct research in a personal capacity or for an external organisation, but store information related to it on university equipment. 10 is important to stress that the FoI legislation applies to information held -not necessarily owned -by a university.Since researchers typically rely on their university networks and servers to store the information they generate, then such information falls under the requirements of the legislation, even if it is held physically off-site (for example, in the cloud), or transferred temporarily to other partners (including those outside the UK).The position with regard to information held on services outside the university's control (for example, personal Hotmail accounts) may not be so clearcut. 11Researchers need guidance on all these matters, as well as on copyright. 9See, for example, the University of Nottingham's Freedom of Information Manual of Office Procedures (2008). 10See the ICO Decision Notices FS5024399 and FER0289351 issued in December 2010 relating to information generated by a lecturer at the Open University in the course of research for an external organisation. 11The ICO Guidance for the higher education sector states that if the information held in personal email accounts is related to the business of the university, or to an employee's contractual role, then it is likely to be held on behalf of the university in accordance with s3(2)(b) of the FoI Act.But establishing the nature of the relationship between contractual and non-contractual work in a university context can be complex, as exemplified in the Open University case noted in the previous footnote.Volume 7, Issue 1 | 2012 Copyright Ownership of copyright does not constitute a check against disclosure under the legislation. 12But nor does disclosure imply the ceding of rights under copyright legislation, including protection against commercial exploitation: it does not give the recipient the right to reproduce or exploit the information in breach of copyright.But universities cannot place any conditions or restrictions before providing access.And in practical terms, the protection of rights may be difficult, but at least one public body issues invoices along with information it provides to commercial organizations in response to FoI requests.

Personal confidentiality and qualitative research
Researchers have especial concerns about responding positively to requests for information gathered from or about individuals.Personal data is generally exempt from disclosure under FoI legislation; 13 but the relationship between that legislation and the Data Protection Act 1998 is complex, and case law on what constitutes personal data exempt from disclosure can be confusing.Some researchers have particular worries about qualitative research, which may involve interviews or diaries, for example.One fear is that interview notes or entries in reflective diaries may be misinterpreted.Another is that for small-scale studies, fully-effective pseudonymisation or anonymisation may be difficult; that may cause particular problems where highly-sensitive information is provided by individuals whose identity may need to be protected to save them from potential harm.The scope for editing or redaction to protect identity in such circumstances needs careful consideration.
More generally, both researchers and administrators perceive a need for greater procedural rigour and formality when research projects involving the use of personal data are being designed and set up, with full analysis of the risks of disclosure.Some such risks relate to detailed analyses of datasets, or the integration of a number of datasets, which may lead to the identification of individuals whose names are not otherwise disclosed.All this poses challenges not only for researchers, but for institutional ethics committees.There are also concerns that the very process of securing informed consent from potential subjects of qualitative research may deter them from participating.For all these reasons there is a perceived risk that useful research will not get done, for fear of the risks of disclosure.

Commercial interests and confidentiality
Although universities are defined as 'public authorities' in the legislation, and thus subject to its requirements, they are also bodies with commercial interests, which compete with each other and in an international environment.Information relating to those interests may be protected from disclosure.Universities may thus wish to commercially exploit the results of their research, and refuse to disclose information that would jeopardise their commercial interests.In other cases, they may decline to disclose information about a project proposal they are planning to submit to a funding 12 Under the EIR, regulation 12(5)(c) provides an exception to the duty to disclose environmental information if disclosure would adversely affect intellectual property rights.But an Information Tribunal decision (EA/2006/0078) makes clear that this exception can be called in aid only if disclosure would cause in a significant degree of loss or harm to the rights holder. 13For example, under Section 40 of the FoI Act.

The International Journal of Digital Curation
Volume 7, Issue 1 | 2012 body or commercial partner.However, in order to do so under Section 43 of the FoI Act, in both cases they would have to demonstrate that disclosure would indeed prejudice their interests and that such prejudice would outweigh the public interest in disclosure.The ICO Guidance has almost nothing to say about the commercial interests of universities themselves.
Universities may also hold information they have received from other commercial organizations.Some information may have been provided in confidence, in which case it is protected by an absolute exemption under Section 41 of the FoI Act.In other cases, where releasing information would prejudice the commercial interests of the third party, the university may refuse to disclose under Section 43, again, so long as such prejudice would outweigh the public interest in disclosure.
Since the protection under Section 43 is conditional, it is therefore important that researchers and universities planning to work with commercial partners address such issues before they begin work, clearly identifying any information that is to be provided in confidence.There is otherwise the risk that commercially-sensitive information might be released into the public domain; and the wider risk that commercial partners may lose confidence in the HE sector's ability to work productively with them.Failure to achieve clarity on these issues lay behind the ICO's recent decisions (June 2011) in relation to data held by the University of East Anglia but received from National Meteorological Services and others from across the world.The University contended that the data had been provided on the understanding that it would not be passed on to third parties; but it was unable to provide clear documentation to that effect, and the ICO ordered that the data should be disclosed. 14ch difficulties may arise at local, national or international level, with a range of different kinds of organizations with commercial interests; and it may take time to resolve them.At one of the workshops, it was reported that concerns about FoI led to four years of negotiation before agreement was reached between a university and an industrial partner.The issues may be especially difficult to resolve with international partners, operating under different legal systems; and the lack of case law brings unhelpful uncertainty.The only relevant decision on universities' relations with commercial organizations relates to information about grants made to the University of Nottingham by military and commercial organizations both in the UK and overseas.15

Data and records management
Effective records management involves processes and systems to control how the activities of and information generated by a university are received, organized, used, updated, maintained and eventually preserved (or disposed of) as evidence of those activities.It is critical to the ability of universities -as other organizations -to meet their obligations under FoI legislation.But effective management of information generated by or relating to research can pose a number of challenges.Individual researchers and research teams move from one project (and one university) to another, and relatively few of them give high priority to records management.It is difficult, if not impossible, for universities to exert tight control over how individual teams organize and manage their records to meet their particular needs, which may be highly specific or idiosyncratic.The constant flux of individuals and teams means that records may be scattered and/or move from one part of the university to another, or may be lost altogether.Lack of effective control systems may make it difficult to track down relevant information when FoI requests are received.
The problems are exacerbated because researchers are often especially idiosyncratic in managing and preserving the data they create or collect.Research funders and universities are beginning to address this issue by introducing requirements for researchers to produce data management plans before they begin a new project; 16 and some scholarly publishers now require researchers to indicate whether the data underlying their reported findings is accessible, and where, before they proceed to publish those findings. 17As the volume of data generated in the course of research projects continues to increase, both researchers and universities face an increasing need to formalize the arrangements for managing and storing data that retains a value over time, and for making it more readily accessible.
Partly as a result of such developments, some universities and larger research teams are now employing specialists to help with curating and managing their data.But there is a long way to go before good data management practice is embedded across the research community, and before researchers secure appropriate incentives and rewards for adopting such practices and for sharing their data with others.

Quality assurance and scholarly discourse
Quality assurance is at the heart of the research process and the communication of research results.Researchers examine their results as they proceed, and check for flaws in their findings and conclusions before they seek to publish them.Publishers use editors and peer reviewers to determine whether or not research papers meet appropriate standards and are worthy of publication.FoI requests can short-circuit or undermine these processes.Data and information can be sought before they have been checked and tested by the researchers or by external peer review.There may be errors, or the data may require detailed manipulation and analysis before it can be readily interpreted.Researchers may well be wary of releasing data or information in such circumstances.
Discourse between scholars is a further related and key feature of the research landscape.Researchers commonly seek help and critical responses to their findings from colleagues they trust, long before they are published.Such exchanges may be conducted in frank terms and may lead to important changes in how the research is conducted, or in how findings are analysed or presented.Similarly, peer review, which may be conducted in open or closed ways depending whether the identities of researcher and reviewer are made known to each other, may involve significant 16 See BBSRC (2010). 17For example, PLoS journals require that all appropriate datasets should be deposited in public resources, and accession numbers cited in the submitted manuscript; and for Nature journals it is "a condition of publication … that authors are required to make materials, data and associated protocols promptly available to readers without undue qualifications in material transfer agreements.Any restrictions on the availability of materials or information must be disclosed to the editors at the time of submission."(Nature, n.d)The International Journal of Digital Curation Volume 7, Issue 1 | 2012 critiques and result in modifications to research projects or the reporting of their results.
Many researchers fear that the norms of scholarly discourse, and frank exchange between scholars, would be put at risk if such exchanges were to be disclosed and put in the public domain.Some argue that the peer review system itself would be put at risk, though others are more comfortable with the idea that the peer review process should be completely open and visible to all. 18 is not clear whether requests for information generated in the course of normal scholarly discourse could be refused under the exemption (Section 36 of the FoI Act) covering information where disclosure would prejudice the effective conduct of public affairs.The ICO Guidance argues that the legislation recognizes the importance of processes that enable free and frank discussion, and the exchange of views; and that tribunals have accepted the need for 'safe spaces' where policy may be formulated, live issues debated, and decisions reached without hindrance from external sources.On the other hand, it notes that such matters must be weighed against a public interest test, and that "there will be particular public interest considerations in favour of disclosure related to publicly-funded research or research that may have a particular impact on the public."Nevertheless, the only case to date covering formal peer review seems to offer some comfort to those concerned to preserve confidentiality.In 2008, the ICO upheld the Medical Research Council's refusal to disclose review reports and assessments of applications for funding, on the grounds both that the applications and reviews had been provided in confidence, and that making the information public would result in less-constructive comments about applications in the future.19

Interpretation and misinterpretation
Many researchers are also worried by the possibility that the data or information they provide could be misinterpreted.This could result in damage to his or her reputation, or to that of the university.The risk of misinterpretation provides no legal ground for withholding the information requested.However, the provision of contextual information is an important issue for researchers and universities to consider.Although there is no requirement under the FoI Act to provide guidance on how to use or interpret the information requested, it may be wise to do so, even when that adds to the costs of meeting the request.Again, informal dialogue with the requester may help.
Ensuring that datasets provided to those who request them are more easily interpreted is the aim of a clause in the current Protection of Freedoms Bill currently proceeding through Parliament. 20The clause would require public authorities, when applicants seek information that is in a dataset and seek it in electronic form, to provide it in a form which makes it capable of re-use.This may require researchers to recast the ways in which they record or hold information, and may add to the costs of research.

Information intended for future publication
Publication is a key stage in the research process, and under the UK FoI Act, there is a qualified exemption (Section 22) for information held with a view to publication "at some future date (whether determined or not)."However, the nature, scope and timing of what is published is often difficult to determine in advance; and since the exemption is subject to a public interest test, it is not clear that the intention to publish results at some point in the future (including some, but not all, material generated in the course of the research) means that it can be withheld until then.Since formal publication of research results may be delayed by several months or even years beyond the end of a research project, it may be that refusal to disclose before publication may be deemed unreasonable.However, in a recent case the ICO has upheld a university's decision to refuse to release a PhD thesis on the grounds that it was due to be published in book form. 21provements in data management and the sharing of data, either in the form of supplementary material associated with journal articles, or through the use of data centres or other mechanisms, means that more of the material generated in the course of a research project is now 'published' than used to be the case.But when an FoI request is received, it may well be unclear, even with the help of a data management plan, whether the information requested will eventually be published, still less when.Some researchers have expressed the fear that disclosure of information in response to an FoI request may reduce the chances of their being able to publish their results formally in a scholarly journal or otherwise; but there is no evidence to suggest whether that is or is not the case.
Under the FoI Scotland Act, the exemption for future publication (Section 27 (1)) requires publication within twelve weeks.But this is balanced by a parallel exemption (Section 27 (2)) for information obtained in the course of, or derived from, a continuing programme of research where it is intended to publish a report of that research.The exemption is qualified, so it has to be shown that disclosure would result in substantial prejudice to the research, those conducting it, or the university.It was reported at the Glasgow workshop that some requesters have accepted a refusal to disclose on these grounds; but use of the exemption has not been tested with the Scottish Information Commissioner's Office, or in the courts.Nevertheless, some researchers and administrators would welcome the Scottish research exemption's being extended to the rest of the UK; and an amendment has been tabled to the Protection of Freedoms Bill which would essentially replicate the Scottish exemption for the rest of the UK.

Publication schemes
Some have urged that some of the burdens of dealing with FoI requests could be eased by the greater use of publication schemes.These are a unique feature of the UK legislation, and commit public authorities to publish information of certain kinds proactively and routinely.The information may cover such issues as organization and governance; finance; strategy and performance; policies and procedures; activities and services.The suggestion is that universities' publication schemes might include at least some information generated in the course of research.There is no evidence of universities currently extending their publication schemes in this way.But there is 21 See ICO Decision Notice FS50439323 issued in June 2011.

The International Journal of Digital Curation
Volume 7, Issue 1 | 2012 doi:10.2218/ijdc.v7i1.214perhaps scope for them to publish information about research projects currently under way; and to make clear that valuable data arising from the projects will be made available through institutional repositories or services such as the University of Southampton's datastore, or other national or international data centres.

Controversial research
There are some concerns that the risks of early disclosure may add to the pressures that lead some researchers to avoid certain kinds or areas of research which may give rise to controversy.Examples might include stem cells, GM crops, sexuality, key areas of public health, studies of criminal or deviant behaviour, or investigations of toxicity or pollution, where commercial interests might be at stake, and so on.There is no evidence at present of reluctance to engage in such areas; but a perceived need to monitor the situation.Such concerns may have been exacerbated Philip Morris International's request to Stirling University under FoI legislation for access to data relating to research into the smoking habits and attitudes of teenagers. 22otential for adverse effects on researcher behaviour While there was much discussion at the workshops about the need to encourage researchers to adopt more rigorous records management practices, fears were also expressed that FoI considerations might have precisely the opposite effect.Researchers might deliberately keep information away from university systems and servers, or delete it as soon as possible, in order to guard against any risk of disclosure.Again, there is no evidence of such behaviours at present.
A different kind of concern is that unscrupulous researchers might seek access to rivals' research data or other information either to 'scoop' them by making use of their findings before they can be published, or to undermine their reputations by revealing flaws in their work.It is difficult to judge how real such fears are.Research cultures are strong, and norms have been codified in various Codes of Ethics, but researchers are also aware of cases of misconduct arising in an increasingly competitive research environment.

FoI as a tool for researchers
From a different perspective, judicious use of FoI enquiries can be a useful tool for researchers working in certain disciplines, such as contemporary history, politics or public health.Beyond such domains, there is little awareness of the potential that it might represent as an aid to research.Nevertheless, the researchers who have requested information can provide useful insights for colleagues on the receiving end of FoI requests.For instance, they stress the importance of establishing a good rapport with the individual or organisation from which information is sought, that confrontational approaches are likely to be counter-productive, and that informal dealings may well be fruitful.However, informality is not always appropriate and on occasion it is more effective to use the services of institutional FoI officers.

Support and training
It is universities that are responsible, under the FoI legislation, for establishing effective policies and procedures to ensure that they comply with its requirements.In effect, FoI sits alongside a range of other regulations and requirements -data 22 See the BBC news report at http://www.bbc.co.uk/news/uk-scotland-tayside-central-14744240 The International Journal of Digital Curation Volume 7, Issue 1 | 2012 protection, financial regulations, health and safety and so on -that must be taken into account to ensure good governance of research across the institution.Universities are increasingly aware of the risks to their reputation if such arrangements are not in place.With regard to FoI, the risks can be acute if universities are perceived as being obstructive or unreasonable in refusing to release information.
It is important that researchers are made aware of the legislation and its requirements, and of the need to manage and organize their information effectively.The JISC Q&A guidance is a good example of a UK-wide initiative in this area, but more information is also needed at local level, including guidance on the practical problems and challenges that are arising.There is a particular need for guidance on relevant exemptions to the general requirement to disclose information, their nature and scope, the public interest test, and the use that has been made of the exemptions in dealing with requests.Researchers also need guidance on such issues as copyright and ownership of information.The ICO Guidance is a start, but universities will need to do more to make researchers aware of the issues.
Universities also need to provide appropriate facilities and support to enable researchers to adopt good practice.The current JISC research data management programme should help to spread such practice more widely across the HE sector.How such support is organized will vary.Some universities adopt centralized approaches to research support, others a more devolved approach focused on faculties and departments, in at least one case with designated 'information champions.'It is important that guidance and support is tailored, as appropriate, to the needs of researchers in different subjects and disciplines.
Guidance and support are also needed in the handling of FoI requests when they arise, not least to allay researchers' concerns and uncertainties.Effective arrangements depend on good lines of communication between researchers, records managers, compliance officers and, when necessary, senior officers of the university.Close cooperation and trust are essential if requests are to be dealt with satisfactorily.

Conclusions
We are, as yet, in the early stages of the impact of FoI legislation in the UK and its implications for the research community; the legislation came into effect only six years ago.The University of East Anglia climate data cases have highlighted a number of issues, but general levels of awareness are low.Both researchers and senior managers in the sector have expressed concerns of the kind outlined here, and the concerns are real.They echo those expressed in the US a decade and more ago, in a context where the impact of FoI legislation is more limited, since it does not apply across the board to universities.
There is, as yet, little concrete evidence as to the impact of the more recent legislation in the UK on the behaviours of researchers, funders and research institutions -to either positive or negative effect.Moreover, some of the impact of the legislation will be difficult to disentangle from that of the wider moves towards more openness in the dissemination of information resulting from research, and the increasing demands for more effective management and sharing of data.However, the lack of guidance from funders on the legal responsibilities that researchers now face is The International Journal of Digital Curation Volume 7, Issue 1 | 2012 striking.Nevertheless, it seems likely that the FoI legislation will add to the pressures leading towards closer management of researchers and their activities.Whether that is a desirable development or not is another question.