Biological Resource Centres and the Use of Microbes

DISC-UK Image © JupiterImages Corporation 2007 Policy-making for Research Data in Repositories: A Guide DataShare project http://www.disc-uk.org/docs/guide.pdf by Ann Green Stuart Macdonald Robin Rice Data Information Specialists Committee -UK May 2009 Version 1.2 Table of Contents i. Introduction 3 ii. Acknowledgements 4 iii. How to use this guide 4 1. Content Coverage 5 a. Scope: subjects and languages 5 b. Kinds of research data 5 c. Status of the research data 6 d. Versions 7 e. Data file formats 8 f. Volume and size limitations 10 2. Metadata 13 a. Access to metadata 13 b. Reuse of metadata 13 c. Metadata types and sources 13 d. Metadata schemas 16 3. Submission of Data (Ingest) 18 a. Eligible depositors 18 b. Moderation by repository 19 c. Data quality requirements 19 d. Confidentiality and disclosure 20 e. Embargo status 21 f. Rights and ownership 22 4. Access and Reuse of Data 24 a. Access to data objects 24 b. Use and reuse of data objects 27 c. Tracking users and use statistics 29 5. Preservation of Data 30 a. Retention period 30 b. Functional preservation 30 c. File preservation 30 d. Fixity and authenticity 31 6. Withdrawal of Data and Succession Plans 33 7. Next Steps 34 8. References 35 The Policy-making for Research Data in Repositories: A Guide is intended to be used as a decision-making and planning tool for institutions with digital repositories in existence or in development that are considering adding research data to their digital collections. The guide is a public deliverable of the JISC-funded DISC-UK DataShare project (2007-2009)1 which established institutional data repositories and related services at the partner institutions: the Universities of Edinburgh, Oxford and Southampton. It is a distilled result of the experience of the partners, together with Digital Life Cycle Research & Consulting. The guide is one way of sharing our experience with the wider community, as more institutions expand their digital repository services into the realm of research data to meet the demands of researchers who are themselves facing increasing requirements of funders to make their data available for continuing access. The guide also can contribute indirectly to efforts to articulate the benefits of sound data management practices, as well as the goals of data sharing and long term access. In addition to setting up data repositories, institutions and libraries in particular can become more active by pursuing some of the following activities: • Raising awareness of data issues within institutions and the benefits of actively managing research data • Assisting in developing policies about data management and preservation • Providing advice to researchers about data management early in the research life cycle; influencing the way researchers will be creating their data, the formats they will use and building a commitment to use the repository to publish/preserve their data • Working with IT service colleagues to develop appropriate local data management capacity • Training and introducing data management and curation concepts to research students • Exploring methods of moving data from work-in-progress storage spaces to repositories in more seamless ways (Lewis, 2008). This guide is largely based upon the online OpenDOAR Policy Tool (SHERPA, 2007), the OAIS Reference Model (CCSDS, 2002) and the TRAC checklist (OCLC, 2007). Although the focus was initially based on social science datasets, research institutions typically produce a vast heterogeneity of data types and so many other research outputs could be considered within the range of requirements listed in the report; other content includes images, texts, audio, video files as well as scholarly publications and ‘grey literature’. The guide does not cover the value-added services that should be offered within a curatorial environment, details of selection and appraisal, nor does it cover 1 http://www.disc-uk.org/datashare.html INTRODUCTION 3 advocacy, researcher requirements and data management considerations surrounding funders’ mandates. Policies should be developed to address the complex issues related to access mechanisms and user support services as part of any service development process. ACKNOWLEDGEMENTS The production of this guide was a collaborative effort by the members of the DISC-UK DataShare project team. The authors would to like to thank Sally Rumsey, Luis Martinez Uribe, and Wendy White for their contributions. Our appreciation goes to editors Harry Gibbs and Jane Roberts, and graphic designer Jackie Clark. HOW TO USE THIS GUIDE Each section contains a set of data-related topics compiled from multiple sources that focus on research data quality, management, and preservation. Repository planners and developers can evaluate each set of requirements and choose the most applicable options for their policy and planning purposes. Examples of how the requirements have been implemented in other instances, or related tools are presented in the shadow boxes. Lists within the text are marked with 2 different symbols: examples are marked with diamonds and points for consideration are marked with arrows. Authoritative references are supplied at the end of the guide for further study. Repository policy examples: Each of the three DISC-UK DataShare partners has a policy web page. See Edinburgh DataShare’s policies: http://datalib.ed.ac.uk/DataShare/index.html#pol See Oxford University Research Archive’s policies: http://www.ouls.ox.ac.uk/ora/ora_documents2/ora_policies See e-Prints Soton policies: http://eprints.soton.ac.uk/repositorypolicy.html 4 INTRODUCTION 4 1. CONTENT COVERAGE 1.a SCOPE: subjects and languages _What subject areas will be included or excluded? _Are there language considerations? Will translations be included or required? (will text within data files, metadata or other documentation in other languages be translated into English, for example?) 1.b. KINDS OF RESEARCH DATA As stated in the Research Information Network’s report, there are “hugely varied kinds of digital research data – from texts and numbers to audio and video streams.” (RIN, 2008.) The 2005 National Science Board (NSB) report entitled Long-Lived Digital Data Collections suggests that data can be differentiated … by their origins: whether they are observational, computational, or experimental. These latter distinctions are crucial when it comes to making choices for archiving and preservation. Observational data (e.g. observations of ocean temperature on a specific date) are essentially historical and cannot be reproduced, and so may be primary candidates for indefinite archiving. Computational data may require archiving complete information about the computer model and the execution (e.g. hardware, software, input data), but not of the data results themselves – which can in theory be reproduced. Experimental data may not be easily reproduced, given both cost considerations and the complexity of all the experimental variables (Gold, 2007). _What kinds of research data will be included? tScientific experiments tModels and simulations, consisting of 2 parts: the model with associated metadata and the computational data arising from the model tObservations: from the astronomical to the zoological – of specific phenomena at a specific time or location, where the data will usually constitute a unique and irreplaceable record (this includes surveys, censuses, voting records, field recordings, etc.) 5 Content Coverage tDerived data: resulting from processing or combining ‘raw’ or other data (where care may be required to respect the rights of the owners of the raw data) tCanonical or reference data: for example, gene sequences, chemical structures etc. tAccompanying material (RIN,2008) Supplemental objects are normally included in the research data ‘information package.’ These accompanying materials may be coding instructions, interviewer guides, graphic flowcharts of data collection, questionnaires, information regarding the methods used and research techniques employed, codebooks, data collection instruments, summary statistics, database dictionaries, project summaries/ descriptions, and bibliographies of publications pertaining to the data. 1.c. STATUS OF THE RESEARCH DATA Data production in science may be characterized in several phases and flavours. To begin with there is ‘raw data’ – data as it comes directly from the scientific instrument. There may be stages of validating and selecting the data. Then the data may be subjected to any number of standard or ad hoc processes that calibrate it – that translate it into a more general schema (e.g. sky maps, catalogues) (Gold, 2007). _Is the inclusion of the data into the repository determined by its status in the research process / lifecycle? Such as: t‘raw’ or preliminary data tdata that are ready for use by designated users tdata that are ready for full release, as specified in access policies tsummary/tabular data (could be associated with a publication) t‘derived’ data. 6 See: ARCHER This is a suite of open-source, production-ready generic e-Research infrastructure components that have been independently tested and documented to provide better management of research data and assist researchers to - collect, capture and retain large datasets from a range of different sources including scientific instruments - deposit data files and datasets to e-Research storage repositories - populate these e-Research data repositories with associated metadata - permit dataset annotation and discussion in a collaborative environment - support next-generation methods for research publication, dissemination and access. For an overview of the ARCHER toolset see http://www.archer.edu.au/products. Content Coverage 1.d. VERSIONS Digital data can be copied, altered or deleted very easily... This makes it very important to be able to demonstrate the authenticity of data, and to prevent unauthorised access to data for ethical, legal and quality reasons. An important related concept is that of the master file, a formalised and checked final copy of the data (or other materials), or copy at a certain stage of development (as opposed to temporary working versions of data and other files) (UKDA, 2008a). Policy considerations for the deposit of multiple versions of a dataset: _The repository uses explicit version numbers which are reflected in dataset names. _The repository records version and status e.g. draft, interim, final, internal2. _The repository stores multiple copies of a dataset in different formats. _The repository keeps the original copies of data and documentation as deposited. _The repository stores supplemental digital objects with the data file/s. _The repository records relationships between items3 , such as ‘supercedes’ or is superceded by’. VERSION CONTROL It is of great importance to ensure that different copies of files, or materials held in different formats, or information that is crossreferenced between files are subject to version control, whereby checks and procedures are put in place to make sure that if the information in one file is altered, the corresponding information in other files is also altered (UKDA, 2008a). Policy considerations for the deposit of multiple versions of a dataset: _The earlier version may be withdrawn from public view. There will be links between earlier and later versions, with the most recent version clearly identified. _The repository ensures that different copies of files or materials held in different formats are subject to version control. 2 The first version might come from the data collection process, the second version might emerge from data cleaning or reformatting, the third might result from composite variable construction, and so forth. 3 The repository maintains the relationship between the parent object and any subsequent child objects (i.e. tracking of unique identifiers for child and parent objects; possible inheritance of parent object identifier to child objects). 7 Content Coverage _The item’s persistent identifier will always link to the latest version. 1.e DATA FILE FORMATS Preferred formats are formats designated by a data repository for which it guarantees that they can be converted into data formats that will remain readable and usable. Usually, these are the de facto standards employed by that particular community. (DANS, 2008 p. 8) _Consider what formats will be accepted for deposit, and which are preferred. In some cases the repository will accept any format, but it won’t necessarily guarantee continued access and preservation of obsolete or obscure formats. It is worth remembering that what is generically called a ‘dataset’ can be made up of data and application specific content in the data file or in a separate program file. ’Raw‘ data can usually be stored without problems, but the accompanying application can present problems for use and may have licensing issues. The management of file formats over time is covered in section 5.c, File Preservation. EXAMPLES OF FILE FORMATS tStatistical (quantitative) data: Application formats: SAS, SPSS, Stata, XML, Excel, etc. _Will it be required that ASCII files be accompanied by data definition statements for a specific software application? _Must Excel files be converted to tabor comma-delimited text? 8 See Versions Toolkit (Versions of Eprints - User Requirements Study and Investigation of the Need for Standards) at: http://www.lse.ac.uk/library/versions/about.html Note: currently focuses on academic papers, not other content types, but principles may be applicable. See examples of preferred formats at: Economic and Social Data Service website: http://www.data-archive.ac.uk/sharing/acceptable.asp ICPSR Guide, see section entitled ‘Final Data Preparation: File Formats’ in: http://www.icpsr.umich.edu/ICPSR/access/dataprep.pdf Content Coverage t Non-statistical data: tQualitative (textual) data: RTF, HTML, ATLAS.ti, NUD*IST, NVivo, XML etc. tSupplemental materials: PDF, Word, image files etc. tDigital audio data: for example, WAV files of human speech signals, Audio Interchange File Format (.aiff), MP3 etc. tPlasmids encoded in genbank format tJava webstart files for software to render data t3D image data: TIFF stacks, RAW, DICOM, and similar format tImage and Digital video data: MPEG-2, JPEG 2000.  Policy considerations for acceptable data file formats from producers and/or depositors: _Will the repository accept only file formats which are welldeveloped, have been widely adopted and are de facto standards in the marketplace? _Will the repository guarantee that specific file formats will be converted into data formats that remain readable and usable? (See section 5.c, File Preservation.) _Will the repository minimise the number of file formats to be managed as far as is feasible/desirable? _Will the repository accept only ‘open’ non-proprietary, welldocumented file formats wherever possible? _Will compression formats be accepted? (e.g.7zip, gzip, winzip) Normalisation: File format conversion by the repository when data are submitted. For consideration: _Will the repository convert digital objects of a particular type (e.g. statistical software system files) to a single chosen file format that is thought to embody the best characteristics of functionality, longevity and preservability? _Will the repository convert proprietary formats to non-proprietary formats? _Will the repository create plain text versions of datasets (encoded in either ASCII or Unicode character sets)? _Will the repository retain the original bitstream (file) with the item, in addition to its converted formats? Dissemination formats: _Will the repository accept formats for the purposes of transfer, storage and distribution to users, which do not meet the conditions of long term access? _Will the repository move data into more current software versions, 9 Content Coverage tMETS8isusedasa‘wrapper’forcompounddigitalobjects, allowingthemtobeidentifiedassuchandactedupon,e.g. importingandexportinginrepositories. tRDF9 provides a simple way to make statements about Web resources, often expressed in RDF/XML, in the form of “triples,” i.e. subject-predicate-object expressions that relate objects to one another. (For example, A is version of B.) 2.d METADATA SCHEMAS Repositories may need to put in place additional metadata schemas to support the ingest, management, and use of data in their collections. Some repositories implement additional or extended metadata schemas for domain specific datasets. For example, creating a new community/ / collection (e.g. for Astronomy or Space Physics) – the SPDML (Space Physics Data Markup Language) schema could be ‘plugged-in’ to DSpace. This would mean that researchers wishing to deposit Space Physics data would be presented with an ingestion interface based on the metadata schema for their particular data type, thus capturing the richness of that particular domain dataset (which otherwise could be lost by the default DC-based schema). The same could be applied to Learning and Teaching Objects (LOM – Learning Object Metadata), biological species observational data (Darwin Core), SPASE Data Model (standard metadata for space science data description) etc. 8 Metadata Encoding and Transmission Standard (METS). http://www.loc.gov/standards/mets/ 9 Resource Description Framework (RDF). http://www.w3.org/RDF/ 16 About OAI-ORE: ‘Open Archives Initiative Object Reuse and Exchange (OAI-ORE) defines standards for the description and exchange of aggregations of Web resources. These aggregations, sometimes called compound digital objects, may combine distributed resources with multiple media types including text, images, data and video. The goal of these standards is to expose the rich content in these aggregations to applications that support authoring, deposit, exchange, visualization, reuse, and preservation. Although a motivating use case for the work is the changing nature of scholarship and scholarly communication, and the need for cyberinfrastructure to support that scholarship, the intent of the effort is to develop standards that generalize across all web-based information including the increasingly popular social networks of web 2.0.’ See http://www.openarchives.org/ore/toc. Metadata 16 17 DDI, the Data Documentation Initiative, is an XML-based metadata schema that can describe not just the dataset as a whole but also descriptive material drawn from the life-cycle of a data resource.This can include information, for example, about the source of funding and methodology used in collecting data to an entire set of survey questions and resultant variables. It was initially created as a metadata schema for codebooks but has developed broader application for time series data, complex hierarchical data files, and tabular data. The DISC-UK DataShare project explored the potential of enhancing institutional data repositories through the use of DDI metadata in a briefing paper (Martinez, 2008). Metadata 3. SUBMISSION OF DATA (INGEST) 3.a ELIGIBLE DEPOSITORS _Will eligibility be restricted by status? Deposits may be made by, for example: tAccredited members, academic staff, registered students, employees of the institution, department, subject community or delegated agents tData producers or their representatives (‘self deposit’) tOnly repository staff. _Will eligibility be restricted by content? For example, eligible depositors: tmay only deposit their own work tmust enter descriptive metadata for all their data tare limited to depositing complete datasets as defined by the repository tmay only deposit data of a certain type or subject. _Will the repository provide a confirmation of receipt to the depositor including a request to resubmit a digital object in the case of errors resulting from the submission? 3.b MODERATION BY REPOSITORY Considerations: _Are submissions checked to ensure that data integrity has been fully maintained during the transfer process? If so, spot checks, or all submissions? _The repository checks metadata records for accuracy. _The repository adds Digital Object Identifiers (DOIs) or another persistent identifier, such as the Handle system. _Does the repository’s administration review items for the following: teligibility of authors/depositors? trelevance to the scope of the repository? tvalid formats? texclusion of spam? 18 Submission of Data (Ingest) 3.c DATA QUALITY REQUIREMENTS 3.c.1 RESPONSIBILITY The repository should have clear and concise depositor agreements written in plain language that are presented to depositors with each acquisition. In most cases, data producers are responsible for the quality of the digital research data. The repository is responsible for the quality of storage and availability of the data. For example, the following statement could be part of a submissions policy: The validity and authenticity of the content of submissions (all materials submitted by the depositor, including full data and metadata) is the sole responsibility of the depositor, and is not checked by the repository (SHERPA, 2007). In these cases, the repository accepts no responsibility for mistakes, omissions, or legal infringements within the deposited object. There may be situations in which the depositor does not guarantee that the dataset is accurate and the depositor indemnifies the repository’s institution against any legal action arising from the content of the dataset. One way to mitigate against legal risks is to have a ‘take-down’ policy for removal of objectionable items. (See section 6, Withdrawal). In some cases, licenses are presented to depositors to cover the range of requirements for reuse of the data (see section 4, Access and Reuse of Data). 3.c.2 QUALITY ASSESSMENT If the repository evaluates data quality in order to make decisions about whether to accept content or not, the repository can choose to determine the quality by assessing the following: Their intrinsic scientific quality through assessment by experts and colleagues in the field. Considerations: _Are the research data based on work performed by the data producer (researcher or institution that makes the research available) and does the data producer have a record of academic merit? _Was data collection or digitization carried out in accordance with prevailing criteria in the research discipline? _Are the research data useful for certain types of research and suitable for reuse? (DANS, 2008 p. 7) _The required contextual information (metadata) has been provided by the data producer. Descriptive, structural and administrative metadata must be provided, or created by the repository, in accordance with the applicable guidelines of the repository. 19 Submission of Data (Ingest) _Data files are inspected to ensure that variables and values are accurate according to the documentation supplied and are sufficiently labelled for secondary use. _Checks are made to verify that metadata in a data file matches metadata in descriptive documentation (e.g. variable names in a dataset match variable names in a codebook). 3.d CONFIDENTIALITY AND DISCLOSURE Repositories will often require that data depositors ensure that data meet requirements of confidentiality and non-disclosure for data collected from human subjects. In some cases, the repository may alter sensitive data to create anonymised data that can be distributed to its user community. In the case of one social science data archive, data collections acquired by the repository “undergo stringent confidentiality reviews to determine whether the data contain any information that could be used - on its own or in combination with other publicly available information - to identify respondents. Only after the completion of those reviews are data made available from the repository. Should such information be discovered, the repository alters the sensitive data after consultation with the principal investigator to create public use files that limit the risk of disclosure” (ICPSR, 2007c). 3.e EMBARGO STATUS Some repository infrastructure systems have the technical capacity to embargo or sequester access to data until the content has been approved for release to the public. Agreements about the embargo – its length and what triggers its ending – need to be made between the repository and its contributors. 20 Submission of Data (Ingest) Data Seal of Approval from DANS (Data Archiving and Networked Services, The Netherlands) ‘DANS was given the task ... to develop a seal of Approval in order to ensure that archived data can also be found, recognized and used in future. Such a seal of approval can be applied for and awarded to research which meets a number of recognizable criteria in the area of quality, durablity and accessibility of the data. The seal of approval can also be requested by and awarded to data repositories that want to store research data perminently and make them accessible. The seal of approval contains guidelines for applying and checking quality aspects of the creation, storage and (re)use of digital research data in the social sciences and humanities. These guidelines serve as a basis for granting a “data seal of approval.’(DANS, 2008). Institutional repositories typically do not permit content to be removed once submitted. However, a variety of legitimate circumstances might require an institution to limit access to particular content to a specific set of users. These circumstances might include copyright restrictions, policies established by a particular research community (limiting access to departmental working papers to members of that department, for example), embargoes that an institution’s Sponsored Programs Office might require to keep the institution in compliance with the terms of sponsor contracts, and even monetary access fees for certain data. Implementing these policy-based restrictions requires robust access and rights management mechanisms to allow or restrict access to content -and, conceivably, to parts of digital objects - by a variety of criteria, including user type, institutional affiliation, user community, and others, (Johnson, 2002). Considerations: _Will the embargo status and length of embargo be determined by repository staff or data producers (or their representatives)? _Will the repository choose to hold materials where the metadata is publicly accessible but the data are embargoed or restricted in some way? _Can the repository software be configured to automatically release the data on the end date of the embargo or will manual procedures be needed? 3.f RIGHTS AND OWNERSHIP The repository service may wish to enter a license agreement with the depositor upon transfer of the data item through a written or click-through Depositor Agreement. _How extensive are the rights given to the digital repository over the deposited material? _Are there limitations to what the digital repository is allowed to do with the material? 21 In the UK, the Data Protection Act defines personal, confidential and sensitive data and sets out parameters under which data processors (researchers) can use them. Furthermore, Research Ethics Committees may approve or disapprove data sharing plans based on the researchers’ methods and what sort of consent has been sought. A thorough overview of ‘consent, confidentiality and ethics in data sharing’ for researchers has been written by the UK Data Archive as part of their suite of web pages dedicated to managing and sharing data (UK Data Archive, 2008d). Submission of Data (Ingest) Considerations: _Are file format changes limited by authenticity requirements? _Is the repository free to change the original submitted material as it sees fit during the preservation processing? (Beedham et al, 2005 p. 105) _Can the repository translate, copy or re-arrange datasets to ensure their future preservation and accessibility, and keep copies of datasets for security and back-up, or can depositors notify the repository that specific restrictions apply? _Can the repository migrate datasets to another repository (e.g. subject-based, institutional) on condition that all metadata are migrated with the dataset, and that no charge will be levied by the destination repository? _Can the repository incorporate metadata or documentation into public access catalogues for the datasets it holds? _Will the repository be under any obligation to reproduce, transmit, broadcast or display a dataset in the same format or software as that in which it was originally created? _While every care will be taken to preserve the dataset, will the repository be liable for loss or damage to the dataset or any other data while it is stored in the repository or a repository to which the dataset is subsequently migrated? _Do depositors retain the right to deposit the item elsewhere in its present or future version(s)? Depositors’ agreements regarding copyright may cover the following: tThe content of deposited dataset does not breach any law. tIt is original and does not infringe the copyright of any other person (e.g. it is not derived from a licensed or commercial product). tIf it contains material that is copyright of a third-party, the depositor has secured permission from the rights-holder or their representative to include such material in the dataset (including a commercial or academic partner in a research project). tAny third-party materials for which the depositor has not secured the necessary permissions have been deleted from the dataset before deposit. tIf the dataset has been sponsored or subsidised by any institution or organisation other than the depositor’s employer, s/he has fulfilled all obligations to that institution or organisation regarding publication. The institution’s legal office may wish to sign off on the final Depositor Agreement. 22 Submission of Data (Ingest) Does data have copyright? Facts are not copyrightable in any jurisdiction, but there are varying levels of protection for data in different countries. In the European Union, the Database Directive has led to greater protection for compiled databases than in the USA which is governed solely by copyright law. The Digital Curation Centre has produced a briefing paper that clarifies IPR in Databases within UK law (McGeever, 2007). 23 Sample copyright statements for inclusion in a Depositor Agreement: Any copyright violations are entirely the responsibility of the authors/depositors. If the repository receives proof of copyright violation, the relevant item will be removed immediately. The repository shall not be under any obligation to take legal action on a depositor’s behalf or on behalf of any other rights holder in the event of breach of intellectual property rights or any other right in the material deposited. Depositors retain all moral rights to the work including the right to be acknowledged. An example of a depositor agreement (for the Edinburgh DataShare repository) is here: http://datalib.ed.ac.uk/DataShare/Depositor-Agreement.pdf. Submission of Data (Ingest) 4. ACCESS AND REUSE OF DATA 4.a ACCESS TO DATA OBJECTS Digital research data should be easy to find, and access should be provided in an environment which maximises ease of use; provides credit for and protects the rights of those who have gathered or created data; and protects the rights of those who have legitimate interests in how data are made accessible and used (RIN, p.10). OPEN ACCESS The open access publishing movement was started by the Budapest Open Access Initiative and its signatories in February 2002. ‘Open Access means access to material via the Internet in such a way that the material is free for all users to read and use.’ (Wikipedia, http://en.wikipedia.org/wiki/Open_access) 24 Open Data: ‘Open Data is a philosophy and practice requiring that certain data are freely available to everyone, without restrictions from copyright, patents or other mechanisms of control. It has a similar ethos to a number of other “Open” movements and communities such as open source and open access’ (Wikipedia - http://en.wikipedia.org/wiki/Open_data). Recent technological advances in web services and infrastructures have democratised the processes allowing data to be used, stored, visualised, analysed in collaborative ways. Examples of collaborative utilites which use ‘community-driven’ Web 2.0 technologies to visualise or ‘mash’ numeric data include Many Eyes, Swivel and Infochimps. There are also a whole range of spatial visualisations or mashups using mapping tools, earth viewers or open geo-browsers such as Google Earth, OpenStreetMap, Open Layers, which capitalise on and utilise the preponderance of location-based information. These utilities allow researchers to upload and analyse their own data in ‘open’ and kinetic environments (Macdonald, 2008 a, b). By opening up their code to repository developers (e.g. through APIs) or by the development of new plug-ins or tools, numeric and spatial data visualization could be enhanced within the repository environment. This would have the potential of engaging potential depositors, to enhance output, and to provide analysis and visualisations as part of ‘value-added’ functionality. See also: non-profit organisations such as the Open Data Foundation (http://www.opendatafoundation.org/), and the Open Knowledge Foundation (http://www.okfn.org/) Access and Reuse of Data Considerations: _Will access to the content in the repository be open to the public? Note restrictions on reuse may apply even though access is allowed. (see section 4.b, Use and Reuse of data Objects.) CONTROLLED ACCESS If access to some or all items is controlled, the repository might be required to limit access based upon: _User type/status (general public, research organization, membership, administrative staff) _Location – access restricted to specific IP location or physical location _The number of concurrent users of an object at a given time. RESTRICTED ACCESS The repository may be required to restrict access to data for a number of reasons e.g. datasets might contain confidential information that could lead to the identification of respondents or datasets may be used to develop a patent or commercial product. How is this implemented? (Dulong de Rosnay, 2008). How will restricted access conditions be implemented? _The repository stores data on a secure non-networked server and has clear policies regarding the physical safekeeping and use of the restricted datasets when in the researcher’s possession. _The repository offers a Data Enclave for restricted data with confidentiality issues for which there is heightened sensitivity to disclosure, as indicated either by the depositor or the repository. The only form of access to such data is through on-site analysis in the repository’s secure data enclave with very controlled conditions (ICPSR, 2007a). REGISTRATION Considerations: _Will registration be compulsory before downloading or accessing data? _Will registration be compulsory for depositors only? _Will a local registration system be implemented or will it interoperate with other systems (e.g. UK Access Federation or campus single sign-on)? 25 Access and Reuse of Data 32 When should fixity information be created and verified? Fixity information can be created or verified at numerous stages in the preservation workflow. The management policy for the archive, its context and the level of confidence required will dictate when and what type of fixity information is created and how often it is verified. Appropriate points in the lifecycle for generating and verifying fixity information include: tAt point of creation tAt point of accession tAt point of ingest tAt point of transformation tAs part of normal maintenance routines tAt point of dissemination (Paradigm Project, 2007). To ensure authenticity of the digital objects and the metadata: _Will the repository establish protocols and audit trails to show who has accessed each dataset, and who has enhanced or annotated it? (RIN, 2008 p.14) _Will existing relationships between datasets and explicit links be maintained? Preservation of Data 33 6. WITHDRAWAL OF DATA AND SUCCESSION PLANS Review of conditions for withdrawal of datasets: _Will items be removed from the repository? _Under what conditions will the repository choose to remove items? Reasons for withdrawal by repository might include: tcopyright violation tlegal requirements and proven violations tnational security tfalsified research tconfidentiality concerns etc. _Will items be removed at the request of the depositor? If items are withdrawn, indicate the terms of the withdrawn items, for example: tWithdrawn items are deleted entirely from the database. tWithdrawn items are not deleted, but are removed from public view. tIdentifiers/URLs for withdrawn items are retained and for how long? For example, indefinitely, transiently, or not at all. tURLs will continue to point to ‘tombstone’ citations, to avoid broken links and to retain item histories together with a link to replacement versions, where available, and with reasons for withdrawal. Metadata for withdrawn items: tShould a dataset be removed by either the repository or the depositor, the repository reserves the right to retain its metadata record in the repository as trace of the dataset. tThe metadata of withdrawn items will / will not be searchable. CLOSURE and SUCCESSION Considerations: _In the event of the repository being closed down, will the data be transferred to another appropriate archive? _Will items be returned to their originators? _Are there other institutional or domain-specific repositories that can offer succession arrangements? Withdrawal of Data and Succession Plans 7. NEXT STEPS Across the globe, institutions are driven by an increasing awareness of research data outputs as highly valued resources within an integrated digital knowledge base. Mandates to share, manage, and preserve this digital heritage present challenges to researchers, their home institutions and national archives, as well as their own disciplines to create trusted, shared, understandable, and usable solutions. The promises of globally open data fuel the demand to build the infrastructure, context, and access methods that further enable a dynamic digital future. This guide proffers a decision-making framework for institutions at a time when rapid innovation for the support of the curation lifecycle of research data is under way (in the UK, this is being promoted by JISC, RIN, and the DCC10). As evidenced by the resources we cite from longstanding social science data archives (for example, ICPSR, DANS, and the UKDA11), expertise in data archiving has been evolving since the 1960s. This guide seeks to bring the best practices and standards employed by these organisations into the context of the accelerating development of institutional repositories. The DISC-UK DataShare project’s interactions between data support specialists and their institutional repository colleagues informed the production of the guide. We propose that this collaborative effort across and among institutions can be a model for future development as we move from discussion to full implementation of policies and practices related to bringing data into repository environments. Among the project’s findings are the importance of policy for data management, as well as shared tools and strategies that support solutions to the many challenges of research data stewardship and persistent access. The guide can move forward conversations and policy development for the range of questions that emerge as institutions plan digital services for data management and sharing. Please consult the references for further authoritative sources of information and advice. 10 Joint Information Systems Committee, Research Information Network, Digital Curation Centre. 11 Inter-University Consortium for Political and Social Research at the University of Michigan, Data Archiving and Networked Services in the Netherlands, UK Data Archive at the University of Essex. 34 Next Steps 35 ARTS AND HUMANITIES DATA SERVICE (2004a) AHDS Guides to Good Practice. London: AHDS. Available from: http://ahds.ac.uk/creating/guides/index.htm ARTS AND HUMANITIES DATA SERVICE (2004b) How to Deposit with the AHDS. London: AHDS. Available from: http://ahds.ac.uk/depositing/how-to-deposit.htm AUSTRALIAN NATIONAL UNIVERSITY (2008) ANU Data Management Manual: Managing Digital Research Data at the Australian National University, v. 1.03. Canberra : ANU. Available from: http://ilp.anu.edu.au/dm/ANU_DM_Manual_v1.01.pdf BEEDHAM, H., MISSEN, J., PALMER, M. and RUUSALEPP, R. (2005) Assessment of UKDA and TNA Compliance with OAIS and METS Standards. Essex: UK Data Archive. Available from: http://www.jisc.ac.uk/media/documents/programmes/preservation/oaismets. pdf CARR, L., WHITE, W., MILES, S. and MORTIMER, B. (2008) Institutional Repository Checklist for Serving Institutional Management, Version 0.2. OR2008, Third International Conference on Open Repositories, Southampton 1-4 April 2008. Southampton: University of Southampton. Available from: http://pubs.or08.ecs.soton.ac.uk/138/1/IRChecklist.pdf CONSULTATIVE COMMITTEE FOR SPACE DATA SYSTEMS (2002). Reference Model for an Open Archival Information System (OAIS). Washington DC: CCSDS. Available from: http://public.ccsds.org/publications/archive/650x0b1.pdf CREATIVE COMMONS (2008) License your work. Massachusetts: Creative Commons. Available from: http://creativecommons.org/about/license/ DATA ARCHIVING AND NETWORKED SERVICES (2008) Data Seal of Approval. The Hague: DANS. Available from: http://www.datasealofapproval.org/ DIGITAL CURATION CENTRE (2006) Glossary of terms. Edinburgh: DCC. Available from: http://www.dcc.ac.uk/resource/glossary/ DIGITAL CURATION CENTRE (2008) The DCC Curation Life Cycle Model. Edinburgh: DCC. Available from: http://www.dcc.ac.uk/docs/publications/DCCLifecycle.pdf References 36 DIGITAL PRESERVATION COALITION (2008) Preservation Management of Digital Materials: The Handbook (section 5.5). York: DPC. Available from: http://www.dpconline.org/graphics/medfor/recommendations.html DULONG DE ROSNAY, M. (2008) Check Your Data Freedom: A Taxonomy to Assess Life Science Database Openness. London: Nature Precedings. Available from: http://dx.doi.org/10.1038/npre.2008.2083.1 ECONOMIC AND SOCIAL DATA SERVICE (2008A) Depositing Data. ESDS: University of Essex and University of Manchester. Available from: http://www.esds.ac.uk/aandp/create/depintro.asp ECONOMIC AND SOCIAL DATA SERVICES (2008b) End User License. ESDS: University of Essex and University of Manchester. Available from: http://www.esds.ac.uk/aandp/access/licence.asp ePRINTS SOTON (2006) University of Southampton Research Repository - Repository Policies. Southampton: University of Southampton. Available from: http://eprints.soton.ac.uk/repositorypolicy.html GOLD, A. (2007) Cyberinfrastructure, Data, and Libraries, Part 1: A Cyberinfrastructure Primer for Librarians. D-Lib Magazine [online], 13 (9/10). Available from: http://www.dlib.org/dlib/september07/gold/09gold-pt1.html INTER-UNIVERSITY CONSORTIUM FOR POLITICAL AND SOCIAL RESEARCH (2005) Guide to Social Science Data Preparation and Archiving: Best Practice Throughout the Data Life Cycle, 3rd ed. Ann Arbor: ICPSR. Available from: http://www.icpsr.umich.edu/ICPSR/access/dataprep.pdf INTER-UNIVERSITY CONSORTIUM FOR POLITICAL AND SOCIAL RESEARCH (2007a) Guidelines for Depositing Data. Ann Arbor: ICPSR. Available from: http://www.icpsr.umich.edu/ICPSR/access/deposit/guidelines.html INTER-UNIVERSITY CONSORTIUM FOR POLITICAL AND SOCIAL RESEARCH (2007b) Levels of Access to Data: Public and Restricted. Ann Arbor: ICPSR. Available from: http://www.icpsr.umich.edu/ICPSR/access/restricted/index.html INTER-UNIVERSITY CONSORTIUM FOR POLITICAL AND SOCIAL RESEARCH (2007c) Confidentiality Review. Ann Arbor: ICPSR. Available from: http://www.icpsr.umich.edu/access/deposit/conf-review.html JOHNSON, R. K. (2002) Institutional Repositories: Partnering with Faculty to Enhance Scholarly Communication. D-Lib Magazine. 8 (11). Available from: http://www.dlib.org/dlib/november02/johnson/11johnson.html References 37 LEWIS, M. (2008) University Libraries in the UK Data Curation Landscape, Keynote. 4th International Digital Curation Conference, Edinburgh 1-3 December 2008. Edinburgh: DCC. Available from: http://www.dcc.ac.uk/events/dcc-2008/programme/presentations/0840_Martin_Lewis.ppt LYON, L. (2007) Dealing with Data: roles, rights, responsibilities and relationships, Consultancy Report. Bath: UKOLN. Available from: http://www.ukoln.ac.uk/ukoln/staff/e.j.lyon/reports/dealing_with_data_report-final.pdf MACDONALD, S. (2008a) Data Visualisation Tools: Part 1 - Numeric Data in a Web 2.0 Environment. DISC-UK, January 2008. Available from: http://www.disc-uk.org/docs/Numeric_data_mashup.pdf MACDONALD, S. (2008b) Data Visualisation Tools: Part 2 - Spatial Data in a Web 2.0 Environment and Beyond. DISC-UK, September 2008. Available from: http://www.disc-uk.org/docs/spatial_data_mashup_V2.pdf MARTINEZ, L. (2008) The Data Documentation Initiative (DDI) and Institutional Repositories. DISC-UK, February 2008. Available from: http://www.disc-uk.org/docs/DDI_and_IRs.pdf MARTINEZ-URIBE, L. (2008) Findings of the Scoping Study and Research Data Management Workshop: Scoping Digital Repository Services for Research Data Management. Oxford: University of Oxford, Office of the Director of IT. Available from: http://www.ict.ox.ac.uk/odit/projects/digitalrepository/findings.xml MASSACHUSETTS INSTITUTE OF TECHNOLOGY LIBRARIES (2009) Data Management and Publishing. Cambridge, MA: MIT Libraries. Available from: http://libraries.mit.edu/guides/subjects/data-management/ MASSACHUSETTS INSTITUTE OF TECHNOLOGY LIBRARIES (2002) DSpace Format Support. Cambridge, MA: MIT Libraries. Available from: http://libraries.mit.edu/dspace-mit/build/policies/format.html MCGEEVER, M (2007) IPR in Databases. Edinburgh: DCC, Oct. 2007. Available from: http://www.dcc.ac.uk/resource/legal-watch/ipr-in-databases/ MCGEEVER, M (2009) Science Commons. Edinburgh: DCC, Mar. 2009. Available from: http://www.dcc.ac.uk/resource/legal-watch/science-commons/ References 38 MOORE, R. W., RAJASEKER, A., and MARCIANO, R. (2007) Implementing Trusted Digital Repositories. DigCCurr2007 International Symposium in Digital Curation, Chapel Hill, North Carolina, April 2007. Available from: https://www.irods.org/pubs/DICE_DigcCur-Trusted-Rep-07.pdf MÜLLER, H (2009) Database Archiving. Briefing Paper. Edinburgh: DCC. Available from: http://www.dcc.ac.uk/resource/briefing-papers/database-archiving/ ONLINE COMPUTER LIBRARY CENTER (2007) Trustworthy Repositories Audit & Certification: Criteria and Checklist. Version 1.0. Dublin, OH: OCLC. Available from: http://www.crl.edu/PDF/trac.pdf OPEN KNOWLEDGE FOUNDATION (2009) Comprehensive Knowledge Archive Network. Cambridge: Open Knowledge Foundation. Available from: http://ckan.net/ OXFORD UNIVERSITY RESEARCH ARCHIVE (2008) ORA Policies. Oxford: University of Oxford. Available from: http://www.ouls.ox.ac.uk/ora/ora_documents2/ora_policies PARADIGM PROJECT (2007) Metadata for Authenticity: Hash Functions and Digital Signatures. Universities of Oxford and Manchester. Available from: http://www.paradigm.ac.uk/workbook/metadata/authenticity.html PURDUE UNIVERSITY LIBRARIES (2007) D2C2, Distributed Data Curation Center. Purdue University Libraries, West Lafayette, IN, USA. Available from: http://d2c2.lib.purdue.edu/ REPOSITORIES SUPPORT PROJECT (2008) Repository Policy Framework, Briefing Paper. Bristol: JISC. Available from: http://www.rsp.ac.uk/pubs/briefingpapers-docs/repoadmin-policyv2.pdf RESEARCH INFORMATION NETWORK (2008) Stewardship of digital research data - principles and guidelines. London: RIN. Available from: http://www.rin.ac.uk/data-principles RURAL ECONOMY AND LAND USE PROGRAMME DATA SUPPORT SERVICE (2006) Guidance on Data Management. Essex: University of Essex. Available from: http://www.data-archive.ac.uk/relu/reluaug2006.pdf RURAL ECONOMY AND LAND USE PROGRAMME DATA SUPPORT SERVICE (2009) Data Management Plan. Essex: University of Essex. Available from: http://www.data-archive.ac.uk/relu/plan.asp References 39 SCIENCE COMMONS. Submit your database policy. San Francisco: Science Commons. Available from: http://shirleyfung.com/mbdb/submit.php SHERPA. (2007) OpenDOAR Policies Tool. Nottingham: University of Nottingham. Available from: http://www.opendoar.org/tools/en/policies.php SMITH, M. and MOORE, R. (2006) Digital Archive Policies and Trusted Digital Repositories - 2nd International Digital Curation Conference Digital Data Curation in Practice, Glasgow, 21–22 November 2006. Glasgow: DCC. Available from: http://pledge.mit.edu/images/6/6f/Smith-Moore-DCC-Nov-2006.pdf TRELOAR, A., GROENEWEGEN, D., and HARBOE-REE, C. (2007) The Data Curation Continuum: Managing Data Objects in Institutional Repositories. D-Lib Magazine, 13 (9/10). Available from: http://www.dlib.org/dlib/september07/treloar/09treloar.html UK DATA ARCHIVE (2008a) Data Formats and Software. Essex: University of Essex. Available from: http://www.data-archive.ac.uk/sharing/formats.asp UK DATA ARCHIVE (2008b) Manage and Share Data - Authenticity and Versioning of Data. Essex: University of Essex. Available from: http://www.data-archive.ac.uk/sharing/version.asp UK DATA ARCHIVE (2008c) Manage and Share Data - Data Documentation and Metadata. Essex: University of Essex. Available from: http://www.data-archive.ac.uk/sharing/metadata.asp UK DATA ARCHIVE (2008d) Manage and Share Data - Consent, Confidentiality and Ethics. Essex: University of Essex. Available from: http://www.data-archive.ac.uk/sharing/confidential.asp US NATIONAL LIBRARY OF MEDICINE (2007) Digital Repository Policies and Functional Requirements Specification. Prepared by the NLM Digital Repository Working Group, Version 1. Bethesda, MD: USNLM. Available from: http://www.nlm.nih.gov/digitalrepository/NLM-Digital-Repository-Requirements-rev032007.pdf US NATIONAL SCIENCE BOARD (2005) Long-Lived Digital Data Collections: Enabling Research and Education in the 21st Century. Arlington, VA: National Science Foundation. Available from: http://www.nsf.gov/pubs/2005/nsb0540/ References 40 UNIVERSITY OF OXFORD (2005). Metadata in the Oxford University Digital Library. Oxford: University of Oxford. Available from: http://www.odl.ox.ac.uk/metadata.htm WITT, M. (2008) Institutional Repositories and Research Data Curation in a Distributed Environment. Baltimore: Library Trends. Available from: http://muse.jhu.edu/journals/library_trends/v057/57.2.witt.html References