7
Página 1 de 7 Why bother with digitisation? Users and using digital resources Content written on: 8th June 2004 by William Kilbride. Content updated: 8th June 2004 by William Kilbride. Why bother with digitisation? Users and using digital resources ....................................1 Proving the argument: digital archives and empirical evidence................................................. 1 Checking the facts: Digital reference resources ............................................................................. 2 Any time any place any where: the digital surrogate.................................................................... 4 From learning to research: digital objects in the classroom. ......................................................... 5 Discussion: interfaces for different types of use................................................................................ 5 Conclusion .............................................................................................................................................. 6 Links and Bibliography .......................................................................................................................... 6 Digitisation is not an end in itself. Successful digital projects begin with a defined set of uses, and an implied set of users. The use to which the digital resource is put relates to real world tasks and behaviours, so it stands to reason that digitisation efforts ought to be cognisant of the real world tasks that users will be expected to fulfil. Consequently, expanding the number of users is not simply a question of marketing: it involves expanding the number of tasks that a resource will support. As will be seen, the fit between a digital resource and its application in these tasks may be influenced by artefacts of the digitisation process and its supporting activities. Furthermore, over-extending a resource risks undermining the core tasks that it was created to fulfil in the first place. Detrimental aspects can be mitigated and positive aspects encouraged if suitable forward planning has taken place. By looking at stereotypes of use associated with different resources this paper informs the planning and development of new resources. Perhaps the most immediately visible opportunity associated with digital resources is the ability to transmit high volumes of information among many people over great distances with relative ease. Networking allows scholars to eschew many of the problems associated with the inflexibility of paper, reaching much wider audiences with novel, interactive forms of publication (e.g. Cunliffe 1996). As user experience and expectation of such resources both grows and hardens, and as the sheer volume of resources grows, so the need to identify and plan for user needs at the outset becomes more evident. But planning for use is not simply a question of maximising the number of 'hits' per month once a project is complete. Ensuring that user needs and expectations are met is a key component of the digitisation process itself. This paper presents four stereotypes of use based on real examples. It looks at digital resources as empirical data to support an argument; as reference resources to support the research process; as surrogates for real world phenomena; and as aids to teaching and learning in classroom and curricular settings. These stereotypes are neither self-contained nor exhaustive but by concentrating on how digital resources are used we explore problems and opportunities associated with their creation. Proving the argument: digital archives and empirical evidence 'Proving an argument' is seldom a matter of presenting uncontroversial facts. There is a complicated relationship in the humanities between empirical data and its interpretation. Currency, consistency, comprehensiveness, replicability, ethics and simplicity are important for high quality humanities research, while methods and the constitution of data are subjects for debate. Digital resources present unique opportunities for humanities disciplines to repeat experiments or to identify previously hidden patterns in data, as well as to question the constitution of the data itself. Distributing and describing empirical data electronically means that fellow researcher can examine, extend or refute conclusions based on data sets that may previously have been difficult to access or assess. Perhaps

AHDS-Why bother with digitisation.pdf

Embed Size (px)

Citation preview

Page 1: AHDS-Why bother with digitisation.pdf

Página 1 de 7

Why bother with digitisation? Users and using digital resources

Content written on: 8th June 2004 by William Kilbride.

Content updated: 8th June 2004 by William Kilbride.

Why bother with digitisation? Users and using digital resources ....................................1Proving the argument: digital archives and empirical evidence.................................................1Checking the facts: Digital reference resources .............................................................................2Any time any place any where: the digital surrogate....................................................................4From learning to research: digital objects in the classroom..........................................................5Discussion: interfaces for different types of use................................................................................5Conclusion ..............................................................................................................................................6Links and Bibliography ..........................................................................................................................6

Digitisation is not an end in itself. Successful digital projects begin with a defined set of uses,and an implied set of users. The use to which the digital resource is put relates to real world tasks andbehaviours, so it stands to reason that digitisation efforts ought to be cognisant of the real world tasks thatusers will be expected to fulfil. Consequently, expanding the number of users is not simply a question ofmarketing: it involves expanding the number of tasks that a resource will support. As will be seen, the fitbetween a digital resource and its application in these tasks may be influenced by artefacts of thedigitisation process and its supporting activities. Furthermore, over-extending a resource risksundermining the core tasks that it was created to fulfil in the first place. Detrimental aspects can bemitigated and positive aspects encouraged if suitable forward planning has taken place. By looking atstereotypes of use associated with different resources this paper informs the planning and developmentof new resources.

Perhaps the most immediately visible opportunity associated with digital resources is theability to transmit high volumes of information among many people over great distances with relativeease. Networking allows scholars to eschew many of the problems associated with the inflexibility ofpaper, reaching much wider audiences with novel, interactive forms of publication (e.g. Cunliffe 1996). Asuser experience and expectation of such resources both grows and hardens, and as the sheer volume ofresources grows, so the need to identify and plan for user needs at the outset becomes more evident. Butplanning for use is not simply a question of maximising the number of 'hits' per month once a project iscomplete. Ensuring that user needs and expectations are met is a key component of the digitisationprocess itself.

This paper presents four stereotypes of use based on real examples. It looks at digitalresources as empirical data to support an argument; as reference resources to support the researchprocess; as surrogates for real world phenomena; and as aids to teaching and learning in classroom andcurricular settings. These stereotypes are neither self-contained nor exhaustive but by concentrating onhow digital resources are used we explore problems and opportunities associated with their creation.

Proving the argument: digital archives and empirical evidence'Proving an argument' is seldom a matter of presenting uncontroversial facts. There is a

complicated relationship in the humanities between empirical data and its interpretation. Currency,consistency, comprehensiveness, replicability, ethics and simplicity are important for high qualityhumanities research, while methods and the constitution of data are subjects for debate. Digital resourcespresent unique opportunities for humanities disciplines to repeat experiments or to identify previouslyhidden patterns in data, as well as to question the constitution of the data itself. Distributing anddescribing empirical data electronically means that fellow researcher can examine, extend or refuteconclusions based on data sets that may previously have been difficult to access or assess. Perhaps

Page 2: AHDS-Why bother with digitisation.pdf

Página 2 de 7

more importantly, they can also examine how the data has been constituted and methods associated withit. This changes significantly the relationship between research publications and research archives.

The Ave Valley Survey project is a good example of how this opportunity has been exploited.This extensive field survey of a river valley in northern Portugal led to various conclusions about therelationship between Iron Age, Roman and early medieval settlement of the area, and thus gave insightsto major themes such as Romanisation in Western Europe. The research team presented its conclusionsin a substantial e-monograph (Millett et al 2000), but is able to support this monograph by allowingreaders to reach into digital archives of the project (Millett 2000). Presenting these unpublished files indigital form allows reviewers to see parts of the project that might otherwise be hidden, and to reassessconclusions based on them.

For example, 'cost surfaces' are used to analyse aspects of settlement in the area. In thepublication these are presented as a series of maps, but in the archive it is possible to test the digitalelevation model upon which these cost surfaces are derived. The elevation model is a triangulatedirregular network based on the detailed digitisation of contour lines. Contour lines are an interpolation ofheight, and thus the cost surfaces derive from an interpolation of an interpolation of real world heights. Ifthe digital elevation model had instead been derived from observed spot heights, then it would be onestep closer to the real world, and thus conclusions based on it would be more compelling. If the contourshad been used to create some other form of digital elevation model such as a raster grid, or if thealgorithm used to specify costs had been different, then the maps of expected cost may too havechanged. By presenting the underlying data, an alterative line of research is possible. This example is notintended as a critique of Millett: by making underlying data available in this way they open up debate anddiscussion that would otherwise be impossible. It shows that by presenting the underlying data, theprocess of research and re-analysis is strengthened.

Sharing data in the arts and humanities is not simply a question of swapping processing andintegrating uncontroversial facts about the world. The interpretative process depends on being able tocontest data and the methods by which it is processed. Retaining and presenting the digital by-productsof research for scrutiny by others is more than just a way to open debate: it is an essential part of thereview process.

The digital resources presented in the Ave Valley archive were created for the purposes ofthe research project itself, so the primary user community is the project team. In such circumstances,attention focuses on the research design and there is little to be gained from an intensive evaluation ofuser needs or expectations: a focus group of the public is unlikely to contribute to the success of thedigital resource. But given that it is expected that these resources can and should be used by fellowresearchers, it is important to document the processes that led to salient conclusions being drawn. Thisincludes not just a description of the data set, but also the data gathering methods - in Ave Valley there isa detailed description of fieldwork methods - assessments of data quality and a description of theprocesses applied to derive relevant conclusions. This supporting metadata can be as important to futurescholars as the data itself, and there is little to be gained from simply presenting the data without anexplanation of what it is and how it was derived. It may even be appropriate to prompt such users byidentifying unexplored research topics that might be supported with the digital resource.

Checking the facts: Digital reference resourcesComprehensiveness is a characteristic of first-rate scholarship in all disciplines. Anyone who

has completed a research project is familiar with the 'literature search' where every single article, bookand thesis on a given topic is listed, acquired and digested. The Internet has transformed 'resourcediscovery'. Whereas previous generations faced a challenge simply compiling a suitable bibliography, theInternet allows scholars to share bibliographic and other information with relative ease (Clarke, Hardmanand Kilbride 2003). Whereas previous generations of scholars constructed and consulted multi-volumeddefinitive editions that described every known example of a specific phenomenon, it makes a lot moresense now to create databases of such resources and then create online interfaces to them. The onlinecorpus is faster to search, often cheaper to produce, easier to update and accessible to a much wideraudience.

An example of an online digital corpus is English Heritage's Excavation Index for England.This index describes and summarises all archaeological research undertaken in England in any givenyear. Reaching back to the early years of antiquarian interest, it is an essential tool for anyone interestedin local history or archaeology. Because it grows year on year, it attracts frequent repeat visits. The

Page 3: AHDS-Why bother with digitisation.pdf

Página 3 de 7

Excavation Index is an exemplary reference resource, but various aspects of it demonstrate some of theproblems associated with digitising all reference works.

Firstly, and perhaps most importantly, the Excavation Index is precisely what it claims to be -an index. The virtue of comprehensiveness is offset by its relatively shallow coverage of each individualrecord. This has two implications for users and developers. On one hand, users can very quickly getaccess to the sorts of data that might otherwise have slipped their notice - especially in terms of 'greyliterature', the small or unpublished reports that are not readily accessible. Perversely, an attempt tocreate access underlines exclusion, especially if there is no mechanism by which users can follow uprequests. Resources that point to offline or rare resources should include some explicit method by whichmore detailed questions can be pursued, such as bibliographic references or contact details. On the otherhand, curatorial staff in libraries archives or museums may become inundated with requests forinformation that they may either not have, or not be able to supply. It is therefore sensible to plan not justdigitisation but to have in mind the implications for increased demand on conventional resources that mayresult, such as interlibrary-loan requests or research trips. Experience shows that users often expectmuch more than can reasonably be supported, so if necessary expectations should be managed: it iswiser to promise little and over-deliver than to promise much and disappoint.

A record set as large as the Excavation Index is created by staff in many different agencies.In these circumstances it is important to develop terminology controls to prevent confusion. Appliedthoughtfully, however, such terminology controls can have three uses beyond speeding up thecompilation of the underlying data set. Firstly, they are a measure of data quality - both in terms ofoperator error (accuracy) and as an implicit declaration of precision. Secondly, terminology controls arepowerful tools for retrieving data: so if users can be taught the terminology, or better still the terminologydraw from familiar subject classifications, then users will be equipped with some of the basic tools tounderstand the data. Finally, terminology controls and thesauri can be difficult to construct, so it mayoften be possible to adopt or adapt an existing vocabulary. At very least, by making your vocabularyavailable one can expedite the work of others. Promoting your own thesauri, or adopting someone else'senhances enormously the ability to cross-search more than one data set, and this can increaseexponentially the user-base.

Cross-searching is also enhanced by familiarity with relevant 'resource discovery' standards(Miller and Greenstein 1997). The Excavation Index fits into a qualified implementation of the Dublin Coremetadata element set. Consequently, the Excavation Index for England can be cross-searched withcognate data sets like the National Monuments Record of Scotland and the National Trust Sites andMonuments Record. Cross-searching makes the Excavation Index much more powerful as a referencetool.

Geography is often overlooked as a classification scheme (Kilbride 2004). Archaeologicalresearch is pre-discursively geographic, so it is little surprise that the Excavation Index contains NationalGrid References. However, the grid references are more than a mere convention. They also present aninstant classification of the records, and create a map-based search interface. This is possible when twoor more resource share a standard geographic base like the National Grid. This makes it possible to builda powerful cross-search facility that requires little prior knowledge from users. Indeed, mobile phones andglobal position satellites are able to relay locational information directly to personal computers and thus,in an experimental interface, connect directly to local records in the Excavation Index. This is possiblebecause the standard geographic base is widely shared and readily understood.

Some reference resources are definitive and need only occasional updating, though evenwhere the number of specific cases of a phenomenon is largely fixed, the bibliographies associated withthem will grow (e.g Sawyer 1968). Frequent updating may have implications for quality assurance andthus for long-term maintenance. The Excavation Index now has an online update tool, but data submittedin this way is only included in the Index after it goes through three different quality checks whichinvestigate its technical competence, its veracity and its adherence to standards (Hardman and Richards2003). This process implies a long-term commitment. Frequent updates to large reference works are anobvious benefit of digitisation, but the implication in terms of long-term maintenance should not beoverlooked.

Digitising reference resources can have a profound impact on scholarship, but theirdigitisation can cause unforeseen frustrations if not handled well. It may point users to resources that theycannot in fact obtain, or it may place a strain on curators and librarians to supply records that are difficultof access. Vocabulary standards are important from a data quality perspective, but they can also greatlyenhance the use and flexibility of a resource. Attention to appropriate metadata standards, especially

Page 4: AHDS-Why bother with digitisation.pdf

Página 4 de 7

when vocabulary standards are consistent or a common geographic base is used, makes cross-searchingpossible, and thus increases the impact that a resource may have. Finally, update tools are attractive, butquality assurance means that they imply a long-term commitment.

Any time any place any where: the digital surrogate'Digital surrogate' is the name given to a digital version of an expensive, fragile or

inaccessible real world object. The digital surrogate is attractive to archivists and librarians because itreduces wear and tear, but it also allows researchers to work on real world models from the comfort oftheir desks, bringing together or sharing material that may be large, heavy, unwieldy or friable. Thismeans that researchers can reduce the number and duration of expensive research trips, can work inconditions of their own choosing, and can return over and again to the same question withoutinconveniencing colleagues. Supplementing real access, the digital surrogate ensures that whenresearchers require access to the original, they are better able to carry out their work.

An example of a digital surrogate is the British Library's Electronic Beowulf (Kiernan et al1999). This ambitious project has brought together the oldest surviving manuscript of Beowulf andpresented it alongside three of its earliest transliterations. This enables readers to trace the origin of allmodern editions, observing where words or letters have been lost since the start of the nineteenthcentury. Moreover, sophisticated scanning has been used to highlight aspects of the manuscript, such asscribal corrections. The digitisation is of a very high resolution so that it is possible to examine the text indetail. Individual files are assembled in the order that they appear in the manuscript, giving theappearance of page turning, and the whole thing is accompanied by a modern edition that can be read inparallel.

There are four points worth noting about this digital surrogate that may impact on othersimilar projects. Firstly, the high resolution is essential for a project like this to work. Electronic Beowulfallowed analysis of the manuscript at much greater detail than the naked eye would allow, and by usingthe invisible ends of the spectrum, drew attention to details that could not otherwise be seen. If themanuscript had not been presented to such a high resolution, then researchers would have reverted toaccessing the original. This high resolution has the consequence of large files, which in turn hasimplications for data delivery. Not surprisingly, Electronic Beowulf is presented on a set of CDs as the filesizes precluded online delivery. Compression algorithms can now be used to render details withoutcreating huge files to download, but this technology was not available at the time. High resolutionscanning and imagery may also become a threat to the original document or object to be digitised. Forexample, the high-powered lamps required to photograph in detail may in fact be more of a threat to theoriginal object than conventional access. Alternatively, if heavy lifting gear is to be used for large objects,then there should be appropriate health and safety planning. A detailed risk assessment should precedeany digitisation of a valuable or unusual original, with the possibility that the resource not be digitised atall. Presuming that digitisation does proceed the resulting digital object should have a preservation plan ofits own. Digitisation may be unrepeatable, expensive or undesirable. It is ironic that a plan to safeguard areal world object creates a need to safeguard a digital one too.

Perhaps one of the most instructive aspects of Electronic Beowulf is the relationshipbetween the modern edition and the images of the twelfth century manuscript. The modern edition is theproduct of an intensive research project, and is to some extent a product of an analysis of the digitisedmanuscript. But there is more to it than this - the modern edition also performs a supporting role to helpunderstand the images. The script of the medieval document is difficult for human readers, and makescomputer reading - optical character recognition - impractical. This makes it impossible to search throughthe images for occasions of specific words or phrases. However, by presenting a parallel modern edition,and by linking this with the images of the manuscript, the developers make it possible to search theimages of the manuscript vicariously. In essence, the modern edition serves as an index to the images,and thus is a surrogate of the surrogate.

Digital objects render the real world systematically, and in so doing they frequently either failto render the irregularities of the real world, or simplify them. This can be a critical problem for digitalsurrogates that seek to represent objects whose irregularities contain significant properties. The choice ofdata model influences significantly the sorts of use that a resource is fit for. The palaeography andarchaeology of the Beowulf manuscript is a case in point. The penmanship and form of the document isnot simply a source of frustrating irregularities that can be systematised: they are crucial to establishing orcontesting the age of the document. A simple computerised transcription would restrict significantly theability of readers to understand these aspects of the document, and the decisions made by the editorwhere established editions are disputed. Though it may create considerably more work for the digitisation

Page 5: AHDS-Why bother with digitisation.pdf

Página 5 de 7

project, the fit between the data model, the real world object and the real world tasks the surrogate isintended to support is essential.

From learning to research: digital objects in the classroom.Digital resources can be effective tools to assist and even supplant classroom teaching. But

as Laurillard (1995) notes, we should not confuse access to digital resources with teaching or learning. Insimple terms learners are not yet ready to be researchers: we cannot simply digitise a resource andexpect student's learning to flourish. Pathways into or through the digital resource, and tasks that helpstudents explore aspects of the resource are essential. Problem-based-learning with digital resources canbe among the most effective forms of active student learning, and in the end can equip students with theskills to undertake their own open ended engagements with the underlying digital resource (Kilbride andReynier 2002). However, the effort involved in creating such resources can be significant, and isconsiderably greater than the effort of digitising. Consequently, when a teaching and learning strategy isquoted in support of digitisation, curriculum design should be as much if not more of a concern thandigitisation.

The PATOIS project provides a good example of where digital resources have been used inteaching and learning (for a discussion of this see Kilbride et al 2002). Five aspects of this project areworth considering. Firstly, the four tutorial packs developed in the project focus on information skills forstudents, so the Internet is an appropriate vehicle to present these tutorials. Secondly, the tutorials selectfrom a much larger set of resources. Students are presented with only a small selection of the resourceson the understanding that, having completed the tutorials they will be equipped with the skills they need toexplore the much larger whole. Thirdly, students, lecturers and subject specialists were involved in theircreation. This means that the resources fit with a demonstrable need, and that they are written withspecific skill-levels in mind. Fourthly, the appreciation of prior skills, learning outcomes, aims andobjectives means that the tutorials are congruent with widely expressed subject needs and disciplinarynorms (such as QAA 2000). This process also involved an understanding of the hardware and softwareplatforms available to the end users, and thus a decision about the levels of software required to renderor process data for users. Open standards were used, maximising the range of platforms that can supportthe tutorials. Finally, the tutorials are written in as consistent a manner as possible, maximising studentfamiliarity and reducing the possibility of the technology getting in the way of the learning that it isintended to support.

The implication of PATOIS and other projects like this is that teaching and learning cannotbe supported by digitisation alone. The development of electronic teaching and learning materialsrequires more than technical skills, and indeed that the technical skills may be relatively trivial as againstthe intellectual challenge that teaching and learning poses.

Discussion: interfaces for different types of useThese examples are instructive because they reveal different types of use that a digital

resource may have. These uses are not and should not be exclusive, and each of the examples quotedabove can be deployed in more than one of the stereotyped modes above. For example, the Ave Valleyarchive is used in teaching and may even be considered a digital surrogate for the archaeology ofNorthern Portugal. The digitised manuscript of Beowulf is to some extent one large archive supporting thenew edition of the text. The Excavation Index is available as a set of downloads to be analysed andprocessed offline, much in the same way as a research resource can be.

Nonetheless, alternative uses are constrained by the nature of the resource. For example,the lack of a translation with Electronic Beowulf undermines claims that it is readily used in teaching andlearning: it is an academic text for an expert audience (Kilbride 2000). The Excavation Index may beexcellent for reference, but it does not yet provide the deep access that, in another context, Beowulf does.The Ave Valley survey data does not pretend to be a comprehensive account of the archaeology ofNorthern Portugal, so it can only act as a reference source in very specific circumstances. PATOIS mightbe excellent for teaching and learning but it is really only an extended interface to data already digitised.

The degree to which a resource can be redeployed in other ways is in part a function of howit is presented, so projects that want to maximise the flexibility of a resource should consider carefully theonline presentation. The uses of the data sets described above feed into the way that they are presented.

As well as being authoritative and comprehensive, reference data needs to be easily andquickly queried. It would make little sense to present the data as a single large file with supportingdocumentation: better by far to let users build complicated queries quickly, get results rapidly, ignore

Page 6: AHDS-Why bother with digitisation.pdf

Página 6 de 7

irrelevant records and repeat the process with ease. The interface to the Excavation Index facilitatesthese sorts of reference searches by allowing complex searches to be repeated, and in addition allows itto be cross-searched.

The approach to research archives may be radically different: to evaluate the resource, aresearcher will need to see the whole data set, will want to understand how parts relate to the whole, andto try out the data with their own tools. For research uses it makes sense to let fellow researchers transferdata directly to their own computers. The implication is that the interface can be relatively simple, and thatdata files need to be presented in widely supported formats. This may be considerably less work than isrequired for other forms of digital resource. However, the academic process implies a degree of stability,and that alterations and additions clearly marked. Academic research is based on the credibility of thedata presented, so it would bring a researcher into disrepute if published data was changed without theoriginal being retained. This does not mean that changes cannot be made: rather that they should beclearly signalled and incremental. The archive and publication of the Ave Valley data are fixed in so far asthe data and discussion published are not altered. Subsequent additions and extensions are possible, butonly as a supplement not a replacement.

Academic referencing is also important for data that is intended to prove an argument.Conventions of bibliographic referencing like the Harvard convention include guidance on how toreference Internet and other digital resources. For these to work, not only should the data set be stable inits content but in access arrangements and location. The chain of references upon which academiccitation is intended to allow subsequent researchers to follow an argument. It stands to reason thereforethat URLs need to be stable, and that they point to the same resources. Without this sense of authority,research into the archives will be inhibited and the digitisation effort undermined.

Teaching and learning is perhaps the most demanding in terms of interface, and certainlyhas the largest amount of research associated with it (inter alia Littlejohn and Higgison 2003, Armitageand O'Leary 2003 Core, Rothery and Walton 2003, Kilbride and Reynier 2002). Such advancedunderstandings of user needs demand the attention of those engaged in teaching and learning projects,not least because of the discourses of professionalism that are now attached to them (Newland andRingan 2003, Jenkins and Hanson 2003).

ConclusionDigitising has many benefits, not least the large number of users that can access and

contribute to a digital resource. But digitisation is not an end in itself, nor should users be taken forgranted. Evaluating the uses that a resource can support, and understanding how those uses can beexpanded is critical to the success of any digitisation project. Decisions made early in the digitisationprocess can both reduce and expand the ability of a resource to meet these expected uses, so it makessense to consider these uses at an early stage.

Links and BibliographyAve Valley

http://ads.ahds.ac.uk/catalogue/projArch/ave_millett/http://intarch.ac.uk/journal/issue9/millett_index.html

Excavation Indexhttp://ads.ahds.ac.uk/catalogue/collections/blurbs/304.cfm

Electronic Beowulfhttp://www.bl.uk/collections/treasures/beowulf.html

PATOIShttp://ads.ahds.ac.uk/project/patois/module1/index.htmlhttp://ads.ahds.ac.uk/project/patois/module2/index.htmlhttp://ads.ahds.ac.uk/project/patois/module3/index.htmlhttp://ads.ahds.ac.uk/project/patois/module4/index.html

Armitage, S and O'Leary R 2003 A Guide for Learning Technologists, LTSN Generic CentreE-learning Series No 4, York

Page 7: AHDS-Why bother with digitisation.pdf

Página 7 de 7

Clarke, JP, Hardman, CS and Kilbride WG 2003 'Comprehensiveness for All: The OASISProject and Research Values in the Digital Age' in CSA Newsletter XV, online at:http://www.csanet.org/newsletter/winter03/nlw0305.html, last visited 12/03/04

Core, J, Rothery A, and Walton G 2003 A Guide for Support Staff, LTSN Generic Centre E-learning Series No 5, York

Cunliffe, B 1996 'Foreward' in Internet Archaeology 1, online athttp://intarch.ac.uk/journal/issue1/foreword.html last visited 06/03/04

Hardman, CS and Richards, JD 2003 'OASIS: Dealing with the Digital Revolution' in M Doerrand A Sarris (eds) The Digital Heritage of Archaeology: Computer Applications and Quantitative Methodsin Archaeology, Archive of Monuments and Publications, Hellenic Ministry of Culture 325-9

Jenkins M and Hanson J 2003 A Guide for Senior Managers LTSN Generic Centre E-learning Series No 1, York

Kiernan K, Presscott A, Solopova E, French D, Cantara L, Ellis M and Yuan CJ 1999Electronic Beowulf, British Library and the University of Michigan Press, London

Kilbride, WG, Fernie KM, McKinney P, Richards, JD 2002 'Contexts of Learning: ThePATOIS project and Internet-based teaching and learning in Higher Education' in Internet Archaeology12, online at http://intarch.ac.uk/journal/issue12/patois_toc.html last visited 12/03/04

Kilbride, WG and Reynier MJ 2002 'Editorial - Keeping the Learning in Computer-BasedLearning' in Internet Archaeology 12 online at http://intarch.ac.uk/journal/issue12/editorial.html last visited12/03/04

Kilbride WG 2004 'From One Context to Another: Building a Common InformationEnvironment for Archaeology' in CSA Newsletter XVI online athttp://www.csanet.org/newsletter/winter04/nlw0402.html last visited 12/03/04

Kilbride, WG 2000 'Whose Beowulf is it any way' in Internet Archaeology 9, online at:http://intarch.ac.uk/journal/issue9/reviews/beowulf.html last visited 12/03/04

Laurillard, D 1995 'Multimedia and the changing experience of the learner', British Journal ofEducational Technology 26, 179-89

Littlejohn A and Higgison C 2003 A Guide for Teachers LTSN Generic Centre E-learningSeries No 3, York

Miller, P and Greenstein, D 1997 Discovering Online Resources in the Arts and Humanities:a practical implementation of the Dublin Core, AHDS and UKOLN, London. Also online at:http://ahds.ac.uk/public/metadata/discovery.html last visited 12/03/04

Newland B and Ringan N 2003 A Guide for Heads of Department LTSNGeneric Centre E-learning Series No 2, York

Millett, M 2000 Ave Valley Survey Project, Porto, Portugal, online at:http://ads.ahds.ac.uk/catalogue/projArch/ave_millett/ last visited 12/03/04

Millett, M Queiroga, F Strutt, K and Willis, S 2000 The Ave Valley, northern Portugal: anarchaeological survey of Iron Age and Roman settlement in Internet Archaeology 9, online athttp://intarch.ac.uk/journal/issue9/millett_toc.html last visited 12/03/04

Quality Assurance Agency, 2000 Archaeology: Subject Benchmarking Statement, QualityAssurance Agency for Higher Education, Gloucester

Sawyer, PH 1968 Anglo-Saxon charters an annotated list and bibliography, Royal HistoricalSociety, London