Joint, N. and Kemp, B. and Ashworth, S. (2000) Information skills training in support of a joint electronic library in Glasgow: the GAELS project approach to library courseware development. Aslib Proceedings, 52(5). pp. 301-312. ISSN 0001-253X

http://strathprints.strath.ac.uk/2349/

Strathprints is designed to allow users to access the research output of the University of Strathclyde. Copyright (c) and Moral Rights for the papers on this site are retained by the individual authors and/or other copyright owners. Users may download and/or print one copy of any article(s) in Strathprints to facilitate their private study or for non-commercial research. You may not engage in further distribution of the material or use it for any profit-making activities or any commercial gain. You may freely distribute the url (http://strathprints.strath.ac.uk/) of the Strathprints website.

Information skills training in support of a Joint Electronic Library in Glasgow: the GAELS project approach to library courseware development.

Susan Ashworth
Glasgow University Library, Glasgow. G12 8QE.

Email: n.c.joint@strath.ac.uk

Published in: Aslib Proceedings, 2000, September.

ABSTRACT: The GAELS Project is a two year project funded by the SHEFC strategic change initiative which promotes collaborative information services to Engineering researchers at Glasgow and Strathclyde Universities. This paper examines the role of user education in this process. We use arguments against the effectiveness of library skills education and evaluative methods learned from human-computer interface design as means of improving information skills training and as part of a general reflection on user education and library services. Such an approach shows how networked learning materials can be an effective tool for promoting a collaborative library service across the Glasgow Metropolitan Area Network.

Introduction

GAELS is a two year research project, started in July 1998 with funding from the SHEFC strategic change initiative. Its remit includes the promotion of a culture change among Engineering academics and researchers at both institutions. This change would see a decreased dependence on the hard copy engineering collections at both sites and increasing electronic use of networked access to full-text resources, possibly, but not necessarily supplied by a local document delivery service from the richer engineering collection at Strathclyde University Library.

The project has conducted an initial information audit on the use of information by engineers at both institutions1 and has designed a set of courseware which provides engineering research academics with networked information retrieval skills that will make them more independent of their local engineering research collection. One possible outcome of this culture change might be the opportunity to rationalise hard copy holdings at both libraries, which display a fair degree of overlap in core material.

Is library and information skills training misdirected?

Proponents of this case such as Pacey (2) argue for a self-explanatory library in which users 'acquire library skills by using the library: all is explained; everything is transparent'; Savenije (3) points out that library instruction 'has become indispensible since the use of library services has been made too intricate by the libraries themselves...Computer interfaces are evolving into user-friendly tools, so enhancing the accessibility of information.' And lastly Herrington (4) argues that 'Library systems must be so easy to use and transparent that there is no need for library instruction.' Both Pacey and Savenije also promote the idea of reintegrating information skills acquisition into the context of users' mainstream coursework or research, rather than stripping out user education into the vacuum of free-standing information skills acquisition classes delivered by librarians. Herrington, however, argues simply that the Ohio State University gateway design minimises the need for user education, the result of refocussing instructional librarians on designing user-friendly interfaces, rather than teaching.

If this argument is correct, its impact on a project such as GAELS would be significant. Firstly, our attention would turn to the refinement of both the interface to the services developed by the project, and also the interfaces to established local services. Secondly, we would focus on the issue of support for lecturing staff as they expanded the information skills content of their own teaching. However, there are important arguments against the 'transparent interface' and 'reintegration into the curriculum' school of thought which make the creation of a set of free-standing learning materials such as the GAELS courseware still valid.

Counter-arguments (1): the myth of total transparency

However, Shneiderman (6), citing Rutkowski, comments:

'Even though increasing attention is being paid to improving user-interface design, the complexity of online systems grows. There will always be a need for supplemental materials that aid users, in both paper and online form...'

Shneiderman cites as examples of these supplemental materials, brief getting started notes and the online tutorial. Other HCI texts also cite help lines and manuals. The equivalent library materials are our familiar library handouts, library computer-aided learning materials, reference desks and library guides.

And what is true of user interfaces, is true of the traditional library. There are specific reasons why transparency of use is difficult to achieve in the traditional library. The usability of the physical library is hard to optimise. The hard copy collection may be well laid out at one point in time, but that arrangement becomes dated, and the physical rearrangement, re-signing and guiding of printed collections is costly or impractical. Acceptable compromises are made, and user education or library orientation sessions make up for the shortcomings.

By contrast, the electronic library may appear to be easily configured and reconfigured for maximum usability, but most library datasets are externally provided by commercial suppliers, and with this comes the proliferation of interfaces experienced in a typical academic library context (as Shneiderman says 'the complexity of online systems grows...'). Prior to advances on the Z39.50 front, the local library cannot control interfaces that resist being homogenised into a single user interface, transparent or otherwise.

Counter arguments (2): the HE context, academic preference and the FTE economy

However, the clear perception among academic Engineering staff at Glasgow and Strathclyde was that their information retrieval skills had lagged behind the growth in electronic library services. Just as they were not in a position to move to electronic browsing for research information, neither were they in a position to provide information skills training through their teaching or research supervision.

Moreover, library user education is a unique form of teaching activity in British Higher Education in that it is the one of the few extensive forms of service teaching that is provided without the requirement of FTE payments to the teaching department (the library). Because of this, it takes place without the evaluative filter of teaching quality assessment. This is unlike, for example, engineering mathematics service teaching, delivered by Mathematics departments to Engineering departments for FTE income within the structure of TQA. This simple fact of Higher Education economics does create a financial incentive to departments to accept library and information skills teaching delivered outwith the departmental teaching or research context. Fewer departmental resources are involved.

Unfortunately, because such user education is exempt from the teaching quality assessment that departmental teaching receives, it makes the anti-user education argument so easy to promote. If such teaching is never formally evaluated by an outside agency, and if libraries themselves never formally evaluate the quality of their teaching (which is the case according to analyses of library user education evaluation carried out by writers such as Bober(8)), it is hardly suprising that a strong argument can be made against the quality and validity of library-based user education. There is no counter argument being made, no attempt to evaluate and justify this teaching on a wholescale basis.

Decisions on a way forward:

Secondly, and most importantly, library-led user education should continue because, if done well, it is worthwhile. The difficulty of creating a largely transparent user interface to library services means that compensatory information skills training will be necessary for the foreseeable future. However, interface design methods also demand that evaluation of the quality of your training programmes and library system constantly takes place.

Bearing these considerations in mind, our challenge was to design a set of learning materials which responded to the following requirements:

• The courseware should be related to the task-content of departmental teaching but still delivered as a course-independent package.
• It should use the goal of interface transparency as a requirement for the interface to the learning materials, regardless of the lack of transparency of the information services which the package described.
• The package should be created on the basis of evaluative feedback from users, in order to replicate in some way the role of departmental teaching quality assessment procedures and to acknowledge the requirement for evaluation of interface quality as demanded by human-computer interface design methodology.

Task-orientation in library courseware

One important principle for making user interfaces as effective as possible is the idea of task orientation, that is the aim of making interfaces task-specific, rather than type- or process-specific. It seems that this principle can give library courseware developers a way of creating a degree of task relevance without embedding learning materials directly into coursework, however desirable that might ultimately be.

Although the literature of task analysis has become complex and rich in the last two decades, for our purposes we could use these methodologies selectively in support of our courseware development. For example, in Lewis and Rieman's (9) description of the interface development process they emphasise the selection and analysis of real, complete, representative tasks for task-centred design as opposed to abstract, partial elements. They also argue that the interface designer should differentiate tasks from processes, that is 'define what the user needs to do, not how it will be done'. What does this mean when applied to library skills training, or library interfaces in general?

If one were to take a WWW-based set of library information resources, the interface to these resources could typically be an opening home page with a choice of hyperlinked paths through these resources. The Strathclyde University Library 'About the Library' WWW pages (10) (Fig. 1) adopt this format. They give paths into the data such as 'Opening hours', 'Libraries at Strathclyde', 'Membership and access' and 'Collections of the University Library'. Although such a WWW resource is useful, its interface is not at all task-centred in Lewis and Rieman's terms and lacks usability. In their terms, these hyperlinks are 'abstract, partial elements' in the activity of using a library, aspects of processes familiar to librarians; or they are comprehensive descriptions of processes, in which are submerged the tasks users wish to perform.

[(Fig. 1) Types, elements and processes: a WWW page from Strathclyde University Library]

Simply renaming these elements would help. For example, the 'Membership and access' link could be re-titled 'Joining the library'. At least then a library user would be able to identify an activity with which they were familiar. But the activity of 'joining the library' is only a small element in the complete and representative task of (for example) 'finding literature on your topic'. Similarly the link entitled 'purchasing stock' is an activity of marginal interest to most library users, and in this instance it leads to a detailed description of the local collection development process. The task of 'finding literature on your topic' may involve some knowledge of this material (for example, it is important to know if a library's collection development policy means that it is likely to collect information on your topic). But it is hard to pick out such information in the overall description of a library process.

Such a brief assessment of the task-centredness of a local library interface does a number of things.

Firstly, it illustrates the power of the anti-user education argument. In this case, a lack of task-orientation occludes the transparency of a local library interface. A user should be able easily to identify the tasks for which an interface is designed, rather than be presented with a variety of smaller elements of whole tasks or comprehensive overviews of library processes that mask task outlines.

Secondly, if this approach does indeed help improve the usability of an information resource interface, it also gives an indication of how to present learning materials in an effective way, with a usable interface.

Developing these insights further, there seemed to be a variety of concrete ways in which task orientation could inform the design of the package. The starting points from which users entered the GAELS courseware could be task-specific. This would mean arranging a gateway of access points to the courseware, each of which is a real, complete, representative task. For example, the Pathfinders Route through the EDUCATE Project's Into Info (11) CAL package gives the user a variety of named tasks via which to approach the learning materials, such as 'Starting to use the library', 'Searching for facts', and 'Starting research'. Then, at the most specific level of use, the course unit, any one page could start by identifying to the user the task or tasks to be mastered once they had completed the unit. Thirdly, and most obviously, the unit exercises, together with the feedback on the results of the exercises, should involve the execution of real, representative tasks. Lastly, analysis of user needs, evaluation and user feedback would indicate whether the content of the package was truly representative of the tasks needing to be performed by its target market, engineering researchers.

Librarians' preconceptions about information retrieval tasks

Two problems emerged as a consequence of this requirement. Firstly, at the level of interface development, there was stakeholder resistance to a task-oriented approach in the presentation of user education materials. Secondly, the librarians who were key institutional stakeholders had clear ideas of what information retrieval techniques should be taught, but this syllabus seemed likely to be problematic when evaluated against likely patterns of actual users' task execution.

As a compromise, task orientation at the page level was retained, but a larger structural architecture that mimicked the sequence of the searching process was used. Thus, moving in a linear sequence through the GAELS12 courseware package, the student first learns to search in Module 1 (e.g. learning Boolean skills, search field qualifiers and so forth). In the next Module they then find out where to go to implement their searching skills (that is, on search tools such as electronic library catalogues and bibliographic databases). Then in Module 3 they find out about the types of information listed by these search tools (theses, patents, grey literature and other types). This structure enacts the sequence of the search process, in that you start by learning how to search, then discover more about where to search and then learn about what you are searching for. (This is only an overview of the project- a full project report will be forthcoming)

[(Fig. 2) GAELS courseware structure diagram]

Evaluation

We noted the following features thrown up by our evaluative work:

• Assessment of learning outcomes is positive. This set of learning materials seems to fulfil the original remit of the project; however, this did not tell the whole picture.
• Evaluation may tell you how well you have taught someone to use your services but still does not show you how usable those services are. The anti-user education school would say that evaluation in this sense misses the point. It is more important to evaluate the usability of the services taught than the usability and effectiveness of the learning materials that facilitate the use of this services. This charge still had not been laid to rest. Indeed, the interest of the project in interface transparency issues had illuminated exactly what writers such as Herrington had described, questionable local interface design.
• Although evaluation of the effectiveness of the learning materials did not show that the larger process-oriented design structure impeded the usability of the learning materials, it did - as we had anticipated - reflect back on the validity of the syllabus of material taught. Feedback from interviews with research students who completed the course led us to question the correctness of the syllabus of skills or task-content which we were teaching.

'I've enjoyed learning the searching skills taught in this package and can definitely search better but I will still get my information the way I did before in preference. I prefer taking a few key references used by my research group and reading the references used at the end of each paper as my bibliography - searching on bibliographic databases gives too many hits and my own way of browsing through a few references means I don't have to work through long lists of less relevant hits.'

This research student was clearly using different information retrieval techniques to gather research information from those we were promoting. He had enjoyed using our courseware, but felt that the skills he had learned and improved on in the process of using the courseware (broadly definable as information searching skills) were not as relevant to the tasks he was engaged in as skills that he had evolved on his own (broadly definable as information browsing skills). He understood his information retrieval task as something requiring a different strategy from the ones we were giving him. The idea of precision and recall was accepted, because he realised he needed to raise recall from an overly precise core set of hits. But he trusted his own 'reference chaining' strategy more than a search technique such as ours.

Feedback such as this showed that we were successfully teaching the syllabus of information retrieval search skills we had chosen to teach, but that we needed to evaluate not only how well we taught our syllabus but also whether the syllabus itself was properly composed. This type of evaluative feedback reinforced our suspicion that our syllabus was rather conservative, in order to create acceptability amongst librarian teachers and get the learning materials accepted in everyday teaching practice.

What did the literature tell us about this problem? Our background reading to the project included writers such as Bates (14) and Marchionini (15) who have described alternative information retrieval techniques as demonstrated in actual user behaviour. However, we were interested to read articles such as those by Dyckman (16), who described a process-oriented approach to teaching end users search skills on the Dialog online host service. This is a valuable and insightful article. It shows a practitioner librarian seeking evidence of actual information retrieval behaviour by users, and re-evaluating their assumptions in the face of this. The article includes a description of end users generating and accepting high recall, low precision search results:

'Another technique used by many of NYU's DIALOG searchers was shocking at first to librarians. Analysis of DIALOG Classmate invoices showed many searches which retrieved numerous results. It seems that many want great recall and are little concerned with precision. The idea of the 'perfect 20 item search', developed in response to systems which penalize large retrievals (i.e. charge by item and time expended), is incompatible with academic end-user searchers' perceptions of their needs.'

Dyckman concludes that searchers need comprehensive recall, not precise recall. Our initial analysis of our user feedback leads us to suggest that perhaps they do want precise recall, but, as in the NYU case study, simply want to browse through search results in order to create a precise set. Browsing large sets enabled them to control personally the selection or de-selection of relevant hits. Had they used search techniques, the reduction of hits would have been done by an impersonal machine process and there would have been a loss of this sense of personal control. This sense of personal control justified the time invested in an extended browse, rather than in a quicker machine-executed search.

Thus, if our research students use browsing techniques to increase recall, why should end users in other studies not browse effectively to reduce recall? It would have been interesting to have asked end user searchers in the NYU case study what they did do with their overly long sets of results. Did they in fact browse them to extract a core of relevant references? The search logs themselves would not have shown this process of browsing to increase precision, the author simply assumed that the search log contained a complete task analysis. Because Dyckman did not perform such interviews, the assumption that comprehensive recall alone was desired may have been correct. But evaluative user feedback would have clarified the matter further, thus underlining its important role.

Both the NYU case study search log analysis and the GAELS project's investigation of actual user information retrieval activity show the same pattern. Users are taught search skills and then fail to search 'properly'. The teacher may then assume that the syllabus of information retrieval skills has been taught badly. However, we suggest that this is probably a misunderstanding of the evaluative feedback. Rather, it is likely that the syllabus has been taught well, but that the syllabus is not correct, or not wide enough. Users are only being taught about searching, but will spontaneously evolve browsing strategies to deal with problems of recall and precision. If this interpretation of the evidence is correct, then the tutor should expand their information retrieval syllabus to give browsing strategies equal status with searching strategies, and should give advice on solving retrieval problems with both techniques. If the findings of the GAELS project are representative, user education practice does not accept such strategies as integral parts of the user education syllabus, but should do so.

If our investigation of user education practice is correct, then British user education practice in fact differs from important statements of library user education theory such as the Model Statement of Objectives in library user education (17) by the Association of College and Research Libraries of the United States. The ACRL outlines as a terminal objective of bibliographic instruction that the 'user understands how...to identify useful information' by knowing 'when it is appropriate to search for information' but also 'understands the importance of browsing'. Our argument that browsing deserves its rightful place in the user education syllabus is thus only a statement of accepted library educational theory.

This suggestion that librarian teachers accept users' task behaviour and consider re-designing their systems (in this case, their user education syllabus) to accommodate such behaviour is an implementation of task analysis ideas for interface design. As Lewis and Rieman say, 'Once you have some people to talk with, develop CONCRETE, DETAILED EXAMPLES of tasks they perform or want to perform that your system should support'. When you find that people perform different tasks from what you originally thought, do not be 'shocked' (Dyckman's phrase) at your discovery. Rather, congratulate yourself and re-design your product to accommodate your improved knowledge of the user.

Ironically, our suggestion for syllabus re-design could also be construed as adding a third argument to the anti-user education school's repertoire of objections. Not only should librarians stop teaching and improve the usability of their libraries, leaving lecturers to teach information skills in the context of coursework. They should also stop teaching inappropriately complex search techniques to users, and should leave them to augment basic search skills with other browsing skills which can be evolved more effectively in the context of departmental teaching and research. For example, the highly information literate technique of reference chaining from colleagues' bibliographies can only take place within the context of departmental group research.

However, this would be a misinterpretation of our argument. We are merely offering a third way of improving the quality of free-standing information skills teaching courses, which we believe remain necessary (perhaps even a necessary evil) in spite of all counter arguments for the foreseeable future.

Interim conclusions:

The GAELS project was set up to encourage electronic information retrieval in place of hard copy journal browsing. Our initial findings indicate that one way to achieve this culture change may be to stop trying to impose electronic searching techniques onto an established pattern of hard copy browsing. Rather, the virtual information retrieval syllabus should look to methods of electronic browsing, and use these as the desired model of new information retrieval activity among researchers.

In practical terms of syllabus content, this has some simple consequences. Many database services offer a 'save search strategy' option, which allows a one-off literature search to become a stored search to be regularly repeated as a current awareness tool. Rather than recommend this as the first option for a current awareness strategy, a teacher librarian may more successfully promote electronic alternatives to hard copy browsing by recommending electronic table of contents browsing through publishers' own web sites. Ultimately, a researcher may feel happier title browsing through such tables of contents of their favourite journals. This is especially true if the sites offer electronic full-text retrieval.

Even without electronic full-text retrieval, there are various other reasons for such a user preference. A publisher's WWW site often has graphic design features reproducing the look of the hard copy equivalent. These features make it possible to keep some of the feel and familiarity of the library browsing experience in an electronic environment. Users can bring such tables of contents together in a cross-publisher set of WWW pages, creating their own researcher-specific WWW environment which is fully under their own control. In the absence of immediately readable electronic full-text, the option of using hard copy commercial document delivery from scholarly publisher services in preference to local library document delivery services may also prove attractive. Table of contents browsing can be more thorough than saving and re-running a bibliographic database search, given that much database journal article indexing is selective. Part of the user preference for browsing to reduce recall is item by item control of the article de-selection process, a control that users are reluctant to cede to a machine process which hides the imposition of unknown factors such as selective indexing. If they browse, they won't miss anything, if they search, they might.

This model of electronic browsing does not use librarians' own preferred services, such as bibliographic databases services purchased with library funds which demand good search skills, and which are normally promoted in tandem with a local library inter-library loan or document delivery service. However, this model may better resemble real hard copy browsing activity and thus may be a model that users actually prefer. Feedback from researchers to the GAELS project does indeed indicate a valuable role for publisher's WWW sites rather than library bibliographic databases in current awareness. We have also found an interest in commercial document delivery where there is a perception that such services offer advantages over the local library document delivery service, above all in terms of delivery speed. These are further examples of evaluative user feedback informing the task analysis process, in order to create a better user education syllabus and thus better forms of user service.

(2) Build other methods into literature searching syllabus

One way forward for the GAELS project courseware is to build this model into the CAL package and test the acceptability of the approach with users. If an increased emphasis on electronic browsing should be introduced into the current awareness sections of the package, then similarly the sections on general search skills may well need re-balancing in favour of other search strategies. The traditional approach is to teach basic search skills and then introduce advanced search skills to raise or lower recall in an initial set of hits. Users might respond better if asked to adjust recall with a variety of strategies. The users themselves could be invited to try a variety of techniques, but left to judge for themselves which strategy they find more successful.

(3) Emphasise objects retrieved more than techniques of retrieval

Beyond this, further developments to the syllabus also remain possible. Although user feedback has suggested extending and varying the task content of the first GAELS Module, users have expressed satisfaction with the content of the Modules on Information Tools and Information Types, implying that they unambiguously need help with knowing where to look for information and advice on what they find when they look there. The sequence in which the GAELS Modules are presented (which reflects existing local teacher librarian preferences) may thus be inappropriate. A re-working of this presentation sequence in favour of the types and tools of information retrieval might further help in countering an over-emphasis on how to do information retrieval in certain prescribed ways. Again, more evaluative user feedback would show us whether this alteration was an enhancement or not.

(4) Use Library WWW pages as task-based interface to courseware

Perhaps most importantly, a significant development of the task-centred focus of the courseware would be to provide a task-oriented interface to the package by plugging appropriate collections of units into local library WWW pages. These collections would be each be equivalent to 'real, complete, representative tasks'. This would necessitate a prior re-working of the local library interface itself to display a significant degree of task-orientation. But in this way the linear narrative structure of the original package interface could co-exist with a more usable interface, thus satisfying teacher librarian need for a classroom or workshop teaching tool at the same time as meeting user need for training in a particular skill at the point of execution of a task. Thus, a researcher needing to complete a literature search would go to their local library WWW pages and on a top level menu find an option flagged 'Do a literature search'. If they lacked the skills to do this, they would have the option of completing the relevant element of the GAELS courseware at this point.

(5) Use departmental courseware as task-based interface to courseware

Taking this a step further, a more complete integration of learning materials into task execution would be possible by using WWW-based departmental teaching and research training materials as, in a sense, the interface to our library courseware. Thus, at particularly appropriate points in a piece of departmental courseware where an information retrieval activity was required, the local library's information resource links and the courseware supporting the use of these links could be invoked. If such a scenario was carried to a conclusion, user education could theoretically cease to exist on a separate free-standing basis, being completely integrated into the teaching of a department.

If it were possible to use departmental courseware as the preferred task-centred context for library courseware, we would have moved full circle, back to the original aims of Herrington and similar writers: library interface transparency and the reintegration of information skills acquisition into the context of users' mainstream coursework or research. If the final GAELS model emerges as one in which library courseware is taken into the texture of academic teaching and research, via a process of library interface evaluation which first embeds that courseware in the texture of the electronic library service front end, then the primacy of free-standing library information skills user education will have been challenged.

However, this goal will not be achieved without starting from existing, successful user education materials which can then be continuously developed to their full potential. Even then, the absence of significant provision of WWW-based departmental research and learning environments means that the ideal, contextual framework within which to propagate networked information skills education has not yet been created. Library user education providers will have to wait on such environments developing, and work proactively with academic departments to promote them.

In the meantime, the GAELS Project will explore these hypotheses further and examine whether they can create a more effective tool for the ultimate project end, a joint electronic information environment in engineering that is found acceptable and usable by researchers at both collaborating institutions.

Fig. 1: Types, elements and processes: a WWW page from Strathclyde University Library

Fig. 2: GAELS courseware structure diagram

References

1 Barton, J., Ashworth, S. and Joint, N. 'The GAELS Information Audit.' In: 3rd Northumbria International Conference on Performance Measurement in Libraries and Information Services, 1999, [in press].

2 Pacey, P. 'Teaching user education, learning information skills; or, towards the self-explanatory library.' The New Review of Academic Librarianship 1995, pp. 95-103.

3 Savenije, B. 'Integrating library instruction into student education.' Education libraries journal 42(1), 1999, pp. 5-10.

4 Herrington, V.J. 'Way beyond BI: a look to the future (the effect of technology on bibliographic instruction).' Journal of Academic Librarianship 24(5), Sept. 1998, pp. 381-386.

5 Rutkowski, C. 'An introduction to the human applications standard computer interface pt. 1.' Byte 7(10), 1982, pp. 291-310.

6 Shneiderman, B. Designing the user interface: strategies for effective Human-Computer Interaction. 3rd edition (USA: Addison Wesley Longman Inc., 1998).

7 Snavely, J. and Cooper, N. 'The information literacy debate.' Journal of Academic Librarianship 23(1), 1997, pp. 9-14.

8 Bober, C. , Poulin, S., and Vileno, L. 'Evaluating library instruction in academic libraries: a critical review of the literature 1980-1993,' The Reference Librarian 51/51, 1995, pp. 53-71.

9 Lewis, C., and Rieman, J. 'Getting to know users and their tasks.' In: Task-centred user interface design: a practical introduction. Ch. 2 (self-published over the internet: 1993). In: Baecker R.M., Grudin J., Buxton W A S, and Greenberg S., Readings in Human-computer Interaction: toward the year 2000. 2nd edition (San Fransisco: Morgan Kaufmann, 1995).

10 Strathclyde University Library 'About the Library' WWW pages http://www.lib.strath.ac.uk/

11 EDUCATE Project's Into Info courseware: http://educate.lib.chalmers.se/demopath.html.

12 GAELS courseware: http://gaels.lib.gla.ac.uk/info_skills/.

13 Kemp, R. 'Training materials for advanced information skills in engineering and architecture: a web-based approach to advanced information skills.' In: Building on the past: Proceedings of the 12th ARCLIB International Conference, University of Strathclyde and Glasgow School of Art,1999, pp. 10-16.

14 Bates, M. J. 'The design of browsing and berrypicking techniques for the online search interface.' Online Review 1989, 13(5), pp. 407-424.

15 Marchionini, G. Information Seeking in Electronic Environments. (Cambridge: Cambridge University Press, 1995).

16 Dyckman, L. M. 'Beyond "First you push this button, then...": a process-oriented approach to teaching searching skills.' The Reference Librarian 1995, 51/52, pp. 249-265.

17 Association of College & Research Libraries. 'Model Statement of Objectives for academic bibliographic instruction: Draft revision.' College & Research Libraries News 1987, 48(5), pp. 256-261.