AARL

Academic Library Website Design Principles: Development of a Checklist

Roslyn Raward

Abstract

The library profession is now heavily involved in providing access to information through library websites and it is a challenge to design a website that has reliable content and a user interface that is easy and intuitive to those who use it. This paper looks at the development of a checklist for best practice design principles for academic library websites using a Usability Index Checklist (UIC) developed from principles derived from the Human-Computer Interface (HCI) literature. As user acceptance and usability are major issues in the design of library websites this paper suggests that the design will be most successful when a user-centred design model is included in the development and implementation of academic library web pages.

Identifying and implementing best practice design principles in World Wide Web pages has been and continues to be a difficult task for web developers. As Still says, 'While anyone can create a web page just as anyone can write a poem or build a house, not everyone can do it well', and Abels, White and Hahn make a comment that is as valid today as it was in 1997:

Both resource providers and users face the problem of determining the hallmarks of a well-designed, useful resource on the Web. The Web's newness and the richness of its media capabilities are both stimulating and daunting and Website designers continue to struggle with the question of how to meld content and form effectively to create useful information resources.

A plethora of guidelines, suggestions and methods are on offer throughout the web, in books and in journal articles to assist in the design of easy to use, visually pleasing, contextually rich websites. Similarly, a number of methods have been offered to evaluate web page designs either during or after their design and implementation. However, few if any guidelines or evaluative methods have been established to assist specifically in the design of academic library websites.

Clyde says:

The best preparation for creating a home page is for library staff to spend some time 'surfing' the home pages of similar libraries or organisations. This activity will give staff an indication of what works and what does not, what is interesting to the online visitor and what is not, and what features make a home page easy and pleasant to use.

Broad comments such as this have their place, but do little to guide the creators of library websites.

A major consideration in the design of academic library websites is the multiplicity of users and information needs that exist in the university environment. A library website requires an interface that can accommodate the different needs, scholarly disciplines and capabilities of the many and varied users within its institution. The primary users within the university environment are not only academic staff, but also library and general staff and students. The secondary user population comes from other academic institutions both nationally and internationally as well as from the wider population. There can be conflict, too, amongst individual library staff and library sections as to the purpose of their library's website. Some librarians look upon the web, as a gateway to additional resources on a variety of subject areas. Others think of the web as a communications tool for their primary users and yet others see that its primary purpose is to provide information about the innovative services, collections, and projects going on in the library. The design of web pages for each of these uses is different. website design is also influenced by the parent institution, that is, the academic library website does not stand alone but is usually designed or restricted by the design conventions of the parent university's site.

The aim of this paper is briefly to present new research and methodology being completed as part of a master's thesis, which is as yet unpublished. References to and discussion of 'international best practice guidelines' and 'established usability principles', referred to in this paper, require lengthy discussion and subsequently have not been defined here but will be discussed fully in the final thesis. Similarly definitions for the modified library usability checklist questions referred to in this paper will also be provided in the final report.

This particular research deals with the choice of a Usability Index Checklist (UIC) as a suggested method for evaluating website designs. It discusses why this method has been selected and proposes a modified usability checklist designed in order to assist librarians in the design and modification of academic library websites.

Background for this Study

In an unpublished study conducted in 1997-98, three focus groups were conducted with 56 academic physicists and chemists from two major Australian universities, to establish their use of the web and in particular the use they made of their institution's library website. At that time the two academic institutions used in the sample had first generation websites, that is, they had just launched websites for the first time.

Some major themes were identified from these focus group discussions and three critical observations were made to indicate at that time that

• limited use was made of library websites
• lifestyle research habits were changing; and
• digital library services needed to cater for these changing needs.

These results, correlating with the user-centred design literature, suggested that the user should be involved more fully in the design and development process of library websites.

The following conclusions were drawn from the researcher's experience after being involved in the design and implementation of an Australian academic library website, and also from the themes identified from the Focus groups:

• Librarians have few tested principles to refer to when designing library websites
• Librarians are often self taught when it comes to web design
• Web design is usually additional to the librarian's usual job
• Web design is often driven by management not the user
• Best practice guidelines for website design need to be established that relate to academic user needs; and
• The design of a simple evaluation tool is needed to assist academic librarians in developing user-centred websites.

This preliminary study set the groundwork for more comprehensive research into the guidelines or principles that were being used in the design of second or third-wave academic library websites.

Design Guidelines

Spool and Nielsen are leading authorities on effective website design and provide extensive practical advice and hints and there are many other authors, such as Rowlands, Constantine, and Morville, who present viewpoints on good practice design. McClements and Becker also suggest useful hints and design features that must be included if a page is to be effective. For example, links connecting all site pages to the site's home page, text identification of the graphics used on each page, a short homepage finishing 'above the fold', limited use of graphics, clear navigation, scannability, and functional consistency.

It has been suggested that there are three distinct, but related, bodies of literature that summarise the larger body of literature that has appeared on effective website design, which may provide effective guidelines for web design. 'While each of these individual bodies may be burdened by some type of myopia, collectively they provide effective guidelines for website designers.'

The first approach, the 'ad hoc approach', is most commonly based on the designer's viewpoint and personal experience. Although there is much in the literature that emphasises the importance of knowing the site's audience and information needs, this approach shows little evidence of actual consultation with users. Abels, White and Hahn11 concluded that 'the extent to which designers' preferences and/or perceptions of well-designed websites match those of users, needed testing and should not be assumed.'

The second approach that they suggest focuses more on the evaluation and not on the design of web resources. By advocating the application of certain criteria for evaluative purposes, this approach indirectly suggests that these criteria should be considered in the design process. Although some of these criteria are buttressed by research, others simply reflect the professional judgment of information specialists who regularly mediate between information systems and users.

However, while there is so much available in the literature to guide good design, many designers are still struggling to come to grips with usability. Comments such as 'You should seek to design pages that are simple, yet elegant and convey information in an easy, user-friendly environment' do little to assist the designer. Stover and Zink13 elicited criteria from various design publications, checked higher education pages against these criteria and then interpreted the relatively low scores of the sites to indicate general poor design of higher education web pages.

Spool, presenting his findings from usability tests with groups of users, comments:

We started wondering what made a Website usable.

We had heard the opinions of experienced designers about what they felt it took to create a good site.

We looked at books and magazines that talked about how to make a 'cool' site. But no matter where we looked, we couldn't find any data - based on real user experience - about what it takes to make a usable site.

The third approach suggested by Abels, White and Hahn includes principles developed from human factors research in Human-Computer Interface (HCI) design and emphasises the importance of consultation with the user. The HCI design literature is research-based, and strong analogies can be drawn between the design of computer interfaces, especially those associated with hypertext systems, and the design of websites. This research does consult users as part of the design process, relying on usability testing and incorporating the users into testing specific design features, usually in connection with particular tasks. This approach is both more rigorous and conceptually grounded, in contrast to the 'ad hoc' approach seen in Stover and Zink.

Why a User-Centred Design?

Neilsen maintains that usability rules the web.

With about 10 million sites on the Web in January 2000 (and about 25 million by the end of the year and a hundred million by 2002), users have more choices than ever. Why should they waste their time on anything that is confusing, slow or that doesn't satisfy their needs?

Why indeed?

As a result of this overwhelming choice and the ease of going elsewhere, Web users exhibit a remarkable impatience and insistence on instant gratification. If they can't figure out how to use a Website in a minute or so, they conclude that it won't be worth their time. And they leave.

James, speaking of digital libraries, says that 'precious little attention, if any, has been paid to the users of such systems, what they want, or the extent to which these new libraries will deliver the goods.' Anderson argued that 'building digital libraries that work will require that systems design, development and deployment be grounded in use and done collaboratively with users'. It is important to consider the steps needed to move from the analysis of what is, to the design of something new, taking time for iterations and the testing of ideas before the architecture becomes fixed.

The Research Problem being Explored

In order to discover to what degree Australian academic library websites meet established international best practice guidelines for website design and to see to what degree they support or violate established usability principles, a number of usability inspection or engineering evaluation methods were explored.

Heuristic evaluation, a 'discount usability engineering' method was originally designed by Neilsen for quick, cheap, and easy evaluation of a user interface design. Heuristic evaluation, 'a method for structuring the evaluation of a system using a set of simple and general rules', is the most popular of the usability inspection methods and is a good method for finding both major and minor problems in a user interface. Nielsen provides information on how to conduct a heuristic evaluation and provides a list of ten recommended heuristics for usable interface design.

The instrument adopted for this research was a modified heuristic usability checklist, based on Keevil's Web Usability Index Checklist, and was designed specifically to assist academic library website designers in the creation and development of their websites. Unlike other heuristic evaluation methods it was designed to be administered by a single evaluator. Keevil developed his Web Usability Index Checklist on the models provided by Ravden and Chignell. The Ravden Human-computer Interface checklist included about 120 detailed questions in a number of categories, while the Chignell model included over 180 questions. The Keevil model included a series of questions that could be answered with a Yes or No answer. It provided 'a more consistent question and answer method' using only one evaluator, who would test the website a number of times in its developmental life cycle.

An in-progress checklist was established based on the Keevil Web Usability Index Checklist and a pilot evaluation of this in-progress checklist was then carried out on a small sample of five Australian academic library websites by five groups of students. The checklist was also examined by three experts who had either international or national experience in web design and development A further pilot evaluation was conducted by a group of research students before the final modified version of the checklist was completed. Only those questions that were specifically relevant to Australian academic library websites were selected from the Keevil usability checklist and some further questions, adapted from guidelines in using the Internet for users with disabilities, were added.

Using a checklist based on guidelines is a valuable way to identify usability problems in websites. Information developers can use the checklist to measure how easy it is to find, understand, and use information displayed on a website. The advantages of using a checklist to measure the usability index of a website are:

• a checklist is a structured method of applying usability research to improve your website
• the checklist can assist the design of a new website or improve an existing site
• the checklist is inexpensive and easy to implement
• the checklist can be administered by one evaluator; and
• a checklist is a living document and can be updated to reflect the latest advances in usability and technical communication techniques.

Nielson warned that in order to ensure an unbiased assessment of a website it is important to have several evaluators conducting independent evaluations. His experience indicated that three to five evaluators could usually identify about 75% of the usability problems of a particular design. However, Keevil's checklist was designed to simplify the process by using one evaluator only. He suggested that the advantages outweigh the disadvantages and that for those with 'limited budget' and 'lack of time' using checklists based on this method is an effective and simple way to measure the usability index of a website.

One disadvantage of using one evaluator is that bias could be introduced into the interpretation of the questions used in the checklist and the answers given. Also the questions that have been included require a varied level of expertise in answering. A novice user of the site could handle some but others could require some expert knowledge of one form or another.

• using questions that required a Yes or No answer, thus reducing the wide range of subjective values that can be introduced when using a scoring system
• by conducting pilot evaluations; and
• by having the questions checked by experts.

For example, in the original Checklist by Keevil the questions introduce some bias. For example, the question, 'Are resources reliable, appropriate?' requires the evaluator to make a value judgement on the reliability and appropriateness of the resources. The question can be asked, 'What resources?' and 'By what criteria are they judged to be reliable or appropriate?' Keevil does not provide definitions for the questions that he has included in his checklist, therefore in the final published report on the modified academic library website checklist (Appendix 1), definitions will be applied to all questions to provide clarification for the evaluator.

The usability checklist developed specifically for academic library website designers to maintain best practice design principles contains 100 questions. These questions are divided into four main sections:

• finding the information
• understanding the information
• supporting user tasks; and
• presenting the information.

The first section focuses on the ability to find information by establishing whether site maps, indexes, contents lists or search tools are included on the site. It further seeks to establish the quality of the information by identifying the currency, reliability, authority and appropriateness of the content that has been included. Questions are also asked about the author responsibility of the page and contact details, as these details are an important element of website design.

The second section focuses on the organisation of the content of the site, that is, the structure of the lists that have been included, the intuitiveness and usability of headings, the scannability of the page, the conciseness and clarity of the text and the consistency of the terminology used.

In the third section the presence or absence of 'user tasks' is established by asking questions relating to forms, help screens, FAQs and comments pages.

In the fourth and final section the presentation of the information is evaluated. Download times, file and graphic sizes, cross platform compatibility, the use of tables and/or frames, image maps, metadata and consistent HTML formats are examined. Questions are included to examine the workability of links, the presence and clearness of navigation tools, visual appeal, the use of colours, graphics and the effective use of white space.

Currently testing is being conducted on this modified usability checklist by using a random sample of 50% of Australian academic library websites, and its degree of success will be reported on when the results are known.

Conclusion

As academic librarians are increasingly responsible for the design and maintenance of their institution's web page, there is a need for an evaluative tool that will assist them to ensure that best practice principles and guidelines are followed during the design process. This paper suggests that the use of a Human-Computer Interface usability checklist, such as the one designed by Keevil, if used during the design process of a website will ensure that best design practice principles are supported. To further assist academic library website designers a modified version of the Keevil checklist, that has been designed specifically to address the issues involved in an academic library website, has been suggested as the tool best suited to ensure user-centered best practice principles.

Notes

J Still Designing Outstanding Library Web Pages In IOLS97: Proceedings of the 12th Integrated Online Library Systems Meeting New York 14-15 May 1997 pp127

E G Abels, M Domas White and K Hahn Identifying User-Based Criteria for Web PagesInternet Research: Electronic Networking Applications and Policy vol 7 1997 p127

L A Clyde The Library as Information Provider: The Home Page The Electronic Library vol 14 no 6 1996 pp549-558

J Spool Web Site Usability: The Big Picture 1998 Available
http://webreview.com/pub/web98east/23/spoolx.html [5 October 2000]

Jacob Nielsen How to Conduct a Heuristic Evaluation Available
http://www.useit.com/papers/heuristic/heuristic_evaluation.html

[19 September 2000]

C Rowlands What is Good Design? 2000 Available
http://webreview.com/pub/2000/03/10/feature/index3c.html

[5 October 2000]

L Constantine Usage-Centered Web Design 1999 Available
http://webreview.com/pub/1999/06/25/feature/constantine.html

[5 October 2000]

P Morville Information, Architecture, and Usability 1999 Available
http://webreview.com/pub/1999/03/12/arch/index.html [5 October 2000]

N McClements and C Becker Writing Web Page Standards College & Research Libraries News vol 57 no 1 1996 pp16-17

E G Abels, M Domas White and K Hahn op.c it.

Abels, loc.cit.

Abels, loc.cit.

M Stover and S D Zink World Wide Web Home Page Design: Patterns and Anomalies of Higher Education Library Home PagesReference Services Review vol 24 no 3 1996 pp7-20

J Spool op.cit.

Abels, loc.cit.

K Instone HCI and the Web: A CHI 96 Workshop SIGCHI Bulletin vol 28 no 4 1996 pp42-5; J Nielsen and D Sano

SunWeb: User Interface Design for Sun Microsystems Internal Web Paper presented at the Second World Wide Web Conference

94 Microsoft Improving Web Site Usability and Appeal Available
http://msdn.microsoft.com/workshop/management/planning/improvingsiteusa.asp

[5 October 2000]; B Shneiderman Designing Information-abundant Websites Available
http://www.cs.umd.edu/projects/hcil//Research/tech-report-list.html#1996

[5 October 2000]; S B Shum The Missing Link: Hypermedia Usability Research and the Web SIGCHI Bulletin vol 28 no 4 1996 pp68-75

Stover, loc.cit.

Nielsen,loc.cit.

J James Migrating Foundational Study Approaches to the Virtual EnvironmentIn Session 1: How We Do User-Centered Design and Evaluation of Digital Libraries: A Methodological Forum Available
http://edfu.lis.uiuc.edu/allerton/95/s1/janes.html [21 October 2000]

W L Anderson Socially Grounded Engineering for Digital libraries In Session 2: How We Do User-Centered Design and Evaluation of Digital Libraries: A Methodological Forum Available
http://edfu.lis.uiuc.edu/allerton/95/s2/anderson.html [20 October 2000]

J Nielsen HCI: Using Discount Usability Engineering to Penetrate the Intimidation Barrier Available
http://www.dcc.unicamp.br/~hans/mc750/uidesign/guerHCI/guerrilla_hci.html

[5 October 2000]

B Keevil, op.cit. p276

J Neilsen Ten Usability Heuristics Available
http://www.useit.com/pa pers/heuristic/heuristic_list.html [19 September 2000]

B Keevil Measuring the Usability of Your Web Site: Usability Index Checklist for Web Sites Available
http://www3.sympatico.ca/bkeevil/sigdoc98/checklist/WebCheck_Sep13.html [23 February 2000]

S J Ravden and G I Johnson Evaluating the Usability of Human Computer Interfaces: A Practical Method West Sussex Ellis Horwood 1989

M H Chignell and B E Keevil Developing Usable Online Information for a Web Authoring Tool Presented at the annual SIGDOC96 Conference in Research Triangle Park North Carolina USA on 21 October 1996

Ibid, Measuring the Usability Index of your Web SiteACM SIGDOC 1998: Conference Proceedings 98: the 16th Annual International Conference on Computer Documentation Scaling the Heights: The Future of Information Technology 23-26 September 1998 Quebec City pp271-277

J Nielsen and D Sano SunWeb op.cit.

B Keevil op.cit.

Ibid

E-mail to author from B Keevil 28 August 2000 Subject Re: definitions for usability checklist

Appendix 1

Usability Index Checklist for Academic Library Web Sites

Usability Index

• What is the name of the web site?
• What is the location of the web site?
• What is the main purpose of the site?
• When was the usability index measured?

Checklist questions