Database Interface Design – washing powder optional

As a librarian supporting students and academic staff over the last 20 years, I have substantial experience of working with digital content and currently work with around 30 electronic databases. My ability to work productively and assist students in gathering information or data is influenced significantly by interface design – a critical component of the user experience (‘UX’) of these databases.

The following represent my personal views on interface design.

 

Why is interface design important?

The ability of a user to work well with an interface influences their productivity. Design changes to the front end of any database system are comparable to the challenges faced with a new computer operating system. For example, changing from: Windows 3.1 to 95 to 98 to 2000 to XP to 7 to 8 to 10 – meaning the familiar is no longer so, resulting in one’s productivity going down, at least initially.

 

Changing search interface / results display

I am not advocating that interface design is set in stone and appreciate that carefully considered changes can generate substantial improvements. What is important is that those changes are driven by user needs and then tested to ensure they have met those needs. Otherwise they can cause major difficulties for students. The Passport database is such an example, where a simple and intuitive process appears to have been replaced in such a way that it is far less clear and valuable functionality has disappeared altogether. Previously, when selecting the ‘Menu Search’ option, a three-step process was revealed:

Topics →  Geography →  Results

A helpful feature was the ‘predefined selections’. Here, many country groupings (e.g. EU, G20) are clearly visible and easy to select.

The ‘Results List’ screen displayed a variety of categories available to select (e.g. Industry Overview, Local Company Profile).

 

The look and feel of the interface has now changed, in such a way as to reduce the effectiveness of the interface. I find it very confusing, with the results screen not showing the options of the old version.

I’m not alone, if the comments on the Business Librarians Association (BLA) bulletins are a guide. So if I and my professional colleagues are having difficulty, students must be in an even worse position.

It is hard to see these changes as being driven, as they should be, by attention to the user and how they work. Instead they feel to me more like change for the sake of change. Just as washing power might be branded ‘New and Improved!’ so these changes  to the look and functionality of a database must be a step forward over what went before. Mustn’t they? Well, no.

 

The user experience (UX)

The Japanese concept of Kaizen, which translates roughly as ‘continuous improvement’ is pertinent here.

 

This principle runs through effective UX developments. With the emergence of ‘the perpetual beta‘ and web 2.0, users can be seen as co-developers, through an ability to feed into service changes.  By monitoring the take up (or not) of new features, service providers can reflect user needs.

Such an approach drives a user-centred evolution of a system interface, introducing incremental and necessary improvements over what has gone before. This is far preferable to a revolutionary approach, resulting in drastic changes to the interface design which risk causing significant disruption to its users.

This reflects the position taken by Joe Hardenbrook, that a state of permanent beta is OK, when feedback is received and adapted to, enabling improvements to flow from this.

For example, in undertaking a review process, the provider of the Factiva database has reassuringly put the user experience at the heart of the exercise. This includes undertaking interviews, with the goal of improving productivity, through ‘compact, clear and consistent design’.

 

Why do effective design features matter?

Stability

Does the database have a look which persists over time? Some good examples, in my opinion, include Thomson Research and Factiva. Both the search interfaces for these databases have remained similar for over five years. The search screen is a single page, with no scrolling down required and all options clearly visible. These can be selected by typing in a free text box or using a ‘look-up’ option from a magnifying glass icon. Other selection methods are equally clear, from clicking to select options (e.g. Report Type: [Company, Industry] in Thomson Research) or using a drop down menu (Date in Factiva [e.g. In the last week]). In Factiva, search options flow from top to bottom, in a single column, on the left, culminating in a blue colour-coded ‘search’ button, at the lower right of screen. Thomson Research has two columns of search options and again a blue colour-coded ‘search’ button, bottom left of screen.

Familiarity contributes to ease of use. If a database has reached a clear and effective design, there is value in maintaining this appearance. Why is this important? Because students (and library staff) have many demands on their time. When faced with a radically altered search interface, time is wasted trying to get familiar with the new ‘look’. So I would urge our suppliers to ensure such changes respond to user needs and behaviours rather than simply to refresh the brand so that it looks ‘new and improved’.

Some suppliers are seeking insights through user engagement, to improve their products. For example, EBSCO Information Services seeks insight into today’s student as an ‘information seeker’. Kate Lawrence, Vice President for User Experience Research, noted students have an ever increasing desire for efficiency, approaching their studies by looking at priority (which deadline is first) and looking at the return on investment in terms of getting the best results for the amount of time invested. [1] Being aware of such requirements allows a provider to meet these user needs.

 

Clear search options

By providing options clearly visible on the search screen, alternatives can be selected without confusion. A prime example of this is the Factiva database (Business News). Using the ‘Search – Search Builder’ option from the drop down menu, many alternatives become apparent on the left of the screen. For example, Date, Source, Subject – which can be easily explored and used as part of a search. Should it need to be refined, after viewing the results, this can be achieved by simply clicking on the ‘Modify Search’ button to retrieve the original search options.

I sought comments from a colleague who works in a different area of the library, relating to database design. He noted that a lack of familiarity makes usage more troublesome – in his case, identifying financial databases, which I often work with. Further, the examples of Compendex, IEEE Xplore, Scopus and Web of Science were noted as being ‘easy to use’, having good basic and advanced search options. I think this again reflects a point made earlier – to be able to proceed in a search, without confusion.

 

Enhancements

‘Evolution rather than revolution’ can be translated to mean small improvements, which  leave the basic search design unchanged. Again, in Factiva (in the past), to indicate a phrase search in the text box at the top of the screen, two or more words had to be included within quotation marks. For example, “Tesla Model S”, which searches for these terms in this exact format, rather than as keywords scattered throughout the documents being searched. Now, however, this is done automatically. If you attempt to enter quotation marks, a message appears on screen noting: ‘Please check your query, we have identified unbalanced quotes’. This is a good example of an enhancement, which  maintains the basic search interface.

 

 

Additional challenges faced

Name changes

This is a hindrance, in that students are no longer aware of a particular title and the location, when accessing from an A to Z listing. Some examples include:

Global Market Information Database  →  GMID  →  Passport.

Global Insight  →  IHS Connect.

Removing content / features

KeyNote:  this market research database previously showed a comprehensive listing of sections for an individual report on the left hand menu (e.g. Market Size, Competitor Analysis) which I found really helpful. This no longer appears, which I believe is a backward step.

Mintel:  by splitting up and scattering the constituent sections of a market research report on a results page, rather than offering a PDF link to display as a single report (available for KeyNote reports) it is far less user-friendly. I have witnessed students and library staff finding the layout  confusing.

 

Summary

I appreciate excellent database design when faced with a substantial enquiry and a tight deadline – reflected in stability and a clear, intuitive search interface.

Incremental improvements to a database are preferable to radical changes, when seeking to support students and academic staff in their research. This process can be enhanced through testing and should be driven by user needs, such as the Factiva review noted above.

In conclusion, I can do without the washing powder mentality – ‘new and improved’ – as it isn’t always so.

[1] Lawlor, B.  An Overview of the NFAIS 2015 Annual Conference: Anticipating Demand: The User Experience as Driver. Information Services & Use. 2015, v.35(2), p.15.