UX

Mar 21 2013

Real world user-centred design (UCD): case study from the European Bioinformatics Institute

Today’s news is that the latest article I have co-authored has just been published in BMC Bioinformatics (e-published before final formatting):

de Matos P, Cham JA, et al. (2013) Enzyme Portal: a case study in applying user-centred design methods in bioinformatics. BMC Bioinformatics 14: 103

It’s the first ‘how to’ guide for applying user-centred design (UCD) to websites for bioinformatics. Hurrah!

So what is the article about?

I believe that as a community, bioinformatics is notorious for producing hard-to-use, often technical, software that turns off the wet-lab biologist.   There are lots of reasons for this, including the way projects are funded, and the fact that software developers decide how the software looks and behaves, amongst other things.  Bottom line is – software isn’t generally designed based on the needs of users (who in this case are usually academic researchers in bio-sciences).   Our article aims to address this situation by suggesting how we can use UCD to improve the usability of bioinformatics resources.

What does it add to the state of the art?

As you will know, UCD is an approach for designing user interfaces/ websites, where the needs of users dictate the design.  The aim is to create services that are both useful and easy to use for the specific target audience. It has been successfully used in e-commerce websites, retail sites, gaming, mobile apps and other domains, however, case studies about how to increase usability of scientific software are scarce in the literature, and virtually non-existent for complex data domains, such as bioinformatics.  So, guess what – we wrote something about it – taking Enzyme Portal as a real life example.

What exciting stuff will you find in our article?

You will get (for free!):

  • step-by-step account of the UCD process we used; see Figure 1 for a heads-up on this and Additional file 1 for the full algorithm ;-)   Enjoy!  (Remember each UCD process will be different depending on your project, how you can access users, the time and resources you have).
  • toolkit of templates so you can get started NOW for your own bioinformatics (or other) project.  Including: consent form, card sorting canvas (Table 1), usability testing scenarios and tasks, personas and more!
  • ammunition to tell your boss/PI that you should start doing UCD!
Persona

Cartoon from the biochemist persona ‘Eunice’ – Figure 2 in the paper

Why have we bothered?

The dream is that our article will inspire our peers working in bioinformatics, or other scientific fields, to improve the user-friendliness of their resources on the web, so that the usage and impact of them on science may grow.  But what about world peace, you say?  Honestly, I think this is a noble cause – because otherwise we are not getting the value we should be getting from publicly-funded bioinformatics (Yes that’s your taxes).

 

Anyway, we hope you enjoy the paper and get motivated to join in with your own UX stuff soon!  JC out!

 

By • Publications, Research, Uncategorized • 0 • Tags: , , ,

Mar 6 2013

The European Bioinformatics Institute has a new website…

Hello!  Yes, it’s big news this week: the EBI launched its new website on Monday morning!

EBI website Kaizen

It has taken many months to prepare the new website, and the EBI will continue rolling out improvements in the coming months to bring all pages into the new look-and-feel.  It’s such a huge website, including so many different service brands and huge amounts of data, that there will inevitably be some ‘fixing the plane whilst flying’. This is a good thing if you believe, like Lou Rosenfeld (UX guru and publisher), that ‘Redesign’ is a dirty word .  [Quote: "Every time you redesign, God kills a kitten".]

Redesigns tend to be full overhauls.  This is not always great, since they can be driven by politics, have poorly defined scope, and are destined to fall short of solving all your problems ‘once and for all’.  In contrast, continuous improvement/change (a la ‘Kaizen’ of the management world) can lead to a site that can shift to meet the organisation’s and users’ needs.  The result is happier developers and end users, for the long term- even if it means having a ‘harlequin’ website at times.

User-centred design, EBI style

I was going to write a longer post about how we applied user experience techniques to design the behaviour and visuals of the new website, but the EBI’s Associate Director, Ewan Birney has written a lot about it already, see: Ewan’s blog.  One thing to note, is how the EBI looked to the BBC’s experience, in particular the BBC global experience language (GEL), for rethinking their web offering.  We aimed to create an ‘EBI-ness’ about our pages, but still having distinct service brands within it – much like the BBC has specific brands (like Dr. Who) within the overarching BBC brand.

Take a look and let us know what you think of it!

EBI launched the new site on 4 March 2013

EBI launched the new site on 4 March 2013

 

 

 

 

 

 

 

 

By • Uncategorized • 0 • Tags: , , ,

Jul 13 2012

Bioinformatics Meets User-Centred Design: A Perspective

I’m really happy to announce that our latest article on user experience in bioinformatics has just been published in peer-reviewed open access journal, PLoS Computational Biology.

The citation is PLoS Comput Biol 8(7): e1002554. doi:10.1371/journal.pcbi.1002554

Thanks to everyone involved – especially to Katrina, who carried out the interviews with myself and Paula, and put a lot of the manuscript together.  We hope this article is just the start of a trend towards more user experience-related articles in bioinformatics and in the wider life sciences community.  I feel that the lessons learned could also apply in other complex domains, such as engineering, technology and physical sciences.

For PLoS journals you can post comments on the article.  I look forward to hearing what you think, enjoy!

By • Publications, Research, Reviews • 0 • Tags: , , ,

May 24 2012

Careers day at DKFZ Heidelberg

What do I do, and how did I get here?

This is the subject of the talk I am giving today at the careers day at DKFZ (German Cancer Research Institute) in Heidelberg, Germany.  The slides will be available on Slideshare from the organisers soon.

What does a user experience (UX) analyst do?

My way to UX…

My route, from bioinformatics doctorate to User Experience Analyst, wasn’t perhaps typical – but I think no one’s journey ever is.  Other than maybe via HCI (human-computer interaction) or design education, I’m not certain that there is a typical way to get into UX.  Employers seem to look for how well connected you are, the experience you have, and also any transferable skills.  For bioinformatics, an understanding of the science also helps – making it easier, for example, to create plausible usability testing scenarios and tasks, and workshops that work for scientific/computational audiences.

 

By • Uncategorized • 0 • Tags: , , , , ,

May 8 2012

Coming out of your shell (aka UX workshops with life scientists)

InfoQ have now put up my talk from UX Cambridge 2011.  The talk was about workshops and how they are a great way to explore ideas for a user-centered design project.

InfoQ page

The talk includes…

  • At what point would a workshop be useful?  And when would it be harmful?
  • Why have user workshops got a bad reputation with some UX practitioners?
  • Examples of the practicalities of planning and delivering successful UX workshops – the aim being to inspire you to run your own workshops right away.
  • I try to share special insights for those of you that work with ‘geeky’ scientific audiences and/or in an international environment.

Enjoy!

By • Uncategorized • 0 • Tags: , , ,

Sep 5 2011

Usability in complex domains: what’s the optimal route for UX success in bioinformatics?

UX in a complex domain such as biology

To pursue a career in usability practice for a complex domain, such as bioinformatics, what is the best career route to take?  Is it best to train in usability evaluation – more formally known as Human Computer Interactions (HCI) first, then to learn about a specific domain later?  Or is it better to be a domain expert in bioinformatics, then learn the discipline and techniques of HCI later?  Is the order important for success?  You can, of course, study for degrees and doctorates in both, but what’s the most efficient way to become an effective usability practitioner in a complex domain?

One way to think about this situation for the field of bioinformatics is to consider how difficult it is to learn the material in each discrete sphere separately first:  i.e. in HCI or bioscience.  Is it easier to catch up on bioinformatics.…or is it easier to learn usability methods and philosophies ‘crash course’ style?  An added complexity here is that bioinformatics itself suffers from the same problem: should one study biology then computing, or vice versa? (I’ll shelve this issue for now!)

It seems an easy question to answer – obviously catching up on HCI is easier, you might be thinking; short practitioner courses on usability techniques are available after all  (I have been on them myself!), whereas bioinformatics is extremely complicated to learn from scratch.  But what many do not take into account is that HCI is actually a huge discipline too, with vast amounts of research and established best practice.  As an academic field it includes computer science and engineering, psychology and cognitive science, media and design, and social science.  To understand it well, and apply to apply it correctly requires deep study and insight.  In addition to this, I believe a purveyor of usability evaluation must also offer ‘softer’ people/communication skills to operate effectively, and especially so in a complex domain where persuasion and keen listening skills will be needed.

Chilana et al.’s paper from the CHI conference in April 2010 (PDF) addressed the topic of usability in complex domains.  The authors reported summaries of the results from interviewing 21 usability ‘generalists’ (i.e. non domain experts) who have worked on projects involving usability evaluation for a complex data domain.  These domains included genomics, medical imaging, software development, network security, aviation, healthcare, statistical analysis, and others.

Analysis of complex user data such as for a new online bioinformatics service

They found that there was often a lack of credibility for generalist usability consultants within the team/client.  For usability testing, the generalists also reported feeling less confident designing, running, and presenting recommendations when working a complex domain compared to something more mainstream like an e-commerce site.  Another problem was that the expert users they must interact with tended to vary more in their responses, and consensus was hard to find in requirements for new systems/websites.

My view on the matter is that both routes can theoretically lead to success, but a domain expert first, UX analyst second is the best option where possible. I am somewhat biased, because I took this route myself, but to justify it… I believe that having a passion and solid grounding in the complex domain first alleviates the credibility and confidence issues faced by generalists.  More cynically, experts in a specific field usually know less about the specifics of the application of HCI methods compared to other environments, so you can get away with more flexible approaches, which is handy sometimes!  Moreover, if you understand the worldview of the domain expert – but are still somewhat removed from it now you are immersed in UX instead – then this can help you to:

  1. communicate better
  2. empathise better
  3. anticipate problems before they arise

In other words, I think the cliché “better the devil you know than the devil you don’t” applies here.

By • Reviews • 0 • Tags: , , , , ,

Aug 26 2011

The DrupalCon 2011 event has now come to a close

DrupalCon London 2011

It has been great to learn about Drupal at DrupalCon 2011 this week; what it is, what it is not, its history and future plans.  I now understand what it’s all about – and how much there is still to learn!

I know, for example, that the UX of Drupal has two aspects:

  1. UX for the interface used to build Drupal sites – the ‘console’/ ‘dashboard’ interface- basically where the thing is bolted together and where content is inserted.  Users of this interface are the developers, and the client i.e. once the site is handed over.
  2. UX of actual Drupal sites once online – i.e. the resulting output of point 1.  These can be varied and you can only really spot a Drupal site for sure by looking at the code.  The Drupal Museum shows how the look and feel of Drupal sites has changed over the years.

Point 1 above, appears to be the responsibility of the Drupal community (drupal.org), a not-for-profit ground-up initiative, comprising the contributors of the underlying code for Drupal.  As I see it, the main stakeholder, who would be affected by improving the UX of the ‘dashboard’, is the ‘Drupal shop’: that is, web design companies who build and sell Drupal websites/ solutions as their main business.  They will benefit, because when they deliver a flashy site, with lots of functionality and personality – they can also show their client how easy it is to add/modify content to it themselves, as and when they need to, without needing developers to do it, which is unlike other ‘hard-coded’ websites like PHP ones, for example.

Point 2 is up to the site builders themselves, and will depend on how much they choose to engage with a user-centered design process during their site’s inception and development.

Dries on usability of Drupal

The topic of usability/ ease of use was explicitly mentioned several times by different speakers.  Most notably by Dries Buytaert, the creator of Drupal, who talked about the Drupal user survey results, and his company’s (Acquia) Head of UX, Jeff Noyes, who gave a quick tour of the UX techniques he exploits.  There was also a talk about the findings of usability testing of the Drupal 7 ‘console’ carried out at the Usability Lab at the University of Minnesota; this was presented by various Drupal community folks.  Dante Murphy and Angel Brown from UX at Digitas Healthalso spoke about systems thinking for designing user experiences – and how constraints on design are a good thing to have, so you can be free to explore a specific ideas space.  I particularly liked the talk by Todd Ross Nienkerk (Four Kitchens) and Adam Snetman (Thinkso), on ‘Designing Web Systems’, which included a case study of a Drupal website created from scratch using personas, paper prototyping and wireframes.  They talked about mood boards and style tiles too, which was interesting.

Pretty much all the UX material I saw in talks at DrupalCon was not brand new stuff.  I am familiar, for instance, with how to apply most techniques covered in Jeff Noyes’ talk – from material I have seen at UX London or on Syntagm Usability courses.  However, it was probably new for this audience.  Presenters were serving up information on UX specifically for the Drupal code contributors to hear, and for the Drupal community to raise awareness and empower individuals to improve the UX of Drupal 7 and/or 8.  To hear it presented in this way was interesting and served to show me how generally (un-)familiar open source software developers are with the practice of UX.

To sum up, I would say the conference was aimed at developers (as expected), not at designers and UX folks, but I think in future Drupal.org should make a concerted effort to attract more UX people to attend and give talks to continue to raise awareness of the importance of UX.  After all, beauty and efficiency of code is one thing, but only a select few see the code, everyone else sees/has to use the end product!  For me, I think the event was a really helpful for finding out not only about the technology that I will increasingly be exposed to in my current role, but also meeting the people behind the open source phenomenon that is Drupal.

Drupal Tea Lady

On a general note:

The conference host, Fairfield Halls, did really well to cope with the huge numbers of Drupal-ers (1,750 people according to Dries) and the ubiquitous Mac laptops – providing 1,000’s of power supplies, and pretty reliable Wi-Fi for everyone too.  They even managed to serve tea in proper cups and saucers, and lunch with proper crockery – even a Sunday roast.  For the keynote talks, I liked too how they over-layed the live videos on to the PowerPoint slides, and the grungey music between speakers.  The tea lady was a legend too!  Geekiness of the event was on a par with Agile Cambridge, and as a beginner to Drupal, there were plenty of talks I could attend and follow, which was great and not what I was expecting.  So well done all round to the organisers!

I stayed at Selsdon Park Hotel for the duration of the conference – what a find!  I didn’t expect to find such a lovely hotel in Croydon.

Selsdon Park Hotel - Golf

Selsdon Park Hotel - historic building

By • Uncategorized • 0 • Tags: , , , ,