Archive - user experience RSS Feed

Designing for illiteracy – a mass market accessibility challenge

I was in Chicago the other week and out with a friend who has multiple, severe dietary allergies. She can’t eat dairy, eggs, wheat, and a bunch of other stuff. Eating out is a bit of a pain for her but, if she doesn’t get it right, a whole lot more pain later.

We were in one of those posh grocery stores with its own little cafe and, after much deliberation, decided what to eat. The girl who was taking our order had a desk that was positioned in a way that made it easy for me to look over her shoulder at the interface she was using to take the order.

Taking a standard order was pretty easy – you just pressed the button that said ‘thai chicken salad’. Simple.

Then it came time to take my friend’s order. First she had to press the button that said ‘thai chicken salad’ and then my friend asked that she make a special note for the chef about her allergies. To do this, the girl had to press the notes button and then type the special request in. No assistance from the UI whatsoever. And that’s when the trouble struck. Spelling.

Without wanting to ridicule her – she failed to spell ‘dairy’ even to the point that you might guess what she intended. There was no way she was ever going to accurately convey my friends requirements to the chef. I watched, quietly, as she tried and failed to type the instructions and eventually sent the following note through to the chef:

‘eg’

Here’s the thing. Our order taker is far from an edge case. Jonathan Kozol has written extensively about illiteracy in the US (and there are similar problems in many parts of the world). He says:

Twenty-five million American adults cannot read the poison warnings on a can of pesticide, a letter from their child’s teacher, or the front page of a daily paper. An additional 35 million read only at a level which is less than equal to the full survival needs of our society.

Together, these 60 million people represent more than one third of the entire adult population.

The largest numbers of illiterate adults are white, native-born Americans. In proportion to population, however, the figures are higher for blacks and Hispanics than for whites. Sixteen percent of white adults, 44 percent of blacks, and 56 percent of Hispanic citizens are functional or marginal illiterates. Figures for the younger generation of black adults are increasing. Forty-seven percent of all black seventeen-year olds are functionally illiterate. That figure is expected to climb to 50 percent by 1990.

Fifteen percent of recent graduates of urban high schools read at less than sixth grade level. One million teenage children between t velve and seventeen percent cannot read above the third grade level. Eighty-five percent of juveniles who come before the courts are functionally illiterate. Half the heads of households classified below the poverty line by federal standards cannot read an eighth grade book. Over one third of mothers who receive support from welfare are functionally illiterate. Of 8 million unemployed adults, 4 to 6 million lack the skills to be retrained for hi-tech jobs. (more here)

This is a big problem. This is not an edge case. And, before you say it, the answer is not icons. (The number of times people have told me that the solution to designing for an illiterate audience is icons. Now make me an icon for ‘vegan’).

I don’t have the solution, but I do have a couple of guiding thoughts.

  • People are better at recognising words than they are at making them from scratch. My 3yr old can recognise words in books that he is familiar with but he can’t read (no matter what he tells you). This is true for all of us. Recognition is far easier than recall. Think about foreign languages – most of us can read a lot more than we can speak, right?
  • Think about mission critical tasks. Things that, if not done right, could hurt people or have significant negative impact on people or business. Don’t give people a blank box to fill in when you’re designing these tasks. Give options (in words, not icons). Let people recognise and select, don’t make them remember how to spell stuff.

Jan Chipchase has done a lot of design research work with Nokia in the area of device design for illiterate end users and supports the view that making the interface easy to ‘learn’ (which largely means ‘remember’ for people who are less literate), is the best approach – better than icons or audio menus or all other apparently obvious solutions.

This presentation is worth a flip through if you’re interested in his experience and outcomes.

 

None of this is new, granted. But it’s not something I hear us talking about anywhere near enough. Watching that poor girl struggle with that interface and because of the poor design put my friend’s health at risk was a real wake up call and reminder to me of how wide-spread and significant this issue is.

I’m resolving to be more aware of this in the future and I hope you will to.

(And, if you’re in the UK, consider signing the Save Bookstart petition – this invaluable service puts books into the hands of young children – having books in the house in childhood is a key indicator of later literacy).

Personas are for hippies… and transformation and focus

Thanks to London IA, I had the opportunity to share some thoughts on Strategic UX that have been starting to take shape into a book recently.

I happened across this twitter exchange this morning. This is a not surprising response to personas, I’ve shared this response at times and have empathy for both points of view.

Twitter conversation between Tom Coates and Martin Bellam

Here’s the thing… You don’t really want to use personas, do you? They are really a pretty cumbersome way of maintaining your customer’s active presence in the design & product management process.

What you really want is a small, tight team who get who your users are, what they value about your product/service and what is behaviourally significant about them. And you want regular access to them (if you or your boss are representative of your target audience this helps enormously).

Enter reality – the majority of us are not working in this kind of environment. We are working for large organisations who are focussing more on themselves than their users, with people who may not have seen or heard from a customer for years (if ever), whose attention is constantly being focussed on internal KPIs focussed on quantity not quality. Who resist making and decisions in preference to making a sub- optimal decision that can be traced back to them.

Sure, the likelihood of incredible design flourishing in this environment is significantly reduced, but what do we do? Give up?

We can’t all do that, can we? And neither we should.

Many of us have experienced that moment when a team transforms – when they realise what it is like to be their customer and how easy it would be to make that experience better. This most often happens during usability testing. (around the 3rd or 4th participant when the team acknowledges that perhaps we haven’t recruited a bunch of stupid users and maybe we do need to change the design a little).

Well made and well used personas are less able to create this transformation (watching real users will always trump personas) but they can help maintain that transformation and act as a tool to evangelise a customer focus through out the organisation and to create a common language around our users and – possibly my favourite thing – to allow us to reduce usage of the term ‘user’ (so abstract, inhuman and elastic) and replace it with our personas names.

Yes, this does make you feel like a bit of a hippy. I agree. But it helps, a lot, to transform focus from internal processes and priorities to what people actually do, need, want.

You don’t *have* to use personas to do good design. If you make bad personas (made up not researched, focused on demographics not relevant behaviours and attitudes), and if you use then poorly (make them and forget about them, or keep them hidden within the UX team) then you might as well not use them.

But well made personas in day to day use through out the organisation are incredibly useful when you need to gain and maintain focus on the (potential) customer.

Here’s the test:

  • do you have personas for your project/product?
  • are they made of data from real (potential) customers?
  • do they have real names not segment names?
  • do you have fewer than five personas?
  • can you remember all the names of your personas and describe them?
  • do you use them to guide, evaluate and/or explain design decisions?
  • can your boss name your personas?
  • can the developers on your team name your personas?

If you’re not answering yes to the majority of these, there are probably good reasons why personas aren’t really working out for you.

Don’t fret if you didn’t do so well here – most people don’t (out of a room of dozens of UXers last night only one lonely hand remained in the air at the end of this line of questioning last night).

I reckon personas are the best known but most misunderstood and misused tool in the UX toolkit. Don’t throw your personas out necessarily but see how you can incrementally improve how they’re made and communicated.

And if your fortunate enough to work in a project team who doesn’t need personas, well, lucky you – just don’t be too successful or you may find yourself large enough that you’ll windup needing personas after all! ;)

(For help on making good personas, two excellent resources are Designing for the Digital Age or About Face 3)

Ignoring the Android Honeycomb Action Bar Convention

I’m working on the UX for an application for the new Android Honeycomb User Interface (UI). This is the first time I’ve designed for this operating system and we were fortunate enough to have Nick Butcher of Google UK take us for a walk through the new environment and introduce us to some of the UI conventions.

If you’ve not yet had the pleasure, you can take a look for yourself here:

There’s a lot to like about the Honeycomb OS in my opinion, especially once you get to know it a little and learn what the many icons that litter the interface actually refer to, however there is one UI convention that really bothers me, and that in this application we are going to largely ignore and that is the action bar.

This is the action bar (screenshot borrowed from TechCrunch who discuss the UI in some detail here)

Honeycomb Action Bar

So, the action bar is application specific and context specific within an application. The idea is that you put any actions you may need in that part of the application into the action bar where you can then ‘take action’, the more contextually relevant the action, the further to the left, the less relevant and more global, the further to the right.

Now, it is A Good Thing that there is a convention for this, especially in an open source environment where designers/developers play fast and loose with the interface and the overall experience of using the device running this platform suffers as a result (as, having played with a Xoom for several days now, I can report is definitely the case).

The implementation, however, I’m less thrilled with, particularly as I’m confident it will actually encourage designers to ignore or duplicate the action bar’s functionality within the interface… and here’s why.

Tablet Activity Zones - Landscape Here’s a ‘heatmap’ of the activity zones for a tablet in landscape mode from a very useful post by Dan Saffer of Kicker Studios.

Note that the red band labeled ‘reach’ (indicating that you have to reach to access this zone and it is not an easy or comfortable activity) corresponds almost exactly with the Action Bar.

Although the TechCrunch article suggests that you instinctively look to the action bar (which, given that it is visually quite pushed back, seems to me to be an easily learned behaviour rather than a natural one), it qualifies this immediately by saying that you do this only if you can’t locate the action in the applications main UI.

This means that, if you follow Honeycomb’s UI conventions (in the hope of making the ecosystem of applications a good user experience with some potential hope of ever rivalling Apple’s iPad), you have three options: make the actions people need to perform ergonomically awkward (the xoom, for example, is no lightweight and unless you’re resting it on a table or your lap, requires some juggling in order to hit an item on the action bar, this juggling also makes the selection significantly less accurate), duplicate the actions in the bar and in the application UI, or ignore the convention and put the actions that are most contextually relevant in a place that falls with in the ‘easy’ activity zones.

I doubt that the designers of Honeycomb intended the action bar to replace in context calls to action, and neither it should, but it seems to me that it means that the actual use of the action bar is now very ambiguous. The more I design for this environment, the more I’m finding that the action bar actually feels much more suitable to application-wide actions rather than contextual – not the least because it is often visually quite removed from the content it is acting on.

This is a significant iteration on the previous Android software and, as our friends at TechCrunch say: ‘…at least people will actually be able to find these options, which is more than can be said about the options hidden behind the ‘Menu’ button on current versions of Android (which many people never hit).’

At the same time it feels like a rather awkward solution to this problem and certainly one that wasn’t really thinking about the mechanics of using a tablet in the wild.

As someone with an existing interest in design and UX for open source, I’d be really interested to hear what other designers who are encountering this convention are making of it. I’d also like to hear about whether the option of placing the action bar nearer to the ‘easy’ activity zones was considered and how that played out.

I’d really like to see Honeycomb evolve, and for good and clear UX conventions to emerge in this space so that we have a really solid alternative to the closed and ever more controlled world of Apple

(disclaimer: I own Apple gear. I know it’s great).

A template for intensive design

I was recently speaking with a potential client about a project that I very much wanted to work on. Due to scheduling issues (theirs and mine) we ended up with one week in which we could both be available to work on the project. At first, it seemed like the logical thing to do was to walk away and hopefully refer them to someone else with more time on their hands… but we really wanted to work together. We started thinking… could it be possible?

And so it was that we ended up working on one of the most intensive design and research projects I can remember working on. It was hard work, but good work and – in the end, we got the job done. I thought it might be useful to share the format we used so you can consider potentially this approach if you find yourself in a similarly time challenged situation some day!

Important note: this is not a sustainable way of working. You can do this for a week, you can’t do this week in, week out for a year.

The challenge:

We started the week with lot of data/content in a database, a target audience (digitally excluded), a content management system (SharePoint 2010), an accessibility goal (triple A) and a logo.

We needed to end the week with a high fidelity prototype that could be taken into production the following week.

We wanted to do this taking a user centric approach, ensuring that our concepts were evaluated by members of our target audience throughout the design process.

The team:

It was evident from the get-go that this was going to take more than me. I’m a freelancer, so this mean that I had the pleasure of hand picking a team to work with me on this.

I asked Mark Boulton to help bring our prototypes up to high fidelity. Mark & I have worked together a bit in the past and he is really comfortable in that grey area between wireframes and finished design – this is an area where designers can butt heads a little, so avoiding that was going to be very important in this project.

I also asked Andrew Travers to be the second UX designer on the project – I knew we’d need two pure UX people on this project as we were aiming to both design and research in the week. Andrew brought some brilliant subject matter expertise and accessibility know how to the project, but more importantly, he was brave enough and flexible enough to contemplate such an ambitious/slightly mad project plan. (Andrew has also written up his thoughts on this project).

We pretty quickly realised this was a great opportunity to invite an intern to work with us. Not only could we really use an extra set of hands, it was a rare opportunity to see pretty much a full UCD project in the space of a week. We were thrilled when Lisa Drake took a week of holidays from her job to join us. It was a great decision – both to invite a mentor and to choose Lisa, who was fantastic.

In addition to this we also had four members of our client team on site for the entire week including decision-making-enabled representatives from marketing, content, technical and their project manager.

Finally, we had a daily call scheduled with the Shaw Trust who were going to review our work each day and make sure we were on top of any accessibility issues that emerged as our prototypes developed.

The venue:

We needed a space that would allow for our team to be onsite, to do workshops, design work and to conduct research. We booked some space at the London User Research Centre: a research lab and observation room and another workshop room with day light. This gave us the research facilities we needed and enough flexibility in the space to be able to accommodate the range of activities and people that we needed to house in the space of that week.

On the final day, we de-camped back to the client’s offices to wrap up our work and prepare for a presentation to the larger client team.

The format:

The general shape of the project was this:

  • 1 day of UX ‘foundations’ and initial concept development
  • 3 days of prototyping, researching, iterating
  • 1 day of completing templates, annotating and preparing presentation
  • a day or two in the following weeks to finalise any outstanding work.
  • 3x UX resources for all 5 days, 1x ‘visual’ design for the last 4 days (+ several extra days in following weeks)

The general shape of the day tended to be:

  • project team ‘kick off’ meeting in nearby cafe around 8am
  • full team kick off in labs around 9am
  • work, work, work, work, work,
  • full team debrief at end of day
  • project team continue work/debrief at pub/over dinner that evening.

Preparation:

There was limited time for preparation, and this largely consisted of agreeing a recruit brief for research, briefing recruiters, reviewing existing materials that the client had (mostly from an aborted previous attempt at this project), and project planning – working out a rough idea of what we were going to do on each day and a fairly specific plan for day one.

Day one: UX Fundamentals

This day had to provide the grounding for the rest of the weeks work – we needed to:

  • clearly articulate the value proposition
  • clearly identify and describe the priority audience(s)
  • understand the primary scenarios of use  that we wanted to support
  • come up with some concepts for how we might present the our content to this audience to support these scenarios in a way that clearly expressed and supported the value proposition.

Our approach to the first three items entailed extensive use of post it notes, individual brainstorming, collaborative affinity sorting and prioritisation. Our approach to the final item involved a lot of group sketching (including our client team, of course), discussion and ranking.

As we left the lab at the end of the first day we did have a couple of concepts we were going to move forward with but we weren’t feeling particularly inspired by them. Upon decamping to a local pub that evening (in preparation for meeting and briefing Mark who was joining the team that evening) it became clear that we had quite a bit of affection for one of the concepts that we’d dropped while in conversation with the client – it was perceived as a little too risky. Over a pint, we did a little more work on this concept and got it into a sufficiently good shape to include as an option to present in research the following day. We then went to get pizza and bring Mark up to speed.

Day two:

After our morning kick off, Mark took the client team off to start work on the ‘look and feel’ of the site, starting with a mood board exercise. Meanwhile, in the observation room, the UX team were frantically building prototypes of 3 concepts (using Flairbuilder, mostly for speed) and preparing a discussion guide in time for the first research session at midday – the first of 14 interviews scheduled in this next three days.

By midday, three very rough prototypes and one very unrehearsed discussion guide in place – the research began. We saw six people in the rest of that day – tag teaming research between Andrew & myself, clients watching every moment of the interviews, and design happening on the fly meaning that no two participants saw exactly the same prototypes.

By the end of that day, we had learned a lot. We’d abandoned one concept entirely, introduced another, were pretty sure two concepts were not right and that the concept we’d rescued in the pub the night previous – the risky one – was going to be the right way forward – but it still felt a little scary. We needed more evidence it was right. We didn’t have much to show the Shaw Trust for them to advise on.

That night, we were all pretty nervous.

Day three:

Four more research participants today. At some point it becomes evident that the ‘risky’ version is definitely the way forward. A whole range of participants have now managed to identify personally with it (beyond our expectations) when our initial fear was that it would be alienating. We leave Mark to grapple with increasing the fidelity of the design and move onto tackling the more content rich templates and, as it turns out, the content itself.

We uncover a range of information architecture issues, particularly around terminology/labelling on a freshly ‘redesigned’ content model, we completely reshape the way the content is presented and in the process get very excited about a fancy faceted navigation system.

The Shaw Trust remind us that people with cognitive disability will struggle to make sense of our fancy faceted interface. We realise we’ve gotten excited about an idea and forgotten about our audience (who are not necessarily cognitively disabled, but who are the least experienced web users). We prepare to kill our darlings.

Day four:

Another four participants today. Having sketched all the way home yesterday and back again this morning, over coffee before our kick off meeting I have a feeling I may have replaced yesterdays darling facets with a much simpler solution that properly matches the needs we’re hearing coming out of research.

Our clients are more energised and excited about this project than they were at kick off and this is in no small part due to them having the chance to actually witness the people they work to help every day, actually using the system we’re designing for them. These people are stepping out from behind stereotypes and suddenly feeling a lot like us – but with the specific needs they have more clearly articulated than ever.

We test the newly simplified data-rich interface and struggle to keep a straight face when the participants describe the hard to make sense of page they’re expecting to see, then react with visible delight when they see our stripped down page, designed to focus specifically on the content they are seeking. (You don’t get those proper delight moments often, we cherish those).

We copy and paste ‘high fidelity’ designs into our prototypes as parts of them are ‘ready’. Headers and footers first, then bits of content as it starts to feel like it’s working.

We’re having all kinds of difficult discussions with the client about corporate colours and logos, but we’re also able to test our variations as we go – to understand which fights are really worth having and which are less so.

Even now, we rarely show the same prototype twice. Constantly refining.

We leave the lab on day four feeling pretty amazed at how confident we feel that we’ve actually, really cracked this. That it’s actually going to work.

Day five:

We’ve got a meeting room at the client’s offices. We make a window full of post it notes of outstanding tasks, we prioritise and allocate the tasks. We make tea. Lisa, miraculously, produces a packet of chocolate biscuits.

We work as fast as we possibly can to work through the details of the templates, to make sure we can map the database to our templates, that we can make any ‘massaging’ that needs to happen to the content relatively painless, that we’ve thought through various states and orders in the flows.

We put together a presentation of the work we’ve done over the past week, our rationale and our designs. We do this so quickly it takes less time to make the presentation than it does to give it.

We kick off our presentation by showing some of the profiles of participants we’d met that week – young single mothers, people suffering from mental illness, people who are now or were recently homeless, or in prison. People who really need us to help make access to services easier to find – especially as more and more of those services go online.

Our client is happy with the work we’ve done, but we’re not really surprised because they’ve been there, with us, helping make decisions and seeing how and why decisions were being made the entire time.

Wrapping up and next steps:

Not everything happens in a week. The following week we put the wireframes into a more formal document with annotations and some notes to capture the general principles of the design approach and content strategy.

Mark has more work to ‘design’ all the wireframes into developer-ready templates. We’re still struggling with the homepage … we know the components that need to be there but getting them to work visually is tricky.

We do a handover meeting with the client to talk through everything including any questions they have outstanding. There’s a bit of work required to properly map the database to the templates. We agree there is a whole other project required to look at the information architecture and bring it into alignment with our new findings and approach.

What we learned:

  • your team is everything – you need a good, flexible, friendly, committed team to work this way
  • having the client on site is invaluable. This approach would probably not have worked (or at least, worked so well) if they hadn’t been there participate, observe, field our questions, respond to our challenges.
  • you don’t need to sacrifice research just because your timeframes are short. You will have to be flexible and not get hung up on process, but you will learn what you need to make good, informed, decisions. Also, you give your client the opportunity to see their ‘customers’ in real life. Both of these are invaluable.
  • although you’re in a hurry, you need to take time to communicate.
  • if you want to work like this you need to be brave and confident, you can’t be a perfectionist, you have to be careful with your client seeing you making mistakes and being wrong (all part of the process)
  • not only does an approach like this work but it works well. As a team, we were inspired and energised and felt we’d probably done some of our best work because of the way we were working not in spite of it. I think we’d all be keen to work this way again. (As soon as we get leave from our families who we saw very little of that week).

Photos by Lisa Drake. Thanks to Start Here for being brave enough to work with us this way!

Page 3 of 19«12345»10...Last »