More Alike Than We Think

Writing for Low Literacy or Low English Proficiency

It might seem counterintuitive to worry about how well people with lower literacy can read a university Web site, but the OU prides itself on its open admissions policy. Its students include people with reading difficulties that have nothing to do with their intelligence or capacity to learn, and students for whom English is a second language are an important audience at the OU.

Michael Summers and Kathryn Summers’ work on designing for low-literacy readers focused on different reading habits. In their article in STC’s Intercom, “Making the Web Friendlier for Lower-Literacy Users,” PDF and in a summary published on UseIt, they stressed such readers’ need to read slowly and carefully along with their tendency to skip information they don’t understand.

• They read word for word more slowly than other readers, focusing exclusively on the line of text they are currently reading.
• They don’t scan, because they can’t read quickly, so they often completely skip over large amounts of information, especially dense text, or accept as “good enough” the first answer they find.
• They have more difficulty in wayfinding, because they can’t scan navigation options easily, and often ignore information on parts of the page outside the main content.

They concluded that linear paths help low-literacy readers move through information with a minimum of distractions.

Finding the Overlap

It’s hard to imagine three audiences that seem further apart than teens, older adults, and people with difficulty reading. We had feared that we would find conflicting advice. But the advice we found was remarkably consistent, even looking closely at specific, diverse audiences. The reasons were different, but the same advice served each of these users in different ways:

• Avoid long, dense blocks of text.
• Create informative headings.
• Provide navigation options within the content.

Avoid Long, Dense Blocks of Text

Impatient readers are likely to skip any “walls of words” as too much work. This is not to say that they never read anything in detail, but without cues to the value of such long paragraphs, they have no reason to invest the time and effort it would take to read them.

Lower-literacy readers will often just give up, skipping over text they find too hard—or too daunting. Long paragraphs often mean complex sentences—just the kind of writing that is hardest to read for someone without good reading skills or a deep familiarity with English.

Older adults are penalized by long blocks of text, precisely because they are likely to read everything. This slows them down, so reading each page takes longer, and can take them off track if the information is not completely relevant.

Create Informative Headings

Informative headings—headings that make a point—are not just decorative. They allow readers to simply scan from heading to heading and still pick up the main points. This helps both impatient teens and older adults.

The research with low-literacy readers suggests that they skip headings along with other text, so informative headings might not help them as much as the other groups. But, if they do read just the beginning of each chunk of text, this group will also learn the main points from well-written headings.

Provide Navigation Options Within the Content

For careful readers—older adults and low-literacy readers—adding navigation links at the end of the content provides them with a next-step option exactly where they need it and lets them avoid hunting around the page for options. For low-literacy readers, this also lets them avoid the need for even more reading to find an appropriate menu option. For older adults, with their cautious clicking behavior, the direct connection between content and links adds relevance and makes the decision easier.

For impatient readers and those with poor Web skills, a good navigational choice, presented at just the right place, provides subtle wayfinding guidance—without eliminating other choices.

Trying It Out

Armed with all this research and our previous usability work, we wrote the content for the new site prototype and tested it. Our participants ranged from a 17-year-old automotive-trades student to someone heading into an MA in Marketing to a 58-year-old looking for career training. Their ages, familiarity with universities, socio-economic status, and prior education were all highly varied.

We saw many of the behaviors predicted in the research:

• No one liked pages with long blocks of text. They preferred and were more successful in reading pages with information broken up into short chunks, each paragraph making a clear point.
• They liked pages and menus that clearly stated what questions would be answered. Putting these questions in their own language worked even better. For example, they wanted to know what a course at the OU would cost. A link that echoed their words, saying “What will it cost?” worked better than one that said “Paying for your studies.”
• They were quick to leave a page that did not seem promising—though the older adults tended to persevere for longer. They wanted pages that answered questions, clearly and directly, not pages that gave them information to sift through.

These results were the same across the board. No one turned to us and said that they were finding these pages too easy to read or too simplistic. They were more likely to follow links to additional information if they had already found part of the answer. If the paragraph preceding a link was not helpful, they seemed to have little confidence that a related link would do any better.

When they did follow links into the main site—for example, to read the requirements for a subject or program that interested them—they were more successful there, because they had the context they needed to make sense of the detailed information.

The best result for our business goals was that the new prototype succeeded in accurately communicating answers to people’s top questions about the university. Where previous sites had failed even to help readers understand the basic fact that the OU is a distance-learning institution, these participants were able to correctly explain not only how the OU works, but identify features of the program that were particularly appealing to them.

Are We More Alike Than We Think?

It might be easy to claim that each different audience needs its own separate user interface—whether we are talking about interfaces for novice versus advanced users, different cultures, or people with differing needs for accessibility—but the danger in this approach is a proliferation of different, specialized versions, adding complexity both to the design process and to readers’ process of finding the right version for them.

Our experience in creating and testing the prototype for the Open University suggests that a more universal approach can work even better. (For an overview of the principles of universal usability, see my UXmatters column, “Why People Matter.”) Instead of forcing visitors to select the right version for their needs, we offered information that was useful and usable to all visitors. Rather than creating a design that satisfied a lowest common denominator and a mediocre user experience for all, our universal design helped everyone take the first step in learning about the OU and deciding what to explore next.

I’d like to hear your experiences in designing for diverse audiences? Have you found the differences or similarities more important? Are there design principles or guidelines that have helped you bridge the gaps?

We expect the Open University site to launch in early summer and will post a link to the site here when it does.