Designing a digital textbook: Readability and design

If you are going to publish a digital textbook, the first thing you ought to ask is whether anybody will be able to read it. This seems sort of obvious, but it’s too often left out of discussions about online textbooks and “e-textbooks.” Most e-book versions of college textbooks are just scans of printed pages stuck into a weak knockoff of a web browser. At the other end of the spectrum, professors slap content into “course management systems” without any thought at all to design (and, to be fair, course management systems don’t always make good design particularly easy). Rarely are the content and its presentation designed from the ground up for the medium of delivery — and when they are, the result is too often a mass of hyperlinks and interactive tools that “take advantage of the medium” but don’t actually aid reading or further comprehension. As a result, these means of delivering content to students are usually a step backwards from a well-designed print textbook. That’s a shame, because I think there’s tremendous potential for online textbooks — if they’re designed thoughtfully.

In working on North Carolina Digital History I’ve put a lot of thought and work into making the content as readable as possible — ensuring not only that people will be able to find the material they’re looking for, but also that they’ll be able to read it comfortably and effectively once they’ve found it. I won’t say I’ve entirely succeeded; the current page design is one step in a process of iterative improvement, but I think we’re making considerable progress. It is, in any case, a thoughtful design, and it’s that thought process I want to talk about here.

I’m working from two key premises:

1. People can and will read online, if the content is interesting, well written, and decently designed; and
2. The design of a page of content — its layout and typography, whatever its medium — has a tremendous impact on the ease with which a reader can comprehend it.

Scanning and reading

The difficulty with reading online is that websites are typically designed as if they were reference materials. In fact, the entire theory and practice of web usability is based on the notion that websites exist to help people efficiently find and use discrete bits of information. Usability testing looks at whether “users” (not readers) can navigate a website and find what they’re looking for — essentially, a scavenger hunt. Why? Not because most websites actually exist for that purpose, or even because most websites existed for that purpose back in the ’90s when usability got its start as a field. The web was barely out of diapers when Suck proved that you could publish long-form content on the web, have devoted readers, and support yourself with advertising. No, usability research focuses on information-finding because usability research costs money, and information-finding — or, more specifically, product-finding — is where the money is.

Jakob Nielson’s usability research, the sine qua non of the field, is well-designed, cleary reported, and valuable to designers, but Nielson and his colleagues have to make a living, which they do by selling reports of their research. Who has money to buy research reports at $100 a pop? Big businesses with big business models — business models predicated on either (a) selling products or (b) pointing people to other websites that sell products. Neither of those business models serve readers, so we shouldn’t be surprised when Nielson still insists that people scan, rather than read on the web. This isn’t to disparage Nielson. I believe he’s an honest researcher; I certainly don’t believe he’s twisting his conclusions to suit the needs of the people paying the bills. I do believe, though, that — like all researchers — he designs his questions selectively, with a particular audience and purpose in mind, and before we accept his conclusions we ought to consider how he asks his questions.

The conclusion that people scan rather than read on the web, which Nielson first published in 1997, has become conventional wisdom, but it’s based on a few assumptions that aren’t universally valid. There’s a certain disparagement of “users” that I don’t appreciate — the comparison of users to cattle, for example, foraging for information. (”Users” is bad enough. As Edward Tufte says, “Only two industries refer to their customers as ‘users’: computer design and drug dealing.”) Foremost, though, is the assumption that people are less likely to read a web page than they are, say, a magazine. In fact, people read magazines and newspapers in a variety of ways that are less than thorough and far from linear. Some people read a magazine from back to front; others jump around looking for content that interests them. Newspaper readers will follow a “jump” to a new page only if they are particularly interested in the article. But when readers find an article that interests them, they’ll sit and read it… at least until they lose interest.

All this sounds like pretty much the way “users” are supposed to behave on the web. In fact, some recent research from the Knight Digital Media Center on eye-tracking suggests that people read newspapers online much the way they read them in print. In fact, “Once people chose what they wanted to read they read more thoroughly online than in print.” (Emphasis mine.) As the researchers conclude, we can get over longing for the “good old days.” People will read online content if it interests them, if it’s well-written, and if it’s designed decently, just as they do in print — maybe even more so.

All this is irrelevant to some degree in designing a textbook, because the audience for a textbook is captive. They can’t simply choose not to read it. Still, we ought to care how easily they can read it, and whether we’re doing everything we can to aid comprehension.

What readers need

Mandy Brown writes that readers — not skimmers or “information seekers” but readers who are reading carefully and deeply and thoughtfully — need solitude above all else. To read, we need to isolate ourselves with a text. Books give us the space we need to read; websites do not. But they could; designers simply choose not to design them that way:

There are, of course, readers who shun the screen—those who print out long articles, or—gasp!—purchase printed books and magazines instead. We often attribute their resistance to those elements over which we have no control: the physical discomfort of sitting at a desk (versus curling up with a book); the as-yet-impossible task of producing a screen that is more comfortable than paper; the attention-deficit nature of so much browsing online that makes the transition to reading seem unattainable. But there are in fact other issues at play here, and we are capable of exerting a great deal of control over them: whether or not the design of the page embraces the reading experience, or merely grudgingly squeezes it in among the looking and searching and skimming.

I should say that most books give us the space we need to read. Novels, for example, and monographs, and most nonfiction books present plain, well-spaced text, black on white, with sufficiently wide margins that we don’t feel we’re falling off the page with the words. Textbooks, though — in fact most books for kids — have grown increasingly busy over the last few decades, usually with the excuse that kids like eye candy and will be bored with plain text.

Well, everybody likes eye candy. But eye candy doesn’t aid comprehension. It attracts the attention and distracts the reader, and distractions prevent us from absorbing information, exploring concepts, making connections between ideas. Busy design is fine for reference books, whose purpose is to help you find and recall discrete bits of information — Dorling-Kindersley’s topical encyclopedias and “Eyewitness” books are great fun to browse, and the photos and illustrations are visual anchors to help those discrete facts take root in the reader’s mind. If you’re reading an explanation of the banking crisis, or Anna Karenina, you need room to concentrate.

Designing for readers

What does it mean, on the web, to give readers their space? It means simplifying the user experience to help readers follow the text and stay focused. It means, sometimes, not taking advantage of the medium of the web by littering the page with media and hyperlinks. It means a minimalist design — not a boring one, because attractive sites are actually more usable, but one that, in the words of the too-liberally quoted Albert Einstein, is as simple as possible but not more so.

In North Carolina Digital History, for example:

• The main text column expands with the width of the web browser, but only to a point — it has a maximum width of 56 ems, or (roughly) the width of the letter m in a given type face. The number of characters in a horizontal line of text is called measure. We want to take advantage of the horizontal space a big monitor provides, but too great a measure — too wide a column — and the eye has trouble tracking across the page.
• There is a great deal of white space around that main column. The right-hand sidebar, which includes navigation links, commentary, and links to supplemental material, is separated from the main column by a sizable gutter, and I’ve tried to keep it as short as possible (which I admit isn’t always as short as I’d like).
• To avoid luring students away from the text, I don’t link from the body of the page to ancillary material, but only to material (potentially) necessary for comprehension of the page itself, such as definitions of terms, comments on primary sources, and footnotes. Links to related material, further reading, and so on are relegated to the sidebar.
• Media such as images go in the main text column only if the student is asked to stop and consider them carefully.
• Images that don’t meet the above criteria are “pullout out” into the left-hand column so that they appear next to the text to which they relate. There, they don’t interfere with reading, but they’re close by relevant passages if they’re wanted.
• A big difference between reading a printed page and a screen is that if you look away from a printed page, you can remember your place. Once you scroll away from a paragraph of text on a computer screen, it can be hard to find again. To help readers keep their place, I try to keep paragraphs relatively short, and I use more headings than I would in print. Left-column images help to mark vertical space, as well. In long documents, especially dull ones like charters and constitutions, I pull out headers into the left-column. (In the last case, I’m tacitly — or sometimes openly — admitting that I don’t really expect anybody to do more than skim the document.)

I’ll write more about several of these points, and others having to do with web typography, in greater detail in further posts. For now, I’ve violated enough of my own readability rules for one day.