Karen Schriver—Plain by design: Evidence-based plain language (PLAIN 2013)

We may be good at the how of plain language, but the why can be more elusive. To fill in that missing chunk of the puzzle, information design expert Karen Schriver has scoured the empirical research on writing and design published between 1980 and 2010. She gave the PLAIN 2013 audience an eye-opening overview of her extensive, cross-disciplinary review, debunking some long-held myths in some instances and reaffirming our practices in others.

Audiences, readers, and users

In the 1980s, we classified readers and users as experts versus novices, a distinction that continues to haunt the plain language community because some people assume that we “dumb down” content for lower-level readers. Later on we added a category of intermediate readers, but Schriver notes that we have to refine our audience models.

What we thought

A good reader is always a good reader.

What the research shows

Reading ability depends on a huge number of variables, including task, context, and motivation. Someone’s tech savvy, physical ability, and even assumptions, feelings, and beliefs can influence how well they read.

Nominalizations

What we thought

Processing nominalizations (versus their equivalent verbs or adjectives) takes extra time.

What the research shows

It’s true, in general, that most nominalizations do “chew up working memory,” as Schriver described, because readers have to backtrack and reanalyze them. However, readers have little trouble when nominalizations appear in the subject position of a sentence and refer to an idea in the previous sentence.

Conditionals

What we thought

Conditionals (if, then; unless, then; when, then) break up text and help readers understand.

What the research shows

A sentence with several conditionals are hard for people to process, particularly if they appear at the start. Leave them till the end or, better yet, use a table.

Lists

What we thought

Lists help readers understand and remember, and we should use as many lists as possible.

What the research shows

Lists can be unhelpful if they’re not semantically grouped. If an entire document consists of lists, we can lose important hierarchical cues that tell us what content to prioritize.

Text density

What we thought

A dense text is hard to understand.

What the research shows

It’s true! But there’s a nuance: we’re used to thinking about verbal density, which turns readers off after they begin reading. Text that is dense visually can make people disengage before reading even begins.

Serif versus sans-serif

What we thought

For print materials, serif type is better than sans-serif. Sans-serif is better for on-screen reading.

What the research shows

When resolution is excellent, as it is on most screens and devices nowadays, serif and sans-serif are equally legible and easy to read. Factors that are more important to readability include line length, contrast, and leading.

Layout and design

What we thought

Layouts that people prefer are better.

What the research shows

We prefer what we’re used to, not necessarily what makes us perform better. This point highlights why user testing is so important.

Impressions and opinions

We thought

It takes sustained reading to get an impression of the content.

What the research shows

It takes only 50 millseconds for a reader to form an opinion, and that first impression tends to persist.

Technology

What we thought

Content is content, regardless of medium.

What the research shows

Reader engagement is mediated by the technologies used to display the content.

Teamwork in writing and design

What we thought

Writing and design are largely solitary pursuits.

What the research shows

Today, both are highly collaborative. We now have an emphasis on editing and revision rather than on creation.

***

Evidence-based plain language helps us understand the reasons behind our principles and practices, allowing us to go beyond intuition in improving our work and developing expertise. We can also offer up this body of research to support our arguments for plain language and convince clients that our work is important and effective. What Schriver would like to see (and what the plain language community clearly needs) is a repository for this invaluable research.

3 thoughts on “Karen Schriver—Plain by design: Evidence-based plain language (PLAIN 2013)”

  1. Iva, I wasn’t present at the conference, but it sounds like it was fascinating. Thanks for recapping this great talk for us here.

  2. When Karen said this at the conference:
    ” We now have an emphasis on editing and revision rather than on creation.”
    I understand why, when creating my list book, I realized I needed a chapter on document revision as well as document creation.

    Great article, Iva. Do you do shorthand? All your reports are so good.

  3. Thank you so much for sharing what was presented! I heard that Karen’s talk was awesome, and now I can see why. I love the “What we thought” vs “What the research shows” format. So important to get the research in there. Thanks.

Leave a Reply

Your email address will not be published. Required fields are marked *