useit.com Alertbox Feb.

2011 Reading Website Copy on

Mobile Devices

Jakob Nielsen's Alertbox, February 28, 2011:

Summary:
When reading from an iPhone-sized screen, comprehension scores for
complex Web content were 48% of desktop monitor scores.

It's more painful to use the Web on mobile phones than on desktop computers for
many reasons:

Slower downloads
No physical keyboard for data entry
No mouse for selection; no mouse buttons to issue commands and access
contextual menus (indeed fewer signaling states, as discussed further in our
seminar on Applying HCI Principles to Real World Problems: a touchscreen only
signals "finger-down/up," whereas a mouse has hover state in addition to
button press/release)
Small screen (often with tiny text)
Websites designed for desktop access instead of following the usability
guidelines for mobile
Whacky app UIs that lack consistency

New research by R.I. Singh and colleagues from the University of Alberta provides
one more reason: it's much harder to understand complicated information
when you're reading through a peephole.

Singh and colleagues ran a Cloze test on the privacy policies of 10 popular websites:
eBay, Facebook, Google, Microsoft, Myspace, Orkut, Wikipedia, WindowsLive,
Yahoo!, and YouTube.

I did a quick analysis of Facebook's privacy policy, which features:

5,789 words, or 35 times the number of words users read during an average
page visit.
13th grade reading level, so only people with a year or more of university
would find the text easy to read.
Nicely formatted for Web reading, including a good use of subheads,
bulleted lists, and highlighted keywords, in keeping with guidelines for writing
for the Web. (That said, these guidelines also call for short text and an
8th-grade reading level when targeting a broad consumer audience and not just
Harvard students.)

(Last week, Facebook posted a draft of a rewritten privacy policy. The content is now
written at an 11th-grade reading level, which is a nice improvement. Even better,
the endlessly scrolling document has been broken up into multiple pages, with a
simple internal navigation system and clear summaries that give users an overview
of the information. Adding structure and navigation to turn a nightmarishly long,
linear document into a tight information space is an example of what I call "mini-IA."
Well done.)

In any case, there's no doubt that privacy policies count as complicated Web
content.

In Singh's study, 50 test participants completed Cloze tests while reading the privacy
policies on either a desktop-sized screen or an iPhone-sized screen. (The study didn't
use an actual iPhone, but since users didn't perform navigation or any interactions
other than reading and scrolling, the specific device shouldn't impact the
comprehension results.)

Results:

Desktop screen: 39.18% comprehension score

Mobile screen: 18.93% comprehension score

Test scores must be 60% or higher for a text to be considered easy to understand.
Even while reading from a desktop screen, users achieved only 2/3 of the desired
comprehension level.

The obvious first conclusion from the study is that major websites have overly
complicated content in their privacy policies. Of course, it's not exactly news that
privacy policies are incomprehensible. Users know this and usually don't read them.

In our previous research, we have found that users treat "user agreements" and
similar site copy with contempt. In approaching such agreement text, users

read 10%
scan 17%
skip 73%

And this is users' behavior during usability studies, where they know they're being
video-recorded. At home, I expect they read even less. Basically, people click "I
agree" without reading what they're "agreeing" to.

User comprehension scores on the Cloze test were 48% of the desktop level
when using the iPhone-sized screen. That is, it's roughly twice as hard to
understand complicated content when reading on the smaller screen.

Why? In this case, people were reading only a single page of information, and they
were shown that page as part of the study without having to find it. Thus, navigation
difficulties or other user interface issues cannot explain the increased difficulty. Also,
users were tested in a lab, so there were no issues related to walking around with
the phone or being disturbed by noises or other environmental events. (In the real
world, such distractions and degradations of the user experience further reduce
people's ability to understand mobile phone content during true mobile use.)
The only reason mobile scored lower than desktop is the screen size, since that was
the only difference in the study conditions.

A smaller screen hurts comprehension for two reasons:

Users can see less at any given time. Thus, users must rely on their highly
fallible memory when trying to understand anything that's not fully explained
within the viewable space.
Less context = less understanding
Users must move around the page more, using scrolling to refer to other
parts of the content instead of simply glancing at the text. Scrolling introduces 3
problems:
It takes more time, thus degrading memory.
It diverts attention from the problem at hand to the secondary task of
locating the required part of the page.
It introduces the new problem of reacquiring the previous location on
the page.

This new research provides striking support for the main conclusion in our usability
studies of mobile websites: websites (and intranets) must design a separate
mobile version for optimal usability. Specifically, complicated content should be
rewritten to be shorter, with secondary information deferred to subsidiary pages.

Full-day seminars on Usability of Websites and Apps on Mobile Devices and

Touchscreen/Gesture UI Usability plus a 2-day seminar on Writing for the Web and 2
days on Content Strategy at the annual Usability Week conference.

Reference

R.I. Singh, M. Sumeeth, and J. Miller: "Evaluating the Readability of Privacy Policies
in Mobile Environments," International Journal of Mobile Human Computer
Interaction, vol. 3, no. 1 (January–March 2011), pp. 55–78.

> Other Alertbox columns (complete list)

> Sign up for newsletter that will notify you of new Alertboxes

Copyright © 2011 by Jakob Nielsen. ISSN 1548-5552