By: Martha L. Stone Poynter Study Shows Online, Print Reader Differences
The new Stanford/Poynter Eyetrack study paints a surprising picture of
the frequent online news reader. For example, she overwhelmingly
prefers text over graphics and photos.
Researchers concluded that it appears 'a provider's first and best
chance to engage the reader is through text,' according to the study. The
research showed that articles and briefs were viewed by participants
much more often (92% and 82%, respectively) than photos (64%), banner
ads (45%), and graphics (22%).
In fact, the 67 participants of the study in St. Petersburg, Fla., and
Chicago, read three times as many briefs as articles (from 24 different
categories of news on an average of six news providers in one sitting,
which lasted an average of 34 minutes).
The 67 subjects called up 211 unique news providers for a total of 426
individual visits. They spent 40 hours looking at news on the Internet,
made 608,063 eye 'fixations,' and made 24,530 mouse clicks. Eye
fixations were tallied by the computerized headset that participants
wore while surfing the Web. The headsets counted how many times the eye
focused distinctly on a new subject on the page.
The study revealed a pattern in which participants sought out specific
text and either skimmed or read it at length. Their eyes systematically
went over more than 75% of the length of almost all articles presented
to them.
How the study was done
The study began in 1996 at Stanford University in California, where
online news users were videotaped in their homes and offices. The more
elaborate eyetracking study began in 1998, when the Poynter Institute
of St. Petersburg, Fla., got involved. The study was conducted by
Stanford communications professor Marion Lewenstein, Poynter Institute
for Media Studies fellow Andrew DeVigal, and a host of team members
from both institutions.
The pilot began in December 1998 with news readers from San Francisco
and the San Jose Bay area, where technical bugs were worked out of the
headgear and computer equipment. After that, 34 participants in Chicago
and 33 participants in St. Petersburg were tested. They wore EyeLink
eye-tracking headgear with small cameras mounted to the headgear that
use infrared light to follow the eye as it scans or reads through a Web
page, measuring even slight head movements. The camera collects images
60 times per second.
Meanwhile, the computer collected the eye movement data, length of time
for each eye pause and mouse click. Later, tallies of eye fixations
were overlaid on clicking tallies and the Web page content to produce a
densely covered dotted screen where eyes, clicks, and content met.
After studying the eye movements against content, and the pattern that
users consumed the news, researchers were able to determine where the
eye first made contact with the pages, and in what order the
information was consumed. In all, 5,963 pages were viewed, including
4,542 unique pages.
The participants of the study were news junkies, recruited via
newspaper ads. The chief qualification for selection was that they read
online news at least three times per week.
Contrary to the way newspaper readers consume news in print, online
readers are not grabbed by photos and graphics. Text in the form of
briefs and captions are read first as Web pages are called up. Readers
would go back to photos and graphics, sometimes after they clicked on a
brief to get a full-text article, according to the study.
What the study means
'The study in essence confirms what we already knew: that content is
important and the reason readers go to online news sites,' DeVigal
says. 'But as we observe that readers go to text first, we as an
industry need to recognize that headlines and text are the driving
mechanism of getting our readers to drill deeper.'
Why were news junkies chosen, instead of a broader sample of news
consumers? Study leaders 'wanted to isolate the study to folks in one
spectrum,' DeVigal says. 'If we studied the entire spectrum from folks
who were news junkies to newbies, we would have had lesser degree of
accuracy. We wanted to see how folks who read the news read the news.
That was the premise of the study.' DeVigal says they didn't want to
include the experience of people new to the Web, who would be dealing
with new interfaces as they read the news.
Other findings
? 30-year-olds were more likely to read local news than 60-year-olds or
20-year-olds. 20-year-olds read more science and sports news than other
age groups.
? 80% of all participants read crime and disaster coverage.
? Sports was read by 70% of males and females, but no female read
heavily in this category while 11% of the men did.
? 48% of all participants read local news, with women more likely to do
so.
? Slightly more men than women read national news. Overall, 67% of all
participants read some national news.
? Subjects often moved back and forth between sites rather than read
them serially. This verifies findings from the first round of this
study made in 1997 with video cameras.
? The maximum time spent with a non-general news provider was 20
minutes, while the maximum time spent on a general news provider site
was 46 minutes. General news provider is defined as national, local,
hometown, and customized Web sites, ranging from CNN.com to
chicagotribune.com to ultra-local sites to MyYahoo.com. Non-general
news provider is defined as foreign or non-traditional news sites, such
as BBC.com or Walgreens.com
Please also see Steve Outing's column on the study.
Comments
No comments on this item Please log in to comment by clicking here