AEJMC Archives

AEJMC Archives


Next Message | Previous Message
Next in Topic | Previous in Topic
Next by Same Author | Previous by Same Author
Chronologically | Most Recent First
Proportional Font | Monospaced Font


Join or Leave AEJMC
Reply | Post New Message
Search Archives

Subject: AEJ 97 TremaynM CTP The Internet: Is the medium the message?
From: Elliott Parker <[log in to unmask]>
Reply-To:AEJMC Conference Papers <[log in to unmask]>
Date:Sun, 14 Sep 1997 12:45:58 EDT

TEXT/PLAIN (794 lines)

Mark W. Tremayne
Department of Journalism
University of Texas at Austin
CMA 6-144
Austin, TX 78712
(512) 469-7417
[log in to unmask]
The unique features of each medium can change the nature of messages sent by
journalists.  Does the Internet have unique features and can those features now
be measured?  This study examines these questions, and provides a comparison of
Internet news sites started by newspaper, magazine, television and radio
companies.  The study found that these sites are making use of interactivity and
nonlinear story-telling.  Further, newspaper and television sites are taking a
different approach to nonlinear storytelling.
The Internet: Is the Medium the Message?--
"Men are suddenly nomadic gatherers of knowledge,
nomadic as never before..."
     These words could well describe many of today's media consumers, who select
from dozens of television channels, and who sit in front of computer screens,
"surfing" the Internet for items of interest.  But surprisingly, these words
were written more than 30 years ago (McLuhan, 1964, p. 358).  Marshall McLuhan's
ideas still deserve our attention today, as many of his "predictions" seem to be
coming true.   International television services can be viewed as a first step
in the direction of a "global village."  The Internet could be furthering this
trend as well.
     McLuhan's insistence that "the medium is the message" (p. 7) focused his
readers on the newest "medium" of that era, television.  Today, we might focus
instead on the Internet.  If the Internet is "the message," then what effects
does this new "message" have?  Is the Internet different in some meaningful way
from the pre-existing media?
     As many differences as the print and broadcast media have, there are some
striking similarities as well.  Both media are linear.  Every consumer receives
a story that is presented in one specific order.  And with both, the messages
travel primarily from communicator to receiver, with limited opportunities for
feedback.  Our existing models of mass communication are drawn with these facts
in mind.
     The limitations of the "traditional media" can be overcome on the Internet.
Internet presentations can be linear or nonlinear.  Communication can be two-way
(interactive). If this happens,  existing models of mass communication will need
to be reworked.
     Of course, it's possible that mass communicators will continue with their
long-standing habits.  The purpose of this study, then, is to find out whether
mass communicators on the Internet are changing their thinking about the
audience, or if they are sticking with presentations that are linear, and flow
in one direction.
Related Studies
     By all accounts, Internet use is on the rise.  As it grows, the Internet
secures its position as a new mass communications medium.  Many researchers
(Morris & Ogan, 1996) are already willing to say that the Internet is, indeed, a
new mass medium.  Further evidence of the Internet's acceptance as a medium is a
recent nationwide survey's (Stempel & Hargrove, 1996) inclusion of the Internet
and other on-line services as two of eleven different media types.  That survey,
conducted in June, 1995, found that 5.3% of respondents were regular Internet
users.  However, the researchers seem to downplay the Internet's potential,
concluding "We can expect growth in the next decade, but we should recognize
that if Internet and on-line services triple their audiences they will lag the
other media" (p. 557).   If it takes the Internet a decade to triple its
"audience" it will be quite a surprise.  It may have already doubled in the time
it took that survey to make it into print.
        Internet sites (or web sites) can be presented in a "multimedia" or
"hypermedia" format.  The user selects the items (called "nodes") that he or she
is interested in.  These "nodes" (text chunks, photographs, audio clips, etc.)
are held together by  "links."  By choosing which links to follow, the user
progresses through an Internet site in a nonlinear manner.  One user may read 3
paragraphs, then view a photograph, then listen to an audio clip, the read 4
paragraphs, then finished with 3 photographs.  The next user may choose
different "nodes" and in a different order.  It has been argued (Delaney &
Gilbert, 1991) that this nonlinear presentation closely mimics the way human
beings think.  The human brain, the argument goes, does not operate in a linear
fashion, but by "association."    McLuhan foreshadowed this argument when he
wrote, "It is an important aspect of the electronic age that it establishes a
global network that has much the character of our central nervous system" (p.
     It can be argued that newspapers are laid out in a nonlinear way.  The
reader can skip from section to section, reading stories in any order he or she
chooses.   But individual newspaper stories are linear.  To make sense, the
reader must start at the beginning, and is rarely faced with options.  Radio and
television broadcasts are completely linear.   As newspaper and broadcast
companies move to the Internet (as they are in large numbers) are they  using
this new nonlinear style, or are they sticking with linear presentations?  Can a
trend be observed?
     Beyond the question of linearity, the Internet challenges the traditional
models of mass communication with its potential for two-way communication.
Interactivity has been identified by researchers (Williams, Rice and Rogers,
1988) as the potential driving force towards widespread use of  the new media.
One study (Newhagen, Cordes & Levy, 1995) found that the mere promise of
interactivity was enough to spark interest.  A request by NBC News for viewer
e-mail prompted thousands of responses.  The authors discovered that many of the
electronic letters were written in an intimate style, and with the assumption
that a response would be forthcoming.  Some web sites offer other means of
interaction.  Users can participate in polls on current events issues with the
click of a mouse.  Users can input search terms to find information they are
looking for specifically.  Some sites offer "chat" areas where users can get
answers from the web site authors or from other users.  Some web sites use
electronic bulletin boards where users can post their opinions.  One study of
these bulletin boards (Rafaeli & LaRose, 1993) found that diversity of opinion
was an important factor in bulletin board success.  Greater diversity is a
possible development if the mass communication process becomes more "two-way."
Are journalism Internet sites using this approach?  Can a trend in this
direction be discerned?
Purpose of the Study
     Here, then, are the specific questions investigated by this study:
     RQ1: To what degree are journalism Internet sites telling stories
                in a nonlinear fashion?
     RQ2: To what degree are journalism Internet sites using two-way,
               interactive, communication?
      For journalists, this is new territory.  How does one tell a story in a
nonlinear fashion?  How can interactivity be coupled with current journalistic
principles?  We should it expect to take some time for news professionals to
adapt to this new medium.  If this is the case, we would expect the older
Internet sites to be using interactivity and nonlinearity to a greater degree
than newer web publishers.  Therefore:
     H1: The older the journalism Internet site the greater will be the use of
             nonlinearity in stories.
     H2: The older the journalism Internet site the greater will be the use of
             two-way, interactive communication.
For McLuhan, the Internet would merely be the latest phase of the "electronic
era."  That era, beginning with radio and television, is characterized by the
immediacy of electronic communication.  The immediacy of electronic
communication has already been mastered by broadcasters, and many see the
Internet as merely the newest "child" of the electronic era (de Kerckhove,
1995).  Derrick de Kerckhove (p. 52) predicts "the entire realm of television
will be swallowed by computers."  He calls the resulting medium "telecomputers."
At any rate, we might expect the broadcast industry to more quickly adapt to
this new electronic medium than the print-based media.  Therefore:
     H3: Internet sites created by broadcasting companies show greater use
             of nonlinearity than Internet sites created by print media
     H4: Internet sites created by broadcasting companies show greater use
            of interactive communication than Internet sites created by print
            media companies.
     Since the Internet is evolving so rapidly, this content analysis only
attempted a snapshot of March, 1997.
The content "universe"
     Like cable television, the Internet lends itself to specialization.  There
are sports sites, business sites, and weather sites, just to name a few.  But it
is an assumption of this project that people will always rely on journalists to
perform the "surveillance" function of mass communication.  So this research
will focus only on journalism sites that attempt to cover all types of news, not
one specialty.  There are also Internet sites which collect news items from
other sites, or merely provide links to them.  We are not interested in these,
instead focusing on creators of original content.  The field is further
narrowed to sites that are updated daily (if not hourly), and cover news in the
United States.  Still, hundreds of sites fitting these criteria were identified
with the Internet search engines Yahoo and Lycos (many are local television
stations and newspapers).  The field was narrowed by restricting our attention
to national news organizations.  We eliminate all local television sites, and
keep only 5 newspapers which have a national focus.  Four of those newspaper
sites correspond with the top 4 newspapers by circulation.   The fifth is an
Internet-only national newspaper which claims to be (and appears to be) the
first national Internet newspaper.   So our content "universe" is limited to
U.S. journalism Internet sites that contain original  material on general news
topics, updated daily , and which are national  in scope.  The 15 sites found
Newspaper Sites
Television Sites
Other Sites   (CBS)
This "census" includes 5 sites created by newspaper companies, 5 by television
networks (MSNBC is joint venture of NBC News and Microsoft Corporation), 2 by
magazines, 2 by radio networks, and one by a wire service.
Unit Sampling
      All sampling took place in March, 1997.  A sample week was created by
selecting every 4th day of March, with a randomly selected start of March 6th.
Each of the 15 sites was visited once on each of those seven days between noon
and midnight eastern time.  Care was taken to vary the time of day for each site
visit throughout the month, although this was not a rigorous probability
sampling.  635 story titles were viewed, although links to 2 were invalid and
were not included in the study.  633 stories were coded.
Theoretical variable definitions
     Coders measured story nonlinearity and site interactivity.   A nonlinear
story is one that goes in a variety of possible directions, according to the
interests and needs of the media consumer.
     An interactive site is one that accepts input from the media consumer, and
creates opportunities for that input.
Operational Definitions & Units of Observation
     The entire Internet site was the unit of observation for interactivity.
Coders looked for the availability of e-mail to the organization, e-mail to
individual journalist, user polls, news search options, question and answer
forums, and the publishing (posting) of viewer comments.  Each counted as one
point (maximum of one point), for a total score ranging from zero to six.
     Individual stories were the units of observation for the measurement of
nonlinearity.  Each site's front page stories were coded.  Some sites put no
stories on their "main" page, offering a menu of sections instead.  In these
cases "top stories" was selected, if available, and if not, the front page of
the national news section was coded.  Stories which appeared in identical form
(usually special reports) on subsequent coding days were counted only once.
Internal links (if any) within each story were counted.  A completely linear
story, with no optional links, received a score of zero.   Every optional link
the story does contain counted as one point, although secondary links (and
beyond) were not counted.  Links which merely continued the story onto another
page were not considered "nonlinear" links and were not counted (this technique
was rarely used).   One site (U.S. News) presented identical stories day after
day, with the only significant change being the number of links.  In this case,
the story was counted only once, and a median number of links recorded.
     At the conclusion of the seven pre-selected sampling days, means were
calculated for nonlinearity and interactivity for each site.
Pre-test and reliability
     A pre-test with two coders was conducted for each variable.    On site
interactivity, coder agreement was 91.7% (by Holsti, 83.3% by Scott's pi).  For
story nonlinearity, coder agreement was 99.9% (by Pearson's r).
      RQ1: To what degree are journalism Internet sites telling stories
                in a nonlinear fashion?
     The 15 sites varied widely in their use of nonlinear stories, but most (12)
of them did use at least some nonlinear stories.  The following chart lists them
from highest to lowest links per story.
Internet Sites
Front Page Stories/Mean
Percentage of Nonlinear Stories
Links Per Story/Mean
Online Newshour
U.S. News
CNN Interactive
Washington Post
NY Times
USA Today
LA Times
Nando Times
Fox News
Time Daily
ABC Radio
Site mean=3.4
The 15 sites averaged 3.4 (SD=4.38) links per story.  But the distribution is
bimodal, with the top six sites average more than 8 links per story, and the
remaining nine averaging less than 0.2 links.   Is this a significant finding?
If traditional print or broadcast stories were measured for nonlinearity they
would have to score at or near zero, since the reader or viewer is rarely faced
with options (within the story).   Similarly, most of the Internet sites studied
score near zero, but some (6 of 15) are making significant use of nonlinear
     There are a wide variety of approaches, from the radio sites which present
all stories linearly, to MSNBC and the Washington Post which present no front
page stories linearly.  The sites with highly-linked stories tend to present
less stories on their front pages than sites with simple, linear stories.
     RQ2: To what degree are journalism Internet sites using two-way,
               interactive, communication.
     As with nonlinearity, the 15 sites have very different approaches to
interactivity.   The following table lists the sites from the most to least
Internet Site
Interactive Features
LA Times
NY Times
Washington Post
CNN Interactive
U.S. News
Online Newshour
USA Today
Fox News
Time Daily
ABC Radio
Nando Times
     On an interactivity scale of 0 to 6, the 15 sites studied score an average
of 3.0 (SD=1.51).  While this is not surprisingly low or high, we must consider
the interactivity of the traditional media.  By themselves, newspapers,
magazines, television and radio would score a zero on our 0 to 6 scale.  There
is no potential for two-way communication with these media.  Another carrier
(telephone, postal service, etc.) is required for the traditional media to
become truly interactive.  Even in that situation, the interactivity is delayed,
and has no impact at the point of message consumption.
Age of the sites
     Many methods had to be used to determine the age of the 15 web sites.  A
few provide the information on the Internet site.  Although every site has an
e-mail address, only about half responded.  The ages of the remaining sites were
determined through direct calls or, in a few cases, through archived news
reports.  Internet staff sizes were also sought, but many companies consider
this a "business secret."  The following chart lists the sites from the oldest
to the newest.  The date of Internet availability was used to determine age (as
of March, 1997).
Internet Site
Staff Size
Months on Web
Nando Times
Time Daily
USA Today
CNN Interactive
US News
NY Times
Online Newshour
LA Times
ABC Radio
Washington Post
Fox News
     The first two hypotheses involve correlations between the age of the 15
sites and the two test variables.  P-values are not reported because the 15
sites do not represent a probability sample, rather it is argued that they
represent a "census" of all sites meeting the study's criteria (however, when
calculated, the P-values are more than .05).
     H1: The older the journalism Internet site the greater will be the use of
     There is little correlation (r=-.24) between the age of the 15 sites and
their degree of story nonlinearity.  What little correlation there is runs
counter to the hypothesis.  The regression line on the following scattergram
depicts this tendency.
  [--- Pict  Graphic Goes Here  ---]
So, hypothesis one is not supported.
     H2: The older the journalism Internet site the greater will be the use of
             two-way, interactive communication.
      The result reported for hypothesis one is more pronounced when we look at
hypothesis two,  the relationship between interactivity and the age of the 15
sites.  In this case we get a stronger correlation (r=-.40), and it also runs
contrary to the hypothesis.  The following scattergram shows that the newer
sites are generally the more interactive sites.
  [--- Pict  Graphic Goes Here  ---]
So, hypothesis two is not supported.
Print versus Broadcast Sites
     H3: Internet sites created by broadcasting companies show greater use
             of nonlinearity than Internet sites created by print media
     There are two ways to run the t-test for nonlinearity.  One is to compare
between media at the unit level of "story," the other is to compare at the unit
level of "Internet site."  The two methods yield similar mean differences, but
with differing significance levels (again, at the site level the test of
statistical significance could be considered inappropriate).  The table shows
results of both measuring methods.
Mean links:
Mean links:
Mean Diff.
367 stories
231 stories
7 sites
    7 sites
The data supports hypothesis one.  The broadcast sites average more than an
additional link than the print sites.  The gap is much wider when comparing the
5 newspaper sites to the 5 television sites.
Mean links:
Mean links:
Mean Diff.
313 stories
5 sites
5 sites
The gap averages more than 3 links per story. The newspaper sites are putting
more stories on their front pages, but using a lesser degree of nonlinearity.
     H4: Internet sites created by broadcasting companies show greater use
             of interactive communication than Internet sites created by print
             media companies.
     This hypothesis is not supported.  In fact, the print sites are slightly
higher in interactivity than the broadcast sites, although the difference is
small.  The print sites average 3.43 on the 0 to 6 scale, while the broadcast
sites average 2.71, for a mean difference of .71.  Comparing the newspaper and
television sites directly yields similar results.  The newspaper sites average
3.80, while the television sites average 3.00, for a mean difference of .8.
Additional Findings
     After the 633 stories were coded for nonlinearity, it was thought that some
investigation of the stories' content might be illuminating.  Initial study
identified 14 story categories.  An intercoder reliability pre-test on those
categories found 88.3% agreement (Scott's pi).  Here are the ten stories most
represented in the sample:
Story Topic
Story Topic
Campaign Fundraising
Midwest Flooding
Mideast Peace
Cult Suicides
OK City Bombing
Clinton's Knee
CIA Nomination
     This list might mirror the March coverage one would find in the traditional
media outlets which run the sites, but that measurement was not part of this
study.  However, it is interesting to note which of these stories receives the
most nonlinear "treatment" by the sites' designers.  The Cult Suicide stories
received the most attention, getting at least one link half the time and scoring
the highest mean number of links.   The following table shows the breakdown:
    % of time
Cult Suicides
   50%  (9/18)
Campaign Fundraising
   46%  (19/41)
Mideast Peace
   46%  (17/37)
Clinton's Knee
   42%  (5/12)
CIA Nomination
   36%  (4/11)
Midwest Flooding
   35%  (7/20)
OK City Bombing
   33%  (9/27)
   33%  (7/21)
   17%  (4/23)
   13%  (2/16)
All Stories
   27%  (170/633)
     We can speculate on the reasons for these findings.  Clearly, the Cult
Suicide story is a good fit for Internet news sites.  The cult itself operated
at least two web sites and many of the Cult Suicide stories in our sample
included links to those pages.  The reasons for other stories being such
"nonlinear favorites" are less clear.  The Campaign Fundraising controversy is
an on-going story and thus provides page designers with dozens of links to
previous stories.   The same is true for Mideast Peace.  It's not true for the
Clinton Knee story, but that event presented numerous angles for linkage.  It's
a medical story (focusing on the injury and operation), a sports story (focusing
on golfer Greg Norman who witnessed the injury), an international story
(focusing on the upcoming summit with Yeltsin), etc.  If Internet journalists
begin to focus on these linkage considerations, nonlinearity could begin to
drive the new medium.
 Interactivity versus Nonlinearity
     One additional finding that was not among of the original hypotheses
concerns the relationship between our two dependent variables, interactivity and
nonlinearity.  There is a strong relationship (r=.628, p<.05) between the
interactivity of a site and the total number of links on the front page.  This
suggests that sites aren't making a choice between interactivity and
nonlinearity, but are pursuing both (or neither).
Summary and Conclusions
      Media companies with Internet news sites are taking different approaches
to the new medium.  Newspaper sites have more stories on their front pages
(about 9), than television sites (about 4).  But television sites are paying
more attention to nonlinearity.  The typical television-run site has stories
with nearly 6 links, while their newspaper counterparts have just over 2 links
per story.  Our hypotheses that experience with the medium will lead Internet
news sites toward nonlinearity and interactivity could not be supported with
this "snapshot" study.  In fact, in this study, the newer sites rated higher,
particularly on interactivity.  A longitudinal study will be necessary to
discern a trend.
     This content analysis was used to determine how media companies are using
this new medium, not to measure their success at pleasing the Internet
"audience."  A study which measures the importance of interactivity and
nonlinearity to news consumers would be useful.
     _It was an assumption of this project that most media companies (in March
of 1997) with an Internet site were taking material from their newspaper or
broadcast newsrooms, and putting it directly on the Internet.  The editorial
content would then be similar to their traditional product. However, if
interactivity and nonlinearity begin to drive the medium, we would expect to see
content affected.
de Kerckhove, D. (1995).  The skin of culture: Investigating the new electronic
     reality.  Toronto: Somerville House Publishing.
Delaney, P., & Gilbert, J. (1991).  In P. Delaney & G. Landow (Eds.)Hypertext
     and literary studies.  Cambridge, MA: MIT Press.
McLuhan, M. (1964).  Understanding media: The extensions of man.  New
      York: McGraw-Hill.
Morris, M., & Ogan, C. (1996).  The Internet as Mass Medium.  Journal of
     Communication, 46(1), 39-50.
Newhagen, J., Cordes J., & Levy, M. (1995).  [log in to unmask]: Audience
     scope and the perception of interactivity in viewer mail on the internet.
     Journal of Communication, 45(3), 164-175.
Rafaeli, S. & LaRose, R. (1993).  Electronic bulletin boards and "public goods"
     explanations of collaborative mass media.   Communication Research,
    20(2), 277-297.
Stempel, G. & Hargrove, T. (1996).  Mass media audiences in a changing
     media environment.  Journalism & Mass Communication Quarterly,
    73(3), 549-559.
Williams, F., Rice, R. E., & Rogers, E. M. (1988).  Research methods and
     the new media.  New York: Free Press.

Back to: Top of Message | Previous Page | Main AEJMC Page



CataList Email List Search Powered by the LISTSERV Email List Manager