Military Review English Edition November-December 2015 | Page 24

his or her views about the candidate’s political party
from personal experience and knowledge. That is, if
individuals do not complete their information pictures
from vertical sources, they must get information from
other sources, namely personal interactions with others, to paint a complete personal picture.
The information from all three sources accounts for
100 percent of what individuals know, or 1.00 in the
ACA formula.
To estimate the relative power of vertical media,
horizontal media, and personal experiences and perspectives, the vertical media correlation is subtracted
from the value of 1.00 (which, by definition, accounts
for everything individuals know) to get the horizontal
media and personal portion of agendas. This figure is
then squared to account for variance. The act of squaring allows formula users to weigh them proportionally.
The square is then added to the vertical media correlation squared and the results subtracted from 1.00. The
resulting number is the ACA. Three examples follow
that illustrate how the formula works.
In the first example (see figure 4), if a public opinion
poll indicated a high correlational agreement between
the vertical media and general public at .80, that
would mean up to .20 of media influence came from

elsewhere, probably from horizontal media, other people, or personal experience. These are assumptions of the
formula that attempt to account for all informational and
experiential knowledge. Thus, the ACA formula for the
group represented by the .80 correlational poll would
look like this equation:
ACA = 1 - [(.80)2 + (1 - .80, or .20)2]
ACA = 1 – (.64 + .04)
ACA =.32
Even with this relatively high hypothetical correlation
(ρ = .80), there is evidence that horizontal media and
personal experience play a role. If, for example, the correlation between traditional media and audiences were
a perfect 1.00, then the traditional media’s agenda would
determine the issues their audiences think about. The
ACA formula would result in zero. That would mean
that if audiences knew what traditional media were saying, analysts could predict what audiences would regard
as important. In fact, that would seldom happen. This is
not North Korea. It is not 1984.14
In the second hypothetical example (see figure 4), let
us imagine the agreement is
.50. In this case,

1.0

ACA = 1 – [(.50)2 + (1 –
.50, or .50)2]

Correlations

0.8
0.6
0.4

Example 1

Example 3
Example 2

ACA = .50

Example 1

Example 3

0.2

Example 2
Example 3

Example 1

0.0

Dominant community
stability

Transitional community

Vertical
agenda

Horizontal
agenda

ACA= 1- (.25+.25)

Alternative community

Finally, in the third
example (see figure 4), if
the traditional correlational agreement was low, at
.20, then,
ACA =1 – [(.20)2 + (1 .20, or .80)2]

Personal
preferences

ACA = 1- (.04 + .64)

Figure 4. Dynamics of Agendamelding and Civic Balance
22

ACA = .32

November-December 2015  MILITARY REVIEW