A content analysis was used to measure the amount of media coverage of the

government, the dependant variable, as effected by the government's use of

deception, the independant variable. The goal is to analyze the communication and

public relations strategies used by the Department of Defense dealing with the Gulf

War syndrome.

Content analysis is the techique for making inferences by objectively and

systematically identifying specified characteristics of messages. There are two

types of messages: latent and manifest. Manifest content is the obvious messages in

a content. Latent content is the implied message within communication (Kaid

et al., 1989). This investigation focuses on the latent content to score the

data. The content analysis has a seven-step procedure that starts with the research

question or hypothesis (Kaid et al., 1989). This approach will be used to address this

paper's hypothesis.

Step 1

The research question is a broad subject suggesting a particular description

of the message. The hypotheses are directional predictions of an outcome in the

form of a question. The hypotheses are generated from theoretical perspectives or

are suggested in previous works or research. Such hypotheses describe the context

of the communications and allows the researcher to draw inferences between the

content of the message and the intentions of the sourceg (Kaid et al., 1989). The

specific hypothesis being addressed by this paper is the use of communicative

deception strategies by the Pentagon to deal with the problems arising from the

Gulf War syndrome.

Step 2

The second step in content analysis is the selection of the sample to be

analyzed (Kaid et al., 1989). The population was all television news media

in the United States. The sample taken from the population was the evening news

for ABC, CBS, CNN and NBC. The sample must ensure two things: representation of

the population and be efficient in size for reliability. Samples in media require

specific requirements to be tested. They must include circulation, region, size and

type. The pre-test included January 1995 through May 1996, and the post test

included June 1996, through December 1997. A split-half test was conducted on the

sample and was found to be reliable (Kaid et al., 1989).

Step 3

The next step is to define the categories to be coded. When defining the

categories, they must be exhaustive and mutually exlusive. Exhaustive means that

all possible characteristics of the category have been explored and mutually

exclusive means that nothing will fit into more than one category (Kaid et al., 1989).

In the area of latent content, the news coverage is either negative, neutral, or

positive towards the Department of Defense. The unit of analysis was a nightly

television news story. The information obtained for the content analysis is located

on the Television News Archives at Vanderbilt University. The sample was set for a

maximum of 100 results. When searching multiple years, the computer program

divides the maximum by the number of youars, i.e., 100/3 = 33 results per year

(rounding down). The search was conducted using the phrase "Gulf War

syndrome." This created non-related results such as "Restless Leg Syndrome" to

"Downs Syndrome" (Television News Archives, Vanderbilt University, 1998).

Step 4

The coders must be trained on how to code for the next step. Because of time

constaints, the coders were selected from this study group. The two coders were

trained on latent context of the forum: negative, neutral and positve. Additionally,

score sheet training and practical exercise was conducted with the two scorers. Had

this been a more involved study, coders not working with this project would have

been selected and used (Kaid et al., 1989).

Step 5

The implementation process is the next step (Kaid et al., 1989). It was

straightfoward with the coders working seperately and using the code book and

code sheet. No individual training was required. The coders were tested after the

first 10 samples using a split half reliability test and found to be 100 percent

reliable.

Step 6

The reliability and validity of the data is calculated for the next step. The data

was found to be 100 percent reliable after a split test was conducted. This is

different than the split test on coder reliability. Validity refers to the degree in which

a procedure produces genuine and credible information. Validity was not

statistically looked at because of the inexperience of the researchers and again, the

time constraints.

Step 7

The last step is to analyze the results. This step in a content analysis is usually

determined by the research question or hypothesis (Kaid et al., 1989). The purpose

of this investigation is to analyze the communication and public relations strategies

used by the Department of Defense dealing with the Gulf War syndrome. The

content analysis data was analyzed by comparing the pre-test and post-test.

Results

The results of the content anaysis was that the media coverage was clearly

against the government, expecially after the Khamisiyah incident. In the years from

January 1995 to May 1996, there were four television news spots: three negative,

none positive, and one neutral. The other news spots were unrelated to the Gulf

War syndrome. From the years of June 1996 to July 1997, there were 47 news spots:

29 negative, nine positive, and nine neutral. Again, the remaining hits did not relate

to the Gulf War syndrome topic (Television News Archives, Vanderbilt University,

1998).

Norm Brewer of the Gannett News Service said it would have been nice to

have more coverage early. Since the briefing on Khamisiyah, the weight of all the

news coverage resulted in the Pentagon and the Veterans Administration taking a

more serious look at what could be happening (as sited in McKenna, 1997).

Aggressive coverage by the news media would not have solved the mystery

surrounding the Gulf War syndrome, but the attention the media brought to the

table forced the government to take the situation seriously (McKenna, 1997).