Discussion

There are two major findings from the pilot survey. The first finding is that the base newspaper is the most significant and credible source for service news and community events. The base newspaper has no role as a source for national and international news, that news source preference is the Internet, television, radio, or local news. Based upon frequency analysis, the second finding is that the Internet is the second primary source for base news.

The theoretical implications of these findings may be explained using a variety of theories. The most significant implications come the dual processing model, which includes the Elaboration Likelihood Model (ELM) and the Heuristic-Systemic Model (HSM). These two models for cognitive information processing say that people with high issue involvement will actively seek information. By definition, ELM and HSM explain that certain messages will be centrally processed such as those in print media and on the Internet will have high issue involvement. Based upon the findings, the survey's sample population seemed to have high issue involvement in all five sources areas: base news, service news, community events, local news, national and international news.

Focusing on the effectiveness of print newspapers to reach the target audience for this study (18-26 year old military), the survey findings were the exact opposite of the expected results proposed by the research team going into the project. The team anticipated that the target audience would prefer other communication media, particularly the Internet, television, or radio, as their news source. The ranking of the Internet as the best source for national and international news is not surprising given the immediacy of the media and the greater availability of access to computers in homes, schools, libraries, work places, and other public domains. People tend to read information on the Internet like they read printed material, so survey results noting the Internet as the second significant source for information makes sense.

In the future, this project could be reworked to include analysis of the survey and its limitations. This project's survey needs to be considered a pretest. The revised survey could be administered to a larger, random sample population to see if more significant findings are identified and if the findings from this project can be repeated. The sample population needs to be expanded to include military bases with strong command access television programs and that use the Internet to disseminate local information. Overseas bases need to also be included because of the unique communication challenges they experience in various parts of the world. At the overseas installations they have local information provided by television, radio and newspaper, which will be effective in determining which of those modalities seem to be most effective in providing command information.

Future researchers need to find a more hassle-free environment for administering the survey. One possible idea is to get permission to administer it to entry and lower level enlisted and officer basic courses at the different military schools or in some type of controlled audience that has the time to take the test. The mall of the Base Exchange was not conducive to a multi-question survey like the one for this project.

Adding research of the current DoD public affairs training would expand the applicability of this study by adding a resource training and allocation dimension. This research could add validity to the current pervasive opinion in the public affairs field that DoD public affairs training is concentrated in print media. A related claim is that public affairs personnel are allocated heavily in print communications. Substantiating these claims is important with the diminishing financial and personnel resources of today's DoD world. If future research provides support for these claims, what impact do they make on the news products offered or provided by each command?

Some limitations were identified during the short time allotted for this project. The lack of survey design knowledge by the research team members was a significant limitation. Precious time was lost during the search for staff with the requisite expertise to guide the team.

The survey used in the pilot study did not make effective use of the variable options. Prior to embarking on further uses of this survey, all independent and dependent variable possibilities should be clearly identified. Along with the identification of the variables, a research plan needs to outline potential data sets for analysis.

The requirement to improve the primary source question design is imperative. The primary source for news was the independent variable in this study. Instead of limiting available responses to source identity alone, the research team recommends dividing the question into separate, scaled questions representing each news source. By separating the questions, the researcher has the opportunity to make more comparisons of the survey data. Within these separate questions, television and radio need to be separate news source choices, because people interact with and learn information from them differently.

Each section of the survey needs to have a question to identify the level of attention paid to primary source for that news source. In order to collect more accurate information, it is necessary to measure both the exposure to and the amount of attention paid to that media source. In the pilot survey, this "attention paid" question was asked only in the base news section. Limiting the "attention paid" question to base news prohibited further research in the area of attention and affected the span of data comparison's for the other four news areas.
Some errors in survey's text went undetected prior to administration. The identified errors must be corrected before the survey is administered again. A good tool to catch survey errors would be to administer the revised survey to an internal test group to determine readability and accuracy, before administering to the desired sample population.