Restricted access to news organizations’ archives limited our ability to gather quantities conducive to scientific analysis. Cost of archive samples was a limiting factor as is the fact that English translations of Arabic news publications were not available in sufficient quantity, nor were past issues archived. While these limitations brought frustration and delays to the initial plan, we believe that they did not impact the overall ability to reject null hypotheses.
The Lexis-Nexis database provided a variety of world news resources, but articles were excerpts of original stories and thus coding is subject to some pre-interpretation. We did not know what part of the story we weren’t seeing, and that determination was made by a Lexis-Nexis employee. Excerpts ranged from 100 to 500 words, and provided a suggested headline along with the body of the story. Dates from the actual dateline sometimes differed from the date of clipping by one day. While the original intent was to solicit three regional newspapers, news services that included the Associated Press for the United States and both Agence France Press and Deutsche Presse-Agentur for France and Germany respectively, were used due to limitations discussed previously.
The three time frames selected, while pertinent to military strategic operations and United States print news coverage, did not pertain specifically to the non-linear element of the Middle-Eastern news cycle. Activities such as protests, which generally occur on weekends, would not have been mentioned in daily news the following week. The cycle of news in the Middle-East, it was discovered, is separated more by category each day, than by the timeliness of the news. Because of this fact, about half of the non-United States research samples were discarded within each time period because of a lack of pertinence to the questions or because the article did not fall within the established time frames.