Media Content Analysis, Article Review Example

A 2010 study by Tumasjan and colleagues examines the prevalence of political sentiment on Facebook, as well as the possibility of using this information to predict elections. The benefits of an online presence to political campaigns have been apparent for several election periods, but it has only been recently that social networking sites (Twitter, Facebook, etc.) have been prominently used in the process. This study examined three aspects of tweets as they relate to political influence using a text analysis. These aspects include indications of political information exchange, validity of displayed political inclinations to real life, and to evaluate the potential use of Twitter, and similar sites, to predict political races. Over 100,000 tweets about the 2009 German election were analyzed, resulting in support for the prevalence of debates, as well as a close correlation of technologically projected political ideals with those held offline. The ability to predict elections through this method is less likely, due to the participation of only a small percentage of the voting population in social networking discussions.

Another 2010 study, this time by Chew and Eysenbach, utilized the surveillance of Twitter to examine the usage of the terms “H1N1” and “swine flu”, to evaluate the content of such tweets, and to test the validity of Twitter as a source of information for the immediate analysis of public sentiment. The statistic results of this research, obtained from testing the information extracted from over two-million archived tweets, support the value of extracting data from online social networks as a means of real-time public trend analyses. The researchers were able to demonstrate a significant increase in the use of the term H1N1 after the WHO recommended it in opposition to the more disturbing swine flu moniker, and that the content of tweets related to the subject was mostly helpful to the dissemination of health information.

Fernandes and associates, again in 2010, focused their study on information in Facebook groups. A relatively small number of group pages were evaluated, and all were comprised of students from universities throughout seven key states. Discussion in these forums was based on the 2008 US presidential election, and included an approximately equal number of posts supporting both John McCain, and Barack Obama. Content analysis revealed a wealth of information, such as a higher rate of activity among Obama supporters, the tendency for McCain supporters to focus more on long-term topics, during primary season, and the importance of several issues (such as policy, civics, campaign material acquisition) in discussions.

Lastly, a 2002 article by Pennebaker and Lay analyzed information from 35 press conferences over an eight year period to examine the supposed change in personality exhibited by then mayor, Rudolph Giuliani. It has been suggested that Giuliani underwent significant changes in demeanor throughout his mayoral career, possibly due to personal crises, as well as the 9/11 attacks. Language analysis of speech material, as well as subjective but consistent qualitative judgements, revealed that Giuliani did indeed undergo significant shifts in the exhibition of empathy, emotions, and cognitive complexity.

The design of these studies could be build upon by including multiple sources of online social discussion, rather than focusing on a single platform. This method would provide insight on the potential differences between sites (or sections within sites), using similar coding to the above studies for linguistic analysis. Additionally, correlations may be calculated with analyses from more traditional forums of discussion, like print and radio.