An example of content analysis as a research method is demonstrated in the work of Pfau, et al. (2005), in which they review the effect of the embedding of journalists in the military units involved in the invasion and occupation of Iraq on the tone and message displayed in television news reports. Content analysis was conducted of television news reports aired on four major networks (ABC, CBS, CNN and NBC), and the results were compared based on whether the coverage for each report depended upon embedded or non-embedded journalists.
Pfau, et al. (2005), indicated that embedded news sources displayed an overall tone that was more favorable toward the military in general and depicted military personnel in a more favorable light. Embedded sources tended to use greater levels of episodic frames, eliciting more positive relational cues. By contrast, unilateral or non-embedded sources tended to minimize the use of episodic framing, and tended to depict the military as a whole, as well as individuals in the military, in a less positive tone. As an aside, the study also found uniform indications that the news stories of the initial invasion and early occupation were more positively reported and employed greater episodic framing than the stories about the Iraq occupation as it drew on.
Shawver (2006) defines content analysis as a procedure that studies natural communications, usi...
questioning, advising, informing, etc.) and in the tone and structure of the presentation. Multiple systems can be applied to the same data over time, as well, allowing additional analysis and investigation.
A striking example of a longitudinal study can be found in the work of D'Angio et al. (2002), in the 15 year follow-up of children born severely premature and treated with surfactant therapy. An initial pool of 213 infants born at ages prior to 29 weeks of gestation across a three year time span (1985-87) had shrunk to a field of 132 of which 127 were evaluated cross a three year span (1992-95), and 126 were re-evaluated in 2000. At each point, the child's physical and mental conditions were evaluated as was the amount of neurodevelopmental compromise.
D'Angio, et al., indicated that 21% of the 127 evaluated at primary school age showed symptoms of at least one severe disability, while 32% were placed in special needs classrooms and 19% showed general cognitive indices below 70. By the time of the secondary school-age follow-up evaluation 41% of the children had no physical or educational impairment, whereas 19% showed at least one severe disability. 15% of the children at both evaluation times were diagnosed with cerebral palsy. D'Angio, et al., concluded that even with surfactant therapy, children born prematurely remain at high risk for neurodevelopmental compromise and significant percentages of these children will require special educational services. Of added note is the fact that the strongest predictors of adverse outcomes related not to the gestational age at birth, but rather to socioeconomic status and presence of neonatal intraventricular hemorrhage.
Keppel, Saufley and Tokunaga (1992) define a longitudinal study as a form of research