Pag 63

The data collected is then printed and the messages numbered. The final count was 310 messages. The printed form helps with a global understanding of the content. After many readings, the messages are analyzed and coded accordingly. The sub-dimension parasocial interaction model proposed by Sood & Rogers (2002) is the platform for the analysis. Each post is coded according to one of the sub-dimensions, as discussed previously; in some cases the messages exhibited more than one sub-dimension. For this academic exercise the messages are coded according to the category that is more strongly emphasized. The intricacies of the messages will be explored in the qualitative analysis results that follow this chapter.

After the messages are coded, and results tabulated, the researcher goes back to the research and recods it. Another set of data is printed to make sure the researcher is not influenced by the first code analysis. This procedure is performed in order to answer questions of reliability, the researcher’s questioning inclusive, since a different coder was not used in this research. The recode has a 90% percent agreement, which certainly reassures the reliability of the analysis. According to Stempel III & Westley (1981), the recoding and the account of the agreement in a percentage form are appropriate and acceptable for reporting reliability. Content validity of the content analysis categories is piloted tested by individuals that have some experience with the telenovela issues and are former viewers of the telenovela themselves (Law, 2000).

Table of Contents

CHAPTER IV - THE STUDY
47-48-49-50-51-52-53-54-55-56-57-58-59-60-61-62-63-64

Top Cities in Brazil for Tourism: