Frequency Counts in Content Analysis

I wanted to add to our group’s discussion about content analysis. As another blogger mentioned, the content analysis approach is systematic and quantitative and therefore presents data that is explicit and transparent. For example, one of the techniques used in this approach involves counting the frequency of recurring content to find patterns and themes. The example that Thomas provides for this practice involves counting motor vehicle accidents in fiction programs (1994, pg.691). When I read this, I was reminded of a media study that I read as an undergrad, and how it tried to support these broad claims through the content analysis of something very specific.

If I remember correctly, the study tried to make a link between the sexual desire of straight men in North America and the number of female redheads that are used in shampoo commercials. I don’t remember what the conclusion of the study was but I remember feeling very sceptical about the results. While it is obvious that the research design was flawed, it left me with a bias against content analysis as a stand alone research method because of its potential to be reductive. I have been thinking about my research project and whether content analysis would be appropriate and I personally feel that it is best used in a mixed methods approach.


1 comment:

Stephanie Lauren said...

I'm reminded of something that Luker has emphasized repeatedly in her book - that social scientists use research methods to help them develop theories. Social scientists don't approach their studies with a particular hypothesis in mind that they are setting out to prove with experiments; instead, they are interested in a particular social situation and want to research it better so that they can begin to establish correlations. Once you compile some research and test your methods, you begin to see flaws in your work and can go back and repeat the process until you've soundly operationalized your concepts and study methods.

I feel as if content analysis can be an important part of this process and that it is possible to use it as your main method. However, it needs to be practiced in a robust manner just like any other research method.

As you mentioned before, the study on shampoo commercials seemed flawed anyway and obviously drew a pretty large correlation based on one type of artifact. I agree that this seems to be a problematic aspect of content analysis, as some researchers may try to draw larger conclusions from their research when the proof isn't really there. Perhaps perusing the case studies for this week will help you change your mind :)

Post a Comment