How I chose methodology for my research

This week I was trying to finalize the proposal for the study. I was reading some of the studies that Becky recommended me to: on content analysis, and discourse analysis. It was very useful for me to get better understanding of the method which I decided to use in my study. First, I realized that content analysis is not really such a good method to begin with ;) and that was kind of frustrating for me. Well, I knew before that content analysis is more about identifying, counting and then providing the analysis of what you found. Content analysis does not really provide any depth, you cannot see any kind of relationships between the patterns present in the text...:( Content analysis does not really look at the meaning side of the written communication, sometimes just scratching the surface. At the same time, content analysis is probably more objective, if a researcher cna actually be obejctive entirely, regardless of the method he is using.

So, I had to make a decision- which method shall I choose, shall I do discourse analysis rather than content analysis, or I still choose content analysis.

There are several things, which finally helped choose the method:
first, the main focus for this research is to identify the most common topics the financial bloggers discuss in their blogs.
One of the parts of my main research will be - most common topics discussed by financial bloggers. If we look at the literature, there are no studies that looked at the topics of the financial blogs. Basically, there is very little studies on financial blogs in general.

I was looking at how people idnetified the topics discussed by other blogs, for instance, library blogs. What they did - is a content analysis.

So, that made me thinking, that probably content analysis is the methodology which I should use to achieve the goal of my research. I decided not to use any of the computer software to identify the most common topics, I will be reading the posting myself.
Also for reliability I will have another coder to code at least 10 percent of the postings.