After completing the process of completing a content analysis assignment, there are a number of things that I learned during the experience. First, I have discovered that in some cases it can be difficult to find the correct documents (in this case newspaper articles) needed to complete the content analysis procedure. Even though we as a class received specific directions on how and where to find the articles we wanted, I still had difficulty accessing Wall Street Journal news articles due to the need for a subscription. I eventually found access to relevant articles, but I didn't find them in the way described in class. In a real academic content analysis research study, I would have to follow exactly what the coding scheme says to get accurate and consistent results. Another thing I learned through this experience is how difficult it can be to code the tone of paragraphs as positive, negative, or neutral. This is an entirely subjective task and it's easy to see how there might be some disagreement in assigning scores between different programmers. What one person sees as a negative-toned paragraph, another person may read the same thing and see it as neutral. I spent quite a bit of time thinking about how to code each paragraph. Since the accuracy of the results depends on coding paragraphs with consistency across coders, I used my best judgment and tried to think about how others would also code each paragraph. I've also found that something as simple as determining what types of people (Republicans, Democrats, Unknown/Other) are cited directly in each article can be challenging. For example, one of the articles mentioned a person who worked in the Clinton administration. However, he did not directly say whether that person was a Democrat or Republican. Even if it can be assumed that the person was a Democrat, Republicans can also work for presidents who are Democrats. So I had to decide whether to count this individual as a Democrat or count him in the “Unknown/Other” category. This demonstrates how something that seems simple can be difficult to determine when conducting a content analysis. If as a class we collected all the articles from the four selected journals regarding our topic, coded them, and generated a database, there are a number of research questions and hypotheses we could formulate and test using this data.