CAMPAIGN BRAIN DRAIN: DESCRIPTION OF RESEARCH METHODS
Textual analysis
During our research, we applied textual analysis method to all sources that we found. This involved asking the correct questions such as is this source primary or secondary, what article is talking about, what is the author’s personal position, is this a reliable source or not, and etc. As a result, this guided us to limit our research material and to focus only on what can be used to support the team’s argument. In some cases, textual analysis helped us to see that some sources do not provide objective and reliable information. In other cases, the method revealed that authors focus on meaningless and trivial information. So, our research will be based on the reliable and unbiased material.
Data mining and analysis
As we know, candidates make a lot of statements and promises in their speeches during their presidential campaign. While doing our research, we came across an archive that contains all speeches that Obama and Bush made during their presidential run. It is a long list of speeches and it would be difficult for us to analyze it in short time period. So, we used data mining application to analyze the large amount of text and identify a pattern. Specifically, we used Voyant (http://voyant-tools.org/) word cloud software in order to generate word cloud of these speeches. It is a great tool to analyze a speech because it immediately shows the key words that are frequently used in the text. Word cloud helps to effectively prove our argument by visually showing the pattern in the speeches of both candidates.
Visual and Aural artifacts
Artifacts such as photo images, videos and audio are very effective in supporting our argument. Just like transcripts of speeches, video recordings of candidates’ public addresses are considered to be primary sources. Videos that our group found on YouTube and C-SPAN will be instrumental to bring attention to the specific statements and promises the candidates made. It’s one of the most effective and reliable proves that later cannot be denied by either candidates or media.
Spatial history
Finally, we came across the detail information about what states candidates visited during their campaigns and how often. We are going to plot this data using Fusion Table and create a map that will visually present all US states visited by candidates using color gradation and label them with the number of times visited. This information emphasises the important role these states play in the final results of the presidential race.
Thanks for laying this out, Brain Drain.
There is no need to entirely rule sources out because they are biased. You just need to handle those sources with special care, taking your understanding of bias in them into account. Given your subject, campaign promises, you will be hard pressed to locate primary sources that are truly unbiased. It is likely to be the most biased documents you find in this realm that tell you the most. As Luke pointed out in his comment on your bibliography, it is essential that you answer the “so what” question — what do differences and similarities in campaign promises mean? What do they tell us about the historical moments in which they were communicated?
Regarding data mining: How many clouds will you be generating with Voyant? For the purpose of analysis, you likely need many of them broken down in different ways. Otherwise you will likely detect only large-scale patterns at the most general level. I strongly recommend disaggregating the information and running some more focused analyses. You can group together speeches depending on factors such as location of the speeches (have you given consideration to this approach since we discussed it in class?), audience demographics, medium used to deliver the promise, etc., and then look for resulting patterns.
Visual and aural: How will you be analyzing these sources? As we have studied in class, they need to be “read” in a different way than text sources. What in particular will you be looking for?
Spatial: You can bring much greater meaning to the patterns of where candidates traveled if you find a way to link those patterns to the content of their promises. This will set you up to draw more informed conclusions about the process of promise-making.