From Satira, Trevor, Guang, and Jeremy
We all know that the media greatly affects how a certain topic or event is perceived by the general public as it certainly did with the BP Oil spill. Our mission was to determine just how certain news agencies portrayed the spill differently. Our specific question was:
Following the initial reports of the spill, what were the differences in the facts reported by Fox News vs. MSNBC?
In order to answer this question, we were going to need to have the data from each source. But how do you get data from articles? We decided that there were three distinct tones that a news agency could project:
To be able to classify the different agencies into these categories we needed similar articles from each source. So we found timelines that included 1-2 paragraph summaries of each day during the oil spill crisis from Fox News, MSNBC, and UK’s The Guardian. (We decided to add The Guardian in order to have a source from outside of the USA to compare with the tones of the USA agencies.)
To classify the articles, it was going to take more than “a gut” feeling of how we felt they were toned. We wanted hard data. Therefore, we came up with 5 key words for each category so that we could count how many times those key words appeared in the article. By determining the amount of times they appeared we could then determine if the source was political, emotional, or “sciency”. The key words are as follows:
In the beginning, we predicted that Fox News would mostly discuss politics and would attack Obama. We expected MSNBC to praise Obama and to take sympathy for the region. We thought that The Guardian, being from another country, would leave out the politics and discuss the science.
Our results were as follows:
From the counts, we put together this graph. To ease confusion I’ll explain its construction. Every time one of the key words was found in say Fox’s time line we would make a tally. We found Obama’s name in the timeline 130 times. We found all of the key words a total of 458 times in Fox’s timeline. Because just using key word frequency would not make sense due to different length timelines we decided to take a percentage. So the 130 Obama’s was divided by the 458 total key words, yielding 28%. Therefore, the graph illustrates the different percentages of the total key words that a certain word was mentioned for each of the sources.
· Fox News focuses little on oil spill. Most of the discussion attacks President Obama. Only one to two sentences discuss the spill itself, and little science is discussed. As predicted Fox News was biased towards Obama.
· MSNBC is less political and more economic. Most summaries are quantitative. More science is discussed, but not much more. MSNBC was not as emotional as we expected and it was fairly neutral and discussed the three categories about the same amount.
· The Guardian was more well-rounded. Most of the summaries contained only updates on the status of the well and surrounding environment. The Guardian, surprisingly, was very focused on politics.
We have determined the pitfalls of this method and have put together a list of what we would do different if faced with a similar task:
· Include more key words
· Have one person doing the research. As humans we all view things differently when read.
· Include more sources.
In the end, we concluded that all of the considered agencies were politically motivated with little science communication. This is expected as mentioned by Jen many times in class, “It is very hard to take the politics out of science.” The mainstream media is more concerned with its ratings/viewership than with exposing the community to science.