A study testing whether or not Google News is biased was recently released. The study found that Google News does not show favoritism, but may contain some residual biases.
The study was conducted by Online Journalism Review and tried to determine whether or not Google News had any political biases in their method of posting news stories. The methods they used to conduct this story are, in a word, surprising.
In order to derive their hypothesis, during the 2004 election, the OJR studied news articles related to two keywords phrases only: George W. Bush and John Kerry… which is most certainly not a very broad method to conduct a study of this nature, as pointed out by Nathan Weinberg and Danny Sullivan.
After they conducted the keyword searches, OJR studied the selected articles sentence-by-sentence in order to determine whether a bias existed. As Danny pointed out, because Google uses “non-traditional” news sources in Google News, sometimes the news reflected the views of the person who was writing the article. This “bias” was found to be true for both sides of the politcal fence.
Danny extended his evaluation quite extensively (this includes keyword searches of his own to practice OJR’s method; Danny includes these results) and found that while the test was indeed interesting, it certainly was nowhere near conclusive.
When you look at the methods used, it’s quite easy to see how he drew this conclusion. If you are testing volatile keywords on a news engine that allows opinionated publications to appear, of course there is going to be a bias… and it will likely be representative of both political views. Had OJR studied Google’s method of news collecting, the rest of the study would have been unnecessary because the conclusions would have been a given.
Perhaps Google should be acknowledged for allowing opposing and opinionated viewpoints in such a popular news resource.
Chris Richardson is a search engine writer and editor for Murdok. Visit Murdok for the latest search news.
