Audience measurement on mobile and social media

I liked Barthel and Shearer's (2015) Pew article on Twitter. How Americans use Twitter for news is discussed but the in-depth exploration into how people use twitter, what they are talking about, who they follow, etc. can be applied to many other disciplines. For example, I took an in-depth look into the topics of conversation through data mining of tweets in a emergency health communication context. I would also like to do textual analysis of tweets in an advertising context in the future because I believe a study of consumer behavior is important. Studying consumer behavior has been limited and rather abstract in the past but advancements in technology allow us to actually observe and learn from real data.

The article mentioned a disappointing aspect of their twitter examination. The sample size yielded was too small. This is a challenge I encountered when I first started poking around with data mining. If there is not a controversy or crisis that is a top topic of discussion, it is unlikely that researchers will find enough data around a specific topic. If a researcher is interested in a particular aspect of advertising, let's say privacy, he/she will not find thousands of comments being generated about audience's privacy rights because it is a general problem and not a current hot topic. On the other hand, researchers can expect to get a lot of data around events such as the Ebola outbreak. Therefore, I agree with Barthel and Shearer that Twitter analysis can be disappointing and limiting. However, compared to the days prior to social media, Twitter allows researchers to observe consumer behavior in real-time which is valuable and extremely useful if used in the right way.