JBJohnson, publicdiplomacycouncil.org
What can public diplomacy staffers learn from a data scientist? A new book by a Google alumnus sheds light on how Big Data differs from opinion surveys, and how statisticians and social scientists find and use digitized public records to tease out insights on human behavior.State Department public diplomacy staffers are learning to pay attention to data, for two reasons. They want to focus on groups that are relevant to bilateral issues and objectives; and they want to know to what degree they are informing or influencing their audiences. So when Will Stevens, director of FSI’s Public Diplomacy Training, put the book on his summer reading list I decided to check it out.
Author Seth Stephens-Davidowitz titles his book “Everybody Lies.” Up front he takes us through a funny and often raunchy trip through the billions of things people type into their Google search box, which they would never reveal to a survey. “Providing honest data” is one advantage of Big Data, he asserts. Here are some others.
- “Offering up new types of data.” Google’s record of searches, and Twitter and Facebook are just for starters. There are vast collections of digitized books and newspapers. The author even found a digital audio recording of a large number of speed dates. These provide insights from sentiment analysis, studies of how words are used over time, and other techniques.
- “Allowing us to zoom in on small subsets of people.” Doppelganger research can search millions of medical records available to find people who share specific characteristics. It can predict, for example, whether a new treatment may help a specific patient.
- “Allowing us to do many causal experiments.” With so many data points on so many subjects, A/B testing can reveal which of many factors is the cause for a change in audience behavior.
Stephens-Davidowitz also covers things that Big Data is not so good for (predicting the stock market, for example) and things it should not be used for (spying on individuals who are not suspected of crimes.)
The book’s scope is very U.S.-oriented and cites almost exclusively sources in the English language. In much of the world, Big Data is less relevant. Moreover, as the author admits, data analysis by itself can sometimes lead to false conclusions, so the researcher is well advised to use methods like key-person interviews or focus groups in addition to computerized analysis.
The book’s key quote is:
“To squeeze insights out of Big Data, you have to ask the right questions.”
I would recommend this book for public diplomacy staffers who are working with State’s experts in Washington to interpret web and social media analytics plus a full range of other social science methods.
In a given situation Big Data may help, or it may not. It all starts with asking the right questions.
Everybody Lies: Big Data, New Data, and What the Internet Can Tell Us About Who We Really Are. By Seth Stephens-Davidowitz. 2017
No comments:
Post a Comment