How was the book?

What a great book! Seth Stephens-Davidowitz really has a point and it is that social science is becoming real science with the help of data science and Big Data.

The book is almost hair-rising experience and an eye-opener, because Stephens-Davidowitz showcases with the help of data from Google and PornHub how Big Data turns aggregated data into action.

What are the key learnings of the book? 

We lie to people, we don’t tell the truth to surveys and in our Facebook or Instagram accounts we might show alternative facts, but for Google Search nobody lies.

The goal of the book is to provide missing evidence of what can be done with the Big Data and demystify data science. The book is built around three major topics or three powers of Big Data:

·      Data reimagined = Offering new types of data is the first power of Big Data.

·      The Digital Truth Serum = Providing honest data is the second power of Big Data.

·      Zooming in = Allowing us to zoom in on small subsets of people and allowing us to do many causal experiments.

In digital age we still hide our embarrassing thoughts, but not from internet and Google searches. Data science is about spotting patterns and making predictions. And for Big Data you don’t need tons of data – you need right data. Data size, by itself, is overrated. People lie to people, but not to Google Search and because people are honest you can get also narrow data! Due to these facts with data analysis from Big Data you are most probably searching for causality or correlation.  

Case New York Times. Stephens-Davidowitz brings up also the sentiment analysis and uses New York Times as an example. NYT has shared an insight that people typically share positive articles. So positive stuff gets viral. The old journalist adage says that ”if it bleeds, it leads”. The new adage could that ”if it smiles, it’s emailed”.

During his research Stephens-Davidowitz developed two cosepts – digital truth serum and lies:

·      Digital truth: searches, views, clicks, swipes

·      Digital lies: social media posts, social media likes, dating profiles

Internet does not generate a political bubble – IRL does. Internet generates more political desegregation than your everyday life. This happens in news sites as well in Facebook. Internet brings people with different views together. On social media sites we don’t have an incentive to tell the truth. We how our cultivated our political views for example in Facebook. With the help of digital truth serum the Internet helps us, because there we are not alone with our insecurities and embarrassing behavior. Internet also provides peer-support.

Some data points from the book:

·      Why people are smiling so much in Social Media? Because Kodak started to advertise that taking pictures is associated with having a good time.

·      A/B testing…. 80 % of all tests fail. Only 20% prove some issue.

·      Google searches related to crimes correlate with actual crimes. Murders of Muslims 1 to 10.000 ”kill Muslims” searches. Data could be used to predict certain behavior.

·      Only 7 % of people finish Kahnemans ”Thinking, fast and slow”.

How should we change according to the book?

Big Data truth: ”Never compare your Google Searches to everyone else’s social media posts.” Google Search:

·      Can alert us to people who are suffering.

·      It has the ability to lead us from problems to solutions.

What should I personally do? 

Imagine my own Big Data dream.

Summary

The book in six words – ”If it smiles, it’s emailed”.