You need to sign in or sign up before continuing.
Take a photo of a barcode or cover
662 reviews for:
Everybody Lies: Big Data, New Data, and What the Internet Can Tell Us About Who We Really Are
Steven Pinker, Seth Stephens-Davidowitz
3.82 AVERAGE
662 reviews for:
Everybody Lies: Big Data, New Data, and What the Internet Can Tell Us About Who We Really Are
Steven Pinker, Seth Stephens-Davidowitz
3.82 AVERAGE
Interesting look at what big data can tell us but with a strong focus on dark or sensational topics - porn, sexuality, racism, violence. While he talks later in the book on differentiating correlation vs causation, it would have been better done upfront as I’d worry about how some of his data and points could be misinterpreted.
This book extrapolates that Google searches reveal "how people really think". People do more searches when they have a problem or are dissatisfied. If there are more searches for dealing with unhappy marriages, that is partly because people in happy marriages do not need to do Google searches on their marriage.
However, this author ignores Occam's razor and instead concludes that he is the only genius to found the hidden meanings. His paper is repeatedly rejected by peer reviewed journals and instead of taking lessons on being more fact based, he concludes that others cannot appreciate his unique insights.
However, this author ignores Occam's razor and instead concludes that he is the only genius to found the hidden meanings. His paper is repeatedly rejected by peer reviewed journals and instead of taking lessons on being more fact based, he concludes that others cannot appreciate his unique insights.
I enjoyed discovering 1/ the datasets he used and 2/ the findings from studies he cites. The main takeaway is that internet data (search data, most notably) can be used to learn about the preferences, thoughts and desires of groups of people with more accuracy than via traditional studies or questionnaires.
If you're in the data science field, some ideas won't be new to you (correlation vs causation, omitted variable bias, sampling issues).
If you're in the data science field, some ideas won't be new to you (correlation vs causation, omitted variable bias, sampling issues).
An interesting book that brings up a good point about different data sources to mine. But this book also demonstrates the danger of bringing preconceived ideas into the analysis. The author's conclusions could be the right ones, but he doesn't spend much time talking about alternative ways the same data could be interpreted. So to me, this book becomes an entertaining but ultimately light way to look at google searches.
informative
medium-paced
Entertaining, but didn't change my mind regarding social sciences as soft sciences.
For social sciences to have the same respect as physics, biology, and mathematics, their conclusions must be repeatable. If anything, Everybody Lies shows that many of the conclusions are only correlated for that moment in time. Ten years from now, those conclusions won't be relevant. Just like much of [b:Freakonomics: A Rogue Economist Explores the Hidden Side of Everything|1202|Freakonomics A Rogue Economist Explores the Hidden Side of Everything|Steven D. Levitt|https://images.gr-assets.com/books/1550917827s/1202.jpg|5397] has aged out of favor, most of Stephens-Davidowitz's conclusions using google data will follow suit.
So why the 4 stars? I was thoroughly entertained! How could I not be when there's a book about using PornHub data?
For social sciences to have the same respect as physics, biology, and mathematics, their conclusions must be repeatable. If anything, Everybody Lies shows that many of the conclusions are only correlated for that moment in time. Ten years from now, those conclusions won't be relevant. Just like much of [b:Freakonomics: A Rogue Economist Explores the Hidden Side of Everything|1202|Freakonomics A Rogue Economist Explores the Hidden Side of Everything|Steven D. Levitt|https://images.gr-assets.com/books/1550917827s/1202.jpg|5397] has aged out of favor, most of Stephens-Davidowitz's conclusions using google data will follow suit.
So why the 4 stars? I was thoroughly entertained! How could I not be when there's a book about using PornHub data?
Overall, I found this book to be an enjoyable and fairly informative introduction to "Big Data" and why data analytics matters and will continue to be absolutely vital to the social sciences (and all other sciences and fields of study, really). The author's main focus of study was data from Google searches and also the website PornHub. If you're squeamish about human sexuality then I would not recommend this book, but if such things interest you (and other crazy truths about humanity and our country that can be revealed from internet searches), then I definitely recommend it. I can also see fans of Malcolm Gladwell enjoying this book. I know that the author kept a lot of the details of the math/stats out of the book to make it easier for a general audience to read, but every now and then it seemed to me that point A did not connect to point B and I was left wondering exactly how he reached a certain conclusion. However, this is probably me being nit-picky. Anyway--I laughed. I gawked. And I seriously have to hand it to data scientists that theirs is probably the most crucial field of all right now and in the future.
The book surely provided an overview on why people should look into big data, however the practical takeaways were missing for me.
Not bad, an interesting overview of what's going on with the analysis possible with some of today's big data sets. There are some stylistic quirks, most notable in the conclusion, which is hopelessly meta. Just write your book. There's also a bit of the usual cishet normativeness of these techie books, unfortunately. The author seems possibly aware of the unbiased perspective but doesn't quite achieve it.
Subjective unsubstantiated garbage speculation