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
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
Seth Stephens-Davidowitz
3.82 AVERAGE
4.5 stars. After reading Freakonomics, the author decided to become a data scientist and loudly shouts that everyone lies.
In surveys any way. The traditional way to figure out what a large group of people will do is to sample a small group of people. This methodology fails because people lie. They are more truthful with computers (namely Google and Porn Hub) and in this world of big data, the lies the truth.
Seth really has fun digging in big data and finding some entertaining facts. I look forward to digging into google trends myself.
The last part of the book dealt with the dark side of big data which really made me think. The beginning of this book, Seth showed me how to use these new Google tools making me want to see what I can find myself ... the last part of the book made me pause because I'm not sure what evils I may unearth. Maybe 1984. Maybe a balrog.
In surveys any way. The traditional way to figure out what a large group of people will do is to sample a small group of people. This methodology fails because people lie. They are more truthful with computers (namely Google and Porn Hub) and in this world of big data, the lies the truth.
Seth really has fun digging in big data and finding some entertaining facts. I look forward to digging into google trends myself.
The last part of the book dealt with the dark side of big data which really made me think. The beginning of this book, Seth showed me how to use these new Google tools making me want to see what I can find myself ... the last part of the book made me pause because I'm not sure what evils I may unearth. Maybe 1984. Maybe a balrog.
Everybody Lies is really interesting - what we can learn from Big Data seems to be almost limitless (except for anything about the stock market, as Stephens-Davidowitz tells us). In particular, it's kind of fascinating to explore what you can learn from people just by considering what they search for on the Internet. That said, this book doesn't go much beyond identifying the subjects you can learn about and providing interesting factoids. With a sub-title that promises to "tell us about who we really are", I think I was just expecting more.
Sorely disappointed that for a book from an author that claims that big data and data analysis is the future of falsifiable research in the social sciences completely neglects the ethics of the research he discusses in his book. A couple times he brings up bias or ethics, such as when discussing natural datasets and why it's impractical to assassinate world leaders for the purposes of A/B testing outcomes. He talks about how big data research doesn't have any of the problems that typical research does. IRBs? Who needs those when you have Internet data! What would I have wanted to add? A discussion of who is missing from research conducted using data from websites, a discussion about whether or not it's ethical to collect and store that kind of data without the consent of users, and then adapting it for second party uses to perform research, a discussion about who has access to these datasets... I could write a longer list but those are the highlights. This falls into the pile of "yet another big data book that insufficiently addresses ethics and data bias" unfortunately.
funny
informative
inspiring
reflective
fast-paced
I recommend this highly with a couple of caveats.
The central insight of this book is that you can get a better idea of what people actually think, despite what they say to others (or even to themselves) by looking at Google and Pornhub searches (among other anonymized big data sets). Things that people won’t admit to other people (thoughts of suicide, to whom they are attracted, homicidal thoughts, racist thoughts, dissatisfaction with a marriage, regrets over having children, etc.) are often clearly revealed to Google through searches. This far I’m definitely willing to go, and would recommend the book on that basis. (Unsurprisingly, Facebook is a far less reliable source for honest data sets than google or pornhub.)
Big data can tell us a lot, but there are limits. In one sense it’s a bigger tree from which to cherry pick data that might support your view. While Stephens-Davidowitz is very open when results didn’t match his expectations, there is a strong ideological bent in which data he chooses to explore. I wish he’d been upfront about that at the beginning. I don’t object to the ends to which he’s working (I agree with most of them), but I wish he’d put in a disclaimer (like Stephen Jay Gould did at the beginning of Mismeasure of Man). He did put in a chapter at the end about what Big Data can’t do, but since it’s at the end, I think it’s less effective (all the more so since one of his findings is that people are less likely to read the final chapters of a book than the early ones).
His conclusion is an argument that Big Data is the best shot for the social sciences to officially move into the realm of “real” science, even claiming that if Karl Popper were alive he’d likely be convinced. I’m not quite so skeptical of the social sciences as Popper was, and certainly no science, even “hard” science, is completely free from the biases of the scientist. But there is definitely more room for ideology to skew results in the social sciences. Stephens-Davidowitz makes a good case that big data (in combination with smaller data) can significantly strengthen the reliability of results. He convinced me that it does, but not to the extent that he believes it does.
On the whole, this is an excellent book. Read it with a bit of skepticism, and it will give you a lot to think about. 3.5, but I'm rounding up.
The central insight of this book is that you can get a better idea of what people actually think, despite what they say to others (or even to themselves) by looking at Google and Pornhub searches (among other anonymized big data sets). Things that people won’t admit to other people (thoughts of suicide, to whom they are attracted, homicidal thoughts, racist thoughts, dissatisfaction with a marriage, regrets over having children, etc.) are often clearly revealed to Google through searches. This far I’m definitely willing to go, and would recommend the book on that basis. (Unsurprisingly, Facebook is a far less reliable source for honest data sets than google or pornhub.)
Big data can tell us a lot, but there are limits. In one sense it’s a bigger tree from which to cherry pick data that might support your view. While Stephens-Davidowitz is very open when results didn’t match his expectations, there is a strong ideological bent in which data he chooses to explore. I wish he’d been upfront about that at the beginning. I don’t object to the ends to which he’s working (I agree with most of them), but I wish he’d put in a disclaimer (like Stephen Jay Gould did at the beginning of Mismeasure of Man). He did put in a chapter at the end about what Big Data can’t do, but since it’s at the end, I think it’s less effective (all the more so since one of his findings is that people are less likely to read the final chapters of a book than the early ones).
His conclusion is an argument that Big Data is the best shot for the social sciences to officially move into the realm of “real” science, even claiming that if Karl Popper were alive he’d likely be convinced. I’m not quite so skeptical of the social sciences as Popper was, and certainly no science, even “hard” science, is completely free from the biases of the scientist. But there is definitely more room for ideology to skew results in the social sciences. Stephens-Davidowitz makes a good case that big data (in combination with smaller data) can significantly strengthen the reliability of results. He convinced me that it does, but not to the extent that he believes it does.
On the whole, this is an excellent book. Read it with a bit of skepticism, and it will give you a lot to think about. 3.5, but I'm rounding up.
informative
fast-paced
This reads similar to Freakonomics where there is a barrage of interest data tidbits. I did not find anything revelationary in the novel and found it leaned too much on Google trends.
Good but read with caution - data doesn't lie but sometimes the conclusions we draw from the data is biased
I liked this book enough to fantasize about being a data scientist for an hour. Then I had other fantasies and had to go watch videos on the internet. Speaking of which, I assumed "what the internet can tell us about who we really are" would be all creepy sex stuff (which is why I read the book). This is a fair part of the book, but there is other interesting stuff too; interesting stuff about health, sports, family violence, and how we present our lives on social media vs the awful shit we google about ourselves and our loved ones. I am convinced by the author's main point that "big data" can provide far better evidence in many cases than the dubious surveys and interviews that social scientists often rely on, even if I was not entirely convinced by all of his examples. And now that I know my searches might be used as evidence in social science studies I will feel compelled to google important issues more regularly.
informative
medium-paced