Take a photo of a barcode or cover
124 reviews for:
Science Fictions: Exposing Fraud, Bias, Negligence and Hype in Science
Stuart Ritchie
4.33 AVERAGE
124 reviews for:
Science Fictions: Exposing Fraud, Bias, Negligence and Hype in Science
Stuart Ritchie
4.33 AVERAGE
quite good summary of issues with publication bias
does contradict himself a bit about p values not being important but being how you should read a paper (per the appendix) also shits a lot on small sample sizes in biomed science without talking about how expensive and inhumane larger sample size experiments are in animal studies where using 10 rather than 100 mice will be a sufficiently powered study larger sample sizes or requiring a higher threshold p value— which would under most cases Require a larger sample size in some case requiring much larger (more expensive and more animals in some research)
the main thing i’m surprised by is that he mentions paying to read papers and that the people writing and peer reviewing the research are basically volunteers and says journals are basically getting paid 3 times in free labor of reviewers, tax payer $ funding most studies, and tax payers/schools buying the articles on the back end but barely mentions that in addition to all these other things that journals themselves are Not Adding value with researchers also have to Pay to get their articles published in most journals like over 3-5k just to have the paper published which is insane.
does contradict himself a bit about p values not being important but being how you should read a paper (per the appendix) also shits a lot on small sample sizes in biomed science without talking about how expensive and inhumane larger sample size experiments are in animal studies where using 10 rather than 100 mice will be a sufficiently powered study larger sample sizes or requiring a higher threshold p value— which would under most cases Require a larger sample size in some case requiring much larger (more expensive and more animals in some research)
the main thing i’m surprised by is that he mentions paying to read papers and that the people writing and peer reviewing the research are basically volunteers and says journals are basically getting paid 3 times in free labor of reviewers, tax payer $ funding most studies, and tax payers/schools buying the articles on the back end but barely mentions that in addition to all these other things that journals themselves are Not Adding value with researchers also have to Pay to get their articles published in most journals like over 3-5k just to have the paper published which is insane.
challenging
informative
reflective
medium-paced
Pokes necessary, well-structured holes in the way people view scientific publishing, with a focus on psychology, the subject of the author. Name drops many famous books such as Why We Sleep by Matthew Walker, questioning the validity of these bestsellers and the science inside them.
A thorough and lively discussion of the ways in which the modern system of scientific investigation works against itself. Too many studies cannot be replicated, replication studies don't get published, and scientists are incentivized to engage in fraud, to succumb to bias, to make sloppy mistakes, and to hype their results. The author suggests that the present system of research employment with its emphasis on publication has created perverse incentives for producing bad work. He suggests preregistration of research projects (to keep researchers from mining their data for something unrelated to the original design, though serendipity can be rewarded by further studies), a more extensive use of preprints, and perhaps a re-evaluation of the for-profit system of sciengtific publications.
As someone who reads a lot of popular science, I found the book enlightened me about some of the things I have enjoyed reading. The research basis for the growth mindset championed by Carol Dweck, for instance, is much less strong than she suggests. Many, many other popular science studies turn out to have very little replicability (you don't necessarily take more food if your plate is bigger, for instance), while others have been seized on and hyped while still preliminary. The most egregious example of fraud is the fraudulent MMR study by Andrew Wakefield, which continues to have terrible effects on vaccination willingness in parents.
Highly recommended.
As someone who reads a lot of popular science, I found the book enlightened me about some of the things I have enjoyed reading. The research basis for the growth mindset championed by Carol Dweck, for instance, is much less strong than she suggests. Many, many other popular science studies turn out to have very little replicability (you don't necessarily take more food if your plate is bigger, for instance), while others have been seized on and hyped while still preliminary. The most egregious example of fraud is the fraudulent MMR study by Andrew Wakefield, which continues to have terrible effects on vaccination willingness in parents.
Highly recommended.
I agree with one of the blurbs saying this is "required reading". It really is. I especially liked the chapter about Hype and scientific press releases running away with minor results to present them as the next big thing.
informative
slow-paced
funny
hopeful
informative
inspiring
medium-paced
hopeful
informative
reflective
medium-paced
This was my favourite book of the year so far. Incredibly informative, well written, exactly the kind of topics I'm interested in. It's about the issues with science, and specifically psychology and medicine; how they brought about the reproducibility crisis; and how they can be solved.
It has chapters on bias, fraud, hype, incentives and other reasons current research has strayed from the ideals we imagine science to have. Along the way he talks about individuals who committed these errors and the people and techniques who are fighting against them. The final chapter, Fixing Science, is an overview of all the techniques used to combat bad science, as well as radical new ideas on how to restructure the very process of science itself. It has a lot about statistics, academia, open science, and other things I'm interested in.
I think it'd be a good read for anyone who's interested in how science works and wants to consume the published literature critically. Feels similar in vibe to Ben Goldacre's books. Very well written, interesting and engaging throughout. I understood it well but not sure how well someone who has no experience with academia would. Also, excellent citations, as well as notes of expansion.
It has chapters on bias, fraud, hype, incentives and other reasons current research has strayed from the ideals we imagine science to have. Along the way he talks about individuals who committed these errors and the people and techniques who are fighting against them. The final chapter, Fixing Science, is an overview of all the techniques used to combat bad science, as well as radical new ideas on how to restructure the very process of science itself. It has a lot about statistics, academia, open science, and other things I'm interested in.
I think it'd be a good read for anyone who's interested in how science works and wants to consume the published literature critically. Feels similar in vibe to Ben Goldacre's books. Very well written, interesting and engaging throughout. I understood it well but not sure how well someone who has no experience with academia would. Also, excellent citations, as well as notes of expansion.
informative
slow-paced
dark
informative
sad
fast-paced