Science and Inquiry discussion

For December 2021, we will be reading Calling Bullshit: The Art of Skepticism in a Data-Driven World by Carl T. Bergstrom.

Please use this thread to post questions, comments, and reviews, at any time.

bump

I've started reading this book. So far, it is very interesting. I had not thought of all the ways people use bullshit, intentionally or unintentionally.

I just finished reading the book. Wow! The author shows how he does research, to unearth the origins of so many unlikely stories and urban myths. I highly recommend it. Here is my review.

Get Viral the hunt for the origin of COVID 19
American Nero (tRUMP) plenty
Rage in the Whitehouse
PERIL!
fire and fury Whitehouse and more!

I just finished the chapter on data visualization and it is enlightening to say the least. Now, I will have a better understanding every time my husband complains to me about the various graphs and charts that his employees produce. (He has a special distaste for 3D charts.) Maybe I will even have some valuable input for him the next time he shows me an ill-formed chart. I'm looking forward to reading the rest of the book.

Jessica, I agree. As the book mentions, 3D charts rarely are needed. 2D charts are usually more understandable.

I had a similar eye-opening experience about graphs and charts after reading How to Lie with Statistics back in grad school.

I just finished reading this as my holiday read and throughly loved it. I've posted a short review here: https://www.goodreads.com/review/show...

In case there are any active Twitter users here - Carl Bergstrom, one of the authors, tweets actively and has been a very reliable source of data and analysis for me over this pandemic period, which is also how I first learnt about this book.

I enjoyed the book and it inspired me to perhaps go to Khan Academy and brush up statistics, which I both suffered through and enjoyed my sophomore year in college. My only negative thoughts are that: 1) the people who could most benefit from this book won't read it or anything like it, and 2) those that want to improve the effectiveness of their bullshit will read it.

Steve wrote: "... brush up statistics, which I both suffered through and enjoyed my sophomore year in college...."

That is the perfect summation of how I got through the little bit of statistics that I took, too. It's both fascinating & mind bending. I've read that screwing up the statistical analysis of experiments by otherwise honest & intelligent scientists is rampant. They really need a mathematician to do that work.

Steve wrote: "I enjoyed the book and it inspired me to perhaps go to Khan Academy and brush up statistics, which I both suffered through and enjoyed my sophomore year in college. My only negative thoughts are th..."

Steve, your negative comments really got to me. I agree with you--the people who might benefit from the book won't read it, and those who want to improve their bullshit will read and study it!

Jim wrote: "I've read that screwing up the statistical analysis of experiments by otherwise honest & intelligent scientists is rampant. "

Jim, this is very true. It is the selective "P-Hacking" that causes a big bias in science articles. This is the argument that allows scientists to write the words "statistically significant".

It reminds me of a famous quote from Richard Feynman:

You know, the most amazing thing happened to me tonight... I saw a car with the license plate ARW 357. Can you imagine? Of all the millions of license plates in the state, what was the chance that I would see that particular one tonight? Amazing!

Finding correlations or results "after-the-fact", that is formulating a hypothesis after conducting an experiment can invalidate the conclusions of the experiment!

Incidentally, I just saw this tweet after reading the above comment on p-hacking :)

https://twitter.com/aarit_ahuja/statu...

My two cents. As a research scientist myself, albeit at a junior level, I completely agree that statistical analysis is a major issue in scientific research currently. Part of the problem is that data being generated has exploded in size now and needs far more statistical competence than, say a biologist, can often handle. Fields like statistics and computational analysis need independent expertise - but funds are rarely or grudgingly spared for them. If we could channel even a small percent of the thousands we pay to publish an article into this, then science would be even more robust. Yup, it's a broken system.