what I'm reading (or rather, have just returned)

[silverflight8]

I was reading Genome by Matt Ridley, just before I had to return it to the library. I like to live dangerously. It was published 1999, so just as the Human Genome Project was entering its wildly successful stage. It tells the tale of 23 chromosomes and "stories" I suppose of selected genes on each

Comments

[cherrytide]

I had to read that book for class a few years ago. It was sort of thrown at us as a sort of 'oh you're an arts student who has to catch up on genetics - here you go!'. I thought it was interesting although the style was oddly dramatised at times, possibly at expense of accuracy? I don't remember specifics though.

(Also: Hi!)

[silverflight8]

I have an innate distrust of popular [field] books in general, because it's not a very good story if you have to hedge things. You know, like when you do confidence intervals for statistics, and you can only say--we're 95% confident that the average lies between 50 to 58% of the population. That doesn't pack the same punch that 54% OF POPULATION BELIEVES X!! I mean, I guess you need to elide things sometimes but :/, and the dramatic style definitely made me wary. Things don't just fall out like stories do!

(Hi!)

[wordsofastory]

Oh, I've been meaning to read 'Questioning Collapse'! I've heard really good things about it.

[silverflight8]

From what I can tell, it's very well researched and written. I liked the diversity of cases--that doesn't hurt :)

[wordsofastory]

I saw the man who wrote the Rapa Nui chapter give a talk on the topic, hmmm, four or five years ago. He was very funny and lively in person! A lot of the other authors I know from their work, as well, though not in person, so it seems like it should be a very good book.

[silverflight8]

Oh, cool! What did you study, if I can ask? You seem to be familiar with the field :)

[cloudsinvenice]

This is why I'm a bit reticent about reading more popular science - I'd love to, but as I don't have any background in that area, I'm wary of swallowing shaky methodology/conclusions...

[silverflight8]

I'm sure they're not all wrongity wrong wrong or anything, but I just don't think facts always make good stories. Like accuracy doesn't always translate to good/popular narratives, and in the end, the point of the book is to make sales.

Of course this is why my browser is full of 20-page reports full of statistical mumbojumbo, because I think that'll be better *facepalm* Obviously this means I need to learn more statistics. We need a better intermediary form--I think that Questioning Collapse is one of those--especially for fields that are frequently opaque for reasons of not only math but also just terminology (my god, economics, whyyyy.)

[cloudsinvenice]

Yes, there's always that tension between narrative and accuracy. I liked how Ben Goldacre explored this in Bad Science - his take (amongst other things) is that people are easily confused because we tend not to have been taught anything about statistics, and statistics are often used unchallenged to hold up wobbly-but-compelling pop science...

[silverflight8]

*nods* And stats, it's just there's rules of thumb but you've got to make decisions, too, you know? Like there's no way you could absorb all the individual pieces of data, not meaningfully, so you've got to synthesize it somehow, but how? Multiple measures I guess is a good start (like if you have a really skewed sample, include both median as well as average) but it's just really very fluid. And yet it's all we've got!

his take (amongst other things) is that people are easily confused because we tend not to have been taught anything about statistics
*nods* I think it's the same reputation that math's got--it's hard, doesn't make sense, etc. Which is damaging; I think (like math) it's a really, really useful tool for understanding all sorts of things.