busy busy busy

[reckless_rex]

As if I don't have enough on my RSS feeds. I like what Seth Roberts is doing with regard to self-experimentation, even if the results need to be taken with a grain of salt. Most interesting is that he's mounted a frontal assault on the dogma that large N studies are always preferable to small N. I would classify him just on the respectable side of

Comments

[ironfrost]

geometric learning at it's finest

[airstrip]

I should adopt the "book a week" strategy with the corollary that, in some fields, one should replace books with the papers that the book was spawned from. Actually, it might not be the case that you can be permanently behind in knowledge because the rate of new knowledge acquisition, the rate of knowledge obsolescence and the marginal value of new knowledge probably conspire in your favor. That is, knowledge in some specific area as opposed to knowledge in all areas.

[tectonic_drift]

I am becoming more convinced of Seth’s point about N-size. Without knowing the bias (i.e. selection bias) that is associated with observational results, the large N may just be reinforcing misleading results. The value of results is not simply in the size of sample but in the overarching method.

If you are interested in more of my thoughts about observational vs. experimental studies, they are here.

Have good trip to SF!

[reckless_rex]

Thanks, that was interesting. This is a pretty ubiquitous problem in policy debates over everything from education to health care to foreign aid: there are few actual experiments, so people just pick and choose whatever observations support their prejudices and ignore/downplay the ones that don't. I agree with Tukey and Fisher (and Seth and you...): a series of small-n randomized trials is worth far more than a whole bunch of large-n observational studies.

[reckless_rex]

On that general methodological topic, I read good things about this book, which I will be reading sometime in the next few months.

[drumnbach]

Good call on the Biology major :D. I'm probably replying to the wrong post here, but now I'm in this post, good call.

I'm taking physics at the mo... I wish I could say I was doing a bachelors but I've actually gone back to college.

Good luck! Oh yeah, my mate told me about some herbal shite that boosts your concentration. He used them for his exams some time ago. I'm going to test them out, as soon as I find out what they're actually called.

[reckless_rex]

Cheers, Ben. Physics is cool, but the current edge of knowledge there requires math skills on a high level. In genetics the Price equations are about as hairy as it usually gets.

Good luck to you too; let me know if the stuff you try does anything for you.