Sunday, April 24, 2016

Resources for statistics in biology

I've recently come across a few resources for learning about the methods, applying, or otherwise studying statistics that I thought others might also find helpful.

This section of the Nature Publishing Group website provides a list of articles for the concerned biologist looking to apply statistics to their work. Topics range from "What is Bayesian statistics?" to the problem of small sample sizes in neuroscience, and testing significance in large scale genetics studies. 

This section of the Nature Publishing Group website provides a list of articles focused on the news and views regarding statistics. I particularly enjoyed the article on "Weak statistical standards implicated in scientific irreproducibility" which suggests that as many as 1/4 of published studies that cite significance based on p values that are less than or equal to 0.05 are actually false. Another interesting op ed piece, focused more on finding solutions for bad stats and lack of reproducibility. 

The soon-to-graduate undergrad in my lab happens to be a math major (which has been very useful!) and he highly recommended the Khan Academy statistics course for any lingering questions I may have or future concerns with applying statistics in the lab. I signed up just to play around with it a bit and it seemed a little basic, but I might just have to get through the first few lessons before it will be more helpful.

This resource I found particularly helpful! It's essentially a textbook on Stats, but it is free and open to the public and starts with the most basic concepts. The left hand side features a list of all of the topics, so it's easy to navigate if you're quickly trying to reference a concept. I also found the writing to be far more accessible than most textbooks or other online stats guides.

Last but certainly not least, I really enjoyed the "Stats Make Me Cry" blog. The author, Jeremy Taylor, runs a consulting company where they will help you with the statistics on your publication or thesis (sorry, due to high demand he's not currently accepting new clients). There are, however, a number of interesting posts on his blog that discuss the qualities of different statistical tests and how to correctly apply them. See: The answer to ice cream is always yes.

I hope you find this (far from exhaustive) list of resources useful: they're the things I've found most helpful and/or amusing through this semester. Feel free to comment and add your own favorite resource!

1 comment:

  1. Thank you for sharing! Just by taking a look of all the sources you mentioned, I think I like the Khan Academy the most-- its homepage is more like a appendix of a textbook and you can start with the topic you have problem with. The blog is also a interesting one. And based on the comments under the posts I can see that people enjoyed reading something fun and well-explained.

    The Nature pages are more serious and formal. But comparing to the other sites that you provided, they provide related papers as result when you do search. The Dell one provides non-relevant PDF mostly; the blog does not show result easily, maybe due to its small capacity. The Khan Academy provides good results-right about statistics, in variation of videos, exercise, etc.

    I found a interesting blog site, not useful as those you mentioned, but it resonate with the BadStats assignment: Many examples of bad statistics that we want to avoid in there.