From Jeremy:
This is old, but I missed it at the time (embarrassingly, since it’s from my own student!) Here, in visual form, is the funny but sad truth on what evolutionary biologists actually do (as compared to what their moms, or their friends, or the media think they do). Anyone want to make the equivalent chart for ecologists?
Here’s the ESA Student Section’s list of questions that prospective grad students and postdocs should ask of their prospective supervisors (and of their prospective supervisors’ lab members). Very useful. In my experience, many prospective students don’t ask these questions.
Here’s how to learn from other academics’ CVs. And while you can also learn a lot from people’s shadow CVs, I’m the only person I know who’s posted their shadow CV, so you’ll have a much bigger dataset if you look at people’s CVs.
Probably I’m the last person in the world to discover this, but just in case I’m not: there’s now a StackExchange advice site for “academia”. It’s like other popular StackExchange sites like Cross Validated (for statistics): ask a question, and wait for answers from other knowledgeable people. It’s strength is that there’s a large user base, so you’re pretty likely to get a useful answer. On the academia site, people have been asking questions about everything from whether to keep an “open” lab notebook, to how much time they should spend on reviewing, to the pros and cons of going back to grad school many years after completing an undergraduate degree.
Just for fun: Normal Deviate suggests new names for common statistical methods. The best ones are for bootstrapping (let’s call it “shotgunning”!) and a regression technique called the lasso (“the taser”!)
From the archives:
All I want for Christmas is a plush version of my study organism
Dynamic Ecology 
Haven’t looked at that link but I do agree that we need some new names for these methods. Indeed, I’ve already begun referring to ANOVA as “the evil one”, per Brian’s conclusive post on the matter recently. Further, Monte Carlo procedures are hereafter “blind crapshoot”.
Monte Carlo methods aren’t just for the select few anymore. So I think they should be called Las Vegas methods. Or Reno methods.
Materials and Methods
Although Einstein has argued that “God does not play dice with the universe”, the roulette wheel has not formally been ruled out, and we applied several variants of the “spin and pray” methods to our data. These methods have been validated by well-replicated visits to Reno and Tahoe to ensure methodological soundness….