A formal model for undergraduate authorship
The standard guidelines for authorship rarely address undergraduate-specific issues. Earlier, there was some interesting discussion about what it takes to make an undergraduate a first author, and the costs and benefits of working to make this happen. Since then, I’ve been made aware of a detailed and thoughtful article specifically addressing this topic, by Burks and Chumchal at Southwestern University and Texas Christian University.
If you’re thinking about investing the time into mentoring an undergrad through the long slog of the writing process, this fuel for thought is worth your own time to read. There is a great list of recommended strategies, which we only touched on in the comments before. Here’s a copy. The paper includes this decision tree:

From: Burks, R.L. and M.M. Chumchal. 2009. To Co-Author or not to co-Author: How to write, publish, and negotiate issues of authorship with undergraduate research students. Science Signaling 2: tr3.
This article was sent to me by a reader who didn’t want himself to be identified. Thanks, anonymous correspondent!
This paper is spot on and provides a very useful way to structure a project even before you start. There are a few tacit assumptions in here, though, of which I’m not wholly convinced.
Publication with undergraduates makes it harder to get into a higher tier journal (potentially because of time constraints)
Lack of institutional support may alter the costs and benefits of involving students in research
The motivation for supporting student authorship will vary with tenure/promotion status
The paper also addresses whether or not students earn any authorship at all, and if so, what position. This part made me feel better, because it looks like my current practice mostly follows the recommendations. However, the authors suggest that if a project couldn’t have been completed without a student, then that students merits authorship, at least somewhere in the paper. Almost nothing in my lab gets done without students. What is the role of an undergraduate student who performs the role of a thoughtful technician? This student didn’t conceive the project, but they spent 200 person-hours working on it. They aren’t in a position to analyze or write (or, at least, I’m not in a position to mentor them on it). They collected nearly all the data but didn’t do much else. Are they coauthors? This is murky. The student has a good deal of ownership and the project would not exist without the student, but you did everything but collect data. I prefer to involve students more deeply, but sometimes this doesn’t happen.
This is still a dilemma for me. One of the pragmatic aspects that enter the equation is the professional trajectory of the student. Would the paper matter for them? This shouldn’t be a part of authorship criteria, but it’s hard to ignore.