calphys wrote:One thing I would be a little worried about with a brand spanking new faculty member is that (if they're an assistant prof), theres a chance that they would be denied tenure sometime while you were in grad school with them. Meaning they would no longer have a job at your school. Likewise, if they got denied tenure after you finished, that would pretty much remove your strongest recommender from the equation for future jobs.
Just a thought.
This is a very important factor. There is a student at my school who's a year or two away from finishing his Ph.D. work and his adviser just failed to make tenure and will be departing from the university soon. There is talk about whether he may continue to pursue the same work under a different adviser in a closely-related field of research, or work from a distance with his current adviser, or some combination of both. Either way it's a logistical mess. Younger prof's tend not to be quite so dogmatic and committed to their preferred approaches and may be able to offer some fresh perspectives on various research topics while remaining very open-minded, but there is a lot of inherent value in working with a senior faculty member with lots of connections in the field and more experience and expertise than you could hope to draw off of in 5 years. But keep the younger prof's close to you, you may just write a paper with them some day!
One other thing to consider is that assistant prof's tend to try extra hard to impress the university enough to make tenure, and you could very well find yourself becoming just another cog in their academic machinery, a means to an end for their own professional success. This isn't anything bad or wrong on their part, they're just particularly distracted with the pressures of the tenure process, and you may find yourself inadvertently in the crossfire. This could result in doing lab work which is more important for them than for your own professional development, and that could lead to a sub-par grad school experience. So I think consider everyone but lean toward experience unless there's a good reason not to do so.