in a perspective written 10 years ago, Barres presents some guidance for choosing an academic advisor based on two main considerations. pick an advisor with good mentorship style and who is good at science.
the best way to find a potential advisor that does science you jive with is to do a quick literature search. how I read academic papers. you dont really need to assess the science, but rather look for red flags such as not publishing for the past 5 years. Barres recommends looking for advisors that publish in high impact journals but there may be intention behind this so tread with caution.
another indicator of someone who will do good science is through their grant support. at the very least, they should have adequate funds to support graduate research. however, new PIs are often an exception to this guidance.
references
Barres, B. A. (2013). How to Pick a Graduate Advisor. Neuron, 80(2), 275–279. https://doi.org/10.1016/j.neuron.2013.10.005
gratitudes to my cohort for helpful lunch discussions about this paper and its limitations