What are your criteria for evaluating potential PhD students?

Question

A more complete description of the question

As the question suggests, I'm wondering if anyone could share the criteria as well as maybe heuristics or questions they use for deciding whether to take on a doctoral candidate. What really matters? What markers do you look for to tell you whether a person could be successful and a good investment of your time?

Background

I'm a full-time Prof in a University-based Business School in the UK. Our PhD programme attracts a) lots of people in general; and b) lots of "post-experience" candidates (i.e., mid- to senior managers in their 30s and 40s). Although there is a structured way to inquire about PhD opportunities, the faculty are also flooded with random applications e-mailed to them directly from various candidates around the world. As usual with the PhD as a degree, very few people understand what this is all about and what they would be letting themselves in for. We have had a number of stellar graduates, but a lot of people struggle and suffer unnecessarily. The problem is, it's very difficult to estimate in advance which category a particular applicant belongs to (except the obvious cases, of course). So I'm wondering if anyone can share what they do to try and estimate a person's chances for success in a PhD programme - criteria, process, or anything else.

Answer 1

I take a two-phase approach (US, computer science Ph.D program). First, I defer the initial judgement to our admissions committee (unless I have identified a specific prospective student I am trying to recruit).

I then look through the admitted students for potential candidates, and focus on research record, statement, and LORs. I'm looking for evidence of (potential for) critical thought and reflection, not just buzzwords.

Once I've identified a candidate or two, I arrange for an interview. A few days in advance of the interview, I send them a recent paper either by my group or a group working on related topics to read. I try to pick one with a low barrier to entry for understanding (feasible in my research area), and then in the interview ask them for (1) a summary of the paper's goals and (2) a question or two it raises for them. At this stage, I am not looking for perfect answers or good research questions, but the first steps of the ability to read something and think about it. I also ask some other things in the interview, such as why they want to do a Ph.D (even if this is in their statement - I want to hear it, and ask questions).

I'm still pretty junior, so I haven't had a lot of time to refine this yet. But so far it seems to be ok, and it lets me see whether we can start to talk about research, and can also surface fundamental incompatibilities in goals or research philosophy (although some of those come up in the LOR too).