Can a research group be formed without a PhD in charge?

Question

In my place of work, which has a very incipient progress in the field of research in Computer Science, has occurred this dichotomic situation:

• It has arrived one lecturer that has finished his PhD students in a foreign country. Because he has some research already on the way from his doctoral studies, he has formed a research group within our Faculty.

• There is another lecturer that has only a Master degree, also in CS, but he has some good ideas but he suffers from imposter syndrome.

In this situation the former group minimizes the attempt of the second group to form a research group; stating things like "a research group to be successful should always be led by a doctoral graduate if not is doomed"

The question is, is it needed to have a doctoral researcher inside a research group (sorry for the redundancy)? or in what ways we should encourage the formation of research groups even though they are not lead by a PhD graduate?

Answer 1

A PhD is a degree supposed to develop (and later demonstrate) the skill of being capable and ready to do independent research. The degree to which that is achieved can vary massively, depending on institution and supervisor.

It is quite possible (and not at all improbable) that a non-PhD person can outshine the PhD person, but, per default, the latter enjoys a priori the expectation that they know more about the process of doing research than the former. However, this is no reason at all not to form a group of interested people, with or without PhD. A PhD member is not a guarantee of success and the lack of one does not presage doom.

The case you mention seems to be of someone who does not want to see a competing group being created (depending on the convention of your location). That being said, make sure you have knowledgeable people in the group. The reason why many countries have a research group led by a professor (not just a PhD) is that the latter not only know how to do research but have had a broad exposure to many methods and perspectives.

Good luck!

Answer 2

My background is in finance, mathematics and software engineering. I have two master degrees, but no PhD. Normally, I would be working in an investment bank as a quantitative developer or something similar.

Somehow, I was able to get into academics as a bioinformatician. I've currently working for a new research project in computational biology with a team member who also doesn't have a PhD. We're expected to publish to a top journal and attend conferences, we work like a team of postdocs.

Personally, I don't think I'm anyway less intelligent than a PhD. I have the qualification for starting a PhD. Unfortunately, I'm struggling:

• I know my goals, but I don't know how to it. It's very difficult for someone like me without any research experience to invent a new method. I thought I was smart, I thought it was just like building on the existing methods, but I overestimated myself. As an engineer, I have years of experience in applying existing (and well-known) methods brilliantly, but I'm lost when I have to do something that nobody has ever done.

• I thought I have good understanding about computation. I was wrong, the postdocs in my institution had much better understanding than me.

• I wished I had someone more qualified (like a professor) in my team. Someone needs to give me guidance and direction.

• Anybody who can't publish results will not survive in academics, but this is making me very frustrating. In engineering, I could get away with no documentation. In academics, I'm expected to publish and the writing must be good. I'm slow at my writing (I hate it).

Good ideas are no guarantee for success. I had my own ideas, but everything failed. Although not strictly necessary, a research team should have a doctoral researcher, because that's what exactly they are good at. A formal research degree like a PhD is a better indicator for success than someone who doesn't.

Answer 3

If the first group is trying to discourage the second, I would take it as a signal that they expect the second group to fail. The difference between a doctorate in philosophy and a masters is more than added coursework. You spend a huge amount of time failing to succeed at things. You learn your field by failing to succeed. Seemingly brilliant ideas are tested, and fail. You get a good sense of what won't work.

It isn't about being bright. It is about discovering the truth. For my own research, I had to spend tons of hours pouring over bibliographies at the back of articles. Read those articles, then pour over their bibliographies and start over. Although Science Citation Index is wonderful, it only goes back to 1975. A person with a masters degree will not have done that.

You also find you have to leave your field to find what you need. You will find techniques in geology or sociology that are exactly what you need. For example, network diagrams come out of sociology, not computer science. They are very old. You need to borrow and steal from other fields. A person with a masters degree is unlikely to do that.

Finally, all fields deal with uncertainty, but uncertainty is usually not taught below a doctoral level, even in finance. What is usually taught is statistical technique. What isn't usually taught is how that technique can break down. As part of becoming a PhD, you may have to build a new statistical technique.

I wouldn't automatically exclude the second group, but I might do a formal evaluation of the problem they are looking at. Some problems are quite tractable at the masters level, others are not.

Research lead by someone with a PhD often fails. It is less likely to fail when compared to someone with a masters because they have less experience at failure. When I finish this post, I will be writing software and I know it won't succeed, but it will tell me what I need to know to write the next piece of software, that will also not succeed, so that eventually I will (probably) write a piece of software that will succeed. I am using well understood techniques, which is why I know it won't work. I don't yet know what will be future well understood technique, but I am working on it.