Thinking of doing a PhD and wondering what might help you succeed?
Supervising PhD students and want the best outcome for them?

Thread Summary of work in @PLOSONE https://doi.org/10.1371/journal.pone.0236327

@AcademicChatter
#AcademicTwitter #phdchat #phdlife #research

In a very interesting analysis of German universities in the 1930s, Waldinger https://doi.org/10.1086/655976  showed that the status of the doctoral student's supervisor influenced:

> whether they would become a full professor
> citations their subsequent scientific work received

Other studies have been done implying
> department reputation https://doi.org/10.1007/BF02017149
> group leader reputation https://doi.org/10.1016/j.respol.2017.02.007
> access to resources/equipment https://doi.org/10.1007/BF02457402
> number of full-professors on staff https://doi.org/10.1023/A:1018705823763
influenced research output

Intuitively, one might think that students with strongest academic scores will do better in PhD.

And scholarships are often apportioned based on the student’s prior performance.

But is this right and is this the right way to go?

We examined the publication numbers, citations and impact factors by students who applied for a PhD some 5 years prior. We compared this to their prior academic performance, research training, supervisor career level, supervisor team quality, and if the student got a scholarship

What did we see?
> the "research environment" (e.g. supervisor quality/experience and student located in a research institute) was the biggest factor in the student's future outputs

> the students' previous academic outcomes and previous research training were not predictors.

Getting PhD scholarship had an important influence on student outcomes.

Not only on outputs, but also on whether they remained in their studies or not.

This is a no brainer (need income to survive!) and has been shown before https://papyrus.bib.umontreal.ca/xmlui/handle/1866/23223

These findings are in line with the idea of "accumulative advantage" in science where differences between scientists at an early stage of their career become reinforced over time

https://doi.org/10.1126/science.159.3810.56
https://doi.org/10.1016/j.respol.2019.03.009

Caveats: citations and pub numbers are important metrics for university rankings, but they are not "EVERYTHING". There is also student satisfaction / mental health, job outcomes, which we couldn't look at

Summary:

If you are looking to get into research: search for the best supervisor you can
(but maybe make sure they do have time for you and are supportive)

For supervisors: be the best researcher you can be and support students!