Letters of recommendations for grad school - who reads them?

[Kron3007]

Random thoughts:

1) 20 is really young.   When I was just that age, I decided, at the beginning of my senior year, to pass on a planned but unrequired senior honors thesis at dear alma mater.   I just was not up for significant independent research at that time of my life.   I still today, however, regularly read scholarship related to the topic I was going to write on, which remains a high interest of mine and one I have incorporated into ancient history teaching I do.  But then, all I wanted to do as a senior was continue my language studies-- I was taking two or three languages a semester.   I got essentially an A average in major, graduated magna cum laude, and then went to an excellent MA program and wrote, two years later, a really pretty adequate MA thesis.   Of course later I went elsewhere and did complete a PhD, which in classics/ ancient history is nothing like the PI-led projects STEM students do-- very very independent.   I stand by my assessment that many students are late bloomers, and should be judged on their merits at the time, especially given the IMNSHO vastly overemphasized research expectations that for whatever stupid reasons have become normative for undergrads seeking PhD admission, and generally at the obvious expense to these students' broac-based well-rounded undergrad educations, especially outside of STEM.

2) Never having written a rec for a PhD applicant, what kinds of questions are asked of the referrers?   Does 'intellectual curiosity' come into play?

3) I would of course be interested in any suggestions as to how to evaluate students' intellectual curiosity, suggestions which of  course ought to be able to be adapted towards the goal of fostering that curiosity...?

4) Risk taking is another issue altogether, not really the same as curiosity.   I for one am just not a risk taker, never have been, for a variety of personal and background reasons, probably similar to the reasons why, in all those career aptitude tests I have taken over the years, 'sales' pretty much ranks dead last, as low as can be.   Not sure that this trait had anything to do with my fittedness to complete a PhD in classics, which I obviously did....?

Every field is quite different.  In mine (at least in my specific university etc.), we do not accept undergraduate students directly into the PhD program.  You have to start a MSc program and either complete that first, or there is an option to transfer directly into a PhD program if you excel and your advisor supports it.  There is a lot of value in the MSc, and I generally prefer doing this first, which addresses the late bloomers to some degree.

Regarding grades, they are a good indicator of some things but dont tell me if a student will be able to make things work on the ground.  In my field, you need to be able to troubleshoot experimental setups, which requires a number of skills that are just not tested in an undergraduate program.  My first lab supervisor said he liked to recruit students that grew up on farms for this very reason, they are resourceful and can independently make things work on the fly.  I largely agree with this.  Not specifically farm kids, but I do like to recruit students who have had a variety of experience outside of academic studies.  I have met several bright students/postdocs that were academically bright, but had difficulties when things went wrong because they had never turned a screw driver (for example).   Again, this is very much field dependent, but my students need to be able to build out experimental setups and fix technical problems as they occur.  This often happens in the field, where they need to be able to address it quickly and independently. 

An A average does not correlate with this from my experience and I would generally prefer to recruit a B student that worked through their program than an A student who didnt.

     

[mleok]

When we judge student's potential ability to work on hard problems where a solution might or might not exist, we must be judging it on some basis.  I think such assessment (most of which *is* linked to a course somewhere) should be part of the basis for a grade in some courses.  One of the reasons is because those who don't go to grad school are also hired for their ability to work on problems where the desired solution might not exist; it's not like that skill is only useful for those going on the graduate school.

Which course, in a typical mathematics curriculum at either the undergraduate or graduate level do you think actually assesses a student's potential to work on hard problems for which a solution might not exist? Sure, real analysis is "hard," but we tend not to assign problems or exams which are sufficiently open-ended that it would be able to assess that potential. Short of a senior thesis or some sort of extended research experience, where would you say this arises? It's not like we're a subject like engineering, where a senior design and capstone project course might serve that role.

[marshwiggle]

When we judge student's potential ability to work on hard problems where a solution might or might not exist, we must be judging it on some basis.  I think such assessment (most of which *is* linked to a course somewhere) should be part of the basis for a grade in some courses.  One of the reasons is because those who don't go to grad school are also hired for their ability to work on problems where the desired solution might not exist; it's not like that skill is only useful for those going on the graduate school.

Which course, in a typical mathematics curriculum at either the undergraduate or graduate level do you think actually assesses a student's potential to work on hard problems for which a solution might not exist? Sure, real analysis is "hard," but we tend not to assign problems or exams which are sufficiently open-ended that it would be able to assess that potential.

And on an exam, some students might spend a lot of time thinking, not writing, while another scribbles random things on paper. How would someone grading the exam have the slightest clue which of those indicated the student who was making the most progress?

[Kron3007]

When we judge student's potential ability to work on hard problems where a solution might or might not exist, we must be judging it on some basis.  I think such assessment (most of which *is* linked to a course somewhere) should be part of the basis for a grade in some courses.  One of the reasons is because those who don't go to grad school are also hired for their ability to work on problems where the desired solution might not exist; it's not like that skill is only useful for those going on the graduate school.

Which course, in a typical mathematics curriculum at either the undergraduate or graduate level do you think actually assesses a student's potential to work on hard problems for which a solution might not exist? Sure, real analysis is "hard," but we tend not to assign problems or exams which are sufficiently open-ended that it would be able to assess that potential.

And on an exam, some students might spend a lot of time thinking, not writing, while another scribbles random things on paper. How would someone grading the exam have the slightest clue which of those indicated the student who was making the most progress?

Some students also do poorly on exams (where a lot of undergraduate grades come from) due to anxiety or other issues, but perform well in a more real world setting excel in grad school.   

I had a MSc student a while ago that had a learning disability of some sort (I didnt pry).  His grades were not stellar, but he did a great job in grad school because he worked hard, asked for help when needed, took initiative, and was determined.  He just had a good balance of traits that do not appear on a transcript.  He graduated with a strong thesis and has since joined a company where he established and is now operating a lab.  Back to the OP, I accepted this student based on a reference that identified some of these less obvious characteristics. 

[Puget]

When we judge student's potential ability to work on hard problems where a solution might or might not exist, we must be judging it on some basis.  I think such assessment (most of which *is* linked to a course somewhere) should be part of the basis for a grade in some courses.  One of the reasons is because those who don't go to grad school are also hired for their ability to work on problems where the desired solution might not exist; it's not like that skill is only useful for those going on the graduate school.

Which course, in a typical mathematics curriculum at either the undergraduate or graduate level do you think actually assesses a student's potential to work on hard problems for which a solution might not exist? Sure, real analysis is "hard," but we tend not to assign problems or exams which are sufficiently open-ended that it would be able to assess that potential.

And on an exam, some students might spend a lot of time thinking, not writing, while another scribbles random things on paper. How would someone grading the exam have the slightest clue which of those indicated the student who was making the most progress?

Some students also do poorly on exams (where a lot of undergraduate grades come from) due to anxiety or other issues, but perform well in a more real world setting excel in grad school.   

I had a MSc student a while ago that had a learning disability of some sort (I didnt pry).  His grades were not stellar, but he did a great job in grad school because he worked hard, asked for help when needed, took initiative, and was determined.  He just had a good balance of traits that do not appear on a transcript.  He graduated with a strong thesis and has since joined a company where he established and is now operating a lab.  Back to the OP, I accepted this student based on a reference that identified some of these less obvious characteristics.

Likewise, I currently have an MA student who had terrible grades his first two years of undergrad, resulting in a low overall GPA. He was a first gen URM student and it took him awhile to get the hang of college, plus he had a lot of life stuff going on. Then it clicked for him and he got passionate about research and went on to work full time in a lab after graduation. He couldn't get into PhD programs though because no one was looking past his GPA. I actively recruited him as an MA applicant for my lab based on stellar letters from his undergrad and postdoc PIs about his research experience, skills, and aptitude. He's been proving himself very well in the lab and getting straight As here. Hopefully the MA transcript will get PhD programs to look at him more holistically next year.