If graduate students were empowered to re-envision graduate training, what new elements would be included, what existing elements would be discarded or reshaped?
22 thoughts on “If graduate students were empowered to re-envision graduate training, what new elements would be included, what existing elements would be discarded or reshaped?”
I am in my 6th year of a PhD in biology in the United States. In the first year, I took a dozen courses and rotated in three labs. In the second year, I also took ten courses, worked in my thesis lab, and presented my research plan to my thesis committee of four professors in the form of an oral qualifying exam. I could have written up a master’s thesis and left then, but hardly anybody does, and it’s looked down on. Since then I have been doing research in my thesis lab with an occasional workshop, some tutoring, and also mentoring of a couple younger students.
Here is what I would change.
1. Replace survey courses with workshops about specific topics. I really did not need to pass another survey course on genetics after taking one in undergrad. If a person with a physics degree, etc. comes into a biology PhD program, let her take undergrad courses if she needs to. I felt and still feel that almost all my courses, except one course that was really a long programming workshop, were a waste of time. But it’s great to have workshops that teach me confocal microscopy, Perl, etc.
2. Remove the stigma from “mastering out” and present it as a viable option for people to get a taste of research and then, if they so choose, move on with life. If a person hates research, he should have a dignified way to leave.
3. Market a biology PhD program not as education but as a temporary research job in academic science. A PhD is so different from all the previous education I had, where I had rigid class schedules, and I got rewarded for following the rules. If I had gone into my PhD with more of a “job” mindset and less of a “student” mindset, I think I wouldn’t have worried so much about class grades and the qualifying exam, and would have thought earlier about how I can build a professional future. My principal investigator is not so much my teacher (which I thought at first with my “student” mindset) but more my supervisor and collaborator, which is how it is in a knowledge-worker workplace. Recruiting people to PhD that will treat it as a (professional and creative) job and not as a continuation of school will reduce the stress of adjusting to how a PhD really is and will improve outcomes.
“Market a biology PhD program not as education but as a temporary research job in academic science.”
I agree that it is important for prospective PhD students to have an accurate understanding of what their role will be in grad school, but we should remember that not all PhD students are actually employed by their professor as a research assistant. So, telling prospective students that getting a PhD is like having a job, can be misleading in some cases. (Students who enter PhD programs expecting that they will have the opportunity to get paid a living wage to conduct their dissertation research are sometimes disappointed, an unfortunate misunderstanding.) It’s more complicated than that. Students should be given a clear understanding of what work will be expected of them and what opportunities for funding they will have, and that these two factors may add up to a job-like situation, or may not, depending on the specific case.
Suggested New Elements:
1. In the STEM education area, especially in educational leadership for STEM education in K-12 settings, dissertation-in-practice would be a more useful approach to graduate research outcome than traditional theoretical research approach.
2. Prominence of cross-curricular connections in graduate program work, focused on leading STEM education in K-12 settings.
Suggestion for the NSF:
Change the eligibility requirement for the GRFP that prospective students cannot hold a master’s degree in their chosen field. This requirement penalizes students for having more experience, which seems counterintuitive and counterproductive to me.
In every job outside of academia I have had the leaders and management are routinely scrutinized for their effectiveness and receive direct evaluations from every employee tier including subordinates (e.g. 360 reviews). This is not the case in academia. In grant evaluations, promotions etc. the people most closely involved with a principal investigator and that probably have the most pertinent information (e.g. students and postdocs) are almost never asked about the effectiveness of their boss. My suggestion would be to include recommendation letters or questionaires from students and postdocs as part of the biosketch for grant submissions. I’ll bet study sections would learn a lot!
As it stands now, all of the power in a lab rests in the PIs hand. Graduate students rely so heavily on written recommendations that there is little incentive to approach PIs on any sensitive topic (e.g. alternative careers) and there is almost zero opportunity for pushback once a person has joined a lab. Just visit any chemistry department. Medical schools seem to be more progressive.
As many have suggested, I also think it is a great idea to connect more of the money that goes into a lab to the students through training grants and fellowships.
The incentives are in all the wrong places in PhD and postdoc training.
Stop calling it school and considering young researchers “students”. The perception of a PhD as “grad school” makes one think of text books and studying, when it is absolutely is a full-time job, and far removed in form and level from BSc studies. When I finished my PhD my family were asking me if I finally get a job; PhD researchers do very specialized and advanced work on a full-time basis, and for me, being considered a student, while a classmate who started an industry job with a BSc had a “real job”, was demeaning.
And related to that, more money! I do not expect a PhD to pay as much as an industry job with a BSc, but it shouldn’t pay half as much. Because of the dismal salaries I think many good researchers go directly to industry and forgo getting a PhD (only those really committed go through with it). And on the other hand, many lower quality students who cannot get a decent job, or don’t even know what they want to work in, see a PhD as a prospect, and we end up with low quality people called “Dr.”.
Stipends should be adjusted to cost of living. The stipend through the GRFP and governmental training grants held by institutions aren’t adjusted for this, so far as I can tell. Living on ~30k a year is dramatically different for someone in Boston, NYC, or the bay area, versus someone in the midwest or south. Another relevant factor might be the availability and pricing of grad housing provided by schools.
As someone who recently transitioned from PhD program to Assistant professor Here are a few thoughts:
1. Everyone should be taught about the finances of running a lab. Everything from writing the budget for a grant (how do you calculate overhead?) to actually buying things for the lab is such a foreign world when you’re first starting out. I’ve had several things delayed because I don’t understand how university bureaucracy works. I actually had a large grant that I wrote with my PhD advisor and I still had no idea what was I doing, so I can’t imagine coming into a professorship without even that experience.
2. For students headed into academia, they should be forced to juggle multiple sources of commitment: teaching, research, and service. I understand the need and value of emphasizing research during grad school, but this isn’t realistic preparation for the real world. I think that is one of the main reasons why people are so overwhelmed that first year as an assistant professor.
In my opinion, the single most important thing that should be done to improve the current system of graduate education is to provide a user-friendly source of up-to-date information for prospective grad students, with standardized statistics on graduate program quality, outcomes, and student satisfaction for all US graduate programs. This could take the form of a webpage maintained by NSF, or some other national entity. The information provided should include aggregate summary statistics for each graduate program on completion rates, time to degree, trainee satisfaction, job placements, and salaries (both while in grad school and in subsequent jobs). This would enable prospective graduate students to make well-informed choices about their training and career preparation. It would also create a more competitive market for quality (and career relevant) graduate education, generating strong incentives for graduate programs to improve.
This idea of providing prospective grad students with greater transparency of outcomes data so they can more effectively find the right match for their own needs and goals (and “vote with their feet”, favoring the programs with the best outcomes) has been recommended in several recent reports on graduate education, including the NIH Biomedical Research Workforce Working Group and the American Chemical Society Commission on Graduate Education . Some universities are already providing their own graduate program statistics , but we really need a unified national source for this information, with standardized data collection, so that graduate programs at different universities will be easily comparable.
There is a precedent for this. This could be analogous to the “Hospital Compare” website created by Medicare for reporting of patient perceptions and outcomes re hospital care. The website allows the user to select specific hospitals to compare on a range of variables. Having such a website for data for graduate programs could have two great effects — 1) enable prospective students to choose programs which are strongest on the characteristics most valued by the student, and 2) promote benchmarking between programs, so that the programs come to compete as each program tries to stand out as an excellent program on the ratings so as to attract the best students.
This is important. Currently, we have a system in which grad students cannot find out some of the most important details about the graduate program they are choosing until after they are committed. They have usually already quit their job, relocated, turned down admissions offers from other schools, and worked their way through the first year of their graduate program when they find out that only 50% of the students who enter the program complete it, or that the average time to completion is over six years, etc. Giving them accurate data on certain well chosen metrics of graduate program quality *before* they make their choice, rather than after, would empower grad students to make better selections, so that more entering grad students would be a great fit for the program they are entering and would enter it knowing what to expect. It would also greatly incentivize graduate programs to improve on these metrics of quality.
Some universities seem to feel that they “can’t afford” to collect and maintain this type of outcomes data on their graduate students, but I think that tracking graduate student outcomes should be considered an essential component of professionalism in graduate education. Perhaps organizing this data collection at the national level could provide an economy of scale making it less costly for each participating university and graduate program.
It is important for graduate programs to have some safe mechanism for students to give feedback (without risk to their reputation, or career, if their feedback is negative). Otherwise, the faculty who run the program don’t know a lot about what is happening to students in the program (or, at least, aren’t aware of the students’ piece of the puzzle of why things are happening the way they are) and recurring problems never get fixed.
There should be more opportunities for the best graduate students to conduct research based on their own ideas. The current situation, in which this is not usually possible even for those with the most important and doable ideas, is holding back the progress of science.
In my department students can do research on their own research questions, and I am somewhat baffled as to why this is not done more widely. The main obstacle is whether resources are available, so it works best if the student’s research can be done on a low budget or if the student can acquire her/his own funding, such as an NRSA F31 or NSF fellowship. Also, there needs to be a good research idea, a student who is quite competent, and a faculty who can mentor. With these circumstances, the student-generated research can provide an excellent foundation for the student’s research career and long-term program of research. I recommend this highly — especially for departments and/or faculty which have not previously given this a try..
At the very least, in the case of those graduate programs that are structured in such a way as to make it unlikely that a good student with a good idea (and with their own funding) will be able to carry out their idea, clear information about this should be available to prospective students beforehand. That way there would be no misunderstandings during the admissions process about what the graduate program has to offer and what it doesn’t.
Im a third year grad student at Duke, we have to take a required number of ethics credits through a daylong workshop before the we start the program and then we can pick shorter 1-2hr topical sessions throughout the semester to reach a certain number of credit hours before graduation. I think this has provided us with a flexible way to learn about a wide range of things like publishing rights etc.
In addition, we should do something similar, but in the context of writing, and communication strategies. This can span a wide range of topics from how to put together a powerful presentation, to outreach opportunities, to navigating within lab dynamics. These are the things we aren’t taught, and they are so valuable to the success and integration of the research we do into circles outside of academia.
This comment could also easily fall under Mentoring (a big part of the graduate experience):
Every graduate program in science and engineering should require an “Intro to Research Methods” course that emphasizes broadly applicable research skills centered on critical thinking and the research process. This course should not include discipline-specific topics as experimental design, statistical methods, or how-to training on experimental equipment. The focus of such a course would be to emphasize and teach the process that is common to all research and emphasize skills-based learning over content-based subjects (heat transfer, thermodynamics, statistics, etc.).
The purpose of such a course would be to demystify research and lay out the concrete steps that established researchers use throughout the path between an idea and a publication. The reason that this should be required of all graduate students and not as an elective or just given to those intending to pursue a PhD is to help introduce research as a profession and to empower entering graduate students to decide whether a PhD is right for them.
I am sure that this type of course is already offered in many graduate programs, but is by no means required as part of a technical graduate curriculum on a national scale. I think that some people (both researchers and students) often assume that students will either eventually “get it” or will leave because they are unsuited to a PhD. My personal belief is that this places an unnecessary burden on the advisor-student relationship and that the university can better enable professors and students by providing every entering student with a basic framework for understanding the research process and then allowing the advisor space to help the student interpret that knowledge for their specific project.
My background: I am a 4th year PhD student at UW – Madison in the Materials Science Program. I am on the Graduate Student Advisory Council in my degree program and am heavily involved in training and mentoring in my research group.
To be honest, I think that we, in the sciences, would all be better off if the federal scientific agencies would put more of their funding into creating stable, career positions for scientists beyond the training stage, and less of it into producing such an over-abundance of trainees, many of whom are training for jobs that don’t exist. One way to begin this shift would be to stop providing training funds to graduate programs that do not meet high standards of quality.
Here is an excellent commentary that argues, “More of the work should be done by staff scientists in stable positions and less of it by (fewer) trainees.” Putting an end to the pyramid scheme sounds like a *very* good idea to me.
Funding for graduate students should come from fellowships and training grants (not research grants) for two reasons:
1) Training should be about training, not labor, and this approach to funding graduate study would help keep it that way.
2) This way of funding grad students would give the funding agencies more control over the number of funded grad student positions, so as to balance the number of graduates produced with expected scientific workforce needs.
NSF Graduate Education Forum 
I am in my 6th year of a PhD in biology in the United States. In the first year, I took a dozen courses and rotated in three labs. In the second year, I also took ten courses, worked in my thesis lab, and presented my research plan to my thesis committee of four professors in the form of an oral qualifying exam. I could have written up a master’s thesis and left then, but hardly anybody does, and it’s looked down on. Since then I have been doing research in my thesis lab with an occasional workshop, some tutoring, and also mentoring of a couple younger students.
Here is what I would change.
1. Replace survey courses with workshops about specific topics. I really did not need to pass another survey course on genetics after taking one in undergrad. If a person with a physics degree, etc. comes into a biology PhD program, let her take undergrad courses if she needs to. I felt and still feel that almost all my courses, except one course that was really a long programming workshop, were a waste of time. But it’s great to have workshops that teach me confocal microscopy, Perl, etc.
2. Remove the stigma from “mastering out” and present it as a viable option for people to get a taste of research and then, if they so choose, move on with life. If a person hates research, he should have a dignified way to leave.
3. Market a biology PhD program not as education but as a temporary research job in academic science. A PhD is so different from all the previous education I had, where I had rigid class schedules, and I got rewarded for following the rules. If I had gone into my PhD with more of a “job” mindset and less of a “student” mindset, I think I wouldn’t have worried so much about class grades and the qualifying exam, and would have thought earlier about how I can build a professional future. My principal investigator is not so much my teacher (which I thought at first with my “student” mindset) but more my supervisor and collaborator, which is how it is in a knowledge-worker workplace. Recruiting people to PhD that will treat it as a (professional and creative) job and not as a continuation of school will reduce the stress of adjusting to how a PhD really is and will improve outcomes.
LikeLiked by 2 people
“Market a biology PhD program not as education but as a temporary research job in academic science.”
I agree that it is important for prospective PhD students to have an accurate understanding of what their role will be in grad school, but we should remember that not all PhD students are actually employed by their professor as a research assistant. So, telling prospective students that getting a PhD is like having a job, can be misleading in some cases. (Students who enter PhD programs expecting that they will have the opportunity to get paid a living wage to conduct their dissertation research are sometimes disappointed, an unfortunate misunderstanding.) It’s more complicated than that. Students should be given a clear understanding of what work will be expected of them and what opportunities for funding they will have, and that these two factors may add up to a job-like situation, or may not, depending on the specific case.
LikeLiked by 3 people
Suggested New Elements:
1. In the STEM education area, especially in educational leadership for STEM education in K-12 settings, dissertation-in-practice would be a more useful approach to graduate research outcome than traditional theoretical research approach.
2. Prominence of cross-curricular connections in graduate program work, focused on leading STEM education in K-12 settings.
LikeLike
Suggestion for the NSF:
Change the eligibility requirement for the GRFP that prospective students cannot hold a master’s degree in their chosen field. This requirement penalizes students for having more experience, which seems counterintuitive and counterproductive to me.
LikeLiked by 1 person
In every job outside of academia I have had the leaders and management are routinely scrutinized for their effectiveness and receive direct evaluations from every employee tier including subordinates (e.g. 360 reviews). This is not the case in academia. In grant evaluations, promotions etc. the people most closely involved with a principal investigator and that probably have the most pertinent information (e.g. students and postdocs) are almost never asked about the effectiveness of their boss. My suggestion would be to include recommendation letters or questionaires from students and postdocs as part of the biosketch for grant submissions. I’ll bet study sections would learn a lot!
As it stands now, all of the power in a lab rests in the PIs hand. Graduate students rely so heavily on written recommendations that there is little incentive to approach PIs on any sensitive topic (e.g. alternative careers) and there is almost zero opportunity for pushback once a person has joined a lab. Just visit any chemistry department. Medical schools seem to be more progressive.
As many have suggested, I also think it is a great idea to connect more of the money that goes into a lab to the students through training grants and fellowships.
The incentives are in all the wrong places in PhD and postdoc training.
LikeLiked by 3 people
Stop calling it school and considering young researchers “students”. The perception of a PhD as “grad school” makes one think of text books and studying, when it is absolutely is a full-time job, and far removed in form and level from BSc studies. When I finished my PhD my family were asking me if I finally get a job; PhD researchers do very specialized and advanced work on a full-time basis, and for me, being considered a student, while a classmate who started an industry job with a BSc had a “real job”, was demeaning.
And related to that, more money! I do not expect a PhD to pay as much as an industry job with a BSc, but it shouldn’t pay half as much. Because of the dismal salaries I think many good researchers go directly to industry and forgo getting a PhD (only those really committed go through with it). And on the other hand, many lower quality students who cannot get a decent job, or don’t even know what they want to work in, see a PhD as a prospect, and we end up with low quality people called “Dr.”.
LikeLiked by 1 person
Stipends should be adjusted to cost of living. The stipend through the GRFP and governmental training grants held by institutions aren’t adjusted for this, so far as I can tell. Living on ~30k a year is dramatically different for someone in Boston, NYC, or the bay area, versus someone in the midwest or south. Another relevant factor might be the availability and pricing of grad housing provided by schools.
LikeLiked by 1 person
As someone who recently transitioned from PhD program to Assistant professor Here are a few thoughts:
1. Everyone should be taught about the finances of running a lab. Everything from writing the budget for a grant (how do you calculate overhead?) to actually buying things for the lab is such a foreign world when you’re first starting out. I’ve had several things delayed because I don’t understand how university bureaucracy works. I actually had a large grant that I wrote with my PhD advisor and I still had no idea what was I doing, so I can’t imagine coming into a professorship without even that experience.
2. For students headed into academia, they should be forced to juggle multiple sources of commitment: teaching, research, and service. I understand the need and value of emphasizing research during grad school, but this isn’t realistic preparation for the real world. I think that is one of the main reasons why people are so overwhelmed that first year as an assistant professor.
LikeLiked by 2 people
In my opinion, the single most important thing that should be done to improve the current system of graduate education is to provide a user-friendly source of up-to-date information for prospective grad students, with standardized statistics on graduate program quality, outcomes, and student satisfaction for all US graduate programs. This could take the form of a webpage maintained by NSF, or some other national entity. The information provided should include aggregate summary statistics for each graduate program on completion rates, time to degree, trainee satisfaction, job placements, and salaries (both while in grad school and in subsequent jobs). This would enable prospective graduate students to make well-informed choices about their training and career preparation. It would also create a more competitive market for quality (and career relevant) graduate education, generating strong incentives for graduate programs to improve.
LikeLiked by 2 people
This idea of providing prospective grad students with greater transparency of outcomes data so they can more effectively find the right match for their own needs and goals (and “vote with their feet”, favoring the programs with the best outcomes) has been recommended in several recent reports on graduate education, including the NIH Biomedical Research Workforce Working Group and the American Chemical Society Commission on Graduate Education . Some universities are already providing their own graduate program statistics , but we really need a unified national source for this information, with standardized data collection, so that graduate programs at different universities will be easily comparable.
LikeLiked by 1 person
There is a precedent for this. This could be analogous to the “Hospital Compare” website created by Medicare for reporting of patient perceptions and outcomes re hospital care. The website allows the user to select specific hospitals to compare on a range of variables. Having such a website for data for graduate programs could have two great effects — 1) enable prospective students to choose programs which are strongest on the characteristics most valued by the student, and 2) promote benchmarking between programs, so that the programs come to compete as each program tries to stand out as an excellent program on the ratings so as to attract the best students.
LikeLiked by 1 person
This is important. Currently, we have a system in which grad students cannot find out some of the most important details about the graduate program they are choosing until after they are committed. They have usually already quit their job, relocated, turned down admissions offers from other schools, and worked their way through the first year of their graduate program when they find out that only 50% of the students who enter the program complete it, or that the average time to completion is over six years, etc. Giving them accurate data on certain well chosen metrics of graduate program quality *before* they make their choice, rather than after, would empower grad students to make better selections, so that more entering grad students would be a great fit for the program they are entering and would enter it knowing what to expect. It would also greatly incentivize graduate programs to improve on these metrics of quality.
LikeLike
Some universities seem to feel that they “can’t afford” to collect and maintain this type of outcomes data on their graduate students, but I think that tracking graduate student outcomes should be considered an essential component of professionalism in graduate education. Perhaps organizing this data collection at the national level could provide an economy of scale making it less costly for each participating university and graduate program.
LikeLike
It is important for graduate programs to have some safe mechanism for students to give feedback (without risk to their reputation, or career, if their feedback is negative). Otherwise, the faculty who run the program don’t know a lot about what is happening to students in the program (or, at least, aren’t aware of the students’ piece of the puzzle of why things are happening the way they are) and recurring problems never get fixed.
LikeLiked by 2 people
There should be more opportunities for the best graduate students to conduct research based on their own ideas. The current situation, in which this is not usually possible even for those with the most important and doable ideas, is holding back the progress of science.
LikeLiked by 2 people
In my department students can do research on their own research questions, and I am somewhat baffled as to why this is not done more widely. The main obstacle is whether resources are available, so it works best if the student’s research can be done on a low budget or if the student can acquire her/his own funding, such as an NRSA F31 or NSF fellowship. Also, there needs to be a good research idea, a student who is quite competent, and a faculty who can mentor. With these circumstances, the student-generated research can provide an excellent foundation for the student’s research career and long-term program of research. I recommend this highly — especially for departments and/or faculty which have not previously given this a try..
LikeLiked by 2 people
At the very least, in the case of those graduate programs that are structured in such a way as to make it unlikely that a good student with a good idea (and with their own funding) will be able to carry out their idea, clear information about this should be available to prospective students beforehand. That way there would be no misunderstandings during the admissions process about what the graduate program has to offer and what it doesn’t.
LikeLike
Im a third year grad student at Duke, we have to take a required number of ethics credits through a daylong workshop before the we start the program and then we can pick shorter 1-2hr topical sessions throughout the semester to reach a certain number of credit hours before graduation. I think this has provided us with a flexible way to learn about a wide range of things like publishing rights etc.
In addition, we should do something similar, but in the context of writing, and communication strategies. This can span a wide range of topics from how to put together a powerful presentation, to outreach opportunities, to navigating within lab dynamics. These are the things we aren’t taught, and they are so valuable to the success and integration of the research we do into circles outside of academia.
LikeLiked by 3 people
This comment could also easily fall under Mentoring (a big part of the graduate experience):
Every graduate program in science and engineering should require an “Intro to Research Methods” course that emphasizes broadly applicable research skills centered on critical thinking and the research process. This course should not include discipline-specific topics as experimental design, statistical methods, or how-to training on experimental equipment. The focus of such a course would be to emphasize and teach the process that is common to all research and emphasize skills-based learning over content-based subjects (heat transfer, thermodynamics, statistics, etc.).
The purpose of such a course would be to demystify research and lay out the concrete steps that established researchers use throughout the path between an idea and a publication. The reason that this should be required of all graduate students and not as an elective or just given to those intending to pursue a PhD is to help introduce research as a profession and to empower entering graduate students to decide whether a PhD is right for them.
I am sure that this type of course is already offered in many graduate programs, but is by no means required as part of a technical graduate curriculum on a national scale. I think that some people (both researchers and students) often assume that students will either eventually “get it” or will leave because they are unsuited to a PhD. My personal belief is that this places an unnecessary burden on the advisor-student relationship and that the university can better enable professors and students by providing every entering student with a basic framework for understanding the research process and then allowing the advisor space to help the student interpret that knowledge for their specific project.
My background: I am a 4th year PhD student at UW – Madison in the Materials Science Program. I am on the Graduate Student Advisory Council in my degree program and am heavily involved in training and mentoring in my research group.
LikeLiked by 2 people
To be honest, I think that we, in the sciences, would all be better off if the federal scientific agencies would put more of their funding into creating stable, career positions for scientists beyond the training stage, and less of it into producing such an over-abundance of trainees, many of whom are training for jobs that don’t exist. One way to begin this shift would be to stop providing training funds to graduate programs that do not meet high standards of quality.
LikeLiked by 1 person
Here is an excellent commentary that argues, “More of the work should be done by staff scientists in stable positions and less of it by (fewer) trainees.” Putting an end to the pyramid scheme sounds like a *very* good idea to me.
http://www.sciencemag.org/content/346/6215/1422.full
LikeLiked by 1 person
Funding for graduate students should come from fellowships and training grants (not research grants) for two reasons:
1) Training should be about training, not labor, and this approach to funding graduate study would help keep it that way.
2) This way of funding grad students would give the funding agencies more control over the number of funded grad student positions, so as to balance the number of graduates produced with expected scientific workforce needs.
This blog post articulates it well:
http://scienceforthefuture.blogspot.com/2010/02/why-are-there-still-too-many-graduate_10.html
LikeLiked by 1 person