Do students in your field negotiate scholarship or fellowship amounts?

[Durchlässigkeitsbeiwert]

I'm reminded of a friend in NY who sells commercial real estate, mostly to people who want income properties. When he makes the case for a property based on cap rate, debt service coverage ratio, or some objective metric of profitability, the client "nods" in agreement, taking in the objective data. He is, after all, a "rational" decision maker. Only, however, once the client is able to see the property does a decision occur, presumably on intangible factors. Nonetheless, the client rationalizes their decision based on the numbers. This is apparently less the case when a large corporation is buying the property and sends someone simply to verify the objective data and check of a list of potential difficulties.

Such pattern can be easily explained rationally.
There multiple factors responsible for the deviation of the outcome from the predicted. In terms of commercial property this would things like poorly-maintained adjacent properties, highly-localised peculiarities of parking etc. Large corporations have large enough sample for income to revert to the predicted level (i.e. unexpected losses in one place would be offset by extra income elsewhere). Furthermore, to evaluate such things may be more expensive to them (i.e. if they are going to send an actual decision maker on such verification trip). In contrast, stakes are much higher for individual investors in commercial property warranting additional on-site review.

Graduate schools admissions appear to follow very similar pattern (at least in my personal sample):
- programs with limited pool of both students and advisors (due to high degree of specialisation) and with money coming via advisor's grants (not from department) rely on "soft" factors with GPA used only to filter out some of the original applicants.
- large programs admitting cohorts of nearly-interchangeable grad students funded by department rely on "hard" factors to a much larger extent (as to maintain overall cohort performance).

[financeguy]

I don't doubt these are relevant reasons why buyers (of properties or prospective students) might choose to use this method of decision making. I simply doubt that focusing on the soft factors is more profitable. (Produces more rent overall, or a more successful student, to extend the analogy.) I could be wrong and my assumption would be that in the more objective the field (STEM programs, for example) the more the objective data is likely the more predictive while in fields largely subjective anyway (the arts and humanities, for example) the soft factors may be more predictive relative to the STEM fields.

The conversation is likely to remain in the purely hypothetical. No graduate program is going to "test" if they're right by giving up their ability to make subjective decisions. In a way, the LSAT and it's public reporting on the bar forms has done this in a de facto manner for law schools and the results are pretty clear. If you measure success by passing the bar, the GPA and LSAT numbers for incoming students are highly predictive. My assumption is that letting in "interesting/nice guy who had a conversation point about my home town but managed a 2.3 GPA" would be a less likely indicator of passing on an initial attempt after graduation even if it led a subjective committee to determine him a better candidate.

[Durchlässigkeitsbeiwert]

I simply doubt that focusing on the soft factors is more profitable. (Produces more rent overall, or a more successful student, to extend the analogy.)

I actually concur. The point of "soft" assessment is often to avoid a "lemon", not to get a breakthrough. I.e. it narrows the range for individual outcome with uncertain impact on the average outcome. So, from the perspective of the society it may be detrimental (extra effort for no average gain), but for individual it is much more important to avoid very negative outcome than to get very positive one.
E.g. student wasting 100+ k$ from the grant can undermine professor's ability to get such grants for a while afterwards (in real estate analogy individual getting bankrupt due to defects in the property does not get extra attempts).
Though, I am not sure what are feedback loop in humanities / social sciences (from what I saw professors there are quite indifferent to the progress of their students).

In a way, the LSAT and it's public reporting on the bar forms has done this in a de facto manner for law schools and the results are pretty clear. If you measure success by passing the bar, the GPA and LSAT numbers for incoming students are highly predictive. My assumption is that letting in "interesting/nice guy who had a conversation point about my home town but managed a 2.3 GPA" would be a less likely indicator of passing on an initial attempt after graduation even if it led a subjective committee to determine him a better candidate.

As mentioned by Caracal above, the "soft" assessment for a given program is often about choosing between students in GPA 3.5-4.0 range. So, it has has too narrow range to be a predictor. I suspect the GRE may have helped, but in many fields it is unheard of.

[financeguy]

The GRE is an interesting animal. As my name suggests, I'm in Finance, although my previous degree field was in the arts. Graduate programs in the arts request the GRE, but I have been told by many faculty members and departments chairs that these are requirements of the university or the graduate school within which they are housed, not the department itself. They may affect university wide fellowship opportunities but are generally not taken very seriously by the departments themselves. Further evidence of this is that arts programs housed outside a university (Music, Dance and Theater conservatories, for example) almost uniformly do not require this test.

I totally agree that most humanities faculty are not particularly invested in the success of their graduate students. Why would they be? Almost every humanities role is a hard-money position. Grant performance by an assistant is not only irrelevant, grants in general are. What exactly is the incentive other than cheap labor to teach the 101 course for a few years while they offer esoteric electives in their micro niche? This also explains the outrageous time to degree completion rates in most humanities and arts fields.

[Caracal]

I totally agree that most humanities faculty are not particularly invested in the success of their graduate students. Why would they be? Almost every humanities role is a hard-money position. Grant performance by an assistant is not only irrelevant, grants in general are. What exactly is the incentive other than cheap labor to teach the 101 course for a few years while they offer esoteric electives in their micro niche? This also explains the outrageous time to degree completion rates in most humanities and arts fields.

We've had this conversation before, but I think this is an example of people outside of a field not  understanding the dynamics in that field. I only have a vague sense of how grants work in the sciences and related fields, but my understanding is that grant money is connected directly to almost everything. Hiring, prestige, etc is all mostly dependent on grant money. Correct?

I think that makes it hard for people who have spent lots of time in this world to think about these things being less directly connected to clear rewards and that leads them to conclude that there must be no value for faculty in mentoring grad students and therefore humanities faculty don't bother with it.

Im reality, there are lots of rewards to faculty who do a good job mentoring students. Elite programs actually care a lot about the success of their students. It is how they justify the money spent on the programs. A mid career faculty member at an institution a rung or two down the ladder who has a reputation for being a good advisor often has a better chance of moving to one of these elite schools or making a lateral move with a solid, but not superstar, publication record.

Other rewards are less tangible, but still important to people. Having a bunch of successful grad students is a marker of success. Besides, many faculty enjoy mentoring grad students.

In looking for evidence of the theory that grad mentoring is terrible, there's a tendency to again see differences as being deficiencies. If your model for graduate mentorships is based on the sciences, you're going to think that that humanities mentoring is neglect. It doesn't run on a team model, everyone isn't working in a lab all the time with each other, and sometimes students can not have contact with their advisor for extended periods of time. Mostly these differences are about differences in the professions students are being trained for. The life of a research scientist involves working in labs with a team. A humanities professor's research is mostly going to be done alone. If you can't go off to a library or archive for a month or two and sort through stuff without writing your advisor every step of the way, you aren't going to be able to succeed in the profession.

[financeguy]

My anecdotal evidence is just that, a collection of anecdotes that are reinforced by what one would naturally expect the incentives in place to produce. Those I know who have completed or are in the process of completing programs in arts and humanities fields often have difficulty getting their advisor to do simple things like respond to emails on time. In one case, someone whose funding was going to expire was pulling teeth to get the advisor to respond to messages regarding the scheduling of a defense. When I look at people taking 6, 8, 10 years complete a degree program in a fluff field, I have to assume the primary motivator is low labor cost for mundane intro courses. You don't even have to get into how outrageous it is that most of them are there to begin with given a job market that is unlikely to produce a position for most graduates.

Can I deduce that these people "don't care" about their students? Of course not in the binary yes/no posing of the question. But in a world of trade offs, I can much more easily do so. Every wife, for example, would say yes to "do you want your husband to follow his dream." but given the prospect of that pursuit leading to less likelihood of a white picket fence with the SUV in the garage, limiting her ability to pop out kids and play a life of "what'll the neighbors think" in a suburb environment, her concern for "that dream thingy" goes out the window pretty quickly in favor of wanting the mule to pull. I assume all faculty would "prefer" their students succeed, but it what's best for them involves less students to teach their own 101 intro courses or less people eventually flooding an already crowded market, which would simultaneously limit their own prestige of "having a graduate program," those same faculty would cease to be as concerned. Nothing is easier than caring without cost.

[Caracal]

My anecdotal evidence is just that, a collection of anecdotes that are reinforced by what one would naturally expect the incentives in place to produce. Those I know who have completed or are in the process of completing programs in arts and humanities fields often have difficulty getting their advisor to do simple things like respond to emails on time. In one case, someone whose funding was going to expire was pulling teeth to get the advisor to respond to messages regarding the scheduling of a defense. When I look at people taking 6, 8, 10 years complete a degree program in a fluff field, I have to assume the primary motivator is low labor cost for mundane intro courses. You don't even have to get into how outrageous it is that most of them are there to begin with given a job market that is unlikely to produce a position for most graduates.

Can I deduce that these people "don't care" about their students? Of course not in the binary yes/no posing of the question. But in a world of trade offs, I can much more easily do so. Every wife, for example, would say yes to "do you want your husband to follow his dream." but given the prospect of that pursuit leading to less likelihood of a white picket fence with the SUV in the garage, limiting her ability to pop out kids and play a life of "what'll the neighbors think" in a suburb environment, her concern for "that dream thingy" goes out the window pretty quickly in favor of wanting the mule to pull. I assume all faculty would "prefer" their students succeed, but it what's best for them involves less students to teach their own 101 intro courses or less people eventually flooding an already crowded market, which would simultaneously limit their own prestige of "having a graduate program," those same faculty would cease to be as concerned. Nothing is easier than caring without cost.

I've heard plenty of stories about science labs where the faculty member treats students like indentured labor, behaves abusively and hogs all the credit for the work other people did. I don't assume things like that are the norm, just an abuse of the system. Some advisors in the humanities do ignore their students, but as someone actually in the field, I can tell you that it isn't the norm.

I can also tell you that you're off base about teaching intro courses. What courses grad students teach is rarely going to have much effect on the teaching responsibilities of a faculty member. It just doesn't actually work that way.

Comparing completion times between humanities and science fields doesn't make that much sense. If you added in postdocs I'm pretty sure humanities degrees take about as long as science ones. Six years is completely standard because it takes time to research and write a dissertation. People take longer than that for all kinds of reasons, but advisor inattention isn't a big factor. The crummy job market plays more of a role.