Research grants: how thing work

[virtua]

Hi, guys, I just wondering how things work in research assistantships.

I mean, ok, I understand professor wins grant for some research and he/she hires a student for doing that research, right?

Does the professor have some deadline for doing that research?

Do all money from won grant belong to the professor? I mean he/she is not required to pay the university for something, right? He/she decides himself/herself how to spend this money, right?

So, I just wondering, how this money is payed, to whom it is payed, how it is spent.

[Eigen]

Generally, Universities take 50% or more off the top of the research grant as "overhead" to pay for electricity, water, and lab/office space. Usually, the money comes to and is held by the University, with funds available to the professor  

There arent deadlines, per se, but usually yearly progress reports. Often funding is doled out per year, over a set number of years. 

Grants usually have have budgets that say what the money can and can't be used for- supplies, equipment, salary. 

A a lot depends on the grant and granting agency though. 

[fuzzylogician]

^This. "Overhead" is written into the grant in addition to what the PI expects to pay in direct costs for equipment, salary, etc. This is done automatically by the Office of Sponsored Programs, which is responsible for submitting the grant to the granting agency. So if you wrote a budget for $1K of expenses (to keep things simple) and your institution takes 75% overhead, the grant you'll end up submitting will be for $1,750.

The rest of the money is made available to the university either by year or as a lump sum by the granting agency, and the PI can access the funds through the university. The PI can either make purchases through the university and pay directly with the money allocated to them through the grant, or they can pay with their own funds and be reimbursed from the grant (this is what we often do when we pay participants in our experiments -- we get cash and then get reimbursed against receipts we show our admin person once the experiment is done). This obviously also varies by school and granting agency. 

You write a budget with a description of what the money will be used for. There is some leeway, certainly if the research takes you in a different direction than planned then you can pay your RA to follow whatever lead seems most promising now as opposed to what you'd planned before actually doing the work. You can perhaps move some money around and use e.g. more money for equipment than planned and less on participants, but it's not like you can just do anything. You need to show receipts for purchases and you write yearly and final reports, and if you deviate from the plan you need to be able to justify it. 

[TakeruK]

Fuzzy and Eigen answered almost everything. 

I just want to touch on one question you asked---about the deadline to complete research. Although this could vary from field to field and grant to grant, ultimately, a grant is not directly paying for a set of results or outcomes. For example, a professor might write a grant to compute the orbits of 500 asteroids and ask for funding to pay a graduate student for 1 year to complete this work. If it is funded, the professor will have the money to pay for this.

If the graduate student spends an entire year working on this project but in the end, only is able to determine the orbits of 400 asteroids after one year, then that would be bad for the professor. However, it's not like the funding agency will ask for 20% of the money back or that your professor will take 20% of the graduate student's stipend back. Instead, when the professor writes the annual/end of grant report, they will have to explain why only 400 asteroid orbits were computed. Maybe there is a legitimate reason and maybe there isn't (i.e. just an underperforming student or research group). Either way, this could reflect poorly on the professor if they apply for future grants with this agency. 

[rising_star]

Fuzzy and Eigen answered almost everything. 

I just want to touch on one question you asked---about the deadline to complete research. Although this could vary from field to field and grant to grant, ultimately, a grant is not directly paying for a set of results or outcomes. For example, a professor might write a grant to compute the orbits of 500 asteroids and ask for funding to pay a graduate student for 1 year to complete this work. If it is funded, the professor will have the money to pay for this.

This is true. That said, some grants do ask or require you to submit at least one manuscript for publication or give a presentation on the research they funded within a certain time frame. 

[TakeruK]

This is true. That said, some grants do ask or require you to submit at least one manuscript for publication or give a presentation on the research they funded within a certain time frame. 

Definitely and the grants I've been part of applying for tend to always ask for a list of previous grants awarded and how many papers/presentations etc. came as a result of those grants. And sometimes multi-year grants have annual reviews so if you don't meet stated production goals, you won't get renewed. But I have not yet experienced a grant that require a researcher to give back money already awarded/paid out because they didn't meet publication goals. 

[rising_star]

TakeruK, I haven't heard of anyone asking for money back either but, as budgets get tighter, that could start to happen.