<p>
The general rule is that if your program (or external source, like NIH or NSF) cannot or will not offer you full support, then you are probably going to struggle academically and/or professionally in the field, all while accumulating massive debt. As such, it is not usually recommended to enter a grad program that is not funding you. In some fields (sciences, engineering) the substantial majority will get full funding, in others (humanities, arts) only a few will receive full funding and some will receive partial.</p>
<p>As to the types, the three most common types of funding are (by increasing prestige) teaching assistantships, research assistantships, and fellowships:</p>
<p>Teaching assistantships (TA’s) cover your tuition and fees and pay a stipend in return for 10-20 hours a week helping a professor teach a class - this might entail grading, writing exams, running seminars, leading recitation groups, giving lectures, and/or supervising lab sections. TA positions are usually the worst-paying options, as little as $15k for 9 months.</p>
<p>Research assistantships also cover tuition and fees and pay a slightly more generous stipend, rarely less than $20k in the sciences. In return, you spend 10-20 hours a week supporting your advisor’s research efforts. Hopefully, most of this will support YOUR research as well, so this usually takes up a little less time than being a TA, since you are essentially getting paid to do part of your own thesis.</p>
<p>Fellowships are the best option, giving you decent pay (usually in the $25-40k range) with NO WORK EXPECTATION. This lowers your workload and may allow you to graduate faster, and also gives you more liberty to study (RA and TA positions are generally tied to a particular advisor, so switching advisors may mean searching for new funding… something fellows don’t go through!). The downside is that in most fields they are offered only to a select few.</p>
<p>
Depends on a lot of factors. In my field (EE), I was told that with a 3.5 you could get in at top-5 programs, a 3.2 you could get into top-25 programs, and with a 2.8 you could probably get in somewhere, assuming that you had some research experience as an undergrad, decent GRE’s, and good letters of recommendation. But grad admissions are holistic, and deficiencies in one area can often be made up by relative strength in another area (excluding the GRE, which is all but worthless most of the time).</p>