3

I am a grad student. I am struggling with core statistics courses: probability, statistical inference.

In probability, for example, prove unbiased estimator for uniform. In statistical inference, find the UMVUE for an estimator.

American stat dept has either highly qualified foreign students, or the couple of Americans who are struggling.

I would like to learn this and have a good foundation. Looking to pass the Ph.D. exam in the future.

All I do is study, but still get 60/100 on my midterms.

My problem is, the prof never tests on what he taught, but there are always students that passes, and continue on.

I don't know what to do anymore. What are your suggestions?

user13985
  • 836
  • 4
  • 12
  • 20
  • 1
    Are you going to office hours? Many times the instructor is not telling what is on the test during class, but during sessions in his office. – Ben Voigt Jan 29 '14 at 04:39
  • 3
    Unfortunately, this doesn't appear to be a question about [statistics](http://stats.stackexchange.com/help/on-topic) but about effective learning strategies. Perhaps more importantly, "what are your thoughts" is not a question with a clear, supportable-with-references answer, but simply invites opinion (making it off topic everywhere on the network). Either way, it's off topic here. – Glen_b Jan 29 '14 at 04:44
  • Agree with Glen. We can't help you if you don't provide us with more narrow focus to what your problem is. Moreover, if you're a student in grad school, you already know how to study pretty effectively. – rocinante Jan 29 '14 at 04:53
  • Perhaps more constructively, consider the advice in the [help](http://stats.stackexchange.com/help/dont-ask) on asking questions. If you focus your question appropriately as suggested there, it might be on topic at say the academia stackexchange, or perhaps with additional changes, might come close enough to be on topic here. – Glen_b Jan 29 '14 at 06:56
  • @Glen_b Would "suggestion" be better? I will look up a few examples later today to illustrate what I'm saying. – user13985 Jan 29 '14 at 15:35
  • @BenVoigt I do. I can't get helped on every questions. And in class, all you see is proof, not hw related questions. I really google my way through hw's. – user13985 Jan 29 '14 at 15:37
  • One thought to consider: (leaving aside my reservations about the teaching method), it's not actually up to the professor to teach you the material (you are not a vessel to be filled), but really to help you teach it to yourself. This is doubly the case for mathematics, and much more so again for someone looking to do a PhD. As such it's absolutely necessary to go beyond the immediate material, both in terms of reading and exercises. Staring at the same page of notes or the same exercise for hours, hoping for inspiration, is useless. ... (ctd) – Glen_b Jan 29 '14 at 21:50
  • (ctd)... with topics I have, or had difficulty with (and the list is so long I would be embarrassed to type it out, though much of it can be found on here for all to see, I am sure), I try to find at least six books (either directly on the material if it's not so new as to have no books, or on background material otherwise) and also to read (not always in depth) many many papers relating to the topic. I read notes online if I can find them, and attempt exercises (which can often be found). Actual course content was only ever a signpost of where I needed to be heading in my learning. ... (ctd) – Glen_b Jan 29 '14 at 21:55
  • (ctd) ... now, many many years after completing work on my PhD, there are no courses for me, but I am learning as much as ever. There are some constants - definitions matter (a lot). Skills and practice matter. And there are nowadays very convenient places to ask questions online (like this one). – Glen_b Jan 29 '14 at 22:03
  • @Glen_b I am so thankful to read your words. I will give that a try and report back. In general, I try to dig into the math library. Too often, relevant books are checked out. Also, not everyone in a intro grad class (30 or so people), could obtain multiple books. Most schools don't have the money. – user13985 Feb 02 '14 at 17:39
  • @Glen_b I talked to my peers who are mostly foreign students from Asia or Europe, they have really learned, fed, stuffed with the material as an undergrad. For the Americans who made into Ph.D., either their parents were profs or they have some leg up in the game. – user13985 Feb 02 '14 at 17:41

1 Answers1

1

Theory is important but I find to truly gain insight you need to do exercises and understand data, usually through programming statistics (like R). This link and this link have some good resources, have a dig for something appropriate for you. It is absolutely fine to start at the bottom level (everybody does so sometime), the important thing is to keep going.

Community
  • 1
Hans Roggeman
  • 893
  • 6
  • 9
  • true that, but without getting the theory down, empirical practice could only go so far. For example, I am great R programmer, I am poor at probably theory. I can't publish papers. – user13985 Feb 02 '14 at 17:43