Statistics for Physics/EE Majors: Is it Necessary?

[osnarf]

Hi again everyone. I am revising my schedule for next semester because a schedule conflict came up, and am wondering, if i take statistical theory i (and ii later), would this be useful to me? I am a physics / electrical engineering major w/ math minor, i wish to go to grad school for physics, although i am not sure exactly which field (i am waiting until i get to the higher level physics classes to decide what most interests me). I have already taken the required prob & stats for sci & engineers. I don't particularly like statistics, but I am wondering if going deeper would be more useful and/or interesting?

As a side note, i took this class a while ago and am rusty on it. If you think that I need more statistics, should i go back and reread the book from my prob & stats class? Or will statistical theory basically go over everything we did again, just in more depth?

Thanks again for your help everyone, you all are great!

 

[snipez90]

If the course devoted a huge amount of time to probability then take it. I'm certain probability theory is useful in physics (e.g. in statistical mechanics) but statistics in general is not. The culmination of many statistics courses seems to be hypothesis testing (after various limit theorems) which I imagine would be useless to many sub-disciplines in physics.

On the other hand, knowing statistics could turn out to be helpful if say, you decide you don't want to do physics in the end. It's a just very practical tool useful in a variety of disciplines.

 

[osnarf]

STA 4321 COS-STAT 3(3,0)
Statistical Theory I: PR: STA 2023 or STA 3032;
CR: MAC 2313. Probability axioms, discrete and continuous sample spaces, conditional probability, independence,
one-dimensional random variables, moment generating
functions, transformations, jointly distributed random variables.

STA 4322 COS-STAT 3(3,0)
Statistical Theory II: PR: STA 4321. Conditional distributions, sums of random variables. Chebyshey’s inequality,
central limit theorem, method of movements, maximum
likelihood, confidence intervals, hypothesis testing, transformations of two random variables.

Information taken from ucf.edu/academics.

You tell me :) Sounds like 4321 is about half probability. Not sure about the other

 

[snipez90]

4321 is ALL probability theory. Definitely take it if you've never been exposed to probability theory. It is a math course, though probability theory underlies everything in statistics.

I would go with 4322 later if you liked 4321, since it is unsurprisingly a logical extension of 4321. The topics starting from central limit theorem and ending at hypothesis testing will probably give you an idea of the practical uses of statistics, so it's not that bad even if you don't care at all about that stuff.

 

[osnarf]

So do you think I should re-read as much as i can first or just go into it?

 

[chiro]

Hi again everyone. I am revising my schedule for next semester because a schedule conflict came up, and am wondering, if i take statistical theory i (and ii later), would this be useful to me? I am a physics / electrical engineering major w/ math minor, i wish to go to grad school for physics, although i am not sure exactly which field (i am waiting until i get to the higher level physics classes to decide what most interests me). I have already taken the required prob & stats for sci & engineers. I don't particularly like statistics, but I am wondering if going deeper would be more useful and/or interesting?

As a side note, i took this class a while ago and am rusty on it. If you think that I need more statistics, should i go back and reread the book from my prob & stats class? Or will statistical theory basically go over everything we did again, just in more depth?

Thanks again for your help everyone, you all are great!

If you are interested in information theory and applications (like telecommunications for example) then I would not hesitate in taking it.

 

[symbolipoint]

Statistics or Elementary Statistics was (maybe still is) missing as an official requirement for many undergraduate programs. Best guidance is that if you have an opportunity (or just need to keep attention on some kind of continued Mathematics learning), then study it, even Elementary Statistics such as you'd find at a c.c. You really can use Statistics in many scientific courses and fields - biological, social/behavioral, and certainly physical sciences.

 

[Chaostamer]

Statistical Theory I and II are, in total, about 75% probability theory--at least in my school. I've taken the first part and thought it was a pretty interesting course and probability is useful overall, so I'd say to go for it.

If you don't mind me asking, though, where do you go to school? That course description is identical to that of MY Stat Theory course and I'm curious how many other places use the same--or at least a similar--catalogue.

 

[Rap]

I suspect the reason you found statistics uninteresting is because many statistics courses focus on rote memorization of methods of statistical analysis, without any insight into the theory behind it. Statistical theory will probably provide such an insight, and will be much more interesting. If you plan on doing experimental physics of any sort, I would say definitely take the course. Analysis of experiments cannot be done without it. For theoretical physics, the statistical methods used could be picked up from the particular discipline that uses them (e.g. quantum physics or statistical mechanics).

 

[osnarf]

Chaostamer: University of Central Florida, yourself?

Rap: Dead on. This is how you do this - this is when you do this. I suspect that's the for engineer's part? :) I do plan on doing experimental physics, so it sounds like i should take it. Can i assume then I will be good just taking the course or do i need to go back over it thoroughly?

Symbolipoint: Probability and Statistics for Engineers is a requirement for my EE major. I already took it, i was just wondering if i needed a refresher or if we would be starting from the ground up again, but with more depth.

 

[symbolipoint]

osnarf, I will have to let someone else answer that, "i was just wondering if i needed a refresher or if we would be starting from the ground up again, but with more depth". As long as you took the course that your program required, it must be better than not having the course. Some physical science type programs had not required a specific Statistics course, but merely gave a little bit of Statistics instruction incorporated into one or two of programs' courses. That kind of treatment is not nearly as good as even an Elementary Statistics course.

 

[Chaostamer]

Chaostamer: University of Central Florida, yourself?

Oh, that is funny. Same here. For all I know, then, we could've been in some of the same classes for the last couple semesters.

 

[osnarf]

haha yeah, small world!


Turns out Statistical Theory 1 is fall only, i was looking at 2 :P
This appears to be all probability, so it would then be useful too?

MAP 4113 COS-MATH 3(3,0)
Probability, Random Processes and Applications I: PR: MAC 2313. Elementary probability theory; modes of convergence; martingales, Gaussian, Wiener, and diffusion processes; Brownian motion; applications. Occasional

 

[mtaboga]

Studying probability theory and statistics in depth is always a good investment. By the way, when you go beyond the intermediate level, it is also an excellent way to see real-world applications of advanced maths (e.g. functional analysis, measure theory). Why don't you have a look at some intermediate and advanced textbooks to understand how much you really know already and how much you still need to learn. For a concise advanced treatment of probability theory, have a look at Williams' 'Probability with martingales'. For a more elementary treatment have a look at Ross' 'Introduction to probability models'. There is also very good online material, for example http://www.probability.net"

 

[Chaostamer]

haha yeah, small world!


Turns out Statistical Theory 1 is fall only, i was looking at 2 :P
This appears to be all probability, so it would then be useful too?

MAP 4113 COS-MATH 3(3,0)
Probability, Random Processes and Applications I: PR: MAC 2313. Elementary probability theory; modes of convergence; martingales, Gaussian, Wiener, and diffusion processes; Brownian motion; applications. Occasional

Oh, yeah. With the UCF Math department, basically every class after MAS 3105 is one-semester only. The Probability class sounds really interesting (it is all Probability) and I've been wanting to take it myself, but I hear it's really difficult with Dr. Pensky. I know someone who tried to take it last year and he said it was the bane of his semester.

Of course, it's possible that he was being dramatic.

 

[Bohrok]

Statistics or Elementary Statistics was (maybe still is) missing as an official requirement for many undergraduate programs. Best guidance is that if you have an opportunity (or just need to keep attention on some kind of continued Mathematics learning), then study it, even Elementary Statistics such as you'd find at a c.c. You really can use Statistics in many scientific courses and fields - biological, social/behavioral, and certainly physical sciences.

I was just going to post a question about this...
Would elementary statistics be good to take for someone wanting to focus on math, physics, or computer science? I'm not really going to school now but was thinking about taking an elementary stats class and was wondering if it would be of any use to me later since I'm not sure what I want to get into.

 

[Rap]

I was just going to post a question about this...
Would elementary statistics be good to take for someone wanting to focus on math, physics, or computer science? I'm not really going to school now but was thinking about taking an elementary stats class and was wondering if it would be of any use to me later since I'm not sure what I want to get into.

I would say it would be least useful for computer science, somewhat for math, somewhat for beginning theoretical physics (statistical mechanics and quantum mechanics) and absolutely vital for advanced theoretical physics and absolutely vital for any level of experimental physics.