Looking for a statistics course

[lilihoso]

Hi everyone,

I'm doing a research subject about the correlation between the lunar tides and earthquakes, but I've never study the statistics. Could some one suggest me a online course or something like that to give me some ideas of how to study the correlation between events?

Thank you in advance!


Lilihoso

 

[chiro]

Hi everyone,

I'm doing a research subject about the correlation between the lunar tides and earthquakes, but I've never study the statistics. Could some one suggest me a online course or something like that to give me some ideas of how to study the correlation between events?

Thank you in advance!

Lilihoso

Hey Lilihoso and welcome to the forums.

Do you want to understand only how to calculate quantities and know how to interpret in a basic way them or do you want understand how to calculate, what it means, and a deeper understanding.

My guess is the first one which means first getting a software package and then looking at some tutorials on how to first get your data into the package and then using the right tool to get some quantitative information.

I should point out that you will at the very least have to check the assumptions of your algorithm as well as for correlation in particular, look at the plot of x vs y to see whether the correlation output makes any sense, whether a non-linear as opposed to a linear relationship may exist and whether a correlation seems to exist in general despite low coeffecient output.

This is not a complete answer and I'd probably recommend that you ask someone with strong statistical knowledge about this because the answer will depend on your data, the nature of your data (the context of your data), the assumptions used and also finally the outputs that are generated both visually and numerically.

 

[lilihoso]

Hi Chiro,

Thank you for your reply!

Actually, I'll need to know more than that, I'm looking for some courses (or documents) mathematical about this subject, but not too complicated, just some definitions, applications, examples for the correlation study, and I'll try to write a program to compute it with my data.

Thank you!

 

[chiro]

Hi Chiro,

Thank you for your reply!

Actually, I'll need to know more than that, I'm looking for some courses (or documents) mathematical about this subject, but not too complicated, just some definitions, applications, examples for the correlation study, and I'll try to write a program to compute it with my data.

Thank you!

Well in terms of basic introductory statistics (which includes correlation at a basic level), you can either use a library or google for this.

If you are interested in software packages, some popular ones for statistical analysis include things like SAS and R. SAS is a commercial software package that pretty much does everything but costs heaps, and R is also something that is really really flexible when you look at how many libraries have been written for it, but in some cases you will have to learn how to use the commands (not just point and click and drag and drop like R is).

If you want real understanding of what is going on, and knowing when to use something, then you will need to study this in a lot of mathematical detail.

If you want to just use something and read the help manual for how to use it and when to use it, then just look up 'calculating blah for blah2' where blah is the technique and blah2 is the platform (SAS, R, etc) and for SAS you will get a help page full of this kind of thing. R will give you many hits as well.

The thing is that you could probably do this yourself and even do it carefully and right, but you probably won't understand the subtleties and intracacies involved in the analyses and this is what a statistician needs to do and why they guide researchers through the whole project life-cycle which includes knowing when to use things and when not to use them. This is caution that I wish to give you, and if you are using this for say writing a paper, or a proposal that may impact on you in a big way (reputation smeared, loss of job, etc), then its vital you speak to a statistician.