# Business Statistics: A Synergistic Customer-Focused Experience

May 26, 2009 7:52 PM Subscribe

What are your best practices for learning Business Statistics?

I have just been accepted into my local MBA program, the first step on my path to using the word 'synergy' without irony. One of my first classes is Business Statistics, which I have been advised is largely based in Excel. Apparently Hawkes Learning Systems: Statistics is the source of much of the homework. I took Statistics a long,

I would like to do as much as possible in the next three months to ensure that I do as well as I can. What kinds of things do I need to focus on to achieve this outcome? What do you wish you had been told when you were taking Statistics in your MBA program?

I have already looked up all these questions on the best books. But if you have a great website or resource that you think will help, please feel free to share it.

I have just been accepted into my local MBA program, the first step on my path to using the word 'synergy' without irony. One of my first classes is Business Statistics, which I have been advised is largely based in Excel. Apparently Hawkes Learning Systems: Statistics is the source of much of the homework. I took Statistics a long,

*long*time ago in my undergraduate career, and while I enjoyed it at the time I don't remember very much.

I would like to do as much as possible in the next three months to ensure that I do as well as I can. What kinds of things do I need to focus on to achieve this outcome? What do you wish you had been told when you were taking Statistics in your MBA program?

I have already looked up all these questions on the best books. But if you have a great website or resource that you think will help, please feel free to share it.

First, I wouldn't worry too much. Statistics isn't difficult if it's taught well, and if it's in a MBA program and is taught using excel, chances are it'll be

Second, the key to stats IME, and for that matter most math past algebra, is the problem sets. Solving mathematical problems -- of the sorts that will appear on exams as well as the sorts that will appear in, you know, that real world thing -- is a techne -- and that was the most useless invocation of greek ever -- but it's an evocative concept. What I mean is that it's impossible to learn to do math simply by listening to someone explain the rules. The problem sets are where the real learning happens, where you beat your head against problems that seem far harder than the examples gone over in class, until you've beaten long enough, and then everything clicks, and it all falls into place, and suddenly you actually understand how to do that problem.

So pay close attention to problem sets. That's my best best practice. Don't cop out on them, don't do a half-assed job on them, do them all, right, meet with the prof frequently for help if you can't, etc.

posted by paultopia at 12:54 AM on May 27, 2009

*very*easy relative to the kind of stats class that a specialist would have to learn. Real stats wonks do not use excel. (This is coming from my social scientist's perspective, where the statistical demands are rather heavy.)Second, the key to stats IME, and for that matter most math past algebra, is the problem sets. Solving mathematical problems -- of the sorts that will appear on exams as well as the sorts that will appear in, you know, that real world thing -- is a techne -- and that was the most useless invocation of greek ever -- but it's an evocative concept. What I mean is that it's impossible to learn to do math simply by listening to someone explain the rules. The problem sets are where the real learning happens, where you beat your head against problems that seem far harder than the examples gone over in class, until you've beaten long enough, and then everything clicks, and it all falls into place, and suddenly you actually understand how to do that problem.

So pay close attention to problem sets. That's my best best practice. Don't cop out on them, don't do a half-assed job on them, do them all, right, meet with the prof frequently for help if you can't, etc.

posted by paultopia at 12:54 AM on May 27, 2009

Try to understand the concept of a model and that every statistical analysis (even the so called non-parametric) is based on one. Learn the diffrence of sample and population. Find out what a parameter is. Learn what an estimate is.

Learn the common distributions (normal, binominal, poisson) and try to understand how they link to each other.

posted by ilike at 1:34 AM on May 27, 2009

Learn the common distributions (normal, binominal, poisson) and try to understand how they link to each other.

posted by ilike at 1:34 AM on May 27, 2009

If anyone tries to tell you that stock market returns are normally distributed, run. Fast. And then drop your MBA program.

posted by dfriedman at 10:08 AM on May 27, 2009

posted by dfriedman at 10:08 AM on May 27, 2009

This thread is closed to new comments.

A useful online source is this: http://www.itl.nist.gov/div898/handbook/

posted by a womble is an active kind of sloth at 8:25 PM on May 26, 2009