# These numbers, they vibrate?

July 28, 2010 1:54 PM Subscribe

How do I become a stats and data whiz?

I currently work in marketing analytics. I really like the analytics (both general and marketing) field in general but would like to become more proficient with statistics, data modeling, and data mining.

I know the basics of stats like how to do a t-test in SPSS or the like but am interested in more advanced regressions and using packages like R and SAS.

Thinking about doing an MBA in stats but wondering if there's another route for this.

My goal is both for personal and career development. I'd be interested in doing in-person, online, books, etc... and don't mind spending some money. I'm in NYC for reference.

I currently work in marketing analytics. I really like the analytics (both general and marketing) field in general but would like to become more proficient with statistics, data modeling, and data mining.

I know the basics of stats like how to do a t-test in SPSS or the like but am interested in more advanced regressions and using packages like R and SAS.

Thinking about doing an MBA in stats but wondering if there's another route for this.

My goal is both for personal and career development. I'd be interested in doing in-person, online, books, etc... and don't mind spending some money. I'm in NYC for reference.

I think that if you work through the examples and problems in one (or several) of John Fox's books (An R and S Plus Companion to Applied Regression, Applied Regression Analysis and Generalized Linear Models) or Andrew Gelman's books (Data Analysis Using Regression and Multilevel/Hierarchical Models, Bayesian Data Analysis) you'd be in a pretty good spot.

The thing is that you have to actually *do* the examples and problems. What you get with a class is a) someone who will encourage you to do the work b) someone who will grade your assignments and respond to questions.

I agree that a stat methods class is fine to start, and that a stats MS may be overkill.

posted by a robot made out of meat at 2:37 PM on July 28, 2010 [4 favorites]

The thing is that you have to actually *do* the examples and problems. What you get with a class is a) someone who will encourage you to do the work b) someone who will grade your assignments and respond to questions.

I agree that a stat methods class is fine to start, and that a stats MS may be overkill.

posted by a robot made out of meat at 2:37 PM on July 28, 2010 [4 favorites]

I agree with the previous posts.

I don't know what goes into an MBA with an emphasis on stats. I'd say that an MS in stats, as in from a stats department, wouldn't be so much overkill as irrelevant to your needs.

To be sure, they're likely to approach it with even more math than an econometrician would, but the really big divide is that statisticians often seem to have a radically different orientation to the subject than do people who use statistics as inferential tools. Especially in relation to theories about the world.

You know your field better than we do, but you might ask whether for career advancement you'll need some sort of formal certification of having learned the stuff.

posted by ROU_Xenophobe at 2:57 PM on July 28, 2010

I don't know what goes into an MBA with an emphasis on stats. I'd say that an MS in stats, as in from a stats department, wouldn't be so much overkill as irrelevant to your needs.

To be sure, they're likely to approach it with even more math than an econometrician would, but the really big divide is that statisticians often seem to have a radically different orientation to the subject than do people who use statistics as inferential tools. Especially in relation to theories about the world.

You know your field better than we do, but you might ask whether for career advancement you'll need some sort of formal certification of having learned the stuff.

posted by ROU_Xenophobe at 2:57 PM on July 28, 2010

I'm an unemployed marketing analyst but when employed I had access to "data guys" that were consultants to do the complicated stuff aside from the simple quantitative stats took as an undergrad. These guys would have Masters and PhDs in stats or some stats-esque discipline. I wasn't aware there was an MBA in stats. The MBA people I know are painfully clueless in pretty basic concepts of statistics. These were bright guy from big fancy schools. I gathered the stats course they had at the graduate level were not much more than the intro to stats classes you'd get as an undergrad.

As they mention above, courses at your local university to get your feet wet and explore is you want to go back and get a masters.

I've noticed in this market that almost all but entry level positions as marketing analysts require an MBA (in anything) is "preferred" or "a plus" so that might help you advance your career. When jobs want a degree in stats, it is always a masters.

posted by birdherder at 3:44 PM on July 28, 2010

As they mention above, courses at your local university to get your feet wet and explore is you want to go back and get a masters.

I've noticed in this market that almost all but entry level positions as marketing analysts require an MBA (in anything) is "preferred" or "a plus" so that might help you advance your career. When jobs want a degree in stats, it is always a masters.

posted by birdherder at 3:44 PM on July 28, 2010

I have an MBA in corporate finance, not stats. Close enough though., I wouldn't go the MBA route - it's too general. A MBA "major" is only going to consist of 3 or 4 extra classes in your chosen specialty. I'd go the MS in statistics , or maybe math with a statistics concentration, route. Unless maybe there is some speciality MS in Marketing Analytics, or something like that.

posted by COD at 4:50 PM on July 28, 2010

posted by COD at 4:50 PM on July 28, 2010

Stanford Center for Professional Development offers a decent selection of graduate level statistics classes through distance learning. It's not cheap but a lot of employers will pay for this sort of course, and my experiences have been mostly good with their stats professors. Then if you decided you wanted to get a degree in statistics you could hopefully get letters of recommendation from in some cases quite big names in academic statistics.

posted by little light-giver at 8:47 PM on July 28, 2010

posted by little light-giver at 8:47 PM on July 28, 2010

Also check Cosma Shalizi's notebooks, which contains a ton of books/articles on stats (ctrl+F and search for statistics, Cosma is a Stats Professor).

Key notebooks include:

Teaching Statistics (has a number of beginner stats books)

Probability Theory

Computational Statistics (various books on R, and others)

And finally, Statistics, which contains a load of other links to more specialized areas of stats with numerous books.

posted by ollyollyoxenfree at 10:55 PM on July 28, 2010

Key notebooks include:

Teaching Statistics (has a number of beginner stats books)

Probability Theory

Computational Statistics (various books on R, and others)

And finally, Statistics, which contains a load of other links to more specialized areas of stats with numerous books.

posted by ollyollyoxenfree at 10:55 PM on July 28, 2010

R is fun, just start using it.

posted by jeffburdges at 5:24 AM on July 29, 2010

posted by jeffburdges at 5:24 AM on July 29, 2010

you could check out Open University in the UK - stats choices website is quite a good resource and Open Lab (open educational resources space) has some pretty decent stuff too if you just want to use the materials without getting formal qualifications. Although I'd thoroughly recommend Open University, doing a Diploma in Statistics as we speak.

posted by coffee_monster at 1:19 PM on July 29, 2010 [1 favorite]

posted by coffee_monster at 1:19 PM on July 29, 2010 [1 favorite]

Thanks for the answers all. Looking into some local options as well as the resources in the linked answers. Let me know if you think of anything else.

posted by jourman2 at 2:16 PM on July 30, 2010

posted by jourman2 at 2:16 PM on July 30, 2010

This thread is closed to new comments.

posted by paultopia at 2:23 PM on July 28, 2010