> I've recently come into web analytics and I don't have any kind of

basics?

> statistical background. I'd like to be able to play around with some

> modeling and try to make some predictions and such with the more

> advanced tools in Excel and open source tools like Scilab.

>

> Can anyone recommend a good book that would help me get to grips with the

It's not clear to me what you are asking for help with. If you want to

understand statistics-based modeling, get a stats book "for Dummies" or the

like, they usually explain things fairly clearly and will tell you what is

going on behind the "black box" of some of the analytical engines. If you

want something more robust, take a look at "Data Mining Your WebSite" by

Jesus Mena. It's a bit dated but it does a good job of explaining the guts

of machine modeling and how it applies to the web.

Also, search for "CHAID" and "CART" which are two decision tree models that

are a great introduction to machine modeling, and there is free software

around that will run the models, see:

http://www.kdnuggets.com/software/classification-tree-rules.html

This will allow you to begin understanding the models, and that may be

enough for you. If you want to learn what to *do* with the information

from the models once you get it, how to think about the analysis and turn

it into marketing action, (Eric permitting) I might humbly suggest my book:

http://www.jimnovo.com

Subscribe to the newsletter and get the first 9 chapters by e-mail. This

material will give you a very good idea of what the book is about.

There are tons of models out there; some are good for some things and

others for other things. The biggest stumbling block people usually have

is figuring out what to analyze, and then doing something sensible with the

results. That's what my book is about; very simple models anyone can use

to improve their marketing results, especially in online retailing.

Jim

jim@...

http://www.jimnovo.com- I guess I'm asking two questions.

The first, how can I become familiar with the basics of statistics, is

probably not a question suitable for this forum. But it is answered by

buying a "for Dummies" book. Eric Hansen recommended "Reading Between

The Numbers - Statistical Thinking In Everyday Life" by Joseph Tal

offlist.

The second question is really where can I go after using Excel and web

analytics/measurement packages to learn interesting things about my

client's sites? I'm really just interested in what's possible, what

services I can develop. This question looks like it would quite

probably be answered by your book and "Data Mining Your Website".

So my original question is answered, even if I phrased it poorly.

Thanks very much,

Ben

On 21 Dec 2004, at 16:35, Jim Novo wrote:

>

>> I've recently come into web analytics and I don't have any kind of

>> statistical background. I'd like to be able to play around with some

>> modeling and try to make some predictions and such with the more

>> advanced tools in Excel and open source tools like Scilab.

>>

>> Can anyone recommend a good book that would help me get to grips with

>> the

> basics?

>

> It's not clear to me what you are asking for help with. If you want to

> understand statistics-based modeling, get a stats book "for Dummies"

> or the

> like, they usually explain things fairly clearly and will tell you

> what is

> going on behind the "black box" of some of the analytical engines. If

> you

> want something more robust, take a look at "Data Mining Your WebSite"

> by

> Jesus Mena. It's a bit dated but it does a good job of explaining the

> guts

> of machine modeling and how it applies to the web.

>

> Also, search for "CHAID" and "CART" which are two decision tree models

> that

> are a great introduction to machine modeling, and there is free

> software

> around that will run the models, see:

>

> http://www.kdnuggets.com/software/classification-tree-rules.html

>

> This will allow you to begin understanding the models, and that may be

> enough for you. If you want to learn what to *do* with the information

> from the models once you get it, how to think about the analysis and

> turn

> it into marketing action, (Eric permitting) I might humbly suggest my

> book:

>

> http://www.jimnovo.com

>

> Subscribe to the newsletter and get the first 9 chapters by e-mail.

> This

> material will give you a very good idea of what the book is about.

>

> There are tons of models out there; some are good for some things and

> others for other things. The biggest stumbling block people usually

> have

> is figuring out what to analyze, and then doing something sensible

> with the

> results. That's what my book is about; very simple models anyone can

> use

> to improve their marketing results, especially in online retailing.

>

> Jim

> jim@...

> http://www.jimnovo.com

((Ben Godfrey) (Software) (see "http://www.cohack.com/?src=eval"))