As an applied statistician, scientist and engineer, I entirely agree with Steve

Messenger and I delete postings by VG as a nuisance of no value.

There is a lot of overt hyperbole and utter nonsense published about the very

dubious value of so-called "data science".

John Thompson

From: Steve Messenger <[log in to unmask]>
To: [log in to unmask]
Sent: Sunday, 16 March 2014, 7:08
Subject: Re: For Moderator and ALL: Do we need [blog announcements] on Allstat ?

Old cynics might argue that getting a degree in Data Science is akin to having a degree in drawing graphs.

I was told recently of a presentation by Facebook that showed how they had undertaken tests to see if ads were more effective when sent to people whose 'friends' had already given the ad differing amounts of 'likes'

This used thousands of data points and one could see the effect of the difference it make to click- throughs etc. And it did make a difference.

So a great example of the use of 'big data'.

Yet all the analyses were exactly the same as test and control experiments using much lower numbers of data points.

What was taken away was the statistical risk because the numbers of data points was so large that you did not so much 'analyse the data' but 'weigh it'.

As far as I can see 'Data Science' uses statistics in the same way that other non-statistical professions would do (i.e. in medicine, economics, psychology, human relations and so on).

So for me VG's posts are largely irrelevant most of the time and just get deleted.

Steve

Sent from my iPhone

On 15 Mar 2014, at 07:02, Nick Sofroniou <[log in to unmask]> wrote:

I have also been uncomfortable with the way Allstat is being used in this way. Do we need to have an announcement every time someone updates their regular blog. Clearly the intention is to generate web traffic to the posters' website. Allstat is not a discussion list so why should we receive these personal viewpoints. In my view this is quite different from someone asking a query and perhaps posting a summary of replies back to the list.

I get the digest version of Allstat to keep my email inbox less cluttered, but instead I have to wade through all this spam within the digest message.

Please moderators, do something about this before we get 10 regular bloggers realising that they can up their web traffic using Allstat.

Nick

_________________________

Date: Fri, 14 Mar 2014 19:27:12 +0000
From: "gilbert.mackenzie" <[log in to unmask]>
Subject: For Moderator and ALL: Do we need Granville on Allstat ?

Dear All

Most people recognize that Granville's contributions are a form of
advertising.

The question I raise is simply should this be allowed on Allstat?

My own view is that we do not need these contributions especially given
the frequency
of postings.

I never read his material - which is automatically directed to a junk
directory.

But this is only one view - so what do others think ?

Best

Gilbert
> Vincent Granville:
>
> Your "best-kept" secret is actually quite well-known. The method is
> just Gram-Schmidt orthogonalisation, something I learnt from John
> Tukey at Princeton University more than 40 years ago. I like your
> chutzpah, but I don't think it's superior in any way to the lm()
> function in R.
>
> regards....Don
> You may leave the list at any time by sending the command
>
> SIGNOFF allstat
>
> to [log in to unmask] <mailto:[log in to unmask]>, leaving
> the subject line blank.
>
_____________________________

You may leave the list at any time by sending the command

SIGNOFF allstat

to [log in to unmask], leaving the subject line blank.

You may leave the list at any time by sending the command

SIGNOFF allstat

to [log in to unmask], leaving the subject line blank.