[Gllug] VACANCY: Site Reliability Engineering

[John G Walker]

On Thu, 19 Feb 2009 19:44:39 -0500 Balbir Thomas
<balbir.thomas at gmail.com> wrote:

> 1) It does NOT follow from the data that "income drops at and beyond
> the Ph.D level".
> 
> 2) What does follow (as you yourself indicate) is that
>        a) salaries in academics are lower than in industry
>        b) there is a concentration of the highly educated in academics
> rather than industry
>        c) hence "average" salaries of the highly educated are lower
> than that of other people of
>            say the same age (what else ?) elsewhere. (Actually even
> this does not follow from
>            the data since there is no indication if comparison was
> done in the same age
>            group. Besides was the non academic group a random
> selection from the population
>            or was it selectively CEOs of industry who didn't have a
> Ph.D. Do you see what I am
>            getting at ?)
> 
> To demonstrate that income drops at and beyond the Ph.D level a
> statistician would have to
> compare "average" salaries of Ph.Ds in industry with other levels of
> education in industry (or
> for that matter in academics).
> 
> 1) and 2) above are two different conclusions !


I don't understand this.

My points were:

1) That the story that salaries drop when you get a PhD has a basis in
the data

2) but that this is an artifact of averaging oranges and apples - that
the data are not stratified in technical terms, and that when the data
are stratified then the effect disappears.

I've tried to relate what you wrote to this and failed. I can't see how
you've read that I have two conclusions (nor can I tell, from above,
what you think they are).

Perhaps you'd like to explain it a little clearer?

-- 
 All the best,
 John
-- 
Gllug mailing list  -  Gllug at gllug.org.uk
http://lists.gllug.org.uk/mailman/listinfo/gllug