Re: Emacs Survey: Toolbars

[Christopher Dimech]

That was a criticism on using weighting to account for missing data.  Weighting 
is
customarily used to reduce noise levels or as a sparsity constraint.  One 
possibility
would be a live survey (on the emacs website, a mailing list, ...).  Apologies 
if
the comment was too harsh.  The basic criticism has been that it favoured 
experienced
users.  Another way is to have multiple surveys based on emacs use experience, 
and then
normalise the results between the different experience levels.  In that way 
things
won't get skewed.

I would think that if there are competing views, we must not favour one or the 
other,
but consider them equally valid.  It would then be simply a question of 
priority or
perhaps measured on difficulty.  There is use for toolbars and should not be 
discounted.
Another way is to isolate functionality so those who want to use then can 
install a package.

Ultimately, it is for the maintainers and contributors to decide.  There are 
many books on
statistical inference.   A popular technique is to divide the data into groups 
followed by
a data representation strategy based on the most significant clusters.  One can 
also study
correlations between groups as being statistically significant.

> Sent: Tuesday, December 22, 2020 at 4:28 PM
> From: "Gregory Heytings via Emacs development discussions." 
> <emacs-devel@gnu.org>
> To: "Christopher Dimech" <dimech@gmx.com>
> Cc: emacs-devel@gnu.org
> Subject: Re: Emacs Survey: Toolbars
>
>
> Christopher Dimech:
> >
> > Shitty data obtained without care, attention, and skill is useless data.
> > Forget the weighting!
> >
>
> As the saying goes: "Criticism is easy, and art is difficult."  Could you
> please create and conduct a "good" survey, according to your criteria of
> "good"?  Or at least enlighten us, poor mortals, and explain what should
> have been done, and how?
>
>