[Theforum] survey update?

[Luther, Ron]

Thanks guys!  

I was wondering if I was going a little too far with that.

Hey, sometimes survey questions bomb and ya throw 'em away. La Vie!
Sometimes they tell you a lot all by themselves.  ... and sometimes they
only have value when related (crossed) with other questions --- If we
had a second question where *all* the folks who reported themselves as
'female' answered "a", and *all* the folks who reported themselves as
'male' answered "b" -- if might give us valuable insight that could be
used to improve evolt ... or it may trigger more questions.  {"Why don't
any females have meo accounts?" - to make up a (hopefully) wrong example
to illustrate the point.}

Heck - We might even take an extra step and 'assume' that any non-gender
respondent who answered "a" was 'female' ... yet another way to make a
guess to allocate the responses. We might be wrong - but if you see a
consistent pattern across a number of questions - it might make a better
estimate too!     ;-)


That said, there can be a pretty subtle line between understanding and
reporting what the data is telling you and (excuse me) 'spinning' the
results:

* If we say evolt is 12% female ... what connotation are we
communicating to folks who don't know us?  Is it "<yawn> ... yet another
'geeky guy' site"?  ... or "Hot Dang - that's waaay better than over
here at 'sum_udder_list.com'"? ... or something else entirely?

* If we say evolt is 56% male ... what connotations would that have?

Sometimes there is as much "artsy-craftsy" stuff involved in analyzing
survey results as hard numbers and statistical techniques.


If you *know* your client you can usually figure out how far to go or
not go.  Since we're publishing to the world, I think I might tend to
err a little on the conservative side, yeah? 


Then again ... there was this old rule of thumb that any project was a
success if you didn't have to throw out more than 30% of your data to
insure results that conformed to your expectations!   ;-)

RonL.
(Who actually joined Mensa once ... to meet chicks! ... Yep - I did! ...
{What was that line in 'Crouching Tiger' about smart people doing
incredibly stupid things?} ... Bad idea ... Very bad idea ... [Who?  No,
no ... not me -- I got in on their 'quota' system -- they already had a
bunch of folks who were smart in the head, see ... and they needed a
smart-ass as a counterweight!])

-----Original Message-----
From: spinhead [mailto:evolt at spinhead.com]
Subject: Re: [Theforum] survey update?

Hmmmmmmmmm . . . I was surprised at how high the 'female' portion was.
We're
a high-tech engineering startup, and when I started I expected a large
female population (no, that's not why I'm here) but we're at 15%
overall,
and only 5% in engineering.