On 13 Feb 2009, at 10:36, Florian Anwander wrote:

> Hello Bakis
>
> > and i have to admit that i am one to blame for some misleading
> poll results...
> I think this tells us, that the questions asked in the polls are not
> good.
>
> For future polls you should always provide selections like that:
>
> 1.) I will buy this module
> 2.) I think this module is useful, but I won't buy it (at least not
> now)
> 3.) I think Doepfer should NOT offer this module
> 4.) I have no opinion on this module
>
> Selection 1 should result in the minimum sales Dieter will have
>

Not necessarily.
For instance, I tick this choice simply because I like the idea of
this module and I *might* buy it some time in the future, but because
the way this poll is worded, I have no other choice of selecting this
option if I like this module (even if I'm not gonna buy it, for some
reason or another ~ say, I ran out of money).

> Selection 2 will show a general positive direction. Modules with this
> voting will assumingly have an allover chance of success, though the
> affirmation within the is not that good.
>

This option is rather useless, since it doesn't tell anyone anything
really.
Let's assume that ALL modules are useful (and indeed, in theory, they
are -- there will always be at least one person who'd like to use/have
such a module), but that doesn't necessarily make them financially
interesting for Doepfer...
So, choices 1 and 2 are actually the same: I am interested in this
module [but in fact that is no guarantee that I'm going to buy it].
Reasoning:
- choice 1 is not a 100% guarantee for a sale
- choice 2 *just might* become a sale
So, regardless if 1 or 2 are chosen, in Doepfer's eyes these both say
"maybe a sale".

> Selection 3 will separate the "no"s caused by individual aversion
> ("this
> module is only useful for techno and I hate techno") from more general
> signs of unsuccessfulness.
>

Again, not necessarily.
There can be no real reason in saying that a module should not be
offered -- what's the point, if you know that at least one person is
interested (the one whom suggested the module in the first place), and
you yourself simply don't care about this module "Not caring about"
does not translate to "should not"...

> Selection 4 sounds not very expressive, but this may be the most clear
> signal to not produce this module.
>

Choise 4 is a bit too abstract to use as a viable selection model.
My bet is that quite a few members simply select this choice because
they wanted to show their support as a member by voting, but don't
care about this particular module. So in reality this option does not
tell Doepfer anything, except for the fact that there are members who
are willing to express the opinion that they don't have an opinion.
Hmm...


To me, a good poll is nothing more than a "rating system", a pointer,
to show Doepfer the amount of interest in a particular module.
Therefore, any hints towards "buying" or "not buying" should be left
out, just as any emotional tendencies should be left out as well.

I feel that a simple rating system going from "not interested" over
"mildly interested" to "interested" and arriving at "strongly
interested" will tell Doepfer all that can possibly be extracted from
such polls.

_g

>
>
> We already had some polls with those selections, and they always
> showed
> a much clearer (often less affirmative) picture.
>
> Florian
>