an extract of wishes and ratings for wishlist reports, please
Author: ulrichmerkel@web.de (ulrich-merkel)
Hi Ton,
From the past I got the impression that only a few members rated (some 2 or 3)
so if I and 2 friends rate a "10", I get TopOfTheList no matter what obscure wish I entered.
Not many members commented their rating (especially the 1s).
Saw it from some wishes:
2 * "10" + 6 * "1" adds to 3.3
As a "real uniface developer" I ranked this wish as "mandatory to allow proper coding".
So is it possible to provide an EXCEL extract showing the wishes aand the rating and how many members voted?
TIA, Uli
P.S. would you please add a satus "solved in next Uniface version" to the "planned in next version"
This will give the members more planning security that these function will be available.
4 Comments
Local Administrator
Hi Uli,
At the moment we are updating status of the wishes and provide feedback where possible. This process will take a couple of weeks.
When this is done I will publish a list of wishes per type and priority. The rating will be calculated as avg_rate * number of votes. Based on these results and the existing German and US wishlist statistics we will plan a number of wishes for Uniface 9.5.
In the status of wishes we always state "Planned in the next version". When the version has been offcially released this status changes to "Solved in version x.y.z"
kind regards
Ton
Author: None (None)
Local Administrator
Hi Ton,
thank you for the very fast reply.
on the "solved in"/"planned in" issue: OK, I understand CPWR is cautious giving promises.
But I am a bit confused about your procedure of calculating "TopOfTheList":
If you have some time (I know: wishful thinking)
perhaps you can document the process a bit more detailed
to get some feedback like the following:
given 2 wishes one 100 * '1' = 100, the other 9 * '10' = 90.
The wish where 9 people give highest rating ranks lower
than one where 100 people give lowest rating.
In the german usergroup (c-b-g.org) we had only a "I support this wish" counter
which I think is a better measure than the "uprate/downrate" one above:
One wish will end up with no supports while the other has 10 supports and ranks mile high.
Success, Uli
P.S. I spent a lot of time in university with statistics and
the pros and cons of different scaling measures
I know, it is very hard to find a way which gives minority
demands (like overall support of RTL, more editing support)
the fair chance they deserve.
Author: ulrich-merkel (ulrichmerkel@web.de)
Local Administrator
Hi Ulli, I understand your concern.
Looking at your statistics background, what would be an appropriate formula? We know the average rating and the number of votes.
Would for instance (average rating) * (number of votes)2 (squared) work better?
regards Ton
Author: None (None)
Local Administrator
Hi Ton,
this is a hard question.
For the moment you have the following parameters collected
(please correct me when I'm wrong)
- the number of voters
- the average ranking (sum of all rankings / number of voters)
I will include a list of different goals to be achieved.
Give me some days so I can compile some scenario calculations
and the effect on the different goals to incubate further discussions.
Success, Uli
Author: ulrich-merkel (ulrichmerkel@web.de)