Page 1 of 1
Another survey, another botched sampling methodology
Posted: Wed Jun 08, 2016 3:01 pm
by thekohser
You'll see that the Wikimedia Foundation
wants our help with finding a new Executive Director!
As usual, the WMF has really stepped up its careful sampling methodology:
Our survey is currently open in the top 10 wiki languages. A sample of
editors from various languages have been invited to participate and we are
also sending an invitations to anyone here and through our networks.
So, anyone can participate, and if you want to participate about 12 or 13 times, that's okay, too! (Just use different IP addresses.)
Time is of the essence, though --
the survey is only going to be open another day, unless they extended the one-week plan.
Re: Another survey, another botched sampling methodology
Posted: Wed Jun 08, 2016 3:17 pm
by Kingsindian
I wonder why they even bother with this kind of stuff. It's clear that they have no intention of looking at the data. Another example comes from the Inspire campaign "
Leaderboard":
An idea's placement in this list does not signify that it is better or worse than any other idea, and the number of endorsements an idea receives does not determine whether that idea will be selected for funding..
Then perhaps they shouldn't call it "Leaderboard", with ideas sorted according to number of endorsers? Just a thought.
Re: Another survey, another botched sampling methodology
Posted: Wed Jun 08, 2016 8:48 pm
by Poetlister
thekohser wrote:So, anyone can participate, and if you want to participate about 12 or 13 times, that's okay, too! (Just use different IP addresses.)
For bonus marks, contribute in three different languages.
Re: Another survey, another botched sampling methodology
Posted: Wed Jun 08, 2016 9:04 pm
by Hex
Poetlister wrote:thekohser wrote:So, anyone can participate, and if you want to participate about 12 or 13 times, that's okay, too! (Just use different IP addresses.)
For bonus marks, contribute in three different languages.
Why stop there? You could put a photo of someone from the internet on your user page and pretend that it's you.
Re: Another survey, another botched sampling methodology
Posted: Thu Jun 09, 2016 10:18 am
by thekohser
Hex wrote:Poetlister wrote:thekohser wrote:So, anyone can participate, and if you want to participate about 12 or 13 times, that's okay, too! (Just use different IP addresses.)
For bonus marks, contribute in three different languages.
Why stop there? You could put a photo of someone from the internet on your user page and pretend that it's you.
The survey never asks for the respondent's User page.