Am 03.06.2021 um 20:00 schrieb Bruno Wolff III
<bruno(a)wolff.to>:
Given that badges are being handed out, it is clear the survey is not anonymous.
People that do surveys love to claim they are anonymous when they aren't, to increase
the response. When you are collecting detailed information from someone it is hard to
prevent the person from being identified by that data even if you don't collect their
identity directly.
From a point of social survey: if it is hard to identify a person by the survey data
depends of the data and the sample. Given the questions of the Fedora survey it will be
hard. You would need additional data, e.g. the IP address or other user specific data
stored together (or linked) to the answers (as all those „social“ media try). That would
be a gross breach of trust and malpractice. And the standard software for online surveys
does not include it as an option.
If you have people authenticate or send out links to the survey that
are customised per person (perhaps to prevent someone from filling out the survey
repeatedly) then the survey is clearly not anonymous. Even if you don't do that, web
server logs can often reveal who the responders are.
From a point of social survey: Anonymous refers to the attribution of the response graph
to an individual (rsp. the lack of), not whether someone participated.
In my opinion it is best to state what you are actually doing with
regards to privacy rather than simplify it and make a statement that is clearly false.
It’s not clearly false, but nevertheless a somewhat more detailed note might be useful.