So will we be reviewing the data as anonymous entries or anonymous
aggregates? If I can see that Apache has only one over-65 person, that
doesn't tell me anything other than the single over-65 person filled
out the survey. However, if I can see each entry with an anonymous id
instead of their name, then yes, that is fairly easy to reverse back
to names.

On Thu, 5 Dec 2019 at 22:26, Patricia Shanahan <[email protected]> wrote:
>
> On 12/5/2019 6:17 PM, Patricia Shanahan wrote:
> > On 12/5/2019 3:58 PM, Mark Thomas wrote:
> > ...
> >> Not sure if this is the right venue. Feel free to redirect me.
> >>
> >> Working my way through the survey, I'm finding that - for me at least -
> >> it would be trivial to de-anonymise my response. It isn't going to
> >> affect my answers but I wonder if it will affect the responses of others.
> >>
> >> I'm not sure what could be done about this. We need the demographic data
> >> but that combined with out relatively small population and relatively
> >> detailed data available about each committer's activities makes
> >> de-anonymisation easier.
> >>
> >> I think this is something to file under "things to take into
> >> consideration for the next survey" but I wanted to raise it while it was
> >> fresh in my mind.
> >
> > I have the same concern. One thing that can be done even now is to
> > establish a minimum size of group for which statistics will be broken
> > out for publication.
>
> In a board@ thread, which I am staying away from, Dave Fisher remarked:
>
> > Elsewhere someone suggested that answers may be enough to impute who
> > filled in that set of answers. I would say those are well known board
> > members. The survey is out. Let’s analyze results in a couple
> > months.
> Although inferring identity from one or two answers is probably limited
> to a few board members, combinations can be very powerful. My survey
> answers show that I am over-65, female, PhD, PMC and ASF member, UK
> origin, US resident, and sometimes attend face-to-face meetings but not
> often. That is probably not a very large demographic.
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: [email protected]
> For additional commands, e-mail: [email protected]
>


-- 
Matt Sicker <[email protected]>

---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]

Reply via email to