Hi WSC,
thanks for the detailed feedback. Here's some background on what guided the design of the survey:
Negatively phrased questions All the questions were compiled based on actual input we collected during the pilot. As outrageous as it may sound, editing Wikipedia as a potential threat to one's career is an actual reason some participants reported and we want to assess its importance as a potential barrier. If you google "tenure blogging" you will get an idea of how hot the debate has been around the question whether blogging is a risk or not for one's career (especially for non-tenured faculty). Something similar may happen (or may have happened in the past) for Wikipedia and we want to capture this i the survey.
Starting the section with... The order of the questions in each block is randomised for each participant, so it was just bad luck :)
Individual motivations vs shared perception This is the most delicate issue of the survey, for which we considered several alternate designs. We want to contrast the participants' perception of Wikipedia participation in general with one's individual motivation to contribute or not to contribute. The rationale for this is that we noticed that participants often can dissociate their judgment on statements they would endorse as members of a professional category ("editing Wikipedia does not count towards improving one's CV") from judgments on what represents an individual barrier to participation ("even if editing Wikipedia does not improve my CV, this is not a reason for me not to contribute"). Phrasing questions as regarding one's peers is a way to have a participant focus on shared perception as opposed to individual motivation. The reason why we put general questions upfront and individual-motivation questions at the end is that we believe the noise added by framing questions "from general to individual" is less important (for what we want to study) than if we ordered them "from individual to general". To address this properly we should randomise the order of the two blocks, not just the order of questions within each block, that's something I can look into.
Dario
On 7 Feb 2011, at 19:00, WereSpielChequers wrote:
I appreciate that phrasing questions neutrally can be somewhat difficult. But I think there are too many leading or negatively phrased questions to produce useful information. Starting the section with "Editing Wikipedia may damage one's scientific reputation" rather sets an anti wikipedia tone.
Also I'd switch the sequence between individual and collective perception. Putting the section about the individual before the collective section would start with something that respondents should more easily be able to answer - some people simply won't feel that they can answer questions on behalf of people in their field if they are unaware of those other people's views.
WereSpielChequers
On 7 February 2011 12:16, Dario Taraborelli dtaraborelli@wikimedia.org wrote:
All,
the Expert Barrier survey is ready to run. We took into account most of the suggestions we received during the pilot phase and we are planning to release the survey by Wednesday morning London time, in time for the Imperial College Recruitment Drive [1]. Here is a link to a live preview:
http://nitens.org/ls/?sid=21693
(some blocks of questions are hidden depending on how participants answer in the first screen when asked whether they ever contributed to WP)
If you have a moment to give it a try, comments are welcome by Tuesday night PST.
Best, Dario, Giota and Daniel
[1] http://en.wikipedia.org/wiki/Wikipedia:CONTRIB/Imperial _______________________________________________ RCom-l mailing list RCom-l@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/rcom-l
RCom-l mailing list RCom-l@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/rcom-l