[Talk-us] [HOT] Surveys and studies

Charlotte Wolter techlady at techlady.com
Mon Sep 4 20:41:16 UTC 2017


John,

         Actually, I would have to disagree with you on that point.
Having worked in survey research, I know that each survey is
carefully constructed as to length and order of questions. A kind
kind of grouped survey, as you described, would give an experience
to respondents that would be different from the one the researcher
wanted.
         I just would like to see something that told me that the
supposed research really is, for example, from Zoe at Nottingham.
Using her university email would have been advisable. At least there
would have been a good chance that it was legit.
         And, maybe, requiring an institutional email is the best and
simplest way to go. That way it likely is real academic research.
Otherwise, anyone could post a survey, even someone who did not
wish OSM well, and who wished to use the "honest" opinions for
negative purposes.
         Anyway, that's my take on it.

Charlotte


At 01:20 PM 9/4/2017, you wrote:
>
>
>At Statistics Canada they have a concept of 
>respondent burden. Basically it means you try to 
>limit the number of questions you ask people 
>whilst still trying to get the answers in. Is 
>this information available from another survey?
>May I suggest a more formal arrangement where a 
>survey is organised say every three / six months 
>and researchers submit their questions to be 
>included in the survey? This is done for a 
>number of surveys at Statistics Canada and is a 
>useful way to include one or two additional / 
>supplementary questions to a survey.
>The advantage to the people running the studies 
>is hopefully a wider set of respondents making 
>their surveys more statistically valid.  The 
>advantage to the mappers are fewer messages in 
>the mailing lists and fewer surveys asking to be completed.
>At the very least I'm sure Zoe and Laura could 
>see if the data from their surveys could feed the other.
>There are other things that could be done such 
>as random sampling then following up with the 
>randomly selected sample this reduces self selection.
>Thoughts?
>
>Thanks John
>_______________________________________________
>HOT mailing list
><mailto:HOT at openstreetmap.org>HOT at openstreetmap.org
>https://lists.openstreetmap.org/listinfo/hot

Charlotte Wolter
927 18th Street Suite A
Santa Monica, California
90403
+1-310-597-4040
techlady at techlady.com
Skype: thetechlady

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstreetmap.org/pipermail/talk-us/attachments/20170904/2ee1c7d2/attachment.html>


More information about the Talk-us mailing list