<html>
<head>
<meta http-equiv="Content-Type" content="text/html; charset=UTF-8">
</head>
<body>
<div class="moz-cite-prefix"><font face="Helvetica, Arial,
sans-serif">On 1/10/2021 3:34 PM, Mike Thompson wrote:<br>
</font></div>
<blockquote type="cite"
cite="mid:CALJoUkuTkFWNPcPUr4MAH=HC_ju-cveViBqNX0EpHyvsnOdCDQ@mail.gmail.com"><font
face="Helvetica, Arial, sans-serif">Have we defined the
population (sorry if I missed it)?</font></blockquote>
<div class="moz-signature"><font face="Helvetica, Arial, sans-serif">Please
see my reply to John Whelan, relevant excerpt follows:</font></div>
<div class="moz-signature">
<blockquote type="cite"><font face="Helvetica, Arial, sans-serif">We
will circulate the URL for the survey via talk lists, which
will allow anyone with a valid email address to request a
single-use token that will allow the respondent to participate
in the survey once.</font></blockquote>
<font face="Helvetica, Arial, sans-serif">In effect that means the
population consists of anybody interested in the OSM project.
Any respondent will be able to request a single-use token and
participate in the survey. Will there be biases? Of course
there will be. That's why we are asking for demographic
information, to help us parse the biases. Will the survey be
perfect? No survey is perfect. <br>
</font></div>
<div class="moz-signature"><font face="Helvetica, Arial, sans-serif"><br>
</font></div>
<div class="moz-signature"><font face="Helvetica, Arial, sans-serif">With
regard to this comment:</font></div>
<div class="moz-signature">
<blockquote type="cite"><font face="Helvetica, Arial, sans-serif">The
law of large numbers doesn't correct for selection bias. </font></blockquote>
<font face="Helvetica, Arial, sans-serif">That is why we are
seeking not to select respondents, as some have advocated, but
rather to solicit participation from the broader community by
broadcasting through as many communication channels as possible.
Self-selection will almost certainly be subject to the biases of
a) participation in OSM-related communications media, b) access
to a computer and the Internet, and c) time to spare for
completing a survey. There may also, depending on the sample
size and response, be a bias toward those who care about OSM
Foundation affairs versus those who don't give a rip and would
rather spend their time mapping than answering survey
questions. I don't know what we can do to remediate those
biases, nor do I think they are fatal flaws in the methodology.
Again, awareness of potential biases underlies the demographic
questions in the survey.<br>
</font></div>
<div class="moz-signature"><font face="Helvetica, Arial, sans-serif"><br>
</font></div>
<div class="moz-signature"><font face="Helvetica, Arial, sans-serif">apm</font></div>
-------<br>
<div class="moz-signature">
<i>Allan Mustard, Chairperson</i><br>
<i>Board of Directors</i><br>
<i>OpenStreetMap Foundation</i></div>
</body>
</html>