<div dir="ltr">Ok. Please stand by....<div><br></div><div>It seems like the datatracking library, when used to crawl for a large amount of drafts, pulls an index and then does calls to the datatracker web API for calls the the draft metadata.</div><div><br></div><div>So I've had to write a new data collection script, similar to the script we use for scraping the mailing lists, to get the draft data. It's a slower process. But I should be able to compute these results once I have them downloaded locally.</div></div><br><div class="gmail_quote"><div dir="ltr" class="gmail_attr">On Wed, Aug 26, 2020 at 4:09 PM Joey S <<a href="mailto:joeysalazar@article19.org">joeysalazar@article19.org</a>> wrote:<br></div><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex">
  
    
  
  <div>
    +1 to dnsop, their drafts are also quite numerous and with a very
    active mailing list.<br>
    <pre cols="72">--
Joey</pre>
    <div>On 26-Aug-20 1:25 PM, Niels ten Oever
      wrote:<br>
    </div>
    <blockquote type="cite">
      
      <div dir="auto">Httpbis is the one you're looking for :)<br>
        <br>
      </div>
      <div dir="auto">DNSops is also a nice big one.<br>
        <br>
      </div>
      <div dir="auto">Cheers,<br>
        <br>
      </div>
      <div dir="auto">Niels</div>
      <div class="gmail_quote">On Aug 26, 2020, at 21:17, Sebastian
        Benthall <<a href="mailto:sbenthall@gmail.com" target="_blank">sbenthall@gmail.com</a>>
        wrote:
        <blockquote class="gmail_quote" style="margin:0pt 0pt 0pt 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex">
          <div dir="ltr"> Hmmm.
            <div> <br>
            </div>
            <div> Web mail archives of the http list at  <a href="https://ietf.org/mail-archive/text/http/" target="_blank">https://ietf.org/mail-archive/text/http/</a>
              only go up to 2012. </div>
            <div> Does that make sense to you? </div>
            <div> <br>
            </div>
            <div> It looks like there are several DNS working groups.
              Any one in particular you think would be worth looking at?
            </div>
            <div> <br>
            </div>
            <div> Genericizing the code so that it can loop through many
              groups and compute results is the next step towards
              confirmation. Probably worth looking at a couple other
              concrete and well-understood examples before doing the big
              analysis though. </div>
            <div> <br>
            </div>
            <div> - S </div>
          </div>
          <br>
          <div class="gmail_quote">
            <div dir="ltr" class="gmail_attr"> On Wed, Aug 26, 2020 at
              1:52 PM Niels ten Oever < <a href="mailto:mail@nielstenoever.net" target="_blank">mail@nielstenoever.net</a>>
              wrote: <br>
            </div>
            <blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex">
              <div>
                <div dir="auto"> Very interesting. I'd say the number if
                  drafts and authors in hrpc is too low to make a
                  statement about this though. Could we do this for the
                  HTTP and/or DNS WGs ? </div>
                <div class="gmail_quote"> On Aug 26, 2020, at 19:30,
                  Sebastian Benthall < <a href="mailto:sbenthall@gmail.com" target="_blank">sbenthall@gmail.com</a>>
                  wrote:
                  <blockquote class="gmail_quote" style="margin:0pt 0pt 0pt 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex">
                    <div dir="ltr"> Hello,
                      <div> <br>
                      </div>
                      <div> I'm revisiting the question of whether
                        mailing list gender diversity and draft
                        productivity of working groups are correlated. </div>
                      <div> <br>
                      </div>
                      <div> Putting aside for now all the methodological
                        complications, here is how I am operationalizing
                        the question: </div>
                      <div>
                        <ul>
                          <li>I'm looking specifically at the HRPC
                            working group, with this data:<br>
                            <div> <img alt="image.png" width="418" height="221"> <br>
                            </div>
                          </li>
                          <li>
                            <div> Gender is being detected based on
                              first name birth records. "unknown" is
                              used for cases that cannot with the
                              current data set be determined as either
                              men or women. </div>
                          </li>
                          <li>I'm measuring "diversity" on any day as:
                            (women's activity + unknown's activity) /
                            (men's activity). Because, you know, this is
                            probably close to what most people probably
                            mean by diversity. (Recall that non-Western
                            names are more likely to be categorized as
                            "unknown".)<br>
                          </li>
                          <li>I'm using a 100 day rolling average on the
                            activity counts.</li>
                        </ul>
                        <div> This is the matrix of Pearson correlations
                          between each of these values: </div>
                      </div>
                      <div> <br>
                      </div>
                      <div>
                        <table border="1">
                          <thead> <tr style="text-align:right">
                              <th><br>
                              </th>
                              <th>women</th>
                              <th>unknown</th>
                              <th>men</th>
                              <th>drafts</th>
                              <th>diversity</th>
                            </tr>
                          </thead> <tbody>
                            <tr>
                              <th>women</th>
                              <td><font color="#0000ff">1.000000</font></td>
                              <td><font color="#0000ff">0.910922</font></td>
                              <td><font color="#0000ff">0.804869</font></td>
                              <td>0.008890</td>
                              <td>0.160833</td>
                            </tr>
                            <tr>
                              <th>unknown</th>
                              <td><font color="#0000ff">0.910922</font></td>
                              <td><font color="#0000ff">1.000000</font></td>
                              <td><font color="#0000ff">0.808168</font></td>
                              <td>0.027502</td>
                              <td>0.245059</td>
                            </tr>
                            <tr>
                              <th>men</th>
                              <td><font color="#0000ff">0.804869</font></td>
                              <td><font color="#0000ff">0.808168</font></td>
                              <td><font color="#0000ff">1.000000</font></td>
                              <td>0.015406</td>
                              <td>-0.141915</td>
                            </tr>
                            <tr>
                              <th>drafts</th>
                              <td><font color="#cc0000">0.008890</font></td>
                              <td><font color="#cc0000">0.027502</font></td>
                              <td><font color="#cc0000">0.015406</font></td>
                              <td>1.000000</td>
                              <td><font color="#cc0000">0.061884</font></td>
                            </tr>
                            <tr>
                              <th>diversity</th>
                              <td><font color="#674ea7">0.160833</font></td>
                              <td><font color="#674ea7">0.245059</font></td>
                              <td><font color="#674ea7">-0.141915</font></td>
                              <td>0.061884</td>
                              <td>1.000000<br>
                              </td>
                            </tr>
                          </tbody>
                        </table>
                        <br>
                        Things to note: </div>
                      <div>
                        <ul>
                          <li><font color="#0000ff">The activity of each
                              gender is correlated with the activity of
                              other genders.</font></li>
                          <li><font color="#674ea7">Diversity is
                              anticorrelated with the number of men.
                              This is expected based on how it was
                              defined, and a good sanity check.</font></li>
                          <li><font color="#cc0000">Draft output is MORE
                              correlated with diversity than it is with
                              any individual gender!</font></li>
                        </ul>
                        <div> <font color="#000000">This last point is
                            quite nice. It resonates with the work of
                            Scott Page on the value of diversity to
                            collective intelligence, for example.</font>
                        </div>
                        <div> <font color="#000000"><br>
                          </font> </div>
                        <div> <font color="#000000">These numbers are a
                            bit hard to interpret. How much should we
                            trust them? These are the <i>p</i>-values
                            associated with each correlation:</font> </div>
                        <div>
                          <table border="1">
                            <thead> <tr style="text-align:right">
                                <th><br>
                                </th>
                                <th>women</th>
                                <th>unknown</th>
                                <th>men</th>
                                <th>drafts</th>
                                <th>diversity</th>
                              </tr>
                            </thead> <tbody>
                              <tr>
                                <th>women</th>
                                <td>0</td>
                                <td>0</td>
                                <td>0</td>
                                <td><font color="#cccccc">0.6925</font></td>
                                <td>0</td>
                              </tr>
                              <tr>
                                <th>unknown</th>
                                <td>0</td>
                                <td>0</td>
                                <td>0</td>
                                <td><font color="#cccccc">0.221</font></td>
                                <td>0</td>
                              </tr>
                              <tr>
                                <th>men</th>
                                <td>0</td>
                                <td>0</td>
                                <td>0</td>
                                <td><font color="#cccccc">0.493</font></td>
                                <td>0</td>
                              </tr>
                              <tr>
                                <th>drafts</th>
                                <td><font color="#cccccc">0.6925</font></td>
                                <td><font color="#cccccc">0.221</font></td>
                                <td><font color="#cccccc">0.493</font></td>
                                <td>0</td>
                                <td><font color="#ff0000">0.0059</font></td>
                              </tr>
                              <tr>
                                <th>diversity</th>
                                <td>0</td>
                                <td>0</td>
                                <td>0</td>
                                <td><font color="#ff0000">0.0059</font></td>
                                <td>0</td>
                              </tr>
                            </tbody>
                          </table>
                        </div>
                        <br>
                      </div>
                      <div> Generally, <i>p</i>-values below .01 are
                        considered "statistically significant", i.e.
                        publishable. </div>
                      <div> This correlation between diversity and draft
                        output makes the cut!! </div>
                      <div> <br>
                      </div>
                      <div> <font color="#0000ff">So the verdict is:
                          for HRPC, YES, gender diversity is correlated
                          with draft output.</font> </div>
                      <div> <font color="#0000ff"><br>
                        </font> </div>
                      <div> <font color="#000000">This result is robust
                          to transformations of the activity scores into
                          the log space, which is comforting.</font> </div>
                      <div> <span style="color:rgb(0,0,0)">Further work
                          is needed to see if this result is robust
                          across other IETF working groups.</span> </div>
                      <div> <span style="color:rgb(0,0,0)"><br>
                        </span> </div>
                      <div> <font color="#000000">Nick, what would you
                          say to including a result like this in the
                          paper about IETF and gender?</font> </div>
                      <div> <font color="#000000"><br>
                        </font> </div>
                      <div> <font color="#000000">Cheers,<br>
                          Seb</font> </div>
                      <div> <br>
                      </div>
                    </div>
                    <pre>      <hr>
Bigbang-dev mailing list
<a href="mailto:Bigbang-dev@data-activism.net" target="_blank">Bigbang-dev@data-activism.net</a>
<a href="https://lists.ghserv.net/mailman/listinfo/bigbang-dev" target="_blank">https://lists.ghserv.net/mailman/listinfo/bigbang-dev</a>
</pre>
                  </blockquote>
                </div>
              </div>
            </blockquote>
          </div>
        </blockquote>
      </div>
      <br>
      <fieldset></fieldset>
      <pre>_______________________________________________
Bigbang-dev mailing list
<a href="mailto:Bigbang-dev@data-activism.net" target="_blank">Bigbang-dev@data-activism.net</a>
<a href="https://lists.ghserv.net/mailman/listinfo/bigbang-dev" target="_blank">https://lists.ghserv.net/mailman/listinfo/bigbang-dev</a>
</pre>
    </blockquote>
    <br>
  </div>

</blockquote></div>