<html><head><meta http-equiv="Content-Type" content="text/html charset=iso-8859-1"></head><body style="word-wrap: break-word; -webkit-nbsp-mode: space; -webkit-line-break: after-white-space; ">Nothing wrong with rescaling person measures using multiplication and addition both - everybody does it. I just wanted to stress that one should not use multiplied item locations directly for equating, etc. (shifts up/down are fine - because person estimates can easily be adjusted).&nbsp;<div><br></div><div>Intuitively, this is because when the item parms are multiplied non-trivially, the observed (raw) data of the equating sample that you are going to use do NOT naturally come so multiplied (I don't know how one would do this to observations)</div><div><br></div><div>Rense</div><div><div><br><div><div>On May 3, 2013, at 7:08 PM, Tom Conner &lt;<a href="mailto:connert@msu.edu">connert@msu.edu</a>&gt; wrote:</div><br class="Apple-interchange-newline"><blockquote type="cite">
  
    <meta content="text/html; charset=ISO-8859-1" http-equiv="Content-Type">
  
  <div bgcolor="#FFFFFF" text="#000000">
    Doesn't linear transformation involve both multiplication and
    addition/substraction?<br>
    <br>
    <div class="moz-cite-prefix">On 5/3/13 7:49 PM, Rense Lange wrote:<br>
    </div>
    <blockquote cite="mid:9B8FB3AC-586A-4BC7-A201-08B32184DE30@gmail.com" type="cite">
      <meta http-equiv="Content-Type" content="text/html;
        charset=ISO-8859-1">
      To be sure: multiplications shouldn't be done to Rasch calibrated
      item parameters (adding - subtracting something is fine). Except,
      perhaps, to display them in the same metric as the (transformed)
      person measures. &nbsp;But, don't multiply if you want to use the items
      later for anchoring, say.
      <div><br>
      </div>
      <div>Right?</div>
      <div><br>
      </div>
      <div>Rense Lange<br>
        <div><br>
          <div>
            <div>On May 3, 2013, at 6:04 PM, Mike Linacre &lt;<a moz-do-not-send="true" href="mailto:mike@winsteps.com">mike@winsteps.com</a>&gt;
              wrote:</div>
            <br class="Apple-interchange-newline">
            <blockquote type="cite">
              <meta content="text/html; charset=ISO-8859-1" http-equiv="Content-Type">
              <div text="#000000" bgcolor="#FFFFFF">
                <div class="moz-cite-prefix">Thank you for your
                  question, Felix.<br>
                  <br>
                  You have a "Fahrenheit-to-Celsius" situation,
                  equivalent to <a moz-do-not-send="true" class="moz-txt-link-freetext" href="http://www.rasch.org/rmt/rmt223d.htm">http://www.rasch.org/rmt/rmt223d.htm</a>
                  . You can subtract 0.5 from all the person and item
                  location estimates, then multiply all the resulting
                  estimates by 1.0/1.2. Also adjust discrimination
                  parameters by 1.2/1.0.&nbsp; Verify that these values are
                  correct by computing some raw scores based on the
                  rescaled estimates and comparing them with the
                  original observed scores.<br>
                  <br>
                  OK?<br>
                  <br>
                  Mike L.<br>
                  <br>
                  On 5/3/2013 7:02 AM, Fischer, Felix wrote:<br>
                </div>
                <blockquote cite="mid:C0FE9413C35748478B4DA1614ABCF10B010EE519996D@EXCHANGE41.charite.de" type="cite"><span lang="EN-US">I have been told that
                    it is possible to shift item parameters of IRT
                    models in order to achieve a certain distribution of
                    person-estimates for a given sample. Practically I
                    have a graded response model estimated in a larger
                    sample including a representative sample. now, I
                    want to shift the parameters that the distribution
                    of the person estimates of the representative sample
                    is normalized (has mean = 0 and sd = 1). Actually,
                    the distribution is slightly different, eg mean =
                    0.5 and sd = 1.2.</span><span lang="EN-US"> <br>
                  </span><p class="MsoNormal"><span lang="EN-US">Is anyone
                      aware how to change the item-parameters? Or is
                      there any reference where I could look up the
                      procedure?<o:p></o:p></span></p><p class="MsoNormal"><span lang="EN-US">Best, Felix</span><br>
                  </p>
                </blockquote>
                <br>
              </div>
              _______________________________________________<br>
              Rasch mailing list<br>
              <a moz-do-not-send="true" href="mailto:Rasch@acer.edu.au">Rasch@acer.edu.au</a><br>
              Unsubscribe:
<a class="moz-txt-link-freetext" href="https://mailinglist.acer.edu.au/mailman/options/rasch/rense.lange%40gmail.com">https://mailinglist.acer.edu.au/mailman/options/rasch/rense.lange%40gmail.com</a></blockquote>
          </div>
          <br>
        </div>
      </div>
      <br>
      <fieldset class="mimeAttachmentHeader"></fieldset>
      <br>
      <pre wrap="">_______________________________________________
Rasch mailing list
<a class="moz-txt-link-abbreviated" href="mailto:Rasch@acer.edu.au">Rasch@acer.edu.au</a>
Unsubscribe: <a class="moz-txt-link-freetext" href="https://mailinglist.acer.edu.au/mailman/options/rasch/connert%40msu.edu">https://mailinglist.acer.edu.au/mailman/options/rasch/connert%40msu.edu</a></pre>
    </blockquote>
    <br>
    <pre class="moz-signature" cols="72">-- 
Tom Conner
Professor Emeritus
Department of Sociology
Michigan State University
"What if there were no hypothetical questions?"
</pre>
  </div>

_______________________________________________<br>Rasch mailing list<br><a href="mailto:Rasch@acer.edu.au">Rasch@acer.edu.au</a><br>Unsubscribe: https://mailinglist.acer.edu.au/mailman/options/rasch/rense.lange%40gmail.com</blockquote></div><br></div></div></body></html>