Jan R. Boehnke<p><span class="h-card" translate="no"><a href="https://fediscience.org/@TomJewell" class="u-url mention" rel="nofollow noopener" target="_blank">@<span>TomJewell</span></a></span> </p><p>We have argued that the definition of the target population, <a href="https://mastodon.social/tags/sampling" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>sampling</span></a> procedure, and <a href="https://mastodon.social/tags/ItemContent" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>ItemContent</span></a> of an instrument jointly define the meaning of measures in particular applications (<a href="https://mastodon.social/tags/PsychometricEpidemiology" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>PsychometricEpidemiology</span></a> <a href="https://rdcu.be/dgE5g" rel="nofollow noopener" translate="no" target="_blank"><span class="invisible">https://</span><span class="">rdcu.be/dgE5g</span><span class="invisible"></span></a>).</p><p>Written from the perspective that we were perplexed by psychologists being surprised that psychometric results change when population and sampling procedure are changed, I think it is an illustration of the above from the other side.</p><p><a href="https://mastodon.social/tags/Psychometrics" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>Psychometrics</span></a></p>