https://workforcedatalab.com/Recent content on Workforce Data LabHugoen-usFri, 21 Aug 2026 09:00:00 +0000https://workforcedatalab.com/method/Fri, 21 Aug 2026 09:00:00 +0000https://workforcedatalab.com/method/<p>Workforce Data Lab publishes people-analytics research for readers who will check the method before they quote the finding. This page states the standard we hold ourselves to, so that a reader can tell quickly whether a piece is worth citing.</p> <h2 id="what-every-piece-must-carry">What every piece must carry</h2> <ul> <li><strong>A named source for every figure</strong>, linked to the publisher&rsquo;s own material. If we cannot link it, we do not publish it.</li> <li><strong>Sample size and population</strong>, stated in the text rather than a footnote. A benchmark drawn from 180 respondents is a different object from one drawn from 18,000, and the reader is entitled to know which they are holding.</li> <li><strong>The date the data was collected</strong>, not just the date it was published.</li> <li><strong>Whether a figure is measured, reported or derived.</strong> Measured means we ran the analysis. Reported means someone else did and we are citing them. Derived means we calculated it from someone else&rsquo;s inputs, and the inputs are named.</li> </ul> <h2 id="what-we-will-not-do">What we will not do</h2> <ul> <li>Present a vendor&rsquo;s marketing survey as independent research.</li> <li>Quote a percentage without its denominator.</li> <li>Fill a gap with an estimate and let it read as a measurement.</li> <li>Attribute a statistic to an organisation that has not published it.</li> </ul> <h2 id="corrections">Corrections</h2> <p>Where a figure turns out to be wrong, the correction goes at the top of the piece with the date and what changed. Silent edits defeat the purpose of publishing a method at all.</p>