A rise from 2.9 to 3.8 looks good. Three checks before you put it in your report.
Self-assessment also rises when people have merely become more confident. So we look at three things before drawing a conclusion.
- Is the number of reporters comparable in both measurements?
- Do the open answers rise along, or stay vague?
- Is the rise in the topics that were trained?
If one of the three doesn't hold, the dashboard notes it. Better a figure with a caveat than a figure that turns out wrong six months later.