What does an anchored scale look like?
Take a criterion such as "ownership" on a 1 to 5 scale. A bare scale leaves the numbers to each reviewer. An anchored one writes each point out. The wording below is only an example.
| Score | What the answer shows |
|---|---|
| 1 | Describes a task they were given and nothing they did about it |
| 3 | Names a problem they noticed and fixed, with no follow-through |
| 5 | Found the problem, fixed it, told the people affected and checked it stayed fixed |
Source: an illustrative example written for this page, not Recagn output · 18 Nov 2026
Why do anchors matter?
Without anchors, a 3 from one reviewer is a 4 from another, and scores cannot be compared across candidates or across days. Anchors also give you something to point to when a candidate or client asks why someone scored as they did. They take effort to write, which is why many teams only do it for the few criteria that decide the shortlist. The guide to scoring interviews covers writing them.
In Recagn
Recagn writes anchors on every scored criterion. Each criterion is judged on a 1 to 5 scale, and the AI drafts a description for every score point when it drafts the rubric, Recagn's word for the scoring definition. You edit them before you lock it, and after that the same anchors apply to every candidate. Recagn says anchors, not BARS.
Recagn scores each answer against them and quotes the part of the answer behind the score. Recagn recommends. Only you can reject a candidate. See how anchored criteria fit the scorecard, which Recagn calls a rubric.