The résumé is a claim, not a record: the Evidence Ladder for hiring
A résumé is a claim about the past. An interview gives an account of it. A test produces a result. A work sample is an artefact someone made. Observed work shows conduct: what a person did, in what order, with what in front of them. Each kind of evidence shows something different, and none replaces the others. We call this the Evidence Ladder.
The ladder describes what is observed, not what predicts
The five rungs are a rough ordering of how directly the work itself is seen. They are not a ranking of which rung predicts job performance. Calavion has not shown that any rung, including conduct, predicts performance, reduces bias or improves hiring outcomes. If you take one thing from this page, take the vocabulary, not a verdict.
The five rungs
1. Claim: the résumé. A claim says what a person says they did. It can be specific, relevant and entirely true. What it shows is what the person chose to write. What it cannot show is how the work was done, or what the person did when it went wrong.
2. Account: the interview. An account is a person describing, in their own words and in the moment, what they did or would do. It adds something a claim lacks: you can ask a follow-up. What it cannot show is the behaviour itself. You are hearing how someone tells the story.
3. Result: the test. A result is an output measured against something: an answer key, a rubric, a time. It shows how a person performed on that task. What it cannot show is how the task relates to the job, or how the person got there. Tests are not all alike, and this rung is deliberately called a result rather than a type of test.
4. Artefact: the work sample. An artefact is a thing the person produced: a document, a piece of code, a written reply. It shows the finished work. What it cannot show is the path: what the person looked at, what they ignored, what they changed their mind about.
5. Conduct: observed work. Conduct is a record of what a person did while doing work: which records they opened, what they wrote, what they did and in what order, and what they left open. It shows the path as well as the output. What it cannot show is what the person was thinking, or whether they would behave the same way tomorrow, in a real job, with real consequences.
Why separate them
Hiring decisions often mix rungs without saying so. A confident interview answer gets treated as a result. A tidy artefact gets treated as proof of how the work was done. Naming the rung makes the question easier to ask: what, exactly, am I looking at, and what would I need to see to know more?
Reading any rung
Whatever the rung, it helps to keep what you saw apart from what you conclude from it. We describe that separation in why hiring evidence should separate observation from interpretation.
Where Calavion sits
Calavion records conduct inside a prepared work exercise and gives the employer a record, called a Replay, to read. The employer interprets it and decides. Calavion does not give applicants an overall score, generate an ordered shortlist, or recommend whom to hire. The employer orders its shortlist; the software checks that order against approved criteria and asks for a reason where it conflicts. That describes what the Replay is. It is not a claim that conduct is the best rung.
What we do not know
- Whether any rung predicts job performance better than another. We have not tested this and make no claim.
- Whether the five rungs are the right division. They are a working vocabulary, and other people may cut the evidence differently.
- Whether observed work in a prepared exercise resembles observed work on the job.
Related
- Work samples and take-home assignments:
/resources/work-samples-vs-take-homes-interactive-environments/ - What a Replay shows:
/replay/ - Terms:
/definitions/