Can interviewers change their scores after the debrief?

Yes, but only by adding to the record, never by overwriting it. Keep each interviewer's original score exactly as it was submitted before the debrief, and record any change separately: the new score, the evidence behind it, who changed it, when, and why. A score should move only for evidence the interviewer did not have, a factual correction or a clear misreading of the scale, not because the room leaned toward a verdict. Then check what your applicant tracking system (ATS) does with edits: as of October 2026, Greenhouse can let interviewers edit submitted scorecards and labels them "Edited", and Ashby lets interviewers reopen submitted feedback from a Slack reminder.

The interview debrief template already says a rating changes only when someone names new evidence. This page is the policy behind that line: which reasons count, what the record looks like, how three common ATSs handle edits, and what to do with changes that arrive days later.

Should interviewers be allowed to change their scores after the debrief?

Allow it, under a rule. A debrief produces two different numbers, and most problems come from mixing them up:

• The independent score is what one interviewer concluded from their own interview, written before hearing anyone else. It is a measurement. You use it to check whether interviewers agree, to spot drift and to coach.
• The team score is what the panel concluded after pooling its evidence. It is an input to the decision, and it belongs in the decision record.

The U.S. Office of Personnel Management's structured interview guide (September 2008) keeps the two apart in exactly this way. Panel members rate on their own, then compare notes and reach a consensus. Its sample group rating form has a column for each panelist's individual rating beside the consensus rating, and its instructions say any changes to individual ratings during the consensus discussion "should be initialed by the panel members." Once the ratings are declared final, each panelist signs the form.

Overwriting the independent score destroys the measurement. If every scorecard is edited to match the consensus, your agreement figures look perfect and mean nothing, and you lose the one record of what each interviewer saw before the discussion. That is why the original stays locked and the change is logged next to it. Why scores must be written alone in the first place is covered in independent interview feedback before the debrief.

What counts as a good reason to change a score?

Two questions settle most cases. Can the person asking for the change point to something the candidate said or did? And would the same reasoning change the score of every other candidate scored on that competency? If either answer is no, the original score stands and the disagreement goes into the decision record as dissent.

Watch the direction of changes over time. If revisions after the debrief almost always move toward the hiring manager's view, the debrief is producing conformity rather than evidence, and the speaking-order rules in the debrief template are not being followed.

What should the record of a changed score look like?

Keep one log per candidate, beside the scorecards, filled in by the facilitator:

SCORE REVISION LOG
Candidate / role:      [name], [title, req ID]
Competency:            [competency]
Original score:        [3] by [interviewer], submitted [date, time]  (unchanged)
Revised score:         [4]
Revised by:            [interviewer]      Witnessed by: [facilitator]
When:                  [in the debrief / after it, date]
Reason:                [new evidence | factual correction | anchor recalibration | clerical]
Evidence:              "[candidate's words or observed behavior]"
Source of evidence:    [round, interviewer, notes or transcript]
Applies to others?     [no / yes: candidates rescored]

An invented example of a revision that passes the test:

How do Greenhouse, Ashby and Lever handle edits to submitted scorecards?

From each vendor's own pages, as of October 2026:

• Greenhouse. According to its help article on how to edit a submitted scorecard, a Site Admin can allow users to edit their submitted scorecards in the organization-wide permissions. Depending on the organization's settings, Job Admins can edit scorecards for up to 30 days after submission, and Site Admins can edit them at any time. Edited scorecards display an "Edited" label, and hovering over it shows a message encouraging the hiring team to ask the interviewer why they changed it. The article does not describe a way to view the earlier version.
• Ashby. Its feedback reminders update (September 29, 2026) says interviewers can reopen and edit feedback they have already submitted from an interactive Slack reminder. We did not find an Ashby help article explaining who else can edit submitted feedback or whether earlier versions are kept, so test it in your own account or ask Ashby support.
• Lever. We could not find a public Lever help article on editing feedback after it has been submitted. Ask your Lever admin how edits work in your account before you write a policy that depends on it.

Whatever your system does, set it up so the policy is easy to follow:

1. Decide who may edit and when. A short window for the interviewer's own clerical fixes before the debrief, then no edits in the ATS; revisions go in the log.
2. Copy the original first. If your ATS shows only an "Edited" label, the facilitator copies every original score into the debrief grid before the meeting starts. That grid is your record of the independent scores.
3. Hide other feedback until submission, so the scorecard you are protecting really was independent. Greenhouse and Ashby both offer this setting.
4. Limit admin edits. If admins can edit at any time, agree that they correct only clerical errors, at the interviewer's request.

Without an ATS, the same rule works with a shared folder: interviewers submit a form, the recruiter saves the responses as read-only, and revisions go in the log. See structured interviews without an ATS.

What about changes that arrive days later, or after a complaint?

After the decision, do not edit scores. If an interviewer remembers something or a reference adds information, add a dated note to the file. A score that changes after the decision looks like the decision rewriting the evidence, even when it is not.

A scorecard written after the debrief by someone who missed the deadline is kept, but labelled as written after the discussion. It is not an independent score and should not be counted as one.

After a complaint or charge, edit nothing. 29 CFR 1602.14 requires employers covered by Title VII, the ADA and GINA to keep hiring records for one year from the date the record was made or the personnel action was taken, whichever is later. Once a charge is filed, all records relevant to it must be kept until it is finally resolved, including those of the other candidates for the same position. A scorecard edited after that point is evidence of a different kind. What to keep and how to respond is in what to do about an EEOC charge from a rejected candidate. This is a summary, not legal advice; check your own retention rules with HR or counsel.

Which score should the hiring decision use?

The decision uses the team's view of the evidence, applied through the rule you agreed before the interviews. If that rule says one evidenced score below the bar on a must-have means no hire, the question in the debrief is whether that score survives the pooled evidence, not whether the room likes the candidate. The hire or no-hire decision framework has a rule you can copy.

The independent scores feed everything that looks at interviewers rather than candidates: calibration sessions, drift checks and inter-rater reliability. Run those on revised scores and the numbers will flatter you. A decision record that shows both, original and revised, lets anyone check later that the change was made for evidence.

A policy you can paste into your interview guidelines

SCORE CHANGES
1. Every interviewer submits a scorecard before the debrief. That
   scorecard is the independent record and is not edited once the
   debrief has started.
2. Before the debrief, an interviewer may fix clerical errors in
   their own scorecard within [24 hours] of submitting it.
3. In the debrief, a score may be revised only for evidence the
   interviewer did not have, a factual correction, or a recalibrated
   reading of an anchor. The facilitator logs every revision.
4. An anchor recalibration applies to every candidate scored on that
   competency for this role.
5. Disagreement that is not about evidence is recorded as dissent in
   the decision record. The score does not change.
6. After the decision, scores are not edited. New information is
   added as a dated note.
7. If a complaint, charge or legal hold exists, nothing in the file
   of this candidate or of others for the same role is edited or
   deleted.

Read it aloud at the start of the first debrief for each role, as you would the speaking order. A rule read beforehand feels like process; the same rule applied mid-meeting feels like a correction aimed at one person.