Credibility Assessment in a Workplace Investigation: Looking Beyond Who Seems More Believable

Credibility can be one of the hardest parts of a workplace investigation. Sometimes there is no video, no decisive email and no independent witness who settles the issue. Two people describe the same event differently, and the investigator has to decide what weight to give each account.

That does not mean choosing the person who appears more confident. Nor does it mean deciding who the investigator personally likes or trusts. A credibility assessment should be tied to the evidence in the case.

Confidence is not accuracy

Some people speak easily. Others become nervous during interviews.

A witness who gives quick, confident answers may be mistaken. Another person may hesitate because they are trying hard to remember something accurately.

Demeanor can provide information, but relying heavily on whether someone “looked truthful” is risky.

Start with opportunity to know

Before deciding whether an account is persuasive, consider whether the person was actually in a position to know what they are describing.

Did they see the event themselves? Did they hear it directly? Were they relying on what somebody else told them? Could they see the people involved from where they were standing? Did they join the meeting halfway through?

Firsthand knowledge and secondhand information should not quietly receive the same weight.

Look for consistency, but don't expect perfect memory

People do not normally recount complicated events in exactly the same words every time. Minor differences can be perfectly ordinary.

The more important question is whether the account remains reasonably consistent on material points.

A witness remembering that a meeting occurred at “about three” during one interview and “sometime after lunch” during another may not be particularly significant. A witness first saying they were not present and later saying they personally heard the disputed statement is a different matter.

The investigator has to distinguish normal imperfections in memory from changes that materially affect the account.

Compare the statement with contemporaneous evidence

Evidence created around the time of an event can be particularly useful. That might include email, messages, calendar records, notes, photographs or other records.

Contemporaneous does not automatically mean correct. But a message written five minutes after an incident may deserve different consideration from a recollection reconstructed eight months later.

The investigator should consider when the information was created, why it was created and whether there is any reason to question it.

Corroboration is more than finding another person who agrees

Two people giving the same story can be useful. But agreement is more persuasive when the accounts are genuinely independent.

If Witness B only knows about the incident because Witness A described it afterward, Witness B does not independently corroborate what happened. They may corroborate that Witness A reported the incident. That distinction matters.

Corroboration can also come from records rather than people. A witness's description of when something occurred may be supported by an email, access record or calendar entry even if nobody else observed the event itself.

Consider motive carefully

A person may have a reason to shade an account. The complainant may be involved in a separate dispute with the respondent. A witness may report to one of the parties. The respondent may face disciplinary consequences.

Those facts may matter. They do not automatically make anyone dishonest.

“Motive to lie” can become an easy shortcut when the evidence is difficult. It should instead be one factor considered alongside the rest of the record.

Plausibility matters, but personal experience can mislead

Investigators naturally ask whether an account makes sense. That is reasonable.

But it is worth distinguishing between “This account is inconsistent with known facts” and “I personally would not have behaved that way.”

People behave unpredictably, particularly in stressful workplace situations. An investigator should be cautious about rejecting evidence simply because somebody's conduct does not match what the investigator thinks a reasonable person would have done.

Write down the reasoning

A conclusion such as “Witness A was more credible than Witness B” does not tell the reader much.

A more useful assessment explains why. Perhaps Witness A had direct opportunity to observe the event, gave a materially consistent account across two interviews, and the timing was supported by contemporaneous messages. Perhaps Witness B's statement changed on an important point and conflicted with records created before the investigation began.

The reasoning is what makes the credibility assessment understandable.

Credibility does not need to be all-or-nothing. A witness can be reliable about one part of an event and mistaken about another. It is often more accurate to assess particular portions of an account against the available evidence.

Keep credibility connected to the finding

Credibility assessment should not exist in a separate universe from the allegation analysis.

If a finding depends heavily on one witness's account, the report should make that reasoning visible. What evidence supports the account? What evidence conflicts with it? Why was one version given greater weight?

The reader does not need the investigator's every thought. They do need enough to understand how the conclusion was reached.

Credibility assessment is difficult because there is rarely a mechanical answer. That is exactly why the investigator's reasoning matters.