• The HR Specialist - Print Newsletter
  • HR Specialist: Employment Law
  • The HR Weekly

Investigating Harassment: How to Determine Credibility


White Paper published by The HR Specialist


Most employers understand the importance of doing a fair and thorough investigation when they receive complaints of on-the-job harassment. In-house investigators (usually a human resources manager) often do a good job of interviewing the right people and documenting the interviews but then fall short when it comes time to analyzing the evidence.

For example, many investigators falsely believe they can’t conclude that harassment occurred unless they have independent witnesses to the allegations. This mistake leads to action not being taken when it should. So what’s an investigator to do when confronted with conflicting stories? Here are some guidelines.

Use a lower burden of proof

In every lawsuit, one party has the burden of proving his or her case. In a civil case, the person bringing the lawsuit (the plaintiff) has the burden of proof. However, he or she must meet this burden only by a “preponderance of the evidence,” which means that it is more likely than not that the allegations occurred. This burden is much easier to meet than the “beyond a reasonable doubt” standard that is used for criminal cases.

Investigators of workplace harassment often wrongly believe that the employee claiming harassment must show the harassment occurred “beyond a reasonable doubt.” Instead, the investigator should be using the lesser standard of preponderance of the evidence.

See if ‘baseline’ behavior changes

Determining whether an employee harassed a co-worker is difficult when, as is often the case, there is no direct witness. This is where credibility determinations come in handy. One thing the investigator can look for is changes in a witness’ demeanor. To do this, the investigator must observe the “baseline” behavior of the witness. That is, how does that person behave when being questioned about something neutral.

The investigator should take time to make the employee comfortable and get background information. During that time, the
investigator should be observing the person’s behavior. Then the investigator can compare the baseline behavior with the manner in which the person answers questions about the allegations.

Is there a change in eye contact or body language? Are there significant pauses before the employee answers the question? These are clues to credibility.

Listen for evasive responses

It’s usually difficult to make a determination based solely on the demeanor of the accused person. That’s why it’s vital to listen carefully to responses. Many people will not lie outright. Instead, they’ll give indirect answers to questions that could incriminate

If a witness will not answer directly after being asked the question a few times, this is a strong indication that he or she is not being truthful. Here is an example of such a dialogue:

  • Investigator: Did you ask Megan out?
  • Witness: I’m insulted that you would ask me that question.
  • Investigator: I’m sorry, but it’s necessary for me to ask you these questions, so did you ask her out?
  • Witness: That’s not the sort of thing I would do.
  • Investigator: I understand, but did you ask her out?
  • Witness: Look, I’m a married man.

Listen for important contradictions

An investigator will feel much better finding someone not credible if the investigator can turn up actual contradictions in that person’s statements. For the contradiction to be significant, it should be about something that is both important and not explained by memory problems.

For example, many people are bad with dates. The fact that there is a contradiction about when something happened is often not significant. However, a contradiction about what the witness did, what the witness said or who was present usually would be significant. The investigator should give the witness an opportunity to explain contradictions, and then evaluate whether the explanation makes sense.

A good way to get a witness to contradict him/herself is to ask about the same events multiple times. For example, the investigator can question the witness about the key issue at an initial interview and then again at a follow-up interview a few days later. To really evaluate the information, the investigator should ask for plenty of detail.

Don’t discount contemporaneous witnesses

There may be no direct witness to the harassment itself but there may be people who saw that the complainant was upset or spoke to the person shortly after the alleged harassing event. These people are called contemporaneous witnesses and they can have a big impact on an individual’s credibility. This is especially true if the witness is not a close friend or family member and, therefore, not someone who would typically lie for the witness.

Make a determination and support it

After all the facts are gathered, the investigator should make a determination about whether each allegation happened or not. Investigators shouldn’t take the easy way out and say they can’t make a determination because of conflicting evidence. Instead, they should weigh the evidence, make credibility findings and come to a determination based on a preponderance of evidence. The investigator should support this determination in writing using factors such as the ones explained above.

Of course, there are occasional cases that cannot be decided because both parties truly are equally credible. However, using the factors described here, a determination can and should be made in most cases.
Contributor: Amy Oppenheimer, J.D., who acts as a neutral investigator of workplace harassment, trains employers in how to prevent harassment and testifies as an expert witness on harassment. She is co-author of Investigating Workplace Harassment: How to be Fair, Thorough and Legal.