Does Polygraph Really Work?

Polygraph for the most part does work reliably. There are many good scientific studies that indicate that polygraph testing is much higher than the probability of chance. These studies indicate that polygraph is somewhere between 85 to 95% reliable. I have been doing polygraph tests for over 32 years and can only think of a handful of times that I was doubtful or concerned about my opinion. This post is about a fairly recent case that I had concerns about. The posted article is based on a real polygraph test but the identities and some information has been changed to protect the identity of the examinee. No inferences in this article should be construed to point to any individual. The information is offered to enlighten peoples knowledge about the limits of polygraph testing.

This case starts with a call from a husband that is sure his wife had cheated on him. He said that he already knew that his wife has cheated but she was in denial and would not admit the truth. I told him that I would keep an open mind and I would appreciate if he would consider doing the same until I completed her polygraph examination. I explained the statistical reliability of polygraph and told him that he should not make any major decisions based exclusively on polygraph testing.

Part of my protocol for testing fidelity issues is the following:

  1. On the day of the test I like to meet with both people in the relationship.  I have the pretest discussion regarding the relevant issue to be tested.  This is done very respectfully and in a professional manner to see if agreement can be reached on the type of relevant questions to be asked.  Basic polygraph training dictates that you should never ask a question that would be considered a state of mind. A good example of this is: “Do you love your wife?” or “Do you plan to stay married to your wife?”  The relevant questions have to be on one issue and they have to be about something that did or did not happen. By having this pretest discussion you can eliminate the problem of not asking the right relevant questions that the other person in the relationship is concerned about.

2.  After this discussion I do the test alone and no one else is in the room except the examinee.

3.  After the data collection phase of the polygraph the couple goes on their way. I take a very careful approach to analyzing the polygraph charts. I use several algorithms and I do a manual numerical scoring of the                 qualifying charts. I often even send the charts to another competent examiner to do a blind score review of the data.  I usually have my results completed within 48 hours.

In this particular case, when I met with this couple, I immediately observed that the husband was very dominant and was convinced that his wife was cheating on him.  He insisted that he had evidence of this adulterous act.  He claimed that a recording that he had made surreptitiously depicted and was proof that his wife engaged in sex with someone else.  He did not provide me with a copy of the recording but assured me that it clearly depicted his wife engaged in a sex act with someone else.

When the husband left the room, I began my pretest interview with the examinee. She vehemently denied that she has had sexual relations with anyone else since being married. She came across very honest in her demeanor. Her body language is very consistent with what she was telling and she was very convincing. She advised that her husband had a recording but it was not in substance what he thought it was.  She went on to explain that her husband has had trust issues in the past. She was taking the test to prove to him that the recording was not in fact her or anyone engaged in a sex act.  The examinee reports that her husband is very obsessed and he is extrapolating the recording well beyond reason.  She went on to explain that she played the recording for their minister and he agreed with her that it did not depict her or anyone engaged in the sex act.  After the pretest interview, I collected several good charts that were free of artifacts or other anomalies. I advised the couple that after I completed my analysis work I would send them a report as requested via email.

My conclusion based on my analysis of the polygraph charts was that there was deception indicated.  I issued a report and emailed it to both parties. A short time later, I received an email reply from the examinee’s husband simply indicating that the received a report and he thanked me for my services.  About 1 to 2 days later I received a call from a friend of the examinee who was concerned that a mistake had been made on the test. She advised that the examinee was honest and was telling the truth about her fidelity.  I told her that I do not retest people who are clearly deceptive on their charts. This friend told me that the reason the examinee failed was because her husband was in the other room and she could hear him occasionally talking. I explained to the friend that I would review the video recording to see if I can hear what she had just described. During the test I did not make note of any outside noise that I believed was distracting.  I told the examinee’s friend that if the examinee called and told me that the reason she failed the test was because she could hear her husband talking in the waiting room, I would consider retesting her.

The next day at work I reviewed the case video and determine that there were some sounds that could have been the examinee’s husband talking in the other room. The voice was muffled on the recording but I could hear something clearly. A short time later I received a call from the examinee who insisted that she was being honest about fidelity during the test. I warned the examinee that it was highly unlikely that the result would be anything but what I reported previously.  The examinee insisted on being retested.  I asked her to come to the appointment alone or at least not with her husband.  I did not want a repeat situation as reported on the first test.

On the second test the examinee came in without her husband as requested. She appeared to be very confident and relieved that her husband was not accompanying her. During the pretest interview she again reiterated that the reason she failed the test was because she could hear her husband talking in the waiting room.  I structured a very narrow single issue relevant question test. The issue was, “Did the examinee have sexual relations with anyone during her marriage other than with her husband?” After collecting polygraph charts, I told the examinee that I would analyze the data and send her a second report. However, this time the examinee requested that I send the report to a third-party because her husband had become very verbally abusive since her last test.  She did not want to do anything to aggravate their relationship unless it was good news.

After carefully analyzing the charts and obtaining a second opinion on her charts, I determined that there were indications of deception again reflected in her polygraph data. My opinion was in conflict with my personal intuition about her veracity but good examiners never issue a finding in polygraph based on intuition.  I should note two important things that examiners do pay attention to during the polygraph interview.

1.  All examiners note the body language of the examinee and at least determine if it is congruent with what the person is saying
2.  Most examiners also consider the verbal responses to the questions during the pretest interview as an indicator of veracity.

When the above observations are incongruent with the polygraph charts the confidence level in the reliability of the polygraph opinion begins to diminish.  In other words, when everything is congruent there is a high confidence level in the opinion made on the polygraph charts.  Regardless of the points made above, the examiner’s opinion should not be formulated on anything but supported polygraph data (polygraph charts). Polygraph examiners are trained to be objective and base their final opinions on the polygraph data even if other observable traits are noted during the test that would be in conflict with that opinion. In this case the data very clearly indicated deception but her body language and verbiage appeared to be truthful. This is a very perplexing situation for a polygraph examiner. All examiners want to be correct in their opinion but if opinions were made on anything but the polygraph data the reliability of polygraph will become significantly less than chance. So putting this another way, examiners can issue an opinion and still not have a very high confidence level in that opinion.  This was my dilemma after the second test was analyzed.

After a few days had passed the examinee sent me the following email after she received my second report.

“I was telling the truth, but thank you for all of your efforts. If I thought in a million years that I would fail, I would have never taken this exam. God only knows the truth and I pray that my husband will realize the truth. Even though I failed the exam the audio engineer clearly stated that there was no sexual activity on the tape. Even a pastor listened to the audio and agreed I was innocent. My husband constantly reminds me that I failed the polygraph test and his beliefs were confirmed. He feels justified in all of the physical and verbal abuse he has released upon me for cheating and lying about it. I am in the 5% that the test is not accurate. My life is forever changed because of my reliance upon the polygraph results.”

After receiving this email I have become less confident in my opinion. The one thing that examiners fear the most is that they will call someone a liar when they are telling the truth. These kinds of test results will shake an examiners confidence but one should never lose sight of the fact that polygraph testing is not perfect but it is statistically reliable.  This is one test that will haunt this examiner for many years to come.