“I Did It” – Confession Contamination and Evaluation

“I did it.” Those simple words, uttered by a suspect are the most powerful evidence in the world. Those words almost guarantee a closed case—and a conviction. They wipe away all doubt and overcome any evidence to the contrary. But should they?

The ability to interrogate a suspect has long been one of the most valuable tools of law enforcement, but only relatively recently has the legal community begun to recognize its potential dangers—including the possibility of false confessions leading to a conviction of the wrong person.

False confessions are rarely the result of bad intent or malice by the investigator. Instead, they come about through a combination of factors, including the vulnerability of the suspect, the interrogation tactics and questioning styles that are used, and, most importantly, tunnel vision on the part of the investigator. The overwhelming confidence, referred to as the “Disease of Certainty” in the suspect’s guilt is a major factor in each false confession case.1

The application of procedures such as videotaping confessions and performing confession evaluations, along with training in avoiding the potential pitfalls such as tactics or tunnel vision and awareness of potential “red flags” that might indicate a false confession, can help lead to both more reliable confessions and the ability to better differentiate between true and false confessions.

The Thomas Case

In 1994, Lawrence O’Connell, a 34-year-old employee of the Voice of America in Washington, DC, was kidnapped shortly after leaving work. The kidnappers used both his ATM and credit cards at multiple locations before killing him and leaving his body by the Anacostia River. A grainy bank photo and a composite drawing of a female who was seen using O’Connell’s credit cards was released to the media, resulting in multiple phone calls to the police department—one of which said that Susan Thomas (alias used to protect her privacy) resembled the person in the photo and the composite. This information was supported by a handwriting expert, who said that Thomas signed the credit card slips. Thomas was arrested and interrogated with standard court-approved interrogation techniques. Her confession included numerous details that “only the killer would have known.” But many aspects of her confession did not match the known facts, and she apparently refused to provide police with some of the more mundane details. At the time, this refusal was credited to an attempt to protect someone. A real problem developed when, while conducting additional follow-up, investigators discovered that Thomas had an unshakable alibi. Additional investigation discredited the handwriting expert—there was no way that Thomas had forged the signature on the credit card slips. Though the case fell apart, and the charges against Thomas were dismissed, the Metropolitan Police Department (MPDC) continued to believe in her supposed guilt. After all, she had confessed, and the details that Thomas provided “proved” that she was present when O’Connell was killed.2

In 1994, the officers of MPDC, like most people, believed that unless a suspect was tortured or mentally ill, a false confession was impossible. Today, it still remains counter-intuitive to believe that someone would confess to a crime her or she did not commit. This belief has been shaken with the advent of DNA analysis.

In the Thomas case, the suspect provided a wealth of details about the crime not known to the public. As it turns out, this is typical of most false confessions. In a recent study of 40 confession cases where the confession was confirmed through DNA evidence to be false, 97 percent of the confessions contained “surprisingly rich, detailed, and accurate information” including “inside information” about the crime known only to law enforcement.3 So how is this possible?

Fortunately, the majority of the interrogation had been videotaped, not just Thomas’s final confession. MPDC officers were able to study the interrogation in detail, discovering that the Thomas case could be considered a textbook false confession case.

This article will not attempt to address all of the factors that contribute to false confessions. Instead, through the lessons learned in the Thomas case and numerous other false confession cases, it will provide law enforcement management with some tools that will help them identify problematic confession evidence and minimize its impact on investigations. The policies and practices presented can also be applied to witness and informant evidence. False or mistaken witness evidence and lying informants, two major causes in wrongful convictions, share many of the same causation factors as false confessions.

Videotaping and Confession Reliability Evaluation


Mandatory videotaping of interrogations from start to finish has been promoted as a means to prevent false confessions. Though controversial at first, videotaping is rapidly becoming acknowledged by law enforcement as a best practice even by the leading interrogation schools in the United States. However, videotaping alone does not prevent false confessions. It may prevent the occasional rogue investigator from using improper or illegal tactics when interviewing witnesses or interrogating suspects, but many confirmed false confessions have occurred when the investigator used only standard, court-approved techniques. As seen in the Thomas case, videotaping allows law enforcement and others to critically review the interrogation and evaluate the reliability of the confession evidence. Such reviews are also an important training tool to help investigators improve their interrogation skills.

While debate still exists over what interrogation tactics may contribute to false confessions, some agreement is beginning among practitioners, instructors, and academics. However, the truth remains that even flawed interrogations, more often than not, can result in obtaining reliable confession evidence. In general, interrogation techniques used in the United States are structured to create a situation where suspects reach a point at which they are convinced that confessing will allow them to escape a real or perceived inevitable circumstance or obtain a real or perceived benefit. This is exactly the same mind-set that innocent suspects have described when they have made false confessions. Therefore, law enforcement officers need the ability to distinguish between a reliable confession and a false one. Videotaping the interrogation from start to finish plays a critical role in fulfilling that need.

Confession Reliability Evaluation

Distinguishing between a reliable confession and a false one is not as easy as it sounds. False confessions, much like reliable ones, often contain numerous details that, as is often testified to by the investigator “only the killer would know.” Since an innocent suspect would not have access to such information, it must have somehow been provided to them.

Contamination is a concept well known to law enforcement. Officers wear gloves at crime scenes and separate witnesses so they do not share information. Specially trained personnel interview juvenile witnesses and victims, knowing how suggestible they can be. However, investigators are often not so careful during interrogations. Confession evidence contamination has been identified as the “third step” in obtaining a false confession (the first and second being interrogating an innocent person and using coercive interrogation techniques).4

Contamination during an interrogation is seldom, if ever intentional. It’s usually found in cases where, due to the investigator’s sincere belief in the suspect’s guilt, tunnel vision and the accompanying verification bias kicks in. In other words, the interrogator begins to focus on signs of guilt, ignoring or explaining away any evidence to the contrary. This, combined with a poor understanding of how interrogation contamination can occur, is a recipe for disaster.

When it comes to evidence recovery, be it physical evidence, witness evidence, or confession evidence, some contamination is inevitable. Even the best laboratories recognize this and have DNA samples of their personnel on file to help identify contamination when it happens. In verbal exchanges with a suspect, the longer the exchange, the more likely officers or interrogators are to unknowingly reveal information. The critical examination of confession evidence for possible contamination is a major component of evaluating the confession’s reliability. But in order to identify it, it needs to be understood how contamination can occur.

Sources of Confession Contamination

Contamination from Outside Sources

It is standard practice in an investigation to hold back specific details about a crime from those outside the agency (i.e., media, people of interest, general public). This “hold-back” information is intended to be used to corroborate information from informants, witnesses, and suspects as it should not be known to anyone who does not have some inside knowledge of the crime. Unfortunately, what is often thought of as hold-back is not always as secure as believed. For crime details to be truly considered as hold-backs, the investigator must have a thorough knowledge of not only everything that has been in the media, but also the “word on the street.” Often, some hold-back information is common knowledge in the neighborhood where the crime occurred. Additionally, some hold-back information may be inferred from the details that are public. An example would be a case where a woman was found murdered in an abandoned building. Based on the sex of the victim and the crime scene location, it may be inferred by residents throughout the neighborhood that a sexual assault had taken place—even if that was a detail that investigators believed they had kept from the public.

If the suspect provides information that they learned from one of these sources, the investigator, overly confident in the security of the hold-back information, may conclude that the only way the suspect could have known that detail was from being present at the crime.

Leading Questions

One of the most common causes of contamination during the interrogation is leading questions. Leading questions can take many forms, but all provide the suspect with information or at least a suggestion as to where the interrogator believes the story should go. Some like “forced choice” or “either/or” questions give the suspect up to a 50/50 chance of a correct guess. In a murder case that started off as a burglary, asking the suspect if he got in through the door or the window significantly increases the innocent suspect’s chance of providing the “right” answer, in contrast to the open-ended question of “How did you get in?”

The Investigators’ Response

Even the most subtle response from an investigator to a suspect’s answer can telegraph a lot of hold-back information to an innocent suspect. In the above example of the burglary gone wrong, if the suspect answers incorrectly (e.g., saying that they got in through the door when the investigator knows that entry was made through a window), the investigator’s response of “the door?” tells the suspect that they got it wrong. Often, when the investigator expresses doubt about a response, the innocent suspect can be seen attempting to change it—trying to make their answer more acceptable to the interrogator.

A more drastic response by the investigator is seen when every time the innocent suspect gets an answer wrong the investigator accuses them of lying, and then agreeing with them when they get it “right.” In this way the interrogation resembles the child’s game of 20 Questions. The innocent suspect, desperate to escape an inevitable consequence or obtain the perceived benefit collects clues to determine what will finally end up being an acceptable story to the investigator.

“Cold Reading” Techniques

“Cold reading” is a term used to describe some of the techniques used by psychics in order to convince a person that they have supernatural “inside” knowledge about someone or something. Though these techniques can be studied and learned, they are also intuitive and are unconsciously used by people who truly believe that they have psychic abilities. They work well on an investigator who is convinced of the innocent suspect’s guilt and is willing to ignore or explain away a suspect’s apparent lack of knowledge as evasiveness or an attempt to protect someone else.

One such technique that a suspect may use is providing broad, non-committal answers or statements to hopefully provide the investigator with the information they seek. In the Thomas case, the interrogators asked Thomas to provide details about purchases made at a drugstore with the murder victim’s stolen credit card. Thomas answered “Just a lot of stuff…personal stuff,” inferring the items that would most likely be purchased from a drugstore. This response, combined with our continued and suggestive questioning, allowed her to move on to use another common cold reading technique—guessing.5

A psychic, like the innocent suspect, will make multiple guesses, getting a bit more information even with a wrong guess. Out of 30 guesses, they may get 1 or 2 right or close to right. Verification bias causes one to remember the right answers and not the wrong ones and to “explain away” the wrong answers. Thus, bit by bit, the innocent suspect obtains those hold-back details that will make for a believable but false confession.

Revealing Evidence and Crime Scene Photos

A picture is worth a thousand words, and one crime scene photograph shown to an innocent suspect can provide numerous subtle details that can make a false confession completely believable. Though showing the suspect photos as an interrogation tactic has fallen out of favor, it was recently advocated at a conference on unsolved cases attended by the author. The presenter recommended the creation of a “war room” for the location of the interrogation—complete with crime scene photos posted on the walls.

Contamination through the presentation of evidence is not confined to the use of photographs. While confronting the suspect with evidence information is often necessary when challenging an alibi, it must be used cautiously. Presenting evidence becomes problematic when it is used to “correct” a suspect’s account of the crime (such as when the suspect provides the wrong caliber of the gun used in the murder). Although the use of false evidence is generally permitted by the courts, it has frequently been linked to false confessions and, if not used carefully, can contribute to the contamination of confession evidence.6

Even when the confession is reliable, the excessive presentation of evidence in an interrogation leaves the confession open to attack by the defense attorney since this form of contamination is one that is most easily understood by juries.

Theme Development

Though called different things by different interrogation instructors, the technique of “the investigator express[ing] a supposition about the reason for the crime’s commission” and offering “a possible moral excuse for having committed the crime” is common; however, it can be a source of confession evidence contamination.7 During the employment of a theme such as self-defense in a murder case, the investigator may unintentionally provide the suspect with information such as weapon type, location of injuries, and other details. Once the theme is accepted by the suspect, the investigator must then challenge the theme, often with evidence that contradicts the theme in order to get to the “truth.” If not done carefully, the challenge can cause additional contamination (e.g. “If you say you shot him in self-defense, then how come he was shot three times in the back?”).

In the Thomas case, one of the investigators suggested that Thomas was influenced by others and acted on the spur of the moment. The investigator went on to describe his theory of the case—that there were two others involved, how the victim was first approached, information about the victim’s background and so forth—details that were all later incorporated into the false confession.

Just Telling the Suspect

This form of contamination, where the suspect is handed the details of the crime by the investigator, can creep in with the application of some interrogation tactics. If, while using a false evidence ploy of a non-existent eyewitness, the investigator suggests that the witness saw the suspect and two others creep around to the back of the house, they provided information on the number of perpetrators and the location of entry. But often the contamination is blatant, especially in long interrogations where, blinded by tunnel vision and frustrated with the lack of progress, the investigator begins to push the envelope. Sometimes, when attempting to confront the suspect with the strength of the evidence, investigators will begin a statement with “Let me tell you what we know…” and then go on to outline the details of the crime. Like most actions that contribute to contamination, the investigator is usually unaware this is happening and may not realize it even in recollection until confronted with a videotape.

Guidelines for the Evaluation of Confession Evidence

Fighting Tunnel Vision

Confession evidence is like any other evidence in that it should undergo careful scrutiny before it can be accepted as fact. Tunnel vision, combined with those words “I did it,” can blind investigators to obvious discrepancies. In many false confession cases, if the same evidence that was provided by a suspect through their confession was provided by a purported witness, the evidence would have been dismissed off-hand by the investigator because it simply did not match the known facts. But, as noted above, many false confession cases are rich in corroborated details.

Since confession evidence can be so powerful, and tunnel vision plays such a strong role in many false confessions, the evaluation should be performed by someone outside the immediate investigation. A “devil’s advocate” should be assigned the specific role of challenging any assumptions made by the investigative team and its interpretation of the evidence. They should also be knowledgeable about the causes of false confessions and confession evidence contamination.

The role of a devil’s advocate can be a difficult one. Peer pressure and an organizational agenda that pushes for a quick case closure are just two of the obstacles that can diminish the effectiveness of the position. A strong individual with good investigative skills in a supervisory position and with the support of management is needed. Not only are devil’s advocates useful in the evaluation of confession evidence, but they are invaluable in keeping investigations on track and identifying loopholes that can later be used by the defense. With the proper encouragement, the devil’s advocate approach to testing investigative theories will become part of the normal routine, resulting in better investigators and stronger investigations.

Steps in Confession Evidence Evaluation

First, the devil’s advocate should look for evidence that was provided during the confession that was previously unknown to the investigator—facts which could not have been obtained from outside sources or inference and that can be corroborated through independent means. An example of this would be information from a confession that leads the investigators to a previously undiscovered weapon that is confirmed as the murder weapon through forensic examination. Obtaining new and confirmable evidence should be the goal of every investigator during every interrogation.

Secondly, the devil’s advocate needs to break down the information provided by the suspect that, although known to the interrogator, is believed to be information that only the true perpetrator could provide (the hold-back information). Does the information provided by the suspect include the hold-back information and, most important, was it provided free from contamination by the investigator during the interrogation? This can be shown by breaking the interrogation down on a spreadsheet and comparing when during the interrogation the details were first provided by the suspect to potential contamination sources. This method allows the devil’s advocate to see how the confession developed and determine who is “telling the story”—the suspect or the interrogator.

A leading interrogation school talks about a third form of corroboration—rational corroboration. Acknowledged as the weakest form of corroboration, this is “mundane details or other information that lends credibility to the statement.”8 Such information is usually impossible to independently corroborate. The author has seen such details in his own and numerous other false confession cases and has found that they are often the result of the innocent suspect attempting to make the confession more believable to the investigator. In one case involving a murder and dismemberment, the suspect got every detail wrong—even the tools used to dispose of the body. In defending the “confession,” the investigator pointed to the conversation that the suspect said that he had with the victim before he murdered her as something that “only the killer would know,” but it was a conversation that in no way could be corroborated.

In both reliable and false confessions details are often provided by the suspect that just don’t fit, or the suspect is apparently refusing to admit to known facts. Often this is explained away by investigators by saying that the suspects may be minimizing their involvement or attempting to protect someone else. In a reliable confession, this could be true, especially when (1) there is other evidence that was provided that passes the previous criteria, and (2) the problematic information (or lack of information) points to acts that could increase the suspect’s culpability or connect them to other, unknown or unsolved crimes. In a false confession, the details that don’t fit may occur because the innocent suspect has been unable to obtain the correct information from the interrogator via contamination.

One red flag to a false confession is when the incorrect information provided by suspects actually increases their culpability, although there is no evidence to suggest that what they are saying is true (such as admitting to a sexual assault during a murder when no such assault took place). Another red flag is when the suspect is unable to provide information that is of a mundane nature, and, since the suspect is admitting to the more horrific details of the crime, should not have any problems providing these mundane details. In the Thomas case, Thomas admitted to participating in the kidnapping and murder of the victim, but could not list the individual items she purchased with the victim’s credit card.

Summary and Recommendations

As can be seen, although a valuable tool, interrogations hold potential dangers as well, including the very real possibility of false confessions. Interrogation schools in the United States have slowly begun to acknowledge these dangers, yet their classroom instruction often deals with it on a superficial level, if at all. This practice is analogous to a doctor receiving instructions in the administration of a drug, with no knowledge of the potential side effects, how to recognize them, or the steps necessary to treat them. The U.S. Department of Justice recently addressed this problem in a consent decree with the New Orleans Police Department where they mandated training for investigators in both the causes of false confessions and “criminal investigative failures” (e.g., tunnel vision).9

Along with training in the causes of false confessions, confession contamination, and the critical evaluation of confession evidence, the regular use of devil’s advocates would be an invaluable addition to the investigative process. This process for the review of confession evidence can be applied to both witness and informant evidence as well, since both are potentially subject to unintentional contamination, Bad or false information from witnesses and informants has been found to be a major factor in wrongful convictions.

Finally, law enforcement needs to accept the fact that mistakes have been and will continue to be made. However, by studying the mistakes of others and performing regular autopsies of their own missteps, problems can be identified and corrected and the investigative processes strengthened.
By adopting the scientific approach of conducting critical evaluations in order to test theories rather than just working to prove them, the chance of a wrongful conviction will be minimized, and cases against the guilty will be strengthened. ♦

1Everett Doolittle, “The Disease of Certainty,” FBI Law Enforcement Bulletin (March 2012).
2James Trainum and Diana Havlin, “A False Confession to Murder in Washington, DC,” in Criminal Investigate Failures, ed. D. Kim Rossmo (Boca Raton, Florida: CRC Press, 2009), 205–217.
3 Brandon L. Garrett, “The Substance of False Confessions,” Stanford Law Review; 62, no. 4 (2010):1,051–1,119
4Richard A. Leo and Steven A. Drizin, “The Three Errors: Pathways to False Confessions,” Police Interrogations and False Confessions, eds. G. Daniel Lassiter and Christian A. Meisser (Washington, DC: American Psychological Association, 2010).
5Interrogation transcript, Metropolitan Police Department.
6Saul M. Kassin et al., “Police-induced Confessions: Risk Factors and Recommendations,” Law and Human Behavior 34 (2010): 3–38, http://web.williams.edu/Psychology/Faculty/Kassin/files/White%20Paper%20-%20LHB%20(2010).pdf (accessed May 20, 2014).
7Fred E. Inbau et al., Criminal Interrogation and Confessions, 5th ed. (Burlington, MA: Jones & Bartlett Learning, 2011), 188.
8Ibid., 355–356.
9United States vs. City of New Orleans, Case 2:12-cv-01924, Consent Decree Regarding the New Orleans Police Department (July 24, 2012), 47.

Please cite as:

James L. Trainum, “ ‘I Did It’ – Confession Contamination and Evaluation,” The Police Chief 81 (June 2014): Web-only.