APA: Deception Detection

Research has consistently shown that people’s ability to detect lies is no more accurate than chance, or flipping a coin. This finding holds across all types of people — students, psychologists, judges, job interviewers and law enforcement personnel (Personality and Social Psychology Review, 2006). Particularly when investigating crime, the need for accurate deception detection is critical for police officers who must get criminals off the streets without detaining innocent suspects.

Traditional police practices in deception detection stem from early theories on lying that assume liars will exhibit stress-based cues because they fear being caught and feel guilty about lying. This theory led researchers to search for reliable behavioral indicators of deception. They examined behaviors such as posture shifts, gaze aversion, and foot and hand movements, without much success.

“There really is no Pinocchio’s nose,” says Judee Burgoon, PhD, a professor of communication at the University of Arizona.

Given these early findings, today’s researchers are exploring new methods of deception detection. Instead of looking at people for visual cues that they may be dissembling — such as a lack of eye contact or fidgeting — psychologists are now focused on developing proactive strategies that interviewers can use to elicit signs of deception, says Maria Hartwig, PhD, associate professor of psychology at John Jay College of Criminal Justice.

“The view now is that the interaction between deceiver and observer is a strategic interplay,” she says.

Such research has “enormous potential to revolutionize law enforcement, military and private sector investigations,” says Christian Meissner, PhD, a professor of psychology at Iowa State University, who studies the psychological processes underlying investigative interviews.

Questions and cues

“Liars have a dilemma,” says Ray Bull, PhD, a professor of criminal investigation at the University of Derby, in the United Kingdom. “They have to make up a story to account for the time of wrongdoing, but they can’t be sure what evidence the interviewer has against them.”

Both Bull and Hartwig conduct research on criminal investigative interview techniques that encourage interviewees to talk while interviewers slowly reveal evidence.

Their research consistently shows that being strategic about revealing evidence of criminal acts to suspects increases deception detection accuracy rates above chance levels (Journal of Investigative Psychology and Offender Profiling, 2011; Law and Human Behavior, 2006). For example, Hartwig and colleagues conducted a series of studies to show that withholding evidence until late in the interview leaves room for guilty suspects to blatantly lie, for instance by denying being in the area of the crime. When the interviewer reveals evidence showing the suspect was there — such as surveillance footage — the suspect has to scramble to make up another lie, or tell the truth. The suspect may admit to being in the area, but still deny the crime. If the interviewer then presents more evidence, such as matching fingerprints from the crime scene, the liar will find it increasingly difficult to keep up the deception (Credibility Assessment: Scientific Research and Applications, 2014).

Aldert Vrij, PhD, a professor of applied social psychology at the University of Portsmouth in the United Kingdom, also focuses his research on using strategies to outsmart liars. “Liars are doing more than telling their stories — they need to make a convincing impression,” he says. “If the interviewer makes the interview more difficult, it makes the already difficult task of lying even harder.”

Another way to make lying more difficult is to increase interviewees’ cognitive load by, for example, asking them to tell their stories in reverse order. Truth tellers can rely on their memories to tell their story backwards, often adding more details, but liars tend to struggle. Research shows that liars also often provide fewer details about time, location and things they heard. They also speak more slowly, with more hesitations and grammatical errors (Law and Human Behavior, 2008).

Encouraging interviewees to say more during their interviews also helps to identify liars. “Truth tellers do not immediately say everything they need to say, so when the interviewer encourages them to say more, they give additional information,” says Vrij. “Liars typically have a prepared story with little more to say. They might not have the imagination to come up with more or they may be reluctant to say more for fear they will get caught.”

It’s particularly useful to ask unexpected questions in interviews, Vrij has found. Because liars often prepare their stories, surprise questions can leave them floundering for a response or contradicting themselves (Applied Cognitive Psychology, 2014).

Other avenues of research are examining how liars and nonliars talk. Burgoon studies sentence complexity, phrase redundancy, statement context and other factors that can distinguish truth tellers from liars (Journal of Language and Social Psychology, 2006).

“If liars plan what they are going to say, they will have a larger quantity of words,” she says. “But, if liars have to answer on the spot, they say less relative to truth tellers.”

That’s because trying to control what they say uses up cognitive resources. They may use more single-syllable words, repeat particular words or use words that convey uncertainty, such as “might” rather than “will,” she says.

Examining word count and word choice works well for analysis of text, such as interview transcripts, 911 call transcripts, witness and suspect written statements, and in analysis of written evidence such as emails and social media posts. Research is still needed to understand how well investigators can pick up these cues in real time, says Burgoon.

Research is also examining the communication between co-conspirators by exploring how two or more people interact as they try to deceive interviewers (Human Factors: The Journal of the Human Factors and Ergonomics Society, 2012).

“In field situations, such as checkpoints and street corners, people conspire and collude to get away with crime and terrorist acts,” says psychologist James Driskell, PhD, president of the Florida Maxima Corporation, a company that conducts research in behavioral and social science.

“If two people are lying, they have to concoct a story that is consistent with their co-conspirator so they don’t arouse suspicion,” says Driskell. “If an officer needs to engage them on the street, it would be useful to know what indicators to look for in their responses.”

Compared with truth tellers, when liars tell their story together they tend to not interact with each other and they are less likely to elaborate on each other’s responses, he says. “Truthful dyads are much more interactive as they reconstruct a shared event from memory,” he says.

Culture and context

While recent lie-detection research has centered on verbal reports, there is still a role for behavioral cues in deception detection research, says David Matsumoto, PhD, professor of psychology at San Francisco State University and CEO of Humintell, a consulting company that trains people to read human emotions.

Behavioral cues might change depending on the types of questions asked and the interview circumstances, he says. “Researchers need to take into account different investigative contexts and circumstances that might elicit different behavioral responses.”

One context Matsumoto has studied is culture. In recent research, he found culture-specific differences in tone of voice and vocal characteristics. For example, his research shows that Chinese participants tend to speak in higher pitched voices when lying compared to truth telling whereas Hispanics tend to speak in lower pitches when lying compared to truth telling (International Journal of Psychology, 2015).

Leanne ten Brinke, PhD, a postdoctoral fellow in psychology at the University of California at Berkeley, also considers context in her research, but focuses on how people may unconsciously spot deception.

Ten Brinke conducted preliminary research to explore how indirect measures of deception compare to direct, or conscious, measures. In one study, research participants watched videos of truth tellers and liars and then classified words such as “dishonest” and “deceitful” versus “honest” and “genuine.”

“The trick of the task is that images of the people in the videos were subliminally flashed while participants classified the words,” ten Brinke explains. She found that participants were faster at classifying words associated with lying when flashed images of liars. The same was found for pairs of truthful words and images. In contrast, when making conscious judgments, participants were accurate only about half the time (Psychological Science, 2014).

While the concept of unconscious deception detection is a relatively new direction in research, it highlights one more of the diverse areas psychologists are exploring. These expanding directions in research have resulted in novel investigations that might finally lead to accuracy rates above chance levels.

“A lot of research is flying in the face of law enforcement training and common beliefs,” says Meissner. “As we conduct more research, we will learn more about deception detection. This research has enormous potential to revolutionize law enforcement, military and private sector investigations.”