New Study: Cops Use Wrong Tactics In Questioning Witnesses

Print Friendly, PDF & Email

Officers interviewing witnesses are “potentially reducing the amount of information retrieved” by talking too much, asking too many closed-end questions, and failing to adhere to science-based methods for mining memory, according to new findings by a Canadian research team.

Analyzing a sample of 90 interviews with witnesses to violent crimes, the team discovered that:

  • On average, officers interviewing cooperative subjects talked roughly 36% of the time, compared to a rule-of-thumb 20% that is considered desirable. Indeed, the 80/20 Rule was violated in nearly 90% of the interviews studied.
  • Only about 6% of the interviewers’ questions were considered open-ended; that is, encouraging a broad range of response beyond a simple yes or no or other narrowly restricted replies. “We estimate that between 20 and 30% of all questions asked should be open-ended,” the researchers state.
  • In 12% of the interviews, witnesses were warned in advance about the legal consequences of providing false information, a procedure that appeared to cause “significantly shorter” responses and may have hindered the development of rapport.

“[T]hese mistakes [are] somewhat disappointing given the extensive amount of research devoted to improving police interviewing practices,” the researchers write. Yet the findings are “predictable” because of the “inadequate” training in investigative interviewing afforded to officers, including instruction “on science-supported interviewing practices such as the cognitive interview.”

The study was conducted by Dr. Brent Snook and Kathy Keating of the psychology department at Memorial University of Newfoundland. A report on their results will be published later this year in the journal Legal and Criminological Psychology, issued by the British Psychological Society.

The shortcomings they identified are virtually certain to crop up during interviews of surviving officers after a shooting or other use-of-force incident, in the opinion of Dr. Ed Geiselman, who lectures on interviewing techniques for OIS investigations but was not involved in the team’s recent project. Geiselman has pioneered the use of cognitive interviewing in police work.

Dr. Bill Lewinski, executive director of the Force Science Institute, agrees. “The kind of interviewing style the researchers describe is consistent with what I have found in many officer-involved-shooting investigations,” he says.

The problem, Geiselman told Force Science News, is that “detectives commonly fail to distinguish between interrogating a suspect and interviewing a cooperative witness, and generally take an approach that inhibits the free flow of information needed from witnesses.”

The interviews analyzed by Snook and Keating were evaluated in transcript form and were conducted primarily by constables (and some sergeants) at a police department in Canada’s Atlantic region during a recent 10-year period. Mostly males, the interviewers ranged in age from 34 to 48 and had from 8 to 25 years of law enforcement experience. The interviewees predominately were witnesses to assaults, homicides, and sexual attacks. They were 16 to 63 years old, half were female, and 90% were not victims of the crime in question. In ¾ of the cases, only 1 interviewer was involved in the questioning.

“Although the sample is limited, similar results have been obtained whenever broader studies have been conducted regarding police interviews in the U.S. as well,” Geiselman says.

The researchers noted that the officers in many cases followed good interviewing practices. For the most part, for example, they avoided expounding on their own opinions or statements, asking leading questions that suggested an answer, posing questions that forced a witness to pick between a limited number of responses, voicing multiple questions at one time, and interrupting witnesses in the midst of an answer. “[I]t is encouraging that many of the interviewing officers…appear to be aware of these important interviewing practices,” the study observes.

However, the evaluation results show that “several best practices have yet to become commonplace,” the researchers report. Specifically:

Who Talks

While most officers violated the recommended 80/20 Rule regarding words spoken during an interview, some did so to an extreme. In 16% of the interviews, the “total number of words spoken by the primary interviewer was greater than the total number of words spoken by the witness,” the study reveals. Younger officers tended to talk excessively more often than older interviewers.

Over-talking by officers “can cause interviewers to be less effective at getting witnesses to dictate the pace and structure of their recollections, and ultimately, potentially reduce the amount of information that is retrieved,” the researchers warn.


More experienced officers in the study tended not to issue warnings about false testimony as often than their less experienced colleagues. As evidence of the “detrimental effect” of such cautions, the researchers found that “the average length of responses were significantly shorter” in interviews where they were issued.

“Witnesses may become anxious about the interviewing process and concerned over what they are expected to report,” the study explains. “Poor rapport can also lead to interviewee passivity and foster poor communication.”

If such warnings are considered essential, “efforts should be made to minimize their length, apparent complexity, and severe style.” Damage may be lessened if someone other than the interviewing officer delivers the information.

Question Types

Consistent with findings from previous research, officers in the new study “rarely asked open-ended questions,” which often begin with “Tell me about…” and cannot be answered with a simple yes or no, or other one-phrase reply. Open-ended questions are foundational to cognitive interviewing, which encourages free-form narratives of recollections by interviewees and has been scientifically established to have the greatest potential for thoroughly tapping memory. The dearth of such questioning in this survey “suggests that interviewers obtained less information than what was possibly available from the witnesses,” the researchers write.

“[C]losed-ended questions are problematic,” they explain, “because the elicited information is tied to the specific request and, as a result, no unsolicited information is generated. Consequently, information will not be generated if the interviewer forgets to ask a relevant question.”

“This approach seems to stem from the desire or notion that the officer should rigidly control the interview, treating cooperative witnesses like they’re on the stand in court,” Geiselman observes. “That’s not the optimum way to gather information.”

As examples, Lewinski offers 4 categories of open-ended prompts that can be effective in eliciting useful and thorough responses:

  • DESCRIPTIVE: “Tell me what happened….”
  • CAUSAL: “What do you think brought this about….”
  • HISTORICAL: “Give me time-frames as to when and how this evolved….”
  • COMPARATIVE: “Have you ever seen or experienced anything like this before….” “How did that compare….”

“These approaches can be repeated over and over again—“Tell me some more about this….”—as the witness responds with more information,” Lewinski explains. “I equate it to peeling an onion, layer after layer, probing for more recollections with the same types of open-ended questions. You end up with maximum information in the witness’s own words.”

Again, more experienced officers in Snook’s and Keating’s study “tended to request a free narrative more often than their less experienced counterparts.” Female officers, for reasons that are not explained, asked for free-form responses less often than males.

Given the importance of police interviewing in the pursuit of justice, “one would expect that training on best interviewing practices would be commonplace,” the researchers write. Unfortunately, they conclude, their study adds to “findings from countless studies around the world” that current police training in and use of scientifically grounded interviewing skills is woefully inadequate.

“In some respects, we haven’t made much progress in interviewing since the late 1970s,” Geiselman says. “At that time, studies showed that only about 2% of detectives had any formal training in interviewing, as opposed to interrogation techniques. And that is still a big void that needs to be filled.”

Leave a Reply