Interrogation: Backwards and Forwards

 “OK. Now tell me the story backward.”

Police inquisitors, detectives, and interrogators have long been taught that one method of getting to the truth is just that: to have the subject recount events in reverse order.

The theory appears sound: People like to fill in the blanks in a story with constructs, thoughts or images that did not actually happen, to make a tale run smoother. Remove the linear nature of storytelling and the tendency to confabulate should decrease. The theory is so sound, in fact, that police forces in Australia, Britain, New Zealand, Norway, and Spain, to name a few, have been using reverse recall as a matter of policy for years.

I attended a seminar on interviewing where the speaker actually said that this method of interrogation was the most useful way to elicit a truthful statement from an interviewee. “They don’t have the time or the creativity to make things up in reverse-recall,” my instructor said.

Well, as with many a fine theory, when put to scientific rigor, it falls short.

A brief story in September 2, 2011 issue of The Economist details a study by Lancaster University which basically debunks this theory. The researchers showed a short film depicting a cell phone robbery. Two days later the subjects of the test were separated into three groups:

1 – recall the events freely then in reverse order,

2 – recall the robbery in reverse order first then freely,

3 – (control group) recall the events freely both times.

The researchers found that the control group recalled the events correctly 48.7% of the time. The group that began with reverse-recall and then recounted the story freely scored 42.2 percent accuracy. The group that started with free recall then reverse-recall scored a pathetic 38.7 percent. I think it’s important to note that eyewitness testimony has already been proven to be less than reliable on several occasions. Seriously, none of the groups achieved even 50% correct recall.

The most interesting finding, however, was that the number of mistakes made among the three groups was roughly the same, but the group that recalled events in reverse order first, actually made up (pure confabulation) recollections 600% more often than the control group.

The majority of the confabulations were observed during the reverse recall portion of the exam. This flies directly in the face of what I’ve been taught in seminars and classes about interview and interrogation.

Why people, people who have no reason whatsoever to lie, make up events when using reverse-recall is a mystery. The Economist says that this study, “…does, however, point out the dangers of taking even logically plausible ideas on trust, rather than testing them.”

Those of you who testify as expert witnesses in court proceedings might want to check out the study here. It could come in handy one day.