Imposing cognitive load to elicit cues to deceit: inducing the reverse order technique naturally

Abstract

In two experiments, we tested the hypotheses that (i) the difference between lying and truth telling will be greater when respondents report their stories in reverse order than in chronological order, and (ii) instructing respondents to recall their stories in reverse order will facilitate detecting deception. In Experiment 1, 31 professionals told the truth and lied about a route they took and did this by describing the route in chronological order and reverse order. The reverse-order answers contained more prominent cues to deceit than did the chronological-order answers.

In Experiment 2, 68 observers read the transcripts of the verbal statements given in Experiment 1 and made veracity judgements. Observers detected deception better when judging the routes that respondents had described in reverse order than in chronological order. We recommend the use of the reverse order technique as a tool to detect deceit.

Keywords:

• believability
• credibility
• deception
• deception detection
• interviewer style

Notes

1. To establish which interviewing agents were part of their mission and which ones were not, the participants were requested to ask each interviewing agent the code question: ‘Do you have the time please?’. Participants were told what code answer to expect from interviewing agents belonging to their mission. We used two code answers: ‘No sorry, my watch stopped at 6.38 this morning’ or ‘No sorry, my watch is always at least 20 minutes slow’. Each interviewing agent was informed of only one of these code answers and used this answer to all of the participants they interacted with.

2. This ensured that the truth tellers’ stories would vary. As a result, the interviewers could not describe at the end of the experiment the details of the four truthful missions.

3. Typically in research only two coders are used. However, three good coders were available to us and we therefore decided to use all three.

4. If a the participant had said ‘I walked into Edinburgh Road which is at the top of Commercial Road’, the sentence would have contained three details, reflecting the fact that ‘at the top’ is slightly vaguer than ‘at the top end’. We decided only to code the visual details as other perceptual details that are part of the Reality Monitoring tool (e.g. auditory details), are unlikely to occur in route descriptions.

5. We included Interview Order as a factor in the design to (i) control for its effect on Veracity and Route Recall and (ii) to precise the error term. The Interview Order effects are theoretically not relevant and will therefore not be discussed.

6. These single interview packages were the result of the odd number of participants, N=31, that took part in Experiment 1.

7. Due to an administrative error, three packages in the Chronological Order condition and one package in the Reverse Order condition were read by two rather than by one participant, which is why 68 rather than 64 participants took part in the experiment.

8. Participants were told that senders went from King Henry Building to a certain location via a specific route we had instructed them to take, to collect a package from an agent, and to return with that package via a specific route that we had instructed them to take to St Michael's Building. These senders knew that both ‘friendly’ and ‘hostile’ agents may ask them questions about their mission and that they had to tell the truth to the friendly agents but to lie to the hostile agents, by making up details about what the mission entails. The participants were then informed that the senders were intercepted and interviewed by friendly and hostile agents after they had delivered the package in St Michael's Building.

9. Again we included Order as a factor in the design to (i) control for its effect on Veracity and Route Recall and (ii) to precise the error term. The Order effects are theoretically not relevant and will therefore not be discussed.