Undercover and collective interviewing to detect deception

  • Shyma Jundi

Student thesis: Doctoral Thesis


This thesis aims to examine whether undercover and collective interviewing can elicit cues to deceit. Undercover interviewing is when the suspect is not explicitly informed that they are being interviewed, and collective interviewing involves one interviewer interviewing multiple suspects. In Chapter 1, the thesis is introduced. Then Experiment 1 is described in which participants were covertly interviewed about their plans for an upcoming trip. Findings indicate significant verbal differences in truth tellers’ and liars’ responses.
Experiment 2 is a lie detection study, which is carried out in order to determine whether these differences could be identified by lay observers. Observers were given transcripts of undercover interviews from Experiment 1. They could correctly determine veracity significantly better than chance level.
Experiment 3 examines another short undercover interview. Participants were despatched on a mission to take photographs, with truth tellers aiming to promote the square to visitors and liars surveying the area for a place to plant a decoy device. When they finished taking photographs, each participant was approached by a mime artist who asked them if they had photographed him and if he could see the photos. Results showed that truth tellers were more likely than liars to admit to having photographed him, and to allow him to see the photos. When analysing the photos, truth tellers’ photographs were more open, appealing, included more people, and central than liars’ photographs. Suspicious features were more prominent in liars' photos and liars mentioned them more frequently.
The collective interviewing manipulation is tested in Experiment 4a, in which suspects were interviewed in pairs about their recent activities. Pairs of truth tellers went to lunch in a nearby restaurant, whereas pairs of liars 'stole' money from a purse in an office and were asked to use the truth tellers' activities as an alibi. Results showed that liars looked at the interview more, and exhibited less gaze aversion than truth tellers. More liars than truth tellers developed a strategy prior to the interview. In Experiment 4b, the data from Experiment 4a is analysed to assess the verbal behaviour of the suspects when being interviewed collectively. Truth tellers interrupted each other more, corrected each other more, and added more information to each other’s accounts than liars.
Experiment 5 is a combination study involving undercover and collective interviewing. Participants undertook a mission in pairs, where they photographed an animal enclosure in a park. Truth tellers did this to collect material for a promotional flyer, whereas liars acted as animal rights activists. Participants were interviewed covertly and formally in pairs. Results showed that liars had less overlap than truth tellers when their responses in the covert interview were compared to their responses in the formal interview. Liars were also less likely than truth tellers to mention the undercover interviewer in the formal interview.
Chapter 8 is the General Discussion. Findings are summarised, and implications, future research and limitations are discussed. The overall conclusion is that undercover interviewing and collective interviewing elicit observable cues to deceit.

Date of AwardSept 2013
Original languageEnglish
Awarding Institution
  • University of Portsmouth
SponsorsEngineering and Physical Sciences Research Council
SupervisorAldert Vrij (Supervisor), Lorraine Hope (Supervisor) & Sam Mann (Supervisor)

Cite this