Prison officers and students completed a questionnaire concerning beliefs about 34 verbal and non-verbal behaviours in high- and low-stake situations. They were asked to consider behaviour in four different situations (two high-stake and two low-stake) and were asked to rate each scenario in terms of both the behaviour that they believed that they would show and the behaviour that they believed someone else would show. For each behaviour they were asked to rate on a seven-point scale how frequently they believed it occurred during deception (when compared to truth-telling). While there were no differences within high- and low-stake conditions (i.e. between scenarios of similar stakes), certain behaviours were believed to increase more in high-stake situations compared to low-stake ones. People rated their own behaviour differently to other people's, suggesting that they believed they would appear more credible in deception situations than other people would. Prison officers and students did not differ in their overall beliefs about the cues to deception in line with previous research. Results are discussed in terms of the roles that high-stake, infrequent, easily accessible situations and unsuccessful deceptions may play in the development and maintenance of general beliefs about the cues to deception. Notes
While the prison officers were trainees and so had very little job experience, they had completed courses in non-verbal communication and interviewing skills as part of their training. Although differences between more and less experienced experts have not been explored in terms of beliefs about the cues to deception, there have been no differences found in accuracy of detecting deception between more and less experienced experts (e.g. de Paulo and Pfeifer, 1986). Indeed the evidence from de Paulo and Pfeifer's study with police officers suggests that job experience can lead to the development of inappropriate strategies for detecting deception. Our prison officers had not had an opportunity to develop similar inappropriate strategies. There were no effects of age as a covariate in either analysis and so it was not considered any further. Previous research (e.g. Akehurst et al., 1996) has treated each of the verbal and non-verbal behaviours on the questionnaire as a separate dependent variable and so has used multivariate analyses of variance in order to examine differences in beliefs about the cues to deception. However all behaviours are measured on the same scale, namely an indication of the difference between truth and deception and so it is equally appropriate to treat the verbal and non-verbal behaviours as one independent variable with (in this case) 34 levels. This has been adopted in the current study because of the mixed design used (compared to previous studies that used between-subjects designs). The cues main effect does not really tell us anything about changes in deceptive behaviour from one situation to the next and the other three main effects are concerned with overall changes rather than changes at the level of individual cues, which is something that again conveys little information about the nature of deceptive behaviour. Therefore these effects are not discussed any further.