APPLIED COGNITIVE PSYCHOLOGY Appl. Cognit. Psychol. 18: 877–891 (2004) Published online 16 September 2004 in Wiley InterScience (www.interscience.wiley.com) DOI: 10.1002/acp.1057 The Effects of Training Professional Groups and Lay Persons to use Criteria-Based Content Analysis to Detect Deception LUCY AKEHURST 1 *, RAY BULL 1 , ALDERT VRIJ 1 and GU ¨ NTER KO ¨ HNKEN 2 1 University of Portsmouth, UK 2 University of Kiel, Germany SUMMARY This experiment was designed to assess, for the first time, the effects of training police officers, social workers and students in Criteria-Based Content Analysis (CBCA) in an attempt to increase lie detection accuracy. A within-subjects design was implemented. Participants rated the truthfulness of a maximum of four statements before training in CBCA and rated the truthfulness of a different set of four statements after training. The raters were only exposed to the written transcripts of the communicators. Two thirds of the statements utilized were truthful and one third were based on fabrications. Before training, there were no significant differences in detection accuracy between the police officers (66% accuracy), the social workers (72% accuracy) and the students (56% accuracy). After training, the social workers were 77% accurate and significantly more accurate than the police officers (55%) and the students (61%). However, none of the three groups of raters significantly improved their lie detection accuracy after training, in fact, the police officers performed significantly poorer. Overall, police officers were significantly more confident than social workers and lay persons regardless of accuracy. Further, participants were most confident when labelling a statement truthful regardless of whether or not this was the correct decision. Copyright # 2004 John Wiley & Sons, Ltd. Research on the detection of deception via verbal and nonverbal behavioural cues has shown that the ability of untrained raters to discriminate between true and deceptive statements is generally not very good. Vrij (2000, in press) reported that the average accuracy of lie detection, in studies with a chance expectation of 50%, varies within a range of 45% to 60%. However, even 60% accuracy is insufficient for the judgement of eyewitness statements in real court cases since, at best, four out of ten judgements would be wrong (Ko¨hnken, 1987a). Judgements of credibility based on the contents of a statement have, in general, produced higher accuracy rates (than nonverbal behaviour) ranging from 65% to 90% (Vrij, 2000, in press). One method of assessing the truthfulness of a statement, based on its content, was developed in German forensic psychology by Undeutsch (1967, 1984) and Copyright # 2004 John Wiley & Sons, Ltd. *Correspondence to: Dr. Lucy Akehurst, Department of Psychology, University of Portsmouth, King Henry I Street, Portsmouth PO1 2DY, UK. E-mail: lucy.akehurst@port.ac.uk