Internal seminar: Danni Norman and Owain Ritchie
Contact: Jesse Preston
Owain Ritchie “Investigating the psychological factors influencing evaluations of autonomous vehicle behaviour”
With rapid technological advancements taking place in the transition towards autonomous vehicles, understanding the factors influencing trust and acceptance of autonomy is becoming increasingly important. Much research has concentrated on acceptance of autonomy using survey-based methods, although comparatively little work has examined evaluations of specific vehicle behaviours in driving simulator settings. Three experiments were conducted on evaluations of an autonomous vehicle’s overtaking behaviour. Experiment 1 examined the influence of pull-in distance on trust, finding a rapid linear increase in ratings with increasing distance, before leveling off and dropping towards the largest distances. Experiments 2 and 3 investigated the influence of a following vehicle on ratings of an overtaking vehicle, finding that a safe following distance (2s) did not influence ratings of the overtaking vehicle. However, when the third vehicle followed at 0.5s, the overtaking vehicle was rated more negatively at the largest pull-in distance. This suggests a transfer of the following vehicle’s behaviour onto ratings of the overtaking vehicle, and that evaluations of autonomous vehicle behaviour are to some extent content-dependent.
Danni Norman - Deception detection with the concealed information test – an overview of current application and future opportunities
The Concealed Information Test (CIT, also known as the Guilty Knowledge Test) is a well-established paradigm for memory detection and advocated as an effective method for deception detection. Theoretically grounded, the CIT exploits a specific cognitive process, the orienting response, to detect concealment. Traditionally the CIT is conducted using physiological measures, however, recent research has developed an equally sensitive test based on response times. Over 60 years of research has consistently shown the CIT to be 90%+ accurate at detecting concealed information whilst maintaining a near perfect false positive rate. With the exception of Japan, where the CIT is successfully applied in real criminal cases, the CIT is little-known outside academia. This is probably due to the misconception that it is another ‘polygraph test’. This presentation will: 1) provide an overview of the standard forensically-applied CIT, 2) explore other possible applications within the intelligence and commercial domain, and 3) consider the practical strengths and limitations with its application.