Papers by Matthew B Thompson
Human Factors: The Journal of the Human Factors and Ergonomics Society, 2010
The aim of this study was to assess how background visual motion and the relative movement of sou... more The aim of this study was to assess how background visual motion and the relative movement of sound affect a head-mounted display (HMD) wearer's performance at a task requiring integration of auditory and visual information. HMD users are often mobile. A commercially available speaker in a fixed location delivers auditory information affordably to the HMD user. However, previous research has shown that mobile HMD users perform poorly at tasks that require integration of visual and auditory information when sound comes from a free-field speaker. The specific cause of the poor task performance is unknown. Participants counted audiovisual events that required integration of sounds delivered via a free-field speaker and vision on an HMD. Participants completed the task while either walking around a room, sitting in the room, or sitting inside a mobile room that allowed separate manipulation of background visual motion and speaker motion. Participants' accuracy at counting target audiovisual events was worse when participants were walking than when sitting at a desk, p = .032. Compared with when they were sitting at a desk, participants' accuracy at counting target audiovisual events showed a trend to be worse when they experienced a combination of background visual motion and the relative movement of sound, p = .058. Multisensory integration performance is least effective when HMD users experience a combination of background visual motion and relative movement of sound. Eye reflexes may play an important role. Results apply to situations in which HMD wearers are mobile when receiving multimodal information, as in health care and military contexts.
PsycEXTRA Dataset, 2000
Advances in forensic technologies and procedures seek to produce better and more efficient polici... more Advances in forensic technologies and procedures seek to produce better and more efficient policing for safer societies. Little is understood, however, about how effectively the human forensic professional employs such technologies, or the cognitive and perceptual processes of judgment and decision making the forensic professional engages in during the course of evidence evaluation. For this, experimenters need materials that approximate the realism of crime scene evidence, while ensuring the ground truth about the source of this information. These two goals are often incompatible. We discuss the development of an open-source biometric repository to address the issue of ground truth. This repository contains a range of crime related materials such as fingerprints and palm-prints, shoe-prints, faces, handwriting, voices, and irises. Our goal is to provide a large, open-source repository of forensic information, where certainty of the source in built into the system, to help advance research on identification by humans and technology.
Human Factors and Ergonomics Society Annual Meeting Proceedings, Oct 1, 2009
Multisensory integration is the perceptual process by which the user of a Head-Mounted Display (H... more Multisensory integration is the perceptual process by which the user of a Head-Mounted Display (HMD) combines, into a single object, vision from the HMD with concurrent auditory signals. Because HMD users are usually mobile, visual and auditory information may not always be spatially congruent, yet congruence is a requirement for multisensory integration to occur. Previous research has shown that multisensory integration was less effective when the user was walking and sound was delivered via a speaker in a fixed location. In Experiment 1, we showed that people integrate information less effectively when they hear sound from a speaker while they walk rather than sit, because they experience a combination of sound motion and background motion, not because of any workload associated with walking. In Experiment 2, in which participants' multisensory integration performance did not rely on working memory, their performance is worse when they walk rather than sit when hearing sound with the earpiece, rather than in free-field. These mixed results highlight the difficulty in replicating multisensory integration research in applied contexts.
Advances in forensic technologies and procedures seek to produce better and more efficient polici... more Advances in forensic technologies and procedures seek to produce better and more efficient policing for safer societies. Little is understood, however, about how effectively the human forensic professional employs such technologies, or the cognitive and perceptual processes of judgment and decision making the forensic professional engages in during the course of evidence evaluation. For this, experimenters need materials that approximate the realism of crime scene evidence, while ensuring the ground truth about the source of this information. These two goals are often incompatible. We discuss the development of an open-source biometric repository to address the issue of ground truth. This repository contains a range of crime related materials such as fingerprints and palm-prints, shoe-prints, faces, handwriting, voices, and irises. Our goal is to provide a large, open-source repository of forensic information, where certainty of the source in built into the system, to help advance research on identification by humans and technology.
Tangen, JM, Thompson, MB, McCarthy, D. and Tear, MJ (2010). Ground truth: On certainty in forensi... more Tangen, JM, Thompson, MB, McCarthy, D. and Tear, MJ (2010). Ground truth: On certainty in forensic decision-making research. In: , Proceedings of the 20th International Symposium on the Proceedings of the 20th International Symposium on the Proceedings of the 20th International Symposium on the Forensic Sciences. 20th International Symposium on the Forensic Sciences, Sydney , Australia, (). 5-10 September 2010.
PsycEXTRA Dataset, 2000
Multisensory integration is the perceptual process by which the user of a Head-Mounted Display (H... more Multisensory integration is the perceptual process by which the user of a Head-Mounted Display (HMD) combines, into a single object, vision from the HMD with concurrent auditory signals. Because HMD users are usually mobile, visual and auditory information may not always be spatially congruent, yet congruence is a requirement for multisensory integration to occur. Previous research has shown that multisensory integration was less effective when the user was walking and sound was delivered via a speaker in a fixed location. In Experiment 1, we showed that people integrate information less effectively when they hear sound from a speaker while they walk rather than sit, because they experience a combination of sound motion and background motion, not because of any workload associated with walking. In Experiment 2, in which participants' multisensory integration performance did not rely on working memory, their performance is worse when they walk rather than sit when hearing sound with the earpiece, rather than in free-field. These mixed results highlight the difficulty in replicating multisensory integration research in applied contexts.
Participants Thirty-seven qualified practicing fingerprint experts from five police organizations... more Participants Thirty-seven qualified practicing fingerprint experts from five police organizations (the Australian Federal, New South Wales, Queensland, South Australia, and Victoria Police) participated in the study. In addition, 37 undergraduates from The University of Queensland participated for course credit, providing comparison data on the performance of novices.
Advances in forensic technologies and procedures seek to produce better and more efficient polici... more Advances in forensic technologies and procedures seek to produce better and more efficient policing for safer societies. Little is understood, however, about how effectively the human forensic professional employs such technologies, or the cognitive and perceptual processes of judgment and decision making the forensic professional engages in during the course of evidence evaluation. For this, experimenters need materials that approximate the realism of crime scene evidence, while ensuring the ground truth about the source of this information. These two goals are often incompatible. We discuss the development of an open-source biometric repository to address the issue of ground truth. This repository contains a range of crime related materials such as fingerprints and palm-prints, shoe-prints, faces, handwriting, voices, and irises. Our goal is to provide a large, open-source repository of forensic information, where certainty of the source in built into the system, to help advance research on identification by humans and technology.
Although fingerprint experts have presented evidence in criminal courts for more than a century, ... more Although fingerprint experts have presented evidence in criminal courts for more than a century, there have been few scientific investigations of the human capacity to discriminate these patterns. A recent latent print matching experiment shows that qualified, court-practicing fingerprint experts are exceedingly accurate (and more conservative) compared with novices, but they do make errors. Here, a rationale for the design of this experiment is provided. We argue that fidelity, generalizability and control must be balanced in order to answer important research questions; that the proficiency and competence of fingerprint examiners is best determined when experiments include highly similar print pairs, in a signal detection paradigm, where the ground truth is known; and that inferring from this experiment the statement "The error rate of fingerprint identification is 0.68%" would be disingenuous. In closing, the ramifications of these findings for the future psychological study of forensic expertise, and the implications for expert testimony and public policy are considered.
Proceedings of the …, 2010
Television shows like 'CSI' can give the impression that matching crime-scene fingerprints is ful... more Television shows like 'CSI' can give the impression that matching crime-scene fingerprints is fully automated. But it is actually humans (fingerprint experts) who ultimately decide whether a crime-scene print belongs to a suspect or not. Despite this fact, there have been no published, peer-reviewed studies directly examining the extent to which experts can correctly match fingerprints to one another. In two experiments presented here we aim to determine the factors affecting accuracy using non-expert participants and test (1) whether the advantage found for the sequential presentation of faces applies to prints and (2) whether the amount of information in a print matters.
We discovered an interesting face distortion effect while preparing a set of face images for an i... more We discovered an interesting face distortion effect while preparing a set of face images for an identification experiment. We obtained a set of faces from a Slovakian database (SmartNet IBC, no date) and eye-aligned them using PsychoMorph . To check the consistency of the eye alignment, we started skimming through the images on the computer at a fast pace. After a while, we made a few remarks to one another about the`aesthetically challenged' faces in the set. They began to appear highly deformed and grotesque. But after inspecting the especially ugly faces individually, each of them appeared normal or even attractive.
Psychological Science, 2011
Introduction. We compared anaesthetists' ability to learn six alarm sounds, each in medium and hi... more Introduction. We compared anaesthetists' ability to learn six alarm sounds, each in medium and high priority forms, from three different alarm sets: (1) alarms from the set proposed in the IEC 60601-1-8 standard 1 2 for alarms on medical electrical equipment (IEC), (2) a modification of the IEC set (Modified) with simplified high priority alarms, and (3) a set developed over 20 years ago by psychoacoustician Roy Patterson 3 (Patterson).
Human operators who use head-mounted displays (HMDs) in their work may benefit from auditory supp... more Human operators who use head-mounted displays (HMDs) in their work may benefit from auditory support. It is unclear whether auditory support is better delivered in free-field or via earpiece, and what the effect of walking is. To examine this problem, a novel multisensory integration task was created in which participants identified mismatches between sounds and visual information on an HMD. Participants listened to the sounds either via earpiece or free-field while they either sat or walked about the test room. When using an earpiece, participants performed the mismatch task equally well whether walking or sitting, but when using free-field sound, participants performed the task significantly worse when walking than when sitting. The worse performance for participants using free-field sound while walking may relate to spatial and motion inconsistencies between the sound and vision or because of misperceptions of the time at which the sounds occurred. The results underscore the need for representative design of experiments exploring multisensory integration and they suggest auditory conditions that might influence effective multisensory integration with HMDs.
Human Factors, 2010
Objective: The aim of this study was to assess how background visual motion and the relative move... more Objective: The aim of this study was to assess how background visual motion and the relative movement of sound affect a head-mounted display (HMD) wearer's performance at a task requiring integration of auditory and visual information. Background: HMD users are often mobile. A commercially available speaker in a fixed location delivers auditory information affordably to the HMD user. However, previous research has shown that mobile HMD users perform poorly at tasks that require integration of visual and auditory information when sound comes from a free-field speaker. The specific cause of the poor task performance is unknown. Method: Participants counted audiovisual events that required integration of sounds delivered via a free-field speaker and vision on an HMD. Participants completed the task while either walking around a room, sitting in the room, or sitting inside a mobile room that allowed separate manipulation of background visual motion and speaker motion. Results: Participants' accuracy at counting target audiovisual events was worse when participants were walking than when sitting at a desk, p = .032. Compared with when they were sitting at a desk, participants' accuracy at counting target audiovisual events showed a trend to be worse when they experienced a combination of background visual motion and the relative movement of sound, p = .058. Conclusion: Multisensory integration performance is least effective when HMD users experience a combination of background visual motion and relative movement of sound. Eye reflexes may play an important role. Application: Results apply to situations in which HMD wearers are mobile when receiving multimodal information, as in health care and military contexts.
Human Factors and Ergonomics Society Annual Meeting Proceedings, 2009
Multisensory integration is the perceptual process by which the user of a Head-Mounted Display (H... more Multisensory integration is the perceptual process by which the user of a Head-Mounted Display (HMD) combines, into a single object, vision from the HMD with concurrent auditory signals. Because HMD users are usually mobile, visual and auditory information may not always be spatially congruent, yet congruence is a requirement for multisensory integration to occur. Previous research has shown that multisensory integration was less effective when the user was walking and sound was delivered via a speaker in a fixed location. In Experiment 1, we showed that people integrate information less effectively when they hear sound from a speaker while they walk rather than sit, because they experience a combination of sound motion and background motion, not because of any workload associated with walking. In Experiment 2, in which participants' multisensory integration performance did not rely on working memory, their performance is worse when they walk rather than sit when hearing sound with the earpiece, rather than in free-field. These mixed results highlight the difficulty in replicating multisensory integration research in applied contexts.
Uploads
Papers by Matthew B Thompson