A guide to measuring expert performance in forensic pattern matching.

Decisions in forensic science are often binary. A firearms expert must decide whether a bullet was fired from a particular gun or not. A face comparison expert must decide whether a photograph matches a suspect or not. A fingerprint examiner must decide whether a crime scene fingerprint belongs to a suspect or not. Researchers who study these decisions have therefore quantified expert performance using measurement models derived largely from signal detection theory. Here we demonstrate that the design and measurement choices researchers make can have a dramatic effect on the conclusions drawn about the performance of forensic examiners. We introduce several performance models - proportion correct, diagnosticity ratio, and parametric and non-parametric signal detection measures - and apply them to forensic decisions. We use data from expert and novice fingerprint comparison decisions along with a resampling method to demonstrate how experimental results can change as a function of the task, case materials, and measurement model chosen. We also graphically show how response bias, prevalence, inconclusive responses, floor and ceiling effects, case sampling, and number of trials might affect one's interpretation of expert performance in forensics. Finally, we discuss several considerations for experimental and diagnostic accuracy studies: (1) include an equal number of same-source and different-source trials; (2) record inconclusive responses separately from forced choices; (3) include a control comparison group; (4) counterbalance or randomly sample trials for each participant; and (5) present as many trials to participants as is practical.

The effect of fingerprint expertise on visual short-term memory.

Expert fingerprint examiners demonstrate impressive feats of memory that may support their accuracy when making high-stakes identification decisions. Understanding the interplay between expertise and memory is therefore critical. Across two experiments, we tested fingerprint examiners and novices on their visual short-term memory for fingerprints. In Experiment 1, experts showed substantially higher memory performance compared to novices for fingerprints from their domain of expertise. In Experiment 2, we manipulated print distinctiveness and found that while both groups benefited from distinctive prints, experts still outperformed novices. This indicates that beyond stimulus qualities, expertise itself enhances short-term memory, likely through more effective organisational processing and sensitivity to meaningful patterns. Taken together, these findings shed light on the cognitive mechanisms that may explain fingerprint examiners' superior memory performance within their domain of expertise. They further suggest that training to improve memory for diverse fingerprints could practically boost examiner performance. Given the high-stakes nature of forensic identification, characterising psychological processes like memory that potentially contribute to examiner accuracy has important theoretical and practical implications.

Specific versus varied practice in perceptual expertise training.

We used a longitudinal randomized control experiment to compare the effect of specific practice (training on one form of a task) and varied practice (training on various forms of a task) on perceptual learning and transfer. Participants practiced a visual search task for 10 hours over 2 to 4 weeks. The specific practice group searched for features only in fingerprints during each session, whereas the varied practice group searched for features in five different image categories. Both groups were tested on a series of tasks at four time points: before training, midway through training, immediately after training ended, and 6 to 8 weeks later. The specific group improved more during training and demonstrated greater pre-post performance gains than the varied group on a visual search task with untrained fingerprint images. Both groups improved equally on a visual search task with an untrained image category, but only the specific group's performance dropped significantly when tested several weeks later. Finally, both groups improved equally on a series of untrained fingerprint tasks. Practice with respect to a single category (versus many) instills better near transfer, but category-specific and category-general visual search training appear equally effective for developing task-general expertise. (PsycInfo Database Record (c) 2022 APA, all rights reserved).

The role of anxiety in mediating the relationship between information consumption and COVID-19 protective behaviours.

Throughout the COVID-19 pandemic, concerns have been raised about an 'infodemic', with information and misinformation being spread across multiple channels and mediums. Information consumption has also been associated with increased anxiety throughout the pandemic. Thus, the present study investigates the mediating role of state anxiety on the relationship between information consumption (defined as mean frequency of information consumption multiplied by number of information sources) and COVID-19 protective behaviours. We compare results across Australian and United States samples and account for personal risk perception and belief in misinformation about COVID-19. Cross-sectional data collected between 28 and 30 April 2020 were analysed using Bayesian structural equation modelling among participants from Australia (N = 201), and the United States (N = 306). State anxiety scores were above the conventional clinical cut-off. Information consumption was positively associated with state anxiety, personal risk perception, and COVID-19 protective behaviours in the Australian and the United States samples. Additionally, the relationship between information consumption and COVID-19 protective behaviours was positively mediated by state anxiety in both nations, suggesting some functional benefits of anxiety. Differences in risk perception and belief in misinformation existed between the Australian and United States sample. Findings provide support for current guidance from organisations such as the WHO, APA, and APS on limiting information consumption to reduce anxiety. To effectively communicate critical public health messaging while minimising potential burdens on mental health, there is a need to develop and test interventions that assist people in calibrating the extent and nature of their information consumption.

The effect of expertise, target usefulness and image structure on visual search.

Experts outperform novices on many cognitive and perceptual tasks. Extensive training has tuned experts to the most relevant information in their specific domain, allowing them to make decisions quickly and accurately. We compared a group of fingerprint examiners to a group of novices on their ability to search for information in fingerprints across two experiments-one where participants searched for target features within a single fingerprint and another where they searched for points of difference between two fingerprints. In both experiments, we also varied how useful the target feature was and whether participants searched for these targets in a typical fingerprint or one that had been scrambled. Experts more efficiently located targets when searching for them in intact but not scrambled fingerprints. In Experiment 1, we also found that experts more efficiently located target features classified as more useful compared to novices, but this expert-novice difference was not present when the target feature was classified as less useful. The usefulness of the target may therefore have influenced the search strategies that participants used, and the visual search advantages that experts display appear to depend on their vast experience with visual regularity in fingerprints. These results align with a domain-specific account of expertise and suggest that perceptual training ought to involve learning to attend to task-critical features.

Collective intelligence in fingerprint analysis.

When a fingerprint is located at a crime scene, a human examiner is counted upon to manually compare this print to those stored in a database. Several experiments have now shown that these professional analysts are highly accurate, but not infallible, much like other fields that involve high-stakes decision-making. One method to offset mistakes in these safety-critical domains is to distribute these important decisions to groups of raters who independently assess the same information. This redundancy in the system allows it to continue operating effectively even in the face of rare and random errors. Here, we extend this "wisdom of crowds" approach to fingerprint analysis by comparing the performance of individuals to crowds of professional analysts. We replicate the previous findings that individual experts greatly outperform individual novices, particularly in their false-positive rate, but they do make mistakes. When we pool the decisions of small groups of experts by selecting the decision of the majority, however, their false-positive rate decreases by up to 8% and their false-negative rate decreases by up to 12%. Pooling the decisions of novices results in a similar drop in false negatives, but increases their false-positive rate by up to 11%. Aggregating people's judgements by selecting the majority decision performs better than selecting the decision of the most confident or the most experienced rater. Our results show that combining independent judgements from small groups of fingerprint analysts can improve their performance and prevent these mistakes from entering courts.

An evidence accumulation model of perceptual discrimination with naturalistic stimuli.

Evidence accumulation models have been used to describe the cognitive processes underlying performance in tasks involving 2-choice decisions about unidimensional stimuli, such as motion or orientation. Given the multidimensionality of natural stimuli, however, we might expect qualitatively different patterns of evidence accumulation in more applied perceptual tasks. One domain that relies heavily on human decisions about complex natural stimuli is fingerprint discrimination. We know little about the ability of evidence accumulation models to account for the dynamic decision process of a fingerprint examiner resolving if 2 different prints belong to the same finger or different fingers. Here, we apply a dynamic decision-making model-the linear ballistic accumulator (LBA)-to fingerprint discrimination decisions to gain insight into the cognitive processes underlying these complex perceptual judgments. Across 3 experiments, we show that the LBA provides an accurate description of the fingerprint discrimination decision process with manipulations in visual noise, speed-accuracy emphasis, and training. Our results demonstrate that the LBA is a promising model for furthering our understanding of applied decision-making with naturally varying visual stimuli. (PsycInfo Database Record (c) 2020 APA, all rights reserved).

How low can you go? Detecting style in extremely low resolution images.

Humans can see through the complexity of scenes, faces, and objects by quickly extracting their redundant low-spatial and low-dimensional global properties, or their style. It remains unclear, however, whether semantic coding is necessary, or whether visual stylistic information is sufficient, for people to recognize and discriminate complex images and categories. In two experiments, we systematically reduce the resolution of hundreds of unique paintings, birds, and faces, and test people's ability to discriminate and recognize them. We show that the stylistic information retained at extremely low image resolutions is sufficient for visual recognition of images and visual discrimination of categories. Averaging over the 3 domains, people were able to reliably recognize images reduced down to a single pixel, with large differences from chance discriminability across 8 different image resolutions. People were also able to discriminate categories substantially above chance with an image resolution as low as 2 × 2 pixels. We situate our findings in the context of contemporary computational accounts of visual recognition and contend that explicit encoding of the local features in the image, or knowledge of the semantic category, is not necessary for recognizing and distinguishing complex visual stimuli. (PsycINFO Database Record (c) 2019 APA, all rights reserved).

A visual familiarity account of evidence for orthographic processing in pigeons (Columbia livia): a reply to Scarf, Corballis, Güntürkün, and Colombo (2017).

Scarf et al. (Proc Natl Acad Sci 113(40):11272-11276, 2016) demonstrated that pigeons, as with baboons (Grainger et al. in Science 336(6078):245-248, 2012; Ziegler in Psychol Sci. https://doi.org/10.1177/0956797612474322 , 2013), can be trained to display several behavioural hallmarks of human orthographic processing. But, Vokey and Jamieson (Psychol Sci 25(4):991-996, 2014) demonstrated that a standard, autoassociative neural network model of memory applied to pixel maps of the words and nonwords reproduces all of those results. In a subsequent report, Scarf et al. (Anim Cognit 20(5):999-1002, 2017) demonstrated that pigeons can reproduce one more marker of human orthographic processing: the ability to discriminate visually presented four-letter words from their mirror-reversed counterparts (e.g. "LEFT" vs. " "). The current report shows that the model of Vokey and Jamieson (2014) reproduces the results of Scarf et al. (2017) and reinforces the original argument: the recent results thought to support a conclusion of orthographic processing in pigeons and baboons are consistent with but do not force that conclusion.

Expertise with unfamiliar objects is flexible to changes in task but not changes in class.

Perceptual expertise is notoriously specific and bound by familiarity; generalizing to novel or unfamiliar images, objects, identities, and categories often comes at some cost to performance. In forensic and security settings, however, examiners are faced with the task of discriminating unfamiliar images of unfamiliar objects within their general domain of expertise (e.g., fingerprints, faces, or firearms). The job of a fingerprint expert, for instance, is to decide whether two unfamiliar fingerprint images were left by the same unfamiliar finger (e.g., Smith's left thumb), or two different unfamiliar fingers (e.g., Smith and Jones's left thumb). Little is known about the limits of this kind of perceptual expertise. Here, we examine fingerprint experts' and novices' ability to distinguish fingerprints compared to inverted faces in two different tasks. Inverted face images serve as an ideal comparison because they vary naturally between and within identities, as do fingerprints, and people tend to be less accurate or more novice-like at distinguishing faces when they are presented in an inverted or unfamiliar orientation. In Experiment 1, fingerprint experts outperformed novices in locating categorical fingerprint outliers (i.e., a loop pattern in an array of whorls), but not inverted face outliers (i.e., an inverted male face in an array of inverted female faces). In Experiment 2, fingerprint experts were more accurate than novices at discriminating matching and mismatching fingerprints that were presented very briefly, but not so for inverted faces. Our data show that perceptual expertise with fingerprints can be flexible to changing task demands, but there can also be abrupt limits: fingerprint expertise did not generalize to an unfamiliar class of stimuli. We interpret these findings as evidence that perceptual expertise with unfamiliar objects is highly constrained by one's experience.

Training perceptual experts: Feedback, labels, and contrasts.

Are strategies for learning in education effective for learning in applied visual domains, such as fingerprint identification? We compare the effect of practice with immediate corrective feedback (feedback training), generating labels for features of matching and mismatching fingerprints (labels training), and contrasting matching and mismatching fingerprints (contrast training). We benchmark these strategies against a baseline of regular practice discriminating fingerprints. We found that all 3 training protocols-feedback, labels, and contrasts-resulted in a significantly greater ability to discriminate new pairs of prints (independent of response bias) than the baseline training protocol. We also found that feedback and labels training produced significantly lower rates of bias (i.e., learners in these groups were less likely to overcall matches) compared with baseline training. Our results demonstrate 3 different ways to boost expertise with matching prints, and have direct application to training perceptual expertise. (PsycINFO Database Record

Thinking forensics: Cognitive science for forensic practitioners.

Human factors and their implications for forensic science have attracted increasing levels of interest across criminal justice communities in recent years. Initial interest centred on cognitive biases, but has since expanded such that knowledge from psychology and cognitive science is slowly infiltrating forensic practices more broadly. This article highlights a series of important findings and insights of relevance to forensic practitioners. These include research on human perception, memory, context information, expertise, decision-making, communication, experience, verification, confidence, and feedback. The aim of this article is to sensitise forensic practitioners (and lawyers and judges) to a range of potentially significant issues, and encourage them to engage with research in these domains so that they may adapt procedures to improve performance, mitigate risks and reduce errors. Doing so will reduce the divide between forensic practitioners and research scientists as well as improve the value and utility of forensic science evidence.

The style of a stranger: Identification expertise generalizes to coarser level categories.

Experience identifying visual objects and categories improves generalization within the same class (e.g., discriminating bird species improves transfer to new bird species), but does such perceptual expertise transfer to coarser category judgments? We tested whether fingerprint experts, who spend their days comparing pairs of prints and judging whether they were left by the same finger or two different fingers, can generalize their finger discrimination expertise to people more broadly. That is, can these experts identify prints from Jones's right thumb and prints from Jones's right index finger as instances of the same "Jones" category? Novices and experts were both sensitive to the style of a stranger's prints; despite lower levels of confidence, experts were significantly more sensitive to this style than novices. This expert advantage persisted even when we reduced the number of exemplars provided. Our results demonstrate that perceptual expertise can be flexible to upwards shifts in the level of specificity, suggesting a dynamic memory retrieval process.

Putting bias into context: The role of familiarity in identification.

Previous demonstrations of context effects in the forensic comparison sciences have shown that the number of "match" responses a person makes can be swayed by case information. Less clear is whether these effects are a result of changes in accuracy (e.g., discrimination ability), a shift in response bias (e.g., tendency to say "match" or "no match") or a mix of the 2. We present a series of experiments where we use a signal detection framework to examine the effects of case information (separately) on forensic comparison accuracy and response bias. We also explore the role of familiarity as 1 potential mechanism for case information to sway accuracy. In Experiment 1, case information about crimes perceived to be more severe swayed people to say "match" more, but had little bearing on their ability to discriminate matching and nonmatching fingerprint pairs. In Experiment 2, case information did affect accuracy when it was familiar (i.e., if a previous similar case was associated with a "match" then people were more likely to also rate the current case as a "match," even though it was not). Even when we blinded people to all extrinsic case information in Experiment 3, accuracy was significantly affected by the familiarity of the fingerprints. These results demonstrate that contextual factors can have different (and independent) influences on accuracy and response bias and that even subtle information can affect accuracy if it is sufficiently similar to the case or trace at hand.