Language of lies in prison: linguistic classification of prisoners' truthful and deceptive natural language

Linguistic Inquiry and Word Count (LIWC) software was used to code truthful and deceptive words in prisoners' natural language. Reality Monitoring (RM) and Newman, Pennebaker, Berry, & Richards' (NP, 2003) models were used. NP indicates that lies contain fewer self‐references, other references, and exclusive words, and higher numbers of negative emotion and motion words. Higher sensory, spatial, temporal and affective RM terms were predicted for truths, and more cognitive mechanism words were predicted for lies. The RM model's hit rate was 71.1% and discriminability was 1.11 without spatial words, which were surprisingly higher in lies than in truth statements, and the NP model was 69.7%, d′ = 0.99. The software models were contrasted with humans' hit rates, and younger prisoners had 71% hits, d′ = 0.22, but older prisoners had 50% hits, d′ = 0.90. The software set unbiased criteria (β), but younger prisoners were biased in setting their criteria when judging statement veracity, β = ?0.34. Without other references, found to be higher in truths than in lies, NP classified 59% of statements correctly. Copyright © 2005 John Wiley & Sons, Ltd.     

The effects of length and transposed‐letter similarity in lexical decision: Evidence with beginning,intermediate, and adult readers

Do length and transposed‐letter effects reflect developmental changes on reading acquisition in a transparent orthography? Can computational models of visual word recognition accommodate these changes? To answer these questions, we carried out a masked priming lexical decision experiment with Spanish beginning, intermediate, and adult readers (N=36, 44, and 39; average age: 7, 11, and 22 years, respectively). Target words were either short or long (6.5 vs. 8.5 letters), and transposed‐letter primes were formed by the transposition of two letters (e.g. aminal–ANIMAL) or by the substitution of two letters (orthographic control: arisal–ANIMAL). Children showed a robust length effect (i.e. long words were read slower than short words) that vanished in adults. In addition, both children and young adults showed a transposed‐letter priming effect relative to the control condition. A robust transposed‐letter priming effect was also observed in non‐word reading, which strongly suggests that this effect occurs at an early prelexical level. Taken together, the results reveal that children evolve from a letter‐by‐letter reading to a direct lexical access and that the lexical decision task successfully captures the changing strategies used by beginning, intermediate, and adult readers. We examine the implications of these findings for the recent models of visual word recognition.     

Social Desirability and the Interpretation of Uncertainty Terms in Self‐Report Questions

Uncertainty terms (e.g., possible) are words that are not fixed and hence open to interpretation. This research examined the role of social desirability in how these words are interpreted in self‐report questions. Participants in Experiments 1 (N = 96; MTurk workers) and 2 (N = 96; college students) judged trait (N = 48) and behavior (N = 36) items endorsed by a hypothetical individual to be more likely if they were lower in social desirability. In Experiments 3 (N = 97) and 4 (N = 97) college student participants interpreted four different uncertainty terms (likely, possible, unlikely, and pretty sure) as indicating greater certainty when they referred to socially undesirable (relative to socially desirable) traits (N = 36) and behaviors (N = 24). These results suggest that participants may interpret self‐report items differently, depending on the social desirability of the content.Copyright © 2017 John Wiley & Sons, Ltd.     

Does implementation of reforms authorized in Section 28 of the Youth Justice and Criminal Evidence Act affect the complexity of the questions asked of young alleged victims in court?

This study examined whether the implementation of Section 28 (S28) of the Youth Justice and Criminal Evidence Act and the introduction of mandatory ground rules hearings reduced the complexity of the questions English lawyers asked when examining child witnesses. This study compared cases with (n = 43) and without (n = 43) the S28 special measures and involved children aged 6–15 testifying as alleged victims of sexual abuse. Defense lawyers' questions in the S28 condition comprised fewer words, clauses, false starts, multiple negatives, and temporal and numeric attributes than in the non‐S28 condition. When questioning younger children, lawyers used fewer words, clauses, references to “before/after,” and passive voice. These results demonstrated that S28 successfully reduced the complexity of the questions and that lawyers in both conditions partially adjusted the complexity of their questions to accommodate children's developmental capabilities.     

Gender and age differences in parent–child emotion talk

This study examined gender differences in emotion word use during mother–child and father–child conversations. Sixty‐five Spanish mothers and fathers and their 4‐ (M = 53.50, SD = 3.54) and 6‐year‐old (M = 77.07, SD = 3.94) children participated in this study. Emotion talk was examined during a play‐related storytelling task and a reminiscence task (conversation about past experiences). Mothers mentioned a higher proportion of emotion words than did fathers. During the play‐related storytelling task, mothers of 4‐year‐old daughters mentioned a higher proportion of emotion words than did mothers of 4‐year‐old sons, whereas fathers of 4‐year‐old daughters directed a higher proportion of emotion words than did fathers of 4‐year‐old sons during the reminiscence task. No gender differences were found with parents of 6‐year‐old children. During the reminiscence task daughters mentioned more emotion words with their fathers than with their mothers. Finally, mothers' use of emotion talk was related to whether children used emotion talk in both tasks. Fathers' use of emotion talk was only related to children's emotion talk during the reminiscence task.     

Probing Lexical Ambiguity: Word Vectors Encode Number and Relatedness of Senses

Lexical ambiguity—the phenomenon of a single word having multiple, distinguishable senses—is pervasive in language. Both the degree of ambiguity of a word (roughly, its number of senses) and the relatedness of those senses have been found to have widespread effects on language acquisition and processing. Recently, distributional approaches to semantics, in which a word's meaning is determined by its contexts, have led to successful research quantifying the degree of ambiguity, but these measures have not distinguished between the ambiguity of words with multiple related senses versus multiple unrelated meanings. In this work, we present the first assessment of whether distributional meaning representations can capture the ambiguity structure of a word, including both the number and relatedness of senses. On a very large sample of English words, we find that some, but not all, distributional semantic representations that we test exhibit detectable differences between sets of monosemes (unambiguous words; N = 964), polysemes (with multiple related senses; N = 4,096), and homonyms (with multiple unrelated senses; N = 355). Our findings begin to answer open questions from earlier work regarding whether distributional semantic representations of words, which successfully capture various semantic relationships, also reflect fine-grained aspects of meaning structure that influence human behavior. Our findings emphasize the importance of measuring whether proposed lexical representations capture such distinctions: In addition to standard benchmarks that test the similarity structure of distributional semantic models, we need to also consider whether they have cognitively plausible ambiguity structure.     

How satisfied are you with your job? Estimating the reliability of scores on a single‐item job satisfaction measure

Estimating the reliability of scores on single‐item measures can be difficult because commonly used internal consistency estimates of reliability cannot be calculated. When longitudinal data is available, statistical models can be used to decompose the variability in the latent variable at each wave into trait versus state variance. Then, reliability can be estimated as a ratio of the sum of the trait variance that is captured in repeated assessments over the total variance. The current study used latent trait‐state‐error models on a nine‐year longitudinal data (N = 5,003) to estimate the test–retest reliability of scores on a single‐item measure of job satisfaction. Results showed that job satisfaction scores were somewhat unreliable (r_xx = .49–.59) and amenable to change.     

Are you for real? Exploring language use and unexpected process questions within the detection of identity deception

The current study was to test whether reality monitoring and language use could distinguish identity liars from truth tellers when answering outcome questions and unexpected process questions. Truth tellers (n = 30) and liars (n = 30) discussed their identity in a recruitment interview. No differences emerged between truth tellers and liars in the details they provided. In terms of language use, liars used more positive language than truth tellers, whereas truth tellers used more cognitive process words than liars. However, neither were more pronounced when asking process questions. Overall, process questions elicited more cognitive process and cause words than outcome questions. Therefore, process questions may be able to contribute to the cognitive load approach. The findings suggest that reality monitoring may not be diagnostic when applied to identity deception. We discuss the language use differences in relation to impression management theory.     

States of curiosity and interest enhance memory differently in adolescents and in children

Curiosity – broadly defined as the desire to acquire new information – enhances learning and memory in adults. In addition, interest in the information (i.e., when the information is processed) can also facilitate later memory. To date, it is not known how states of pre‐information curiosity and post‐information interest enhance memory in childhood and adolescence. We used a trivia paradigm in which children and adolescents (N = 60, 10–14 years) encoded trivia questions and answers associated with high or low curiosity. States of high pre‐answer curiosity enhanced later memory for trivia answers in both children and adolescents. However, higher positive post‐answer interest enhanced memory for trivia answers beyond the effects of curiosity more strongly in adolescents than in children. These results suggest that curiosity and interest have positive effects on learning and memory in childhood and adolescence, but might need to be harnessed in differential ways across child development to optimize learning.     

Primary Care Physicians’ Awareness, Experience and Opinions of Direct-to-Consumer Genetic Testing

The purpose of this study was to assess primary care physicians’ awareness, experience, opinions and preparedness to answer patients’ questions regarding direct-to-consumer (DTC) genetic testing. An anonymous survey was mailed to 2,402 family and internal medicine providers in North Carolina. Of the 382 respondents, 38.7% (n = 148) were aware of and 15% (n = 59) felt prepared to answer questions about DTC genetic tests. Respondents aged 50 or older were more likely to be aware of DTC genetic testing than those less than 40 years old (OR = 2.42). Male providers were more likely to feel prepared to answer questions than female providers (OR = 2.65). Among respondents who reportedly were aware, family practitioners were more likely than internists (OR = 3.30) to think DTC testing was clinically useful, and 18.9% had patients ask questions or bring in test results. The small percent of physicians who were aware of DTC genetic testing or felt prepared to answer questions about it suggests that education of providers will be necessary if testing becomes more widespread.     

Body image disturbance and skin bleaching

This study looks at body image disturbance among Jamaicans who bleach their skin. The hypothesis states that there is a positive relationship between skin bleaching and body image disturbance. The study used a convenience sample of 160 participants with a skin bleaching group (n = 80) and a non‐bleaching comparison group (n = 80). The instrument included demographic questions, the body image disturbance questionnaire (BIDQ), and questions about skin bleaching. The results of a t‐test revealed that the skin bleaching group (M = 1.5255, SD = 0.42169) was not significantly different from the non‐bleaching group (M = 1.4938, SD = 0.74217) in terms of body image disturbance, t(158) = 0.333, p = .740. The participants who bleached did not suffer from body image disturbance. Self‐reports revealed that they bleached to acquire beauty, attract a partner, elude the police, and market skin bleaching products. The practice was fashionable and popular and it made some participants feel good, while others were fans of a popular musical artiste who bleached his skin. The majority of participants bleached because of the perceived personal, social, and entrepreneurial benefits of the practice and not because they suffered emotional distress, anxiety, and functional impairment because of their skin colour. However, there was some level of BID among the minority of participants who argued that they bleached because they wanted to be pretty so they were emotionally distressed about there body image and experienced functional impairment.     

Effects of exposure to alcohol‐related cues on racial discrimination

Prior research has shown that exposure to alcohol‐related images exacerbates expression of implicit racial biases, and that brief exposure to alcohol‐related words increases aggressive responses. However, the potential for alcohol cue exposure to elicit differential aggression against a Black (outgroup) relative to a White (ingroup) target—that is, racial discrimination—has never been investigated. Here, we found that White participants (N = 92) exposed to alcohol‐related words made harsher judgments of a Black experimenter who had frustrated them than participants who were exposed to nonalcohol words. These findings suggest that exposure to alcohol cues increases discriminatory behaviors toward Blacks.     

Mapping differential responses to cognitive training using machine learning

We used two simple unsupervised machine learning techniques to identify differential trajectories of change in children who undergo intensive working memory (WM) training. We used self‐organizing maps (SOMs)—a type of simple artificial neural network—to represent multivariate cognitive training data, and then tested whether the way tasks are represented changed as a result of training. The patterns of change we observed in the SOM weight matrices implied that the processes drawn upon to perform WM tasks changed following training. This was then combined with K‐means clustering to identify distinct groups of children who respond to the training in different ways. Firstly, the K‐means clustering was applied to an independent large sample (N = 616, M_age = 9.16 years, range = 5.16–17.91 years) to identify subgroups. We then allocated children who had been through cognitive training (N = 179, M_age = 9.00 years, range = 7.08–11.50 years) to these same four subgroups, both before and after their training. In doing so, we were able to map their improvement trajectories. Scores on a separate measure of fluid intelligence were predictive of a child's improvement trajectory. This paper provides an alternative approach to analysing cognitive training data that go beyond considering changes in individual tasks. This proof‐of‐principle demonstrates a potentially powerful way of distinguishing task‐specific from domain‐general changes following training and of establishing different profiles of response to training.     

Determinants and Predictive Value of Clinician Assessment of Short‐Term Suicide Risk

We examine the interrelations among clinicians' judgment of patients' suicide risk, clinicians' emotional responses, and standard risk factors in the short‐term prediction of suicidal thoughts and behaviors. Psychiatric outpatients (n = 153) with a lifetime history of suicide ideation/attempt and their treating clinicians (n = 67) were evaluated at intake. Clinicians completed a standard suicide risk instrument (modified SAD PERSONS scale), a 10‐point Likert scale assessment of judgment of patient suicide risk (Clinician Prediction Scale), and a measure of their emotional responses to the patient (Therapist Response Questionnaire‐Suicide Form). The Columbia Suicide Severity Rating Scale and the Beck Scale for Suicide Ideation were administered at a one‐month follow‐up assessment (n = 114, 74.5%). Clinician judgment of risk significantly predicted suicidal thoughts and behaviors at follow‐up. Both the standard suicide risk instrument and clinician emotional responses contributed independently to the clinician assessment of risk, which, in turn, mediated their relationships with suicidal thoughts and behaviors. Our findings validate the importance of clinical judgment in assessing suicide risk. Clinical judgment appears to be informed both by concrete risk factors and clinicians' emotional responses to suicidal patients, highlighting emotional awareness as a promising area for research and training.     

The Minds of God(s) and Humans: Differences in Mind Perception in Fiji and North America

Previous research suggests that how people conceive of minds depends on the culture in which they live, both in determining how they interact with other human minds and how they infer the unseen minds of gods. We use exploratory factor analysis to compare how people from different societies with distinct models of human minds and different religious traditions perceive the minds of humans and gods. In two North American samples (American adults, N = 186; Canadian students, N = 202), we replicated a previously found two‐factor agency/experience structure for both human and divine minds, but in Fijian samples (Indigenous iTaukei Fijians, N = 77; Fijians of Indian descent, N = 214; total N = 679) we found a three‐factor structure, with the additional containing items related to social relationships. Further, Fijians’ responses revealed a different three‐factor structure for human minds and gods’ minds. We used these factors as dimensions in the conception of minds to predict (a) expectations about human and divine tendencies towards punishment and reward; and (b) conception of gods as more embodied (an extension of experience) or more able to know people's thoughts (an extension of agency). We found variation in how these factors predict conceptions of agents across groups, indicating further theory is needed to explain how culturally generated concepts of mind lead to other sorts of social inferences. We conclude that mind perception is shaped by culturally defined social expectations and recommend further work in different cultural contexts to examine the interplay between culture and social cognition.     

Linking property crime using offender crime scene behaviour: A comparison of methods

This study compared the ability of seven statistical models to distinguish between linked and unlinked crimes. The seven models utilised geographical, temporal, and modus operandi information relating to residential burglaries (n = 180), commercial robberies, (n = 118), and car thefts (n = 376). Model performance was assessed using receiver operating characteristic analysis and by examining the success with which the seven models could successfully prioritise linked over unlinked crimes. The regression‐based and probabilistic models achieved comparable accuracy and were generally more accurate than the tree‐based models tested in this study. The Logistic algorithm achieved the highest area under the curve (AUC) for residential burglary (AUC = 0.903) and commercial robbery (AUC = 0.830) and the SimpleLogistic algorithm achieving the highest for car theft (AUC = 0.820). The findings also indicated that discrimination accuracy is maximised (in some situations) if behavioural domains are utilised rather than individual crime scene behaviours and that the AUC should not be used as the sole measure of accuracy in behavioural crime linkage research.     

Irrelevant‐action imitation is short‐term and contextual: Evidence from two under‐studied populations

Across the lifespan and across populations, humans ‘overimitate’ causally unnecessary behaviors. Such irrelevant‐action imitation facilitates faithful cultural transmission, but its immediate benefits to the imitator are controversial. Over short time scales, irrelevant‐action imitation may bootstrap artifact exploration or interpersonal affiliation, and over longer time scales it may facilitate acquisition of either causal models or social conventions. To investigate these putative functions, we recruited community samples from two under‐studied populations: Yasawa, Fiji, and Huatasani, Peru. We use a two‐action puzzle box: first after a video demonstration, and again one month later. Treating age as a continuous variable, we reveal divergent developmental trajectories across sites. Yasawans (44 adults, M = 39.9 years, 23 women; 42 children, M = 9.8 years, 26 girls) resemble documented patterns, with irrelevant‐action imitation increasing across childhood and plateauing in adulthood. In contrast, Huatasaneños (48 adults, M = 37.6 years, 33 women; 47 children, M = 9.3 years, 13 girls) evince a parabolic trajectory: adults at the site show the lowest irrelevant‐action imitation of any demographic set in our sample. In addition, all age sets in both populations reduce their irrelevant actions at Time 2, but do not reduce their relevant‐action imitation or goal attainment. Taken together, and considering the local cultural contexts, our results suggest that irrelevant‐action imitation serves a short‐term function and is sensitive to the social context of the demonstration.     

Cognitive Lie Detection: Response Time and Consistency of Answers as Cues to Deception

Purpose  The purpose of this study was to test a new cognitive lie detection method, time restricted integrity confirmation (Tri-Con), which uses response time and inconsistencies across answers as cues to deception. Design/methodology/approach   Data were obtained from two samples of students enrolled in psychology classes (n = 96 for Experiment 1, n = 99 for Experiment 2). The experimental task required students to lie or tell the truth to questions probing biodata under time restriction. The foci of questions (such as Academics or Employment History) were chosen because of their relevance to participants’ lives. Findings  Tri-Con was able to distinguish between truth tellers and liars after controlling for individual differences. In one experiment, liar-truth teller classification accuracies reached 89%. Mean response times and answer consistency can be used to distinguish those who lie from those who tell the truth. Implications  Research on cognitive-based lie detectors, such as Tri-Con, hold the potential for developing reliable and valid methods of screening out employees likely to engage in misconduct and providing deceptive answers to screening questions. A cognitive lie detector would constitute a paradigm shift away from the polygraph, and could be used in tandem with integrity tests. Originality/value   This study was a preliminary test of a cognitive lie detection method based on a model of cognitive events (the Activation-Decision-Construction model) when people answer questions deceptively. It constitutes a step in translating laboratory-based cognitive research into applied technologies for the real world detection of lying, including lying that occurs during pre-employment screening. Received and reviewed by former editor, George Neuman.     

Dimensionality of stress experiences: Factorial structure of the Perceived Stress Questionnaire (PSQ) in a population‐based Swedish sample

We investigated the factorial structure of the Perceived Stress Questionnaire (PSQ‐recent; Levenstein, Prantera, Varvo et al., 1993) in a large (N = 1516; 35–95 years) population‐based Swedish sample (Nilsson, Adolfsson, Bäckman et al., 2004; Nilsson, Bäckman, Erngrund et al., 1997). Exploratory principal components analysis (PCA) was conducted on a first, randomly drawn subsample (n = 506). Next, the model based on the PCA was tested in a second sample (n = 505). Finally, a third sample (n = 505) was used to cross‐validate the model. Five components were extracted in the PCA (eigenvalue > 1) and labeled “Demands,” “Worries/Tension,” “Lack of joy,” “Conflict,” and “Fatigue,” respectively. Twenty‐one out of the 30 original PSQ items were retained in a confirmatory factor analysis (CFA) model that included the five (first‐order) factors and, additionally, a general (second‐order) stress factor, not considered in prior models. The model showed reasonable goodness of fit [χ²(184) = 511.2, p < 0.001; CFI = 0.904; RMSEA = 0.059; and SRMR = 0.063]. Multigroup confirmatory factor analyses supported the validity of the established model. The results are discussed in relation to prior investigations of the factorial structure of the PSQ.     

Pyramidal parent training using behavioral skills training: Training caregivers in the use of a differential reinforcement procedure

Six caregivers participated in a research study in which behavioral skills training (BST) was used within a pyramidal training model to train a differential reinforcement of an alternative behavior (DRA) procedure. Physical prompting was utilized to obtain correct responses across the identified alternative behavior. The caregivers were split into two tiers, comprised of three caregivers each. The experimenter trained tier‐one caregivers, who then trained tier‐two caregivers after meeting a predetermined mastery criterion. A multiple baseline design across participants research design was used to evaluate the effectiveness of correct implementation of the DRA procedure, demonstrating experimental control across participants. During baseline, caregivers did not implement DRA correctly. Following training, tier‐one and tier‐two caregivers demonstrated correct implementation of the DRA and prompting procedure. Intervention score was (M = 96%), from a baseline score of (M = 34.6%), for tier‐one participants. Intervention score was (M = 96.6%), from a baseline score of (M = 33%), for tier‐two participants. A follow‐up maintenance probe demonstrated correct implementation of the DRA procedure with prompting across both tiers of trained caregivers.