It's the deceiver, not the receiver: No individual differences when detecting deception in a foreign and a native language

Marvin K. H. Law; Simon A. Jackson; Eugene Aidman; Mattis Geiger; Sally Olderbak; Sabina Kleitman

doi:10.1371/journal.pone.0196384

Abstract

Individual differences in lie detection remain poorly understood. Bond and DePaulo’s meta-analysis examined judges (receivers) who were ascertaining lies from truths and senders (deceiver) who told these lies and truths. Bond and DePaulo found that the accuracy of detecting deception depended more on the characteristics of senders rather than the judges’ ability to detect lies/truths. However, for many studies in this meta-analysis, judges could hear and understand senders. This made language comprehension a potential confound. This paper presents the results of two studies. Extending previous work, in Study 1, we removed language comprehension as a potential confound by having English-speakers (N = 126, mean age = 19.86) judge the veracity of German speakers (n = 12) in a lie detection task. The twelve lie-detection stimuli included emotional and non-emotional content, and were presented in three modalities–audio only, video only, and audio and video together. The intelligence (General, Auditory, Emotional) and personality (Dark Triads and Big 6) of participants was also assessed. In Study 2, a native German-speaking sample (N = 117, mean age = 29.10) were also tested on a similar lie detection task to provide a control condition. Despite significantly extending research design and the selection of constructs employed to capture individual differences, both studies replicated Bond and DePaulo’s findings. The results of Study1 indicated that removing language comprehension did not amplify individual differences in judge’s ability to ascertain lies from truths. Study 2 replicated these results confirming a lack of individual differences in judge’s ability to detect lies. The results of both studies suggest that Sender (deceiver) characteristics exerted a stronger influence on the outcomes of lie detection than the judge’s attributes.

Figures

Citation: Law MKH, Jackson SA, Aidman E, Geiger M, Olderbak S, Kleitman S (2018) It's the deceiver, not the receiver: No individual differences when detecting deception in a foreign and a native language. PLoS ONE 13(5): e0196384. https://doi.org/10.1371/journal.pone.0196384

Editor: Niels O. Schiller, Leiden University, NETHERLANDS

Received: May 3, 2017; Accepted: April 12, 2018; Published: May 3, 2018

Copyright: © 2018 Law et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

Data Availability: All relevant data are available from the Figshare repository. For the Australian Sample Data and Codebook, please see: https://protect-au.mimecast.com/s/BMPCCk8vAZtg5GK8sQSdk4?domain=figshare.com. For the German Sample Data and Codebook, please see: https://protect-au.mimecast.com/s/9EmLClxwB5CLX4V0iqLpN6?domain=figshare.com.

Funding: The authors received no specific funding for this work.

Competing interests: The authors have declared that no competing interests exist.

1. Introduction

The ability to detect when others are attempting to deceive us is important, but empirical research suggests that we are poor at lie detection. For example, a common finding in lab-based studies is that, on average, people perform no better than chance when judging whether another is lying or telling the truth [1, 2]. Similar to cognitive abilities (intelligence) or personality traits, it has been suggested that the ability to detect deception varies between individuals [3, 4]. That is, some individuals may be capable of detecting deception better than others despite this ability appearing to be negligible at an aggregate level. Individuals who supposedly possess the greatest skill in detecting deception have been referred to as lie detection ‘wizards’ [5]. Some evidence for such individual differences exists. For example, members of the US secret service, a handful of law enforcement agents, and clinical psychologists have been found to perform above chance when detecting lies [6, 7]. Some authors, however, suggest that statistical or methodological artefacts can explain these results [8].

It has been demonstrated that the “cognitive load approach” as a sole explanatory principle is insufficient for lie detection [9]. This has led to a renewed focus on para-verbal and non-verbal cues, and individual's capacity to detect them. The latest integrated model [9] combines newer versions of Baddeley’s [10, 11] working memory model with constructs of mental control [12, 13]. It enables verbal content cues, nonverbal, para-verbal, and linguistic cues to be investigated within a single framework. This framework emphasises the importance of individual differences in language ability and working memory capacity as well as arousal, stress and emotion [14]. Our study examines inter-individual variability of individuals' capacity to discriminate between lies and truths when language comprehension is removed from the equation. This paper presents results of two studies. In Study 1, a verbal message is presented in a foreign language using a cross-language deception technique (see below). In Study 2, verbal messages are presented to native German speakers to control for language comprehension.

1.1. The lie detection paradigm

This research, albeit extending it in several important ways, employed a typical deception detection paradigm in which participants (“judges”) had to determine whether people in video and audio recordings (“senders”) are lying or telling the truth. To investigate individual differences in the ability to detect deception, researchers typically focus on judge ability: the percentage of senders correctly judged to be telling the truth or lying. The chance level is around 50%. Other variables of interest can also be calculated (summarised in Table 1). Judge credulity is measured by the percentage of “Truth” decisions (regardless of their accuracy), and it reflects the tendency of the “judge” to trust others. Sender detectability and credibility refer to information about the people who are being judged as telling the truth or lying. For each sender, their detectability is measured as the percentage of judges who make a correct decision about them, and their credibility is measured by the percentage of judges who decide that they are telling the truth (regardless of accuracy).

Download:

Table 1. Dependant variables investigated in Bond and DePaulo [15].

https://doi.org/10.1371/journal.pone.0196384.t001

1.2. Individual differences in deception detection

Extensive research on these variables suggests that individual differences in the ability to discriminate truth-telling from lying (judge ability) are negligible. Bond and DePaulo [15, 16] best addressed this question in a meta-analysis of 247 samples, including 12704 judges (participants) and 6060 senders (items). Within their meta-analytic model, Bond and DePaulo were able to separate within-study measurement errors from meaningful variance in judge ability by examining patterns of results across studies. Within their findings, three measurement-corrected methods for evaluating the presence of meaningful individual differences in deception detection were empirically reported. One, reasonable internal consistency for judge ability following a ‘truth’ or ‘lie’ response [15, 16, 17] (see Equation A in S1 Appendix). Two, comparable or larger variability in judge ability than judge credulity, sender detectability and sender credibility (see Equation B in S1 Appendix). Three, positive and significant correlations between judge ability when calculated separately for senders who were telling the truth and senders who were lying. Overall, there was null evidence for each of these three methods. The estimated coefficient Alpha for judge ability was .13, and Spearman-Brown coefficient was .22, which are considerably lower than acceptable levels of internal consistency for research (around .70, [18]). Similar estimates for the remaining variables were higher (for judge credulity, sender detectability and sender credibility, the estimated coefficient alpha was .43, .67 and .75 and Spearman-Brown coefficient was .75, .70 and .81 respectively). Examining the standard deviations, variability in judge ability was considerably lower than variability in the other three variables. Finally, the correlation of judge ability computed separately for truth-telling and lying senders was close to zero and negative (r = -.09). These results emerged across all studies despite differences in deception modality (sender presented with audio and video, audio only, video only), sender motivation (getting away with lying would opt a prize), the interaction between sender and judge (e.g. whether the two interacted during the deception), sender preparedness (how much time for preparation of the lie was given), experience of judge as well as baseline exposure (how the sender would act when not lying) (replicated in [19]’s meta-analysis). The overall conclusion made in Bond and DePaulo [15, 16] was that decisions made on deception detection tasks are primarily driven by factors related to judge credulity, and sender attributes influencing detectability and credibility. Thus, if individual differences in deception detection do exist, we are likely able to observe them only when the judges’ access to the information about the sender is limited or removed. This belief has also been supported by [20], who looked at the effects of sender demeanour (appearance of honesty) on deception detection. This study found that honest-looking truth-tellers and deceptive-looking liars (demeanour-veracity-matched) were detected more accurately than senders whose appearance did not match their actual honesty (demeanour-veracity-mismatched). Our research therefore sought to examine the hypotheses postulated by Bond and DePaulo [15, 16] about individual differences in deception detection when access to a critical source of sender information is removed (Study 1) and when it is present (Study 2).

1.3. Language comprehension

The source of sender information to be removed in this research is language comprehension, specifically the semantics. In lie detection tests, the sender can be seen, heard, or seen and heard to be speaking a language that judges can comprehend [21]. Although audio-only deception paradigms produce detection rates that are similar to those observed under the combined audio/video deception presentation, language is considered to be a crucial source of information for detecting deception. To demonstrate, deception can be inferred when a sender makes a false statement about known information. In applied settings, lie detection methods–including those employed within the judicial system and other organisations–often depend on a content-based analysis of the words being uttered and the meaning that surrounds them [22, 23]. The use of such content-based analyses has been shown to increase the accuracy of lie detection to 71% [24], significantly above chance level. This is particularly prominent within forensic contexts, where the use of content-based strategies has been shown to increase deception detection [25]. In particular, strategic use of evidence in interrogation (e.g. late evidence disclosure) has been shown to increase discriminatory content-based cues between liars and truth-tellers [26]. Although strategy use is important, most individuals in the general population would not have been exposed to such training. Moreover, there might be a potentially misleading aspect of language for untrained samples. That is, when the general population is sampled, aphasic patients, whose language processing abilities are impaired, perform significantly better at detecting deception than controls who perform no better than chance [27]. Thus, in our current research, an alternative approach was taken to examine whether the use of non-verbal cues might enhance lie detection when a population of 1st year psychology students is used. In Study 1, we asked English-speaking judges to complete a typical lie-detection task, with the exception being that the senders were speaking a foreign language (German). By having cross-language deception with judges being unable to comprehend the senders’ content, individual differences in lie detection were expected to emerge.

1.4. Additional individual differences variables

The evidence in favour of individual differences being involved in lie detection were assessed via the three measurement-correct methods described above [15, 16], as well as an additional method. The additional method is based on the assumption that should individual differences in deception detection exist, then judge ability scores should correlate in a meaningful fashion with related constructs. Specifically, measures assessing cognitive abilities, personality, and emotional intelligence. Existing literature, however, indicates either weak or no relationship between lie detection accuracy and traditional measures of intelligence (either general or Gf/Gc) and personality variables [28, 29]. Auditory information like changes in pitch and pause durations [30] have been found to relate to lying [31]. No research, however, has examined the relationship between auditory abilities (as assessed by Auditory Processing or Ga measures) and lie detection. Our study aimed at bridging this gap and examining the role of Ga from the Cattell-Horn-Carroll (CHC) theory of Cognitive Abilities [30] in lie detection.

Theory suggests that people who have higher emotion perception, a branch of Emotional Intelligence (EI) defined as the ability to accurately perceive facial emotions, should be better in lie detection [32, 33]. However, the empirical support is limited. A study by Ekman, O’Sullivan, Friesen and Scherer [34] found that when coders applied the Facial Affect Coding System to code the facial expressions of liars, they were able to identify true and false emotional expressions, subsequently helping them to identify when targets were and were not lying. Hill and Craig [35] got similar results for the identification of true and false facial expressions of pain.

This paper consists of two studies. In Study 1, English-speaking judges completed a lie-detection task where the senders were speaking a foreign language (German; cross-language approach). We expect that removing language comprehension cues may enhance evidence in favour of individual differences in lie detection in general population. To control for language comprehension, in Study 2, German-speaking judgers completed the same lie-detection task (using a longer list of same and different senders).

Furthermore, by incorporating novel and conceptually relevant measures of auditory abilities and ability-based measures of emotional intelligence we aimed to extend existing literature, capturing more consistent evidence for the existence of individual differences in lie detection.

1.5. Aims and hypotheses

Overall, our aim is to examine whether individual differences in lie detection exist when the ability to comprehend the linguistic content of the liar or truth teller has been removed. To assess this, participants will complete a lie-detection test in which senders speak a foreign language. Participants will also complete traditional (Gf and Gc) and novel (Ga) measures of cognitive abilities, ability-based emotional intelligence (EI) and personality. A second control study with a German sample will also be examined, to replicate and extend Study 1, and also to compare as a baseline with the Australian sample. Evidence in favour of individual differences is expected to be demonstrated by:

Higher internal consistency for judge ability than those shown in Bond and DePaulo [16].
Increase in the variability of judge ability, and decrease in the variability of judge credulity, sender detectability and sender credibility when compared to Bond and DePaulo [15]
More positive and significant correlation of judge ability scores calculated separately for senders who were telling the truth and senders who were lying than in Bond and DePaulo [15]
Assuming there is evidence for individual differences in judge ability, we expect significant correlations of judge ability with measures of cognitive abilities (especially Ga), and ability-based measures of emotional intelligence.

2. Study 1

2.1. Methods

2.1.1. Participants.

The current study consisted of 126 first-year psychology undergraduates (92 female, M_age = 19.86, SD = 3.53, age range: 17–52). Data for the lie detection task was removed for three participants who indicated that they understood German fluently. Results for three participants in the Talk Masking task and six participants in the BEFKI-Gf task were removed due to data being missing or responses occurring outside the allowed range (evidence that the task instructions had been misunderstood). Additionally, results for the tests administered in the second portion of the study were not available for one participant. The final sample size was 123 for analyses relating to the lie detection paradigm. Missing values were not imputed and treated as pairwise for analyses reported. Ethics approval was obtained at the University of Sydney with Project Number: 2015/229.

2.1.2. Measures.

Table 2 summarises the measures used in this study. This study was part of a larger research protocol. Only measures relevant to this study were presented in this manuscript (full list is available from the corresponding author). Brief descriptions of the lie detection task are included below.

Download:

Table 2. Summary of measures used in this study.

https://doi.org/10.1371/journal.pone.0196384.t002

2.1.3. Lie detection paradigm.

In this test, participants judge whether each of twelve recordings of people speaking German (senders) are lying or telling the truth. Sender recordings include seven lies and five truths spoken. We employed unequal number of truth and lies videos to prevent the participants in this study from anticipating that a half of the stimuli should be lies. Overall ten senders were used to generate twelve videos–four females and six males. Two of the males appear in two recordings, one telling two truths and one telling one lie and one truth. Recordings are presented in one of three modalities: video only, audio only, and combined video and audio. For each modality, senders talked about an everyday topic (e.g., favourite music), their opinion about a controversial issue (e.g., legalisation of certain types of abortion), or their deviant behaviour (e.g., skipping school) which were then categorized into emotional (e.g., abortion) or a non-emotional topic (e.g., popular culture). Overall, participants judged two senders for each modality (3) x emotionality (2) format.

2.1.4. Procedure.

Participants completed the lie detection test, Ga tasks, Emotion Composite task, BEFKI-Gc, Vocabulary task and demographics in a supervised lab. All other measures were completed online in the participants’ own time.

2.2. Results

2.2.1. Hypothesis 1.

Table 3 presents descriptive statistics for variables from the lie detection test. Mean judge ability to detect lies for all items (total) was 45.4%: significantly below chance (t₁₂₂ = -3.47, p < .001), and similar to prior research [21]. For each modality and emotionality condition, mean judge ability was either significantly below, or no different to, chance level. Replicating Bond and DePaulo [16]’s findings, reliability estimates were unacceptably low for judge characteristics and high for both sender detectability and credulity.

Download:

Table 3. Lie detection descriptives.

https://doi.org/10.1371/journal.pone.0196384.t003

2.2.2. Hypothesis 2.

To compare variance of the judge and sender metrics in this study to prior research, we calculated the predicted standard deviations of these variables as recommended by Bond and DePaulo [15] using their meta-analysis data. That is, for each study included in their meta-analysis, standard deviations of the variables were calculated. These were each regressed on the inverse of the square root of n (number of judges or senders as applicable; see [15] for details) for each study. The resulting regression equations were used to calculate predicted standard deviations with confidence intervals for each variable given our sample size (123) and number of sender items (12). Fig 1 plots the predicted standard deviations, with confidence intervals, given Bond and DePaulo’s meta-analysis, beside the results from this study.

Download:

Fig 1. A comparison of the predicted standard deviation from Bond and DePaulo [15] with the actual standard deviation from the current experiment’s estimates of judge ability, judge credulity, sender detectability and sender credibility.

Error bars represent the predicted standard deviations within a 95% Confidence Interval.

https://doi.org/10.1371/journal.pone.0196384.g001

Bond and DePaulo’s meta-analytic results predicted a standard deviation of 14 for judge ability based on 12 senders. The standard deviation of judge ability in this Study was 14.72 (see Table 3), which is within the predicted 95% confidence interval margins. Following Bond and DePaulo’s analysis, we also compared the standard deviation of judges’ ability in determining truths against lies. The standard deviation of judge credulity also fell within the 95% confidence interval of the value predicted by Bond & DePaulo given 12 senders. The standard deviation of sender detectability fell above the 95% confidence interval of the value predicted by Bond & DePaulo (given 123 judges). However, similar to the judge variables, the standard deviation of sender credibility fell within the range predicted by Bond & DePaulo. That is, the standard deviation for three of the four variables examined fell within the range of observed values from prior research not controlling for language comprehension.

2.2.3. Hypothesis 3.

For each participant, judge ability was calculated separately for senders telling the truth or lying. While Bond & DePaulo found a small negative correlation between these separate scores, our data returned a small, marginally significant positive correlation (r = .18, p = .045).

2.2.4. Hypothesis 4.

Within the first three hypotheses, there was only some support for hypothesis 3. Thus, we did not expect any significant correlations between judge ability and other individual difference variables. The limited support for individual differences in judge ability would indicate that even if there were significant correlations, the estimates of judge ability are essentially noise. However, in order to complete our examination of potential individual differences in lie detection, we estimated correlations between judge ability and other individual difference variables. Performance on the Emotional Composite task correlated with total judge ability (r = -.19, p = .03) and judge ability calculated within the video and audio condition (r = -.25, p = .01). However, after controlling for false discovery rate using the Benjamini-Hochberg method [56] with PROC MULTTEST in SAS [57], neither correlation was statistically significant (p_adj = .97 and p_adj = .79 accordingly).

2.3. Discussion

Overall, Study 1 found little evidence to confirm individual differences in the ability to judge the veracity of verbal messages uttered in a foreign-language. The motivation for this work was to extend the findings of Bond and DePaulo [15, 16] by examining whether cross-language deception (removing language comprehension) would reveal evidence of individual differences in judge ability. Extending previous research, we also addressed all constructs suggested by Sporer’s integrated framework [9], including language ability (Gc), working memory (Gf), stress and emotion (EI), as well as abilities specific to detecting para-verbal cues (Ga). We then used a novel approach proposed by Bond and DePaulo [15, 16] consisting of four methods to test the presence of individual differences in judge ability. Judge ability as well as the other judge and sender characteristics were measured on a lie detection task where senders spoke a foreign language.

Firstly, while prior research has typically found judge ability to be at or just above chance levels, judge ability was below chance level accuracy in this study. This may be due to a demeanour-veracity-mismatch in senders as shown in [20].

Compared to prior research, a higher internal consistency for judge ability to detect lies in foreign-language spoken messages was expected in this study. However, this hypothesis was disconfirmed, with spearman-controlled internal consistency estimates for judge ability being so low as to suggest that these scores did not capture anything systematic about individuals’ ability to detect lies.

We expected greater variance accounted for by judge characteristics relative to sender characteristics. This hypothesis received no support with standard deviations of judge ability and credulity not differing significantly from the pattern of results found in previous research. The same result emerged for sender credibility, and only variance in sender detectability significantly increased compared to prior research. This last significant result may be due to larger variation in demeanour-veracity within the senders, with some senders appearing more mismatched and others appearing more matched [20]. The overall pattern of variance was the same as those found in [15]. Sender credibility accounted for the greatest amount of variance, followed by sender detectability, judge credulity and lastly judge ability. Even in the absence of language comprehension, the current results support Bond and DePaulo [15]’s conclusion that sender characteristics are most central for driving decisions on lie detection tests.

We also hypothesised that judge ability scores calculated separately for truth-telling and lying senders would correlate positively. Some support for this hypothesis was found in the form of a small but significant positive correlation observed between the separate judge ability scores. While this provides some support for individual differences in a lie-detecting ability, it should be noted that this correlation was weak (< .20). Thus, this result may be spurious and requires further replication.

Finally, we expected meaningful correlations to emerge between judge ability and theoretically related individual differences constructs. To address this, we assessed a large battery of tasks assessing a range of theoretically related constructs (both traditional and novel). The novel additions included two ability-based tests of emotional intelligence and three tests of auditory processing. After controlling for false discovery rate, no significant correlations between judges’ ability to detect lies and these theoretically related constructs were found. Our results therefore provide no support for the hypothesis that the removal of language comprehension might amplify individual differences in lie detection.

Next, we tested the same hypotheses from Study 1 using a native German-speaking sample, examining deception detection performance as well as standard deviations in judge ability, judge credulity, sender detectability, and sender credibility. One of the novel aspects of this study was to examine whether the use of nonverbal cues might enhance lie detection using a population whom are naïve to the language cues. Thus, it is important to compare the findings between two different populations, one that is completely naïve to the language cues and one that is not, as the explicit comparison between the two provides strong methodological grounds for our conclusions in relation to our hypotheses. Therefore, a second study has been conducted on a population of native German speakers using similar methodology and stimuli. It should be noted that combining the two populations in one common analysis is inappropriate due to two reasons. Firstly, Bond and DePaulo (2008a, b)’s methodology does not offer a way to control for the use of two different populations statistically. Secondly, having the analyses done on two different populations allows this research to be more transparent and easier to replicate—an important consideration given the current replicability crisis.

3. Study 2

3.1. Methods

3.1.1. Participants.

The sample consisted of 117 adults recruited for one of two studies from the German cities Ulm (N = 71) and Berlin (N = 46). The samples did not differ on age (t₍₁₀₀₎ = .59, p = .56; 15 missed the answer) or gender ratio (t₍₁₀₂₎ = -.26, p = .79; 13 missed the answer) or on their performance on the deception detection task (described below) and hence were combined. The total sample was primarily female (53%) and on average 29.10 years old (SD = 8.93, age range: 17–65). Ulm participants were recruited through advertisements posted online (e.g. on Facebook), on radio and face-to-face recruiting in the central city, while Berlin participants were recruited from eBay Minijobs and through paper flyers posted around the community. Participants individually consented to participate in the study and were financially compensated for participating. The study was conducted in accordance with the Declaration of Helsinki.

3.1.2. Measures.

Participants from Ulm were recruited for a study specifically assessing deception detection abilities. In addition to a demographics questionnaire, they completed 54 deception detection trials which presented videos from the same 10 stimulus persons used for the deception detection trials in Study 1. Participants from Berlin were recruited for a larger study on individual differences and completed a subset of 18 deception detection trials selected from the larger set administered to the Ulm participants. Although, the participants also completed several measures of personality and cognitive abilities (full list is available from the corresponding author), only measures relevant to this study were presented in this manuscript. Average performance across the 18 items presented in Ulm and Berlin did not significantly differ between the groups (t_(109.5) = 1.27, p = .21). Three of the 18 trials were also presented in Study 1.

3.2. Results

3.2.1. Hypothesis 1.

Table 4 presents descriptive statistics for variables from the lie detection test. Mean judge ability to detect lies for all items (total) was 55.7%: significantly above chance (t₁₁₆ = 5.82, p < .01). Replicating Bond and DePaulo [16] and Study 1, reliability estimates were unacceptably low for judge characteristics and high for both sender detectability and credulity.

Download:

Table 4. Lie detection descriptive statistics for Study 2.

https://doi.org/10.1371/journal.pone.0196384.t004

3.2.2. Hypothesis 2.

Using the German sample data, we compared the standard deviations in the four lie detection variables with the predicted standard deviations from Bond and DePaulo [16]’s regression equations (see Fig 2). The predicted values and confidence intervals for each variable were adjusted based on the sample size (117) and the number of sender items (18).

Download:

Fig 2. A comparison of the predicted standard deviation from Bond and DePaulo [15] with the actual standard deviations of judge ability, judge credulity, sender detectability and sender credibility in Study 2.

Error bars represent the predicted standard deviations within a 95% Confidence Interval.

https://doi.org/10.1371/journal.pone.0196384.g002

The predicted standard deviations for judge ability, judge credulity, sender detectability and sender credibility were 11.57, 12.76, 9.84 and 13.69 respectively. Given 18 senders and 117 judges, both standard deviations for the judge characteristics fell within the expected 95% confidence intervals. However, the variance of sender characteristics were significantly higher than those predicted by Bond and DePaulo [16]’s regression equations with actual standard deviations of sender detectability and credibility being 23.36 and 23.53 respectively.

3.2.3. Hypothesis 3.

No significant correlation was found between judge ability towards senders lying and judge ability towards senders telling the truth within the German sample, (r = -.14, p > .05).

3.2.4. Hypothesis 4.

Only 46 of the 117 participants completed measures of other variables aside from the lie detection task, which gives us insufficient power (1 – β = .54), assuming a two-tailed p value (α = .05) to detect even medium size (r = .30) correlations. However, similar to Study 1, the lack of support for individual differences in lie detection from the earlier 3 hypotheses suggest that correlations with judge ability would have only been noise.

3.3. Discussion

In Study 2, we aimed to examine the hypotheses examined in Bond and DePaulo [16] using a control sample of German native-speakers. The results of all three hypotheses examined in Study 2 supported a lack of individual differences in deception detection. For hypothesis 1, reliability estimates were low for judge characteristics and high for sender characteristics. Examining hypothesis 2, the standard deviations of judge characteristics fell within the predicted values of Bond and DePaulo [16]’s regression equations whilst both standard deviations of the sender characteristics were significantly higher than predicted. Finally, no significant correlation was found between judge ability towards senders lying and senders telling the truth. Although we were not able to examine hypothesis 4 in Study 2, considering the consistency of findings with both Bond and DePaulo [16] and Study 1, it is likely that Hypothesis 4 in Study 2 would have also supported a lack of individual differences in deception detection.

4. General discussion

Despite the significant extension of the typical lie-detection paradigm, the two studies presented here provide a consistent replication of previous findings, suggesting that lie detection tasks are ineffective in capturing individual differences in the ability to detect lies. To the best of our knowledge, our Study 1 was the first to extend this conclusion to a lie-detection task in which the truth-telling and lying senders spoke in a foreign language to the receiving judges. These findings add to the evidence suggesting that a focus on judge characteristics is likely to be insufficient for understanding deception detection and predicting its outcomes when para-verbal cues are used. Sender attributes appear to account for greater variance in lie detection rates than do receiving judges’ attributes. The importance of sender characteristics in deception detection may also explain considerable heterogeneity across lie detection studies. Many studies do not properly account for sender characteristics. Thus, the high judging ability scores reported in some studies (e.g., [5]) might have been the result of a task performed by easily detectable senders. Future research could resolve this issue by examining sender characteristics or using standardised lie detection measures when comparing between studies. Furthermore, with increased knowledge about sender characteristics, improved measures might be constructed. By examining relations of item difficulty with sender characteristics and, thus by systematically manipulating these characteristics in item development, measures that actually detect individual differences in deception detection ability might be constructed.

The lack of judge characteristics also demonstrates the importance of strategies that support deception detection within training [58]. Without such strategies, reliable deception detection would be substantially more difficult, if not impossible. We did not find any evidence that lie detection ‘wizards’ exist–at least in our general population samples, and the current study should be replicated with specialist populations such as defence and the police force. If results replicate even in such contexts, training in interrogation techniques would appear to be the only effective method of increasing detection, aligning with existing opinions within the current literature [24, 59].

4.1. Limitations and future directions

Despite the convergence with prior research, several limitations of this study must qualify the interpretation of our results. First, the limited number of senders (12 and 18) in the lie detection tasks may have reduced the observed variability. With a greater number of senders, greater variance and less error might be obtained. Still, given the consistency of the results over two studies, it is doubtful that increasing the number of senders will have a drastic effect on the results. Second, the senders in our task were either lying or telling the truth, and only one individual sender did both. Employing the same individuals to present both lies and truthful messages would have been more sensitive to individual differences in sender characteristics—and this method seems worth employing in future studies.

Third, the use of a student and general population sample may have limited the generalisability of our results. Many studies included in Bond and DePaulo’s meta-analysis were also conducted using such samples making our comparisons to previous work valid. However, it also means that generalising our conclusions to specific populations may not be substantiated. Perhaps lie detection is a skill that requires considerable time and experience to develop, making it more likely to emerge in older populations or more targeted populations (e.g., law enforcement officers). Extending future lie-detection studies to such specialised samples would not only improve their generalisability, but also enhance their applicability in practical settings.

4.2 Conclusion

Overall, using the novel selection of individual differences measures, the novel research design and the novel statistical approach, our current studies replicated Bond and DePaulo’s findings [15, 16]. Study 1 replicated the findings under the condition of a foreign language-spoken messaging, where the ability of the judging receiver to comprehend the contents of the message is reduced. Removing language comprehension does not appear to reveal stronger individual differences in judge ability in detecting lies. Importantly, identical results were obtained using native German speakers and a longer list of senders in Study 2. Thus, instead of stressing the role of the judging receiver, our findings strongly highlight a role of sender attributes in predicting the outcomes of lie-detecting performance.

Supporting information

S1 Appendix.

This document contains Equations A and B. Equation A. This is the equation used within Bond and DePaulo (2008b) to calculate the coefficient alphas for judge and sender characteristics. Equation B. This is Bond and DePaulo (2008a)’s regression equation for predicting the standard deviation of judge and sender characteristics within lie detection tasks.

https://doi.org/10.1371/journal.pone.0196384.s001

(DOCX)

Acknowledgments

We would like to thank Prof. Oliver Wilhelm for his advice in the design of the deception detection materials.

References

1. Levine TR, Park HS, McCornack SA. Accuracy in detecting truths and lies: Documenting the “veracity effect”. Communication Monographs. 1999;66(2):125–44.
- View Article
- Google Scholar
2. Wright GR, Berry CJ, Bird G. "You can't kid a kidder": association between production and detection of deception in an interactive deception task. Frontiers in Human Neuroscience. 2012;6:87. pmid:22529790
- View Article
- PubMed/NCBI
- Google Scholar
3. O'Sullivan M. Unicorns or Tiger Woods: are lie detection experts myths or rarities? A response to on lie detection "wizards" by Bond and Uysal. Law and Human Behavior. 2007;31(1):117–23. pmid:17221308
- View Article
- PubMed/NCBI
- Google Scholar
4. Jupe L, Akehurst L, Vernham Z, Allen J. Teenage Offenders' Ability to Detect Deception in Their Peers. Applied Cognitive Psychology. 2016;30(3):401–408.
- View Article
- Google Scholar
5. O'Sullivan M, Ekman P. The wizards of deception detection. In: Granhag PA, Stromwall LA, editors. Deception detection in forensic contexts. Cambridge, UK: Cambridge Press; 2004. p. 269–86.
6. Ekman P, O'Sullivan M. Who can catch a liar? American Psychologist. 1991;46(9):913–20. pmid:1958011
- View Article
- PubMed/NCBI
- Google Scholar
7. Ekman P, O'Sullivan M, Frank MG. A Few Can Catch a Liar. Psychological Science. 1999;10(3):263–6.
- View Article
- Google Scholar
8. Bond CF Jr., Uysal A. On lie detection "wizards". Law and Human Behavior. 2007;31(1):109–15. pmid:17221309
- View Article
- PubMed/NCBI
- Google Scholar
9. Sporer SL. Deception and Cognitive Load: Expanding Our Horizon with a Working Memory Model. Frontiers in Psychology. 2016;7:420. pmid:27092090
- View Article
- PubMed/NCBI
- Google Scholar
10. Baddeley A. Working memory, thought, and action. Oxford, United Kingdom: Oxford University Press; 2007.
11. Baddeley A. Working memory: theories, models, and controversies. Annual Reviews of Psychology. 2012;63:1–29.
- View Article
- Google Scholar
12. Wegner DM. Ironic processes of mental control. Psychological Review. 1994;101(1):34–52. pmid:8121959
- View Article
- PubMed/NCBI
- Google Scholar
13. Wegner DM. How to think, say, or do precisely the worst thing for any occasion. Science. 2009;325(5936):48–50. pmid:19574380
- View Article
- PubMed/NCBI
- Google Scholar
14. Zuckerman M, DePaulo BM, Rosenthal R. Verbal and nonverbal communication of deception. Advances in Experimental Social Psychology. 1981; 14: 1–59.
- View Article
- Google Scholar
15. Bond CF, Depaulo BM. Individual differences in judging deception: accuracy and bias. Psychological Bulletin. 2008;134(4):477–92. pmid:18605814
- View Article
- PubMed/NCBI
- Google Scholar
16. Bond CF, DePaulo BM. Individual differences in judging deception: Reply to O'Sullivan (2008) and Pigott and Wu (2008). Psychological Bulletin. 2008;134(4):501–3.
- View Article
- Google Scholar
17. Pigott TD, Wu MJ. Methodological issues in meta-analyzing standard deviations: Comment on Bond and DePaulo (2008). Psychological Bulletin. 2008;134(4):498–500; discussion 1–3. pmid:18605816
- View Article
- PubMed/NCBI
- Google Scholar
18. Nunnally JC, Bernstein IH. Psychometric theory. New York: McGraw-Hill; 1994.
19. Hartwig M, Bond C. Lie Detection from Multiple Cues: A Meta-analysis. Applied Cognitive Psychology. 2014;28(5):661–676.
- View Article
- Google Scholar
20. Levine T, Serota K, Shulman H, Clare D, Park H, Shaw A et al. Sender Demeanor: Individual Differences in Sender Believability Have a Powerful Impact on Deception Detection Judgments. Human Communication Research. 2011;37(3):377–403.
- View Article
- Google Scholar
21. Bond CF Jr., DePaulo BM. Accuracy of deception judgments. Personality and Social Psychology Review. 2006;10(3):214–34. pmid:16859438
- View Article
- PubMed/NCBI
- Google Scholar
22. Ruby CL, Brigham JC. The usefulness of the criteria-based content analysis technique in distinguishing between truthful and fabricated allegations: A critical review. Psychology, Public Policy, and Law. 1997;3(4):705–37.
- View Article
- Google Scholar
23. Vrij A, Mann S, Jundi S, Hillman J, Hope L. Detection of Concealment in an Information-Gathering Interview. Applied Cognitive Psychology. 2014;28(6):860–866.
- View Article
- Google Scholar
24. Vrij A. Nonverbal Dominance Versus Verbal Accuracy in Lie Detection: A Plea to Change Police Practice. Criminal Justice and Behavior. 2008;35(10):1323–36.
- View Article
- Google Scholar
25. Hartwig M, Granhag P, Strömwall L, Kronkvist O. Strategic use of evidence during police interviews: When training to detect deception works. Law and Human Behavior. 2006;30(5):603–619. pmid:16977348
- View Article
- PubMed/NCBI
- Google Scholar
26. Granhag P, Rangmar J, Strömwall L. Small Cells of Suspects: Eliciting Cues to Deception by Strategic Interviewing. Journal of Investigative Psychology and Offender Profiling. 2014;12(2):127–141.
- View Article
- Google Scholar
27. Etcoff NL, Ekman P, Magee JJ, Frank MG. Lie detection and language comprehension. Nature. 2000;405(6783):139.
- View Article
- Google Scholar
28. Aamodt MG, Custer H. Who can best catch a liar? A meta-analysis of individual differences in detecting deception. Forensic Examiner. 2006;15:6–11.
- View Article
- Google Scholar
29. Porter S, Campbell MA, Stapleton J, Birt AR. The influence of judge, target, and stimulus characteristics on the accuracy of detecting deceit. Canadian Journal of Behavioural Science/Revue canadienne des Sciences du comportement. 2002;34(3):172–85.
- View Article
- Google Scholar
30. McGrew KS. CHC theory and the human cognitive abilities project: Standing on the shoulders of the giants of psychometric intelligence research. Intelligence. 2009;37(1):1–10.
- View Article
- Google Scholar
31. DePaulo BM, Lindsay JJ, Malone BE, Muhlenbruck L, Charlton K, Cooper H. Cues to deception. Psychological Bulletin. 2003;129(1):74–118. pmid:12555795
- View Article
- PubMed/NCBI
- Google Scholar
32. Ekman P. Expression and the Nature of Emotion. In: Scherer K, Ekman P, editors. Approaches to Emotion. Hillsdale, NJ: Lawrence Erlbaum. 1984. p. 319–43.
33. Mayer JD, Caruso DR, Salovey P. Emotional intelligence meets traditional standards for an intelligence. Intelligence. 1999;27(4):267–98.
- View Article
- Google Scholar
34. Ekman P, O'Sullivan M, Friesen WV, Scherer KR. Invited article: Face, voice, and body in detecting deceit. Journal of Nonverbal Behavior. 1991;15:125–135.
- View Article
- Google Scholar
35. Hill ML, Craig KD. Detecting deception in pain expressions: The structure of genuine and deceptive facial displays. Pain. 2002;98:135–144. pmid:12098625
- View Article
- PubMed/NCBI
- Google Scholar
36. Wilhelm O, Hildebrandt A, Manske K, Schacht A, Sommer W. Test battery for measuring the perception and recognition of facial expressions of emotion. Frontiers in Psychology. 2014;5:404. pmid:24860528
- View Article
- PubMed/NCBI
- Google Scholar
37. Wagner HL. On measuring performance in category judgment studies of nonverbal behavior. Journal of Nonverbal Behavior. 1993;17(1):3–28.
- View Article
- Google Scholar
38. Stankov L, Horn JL. Human abilities revealed through auditory tests. Journal of Educational Psychology. 1980;72(1):21–44. pmid:7372913
- View Article
- PubMed/NCBI
- Google Scholar
39. Baayen RH, Piepenbrock R, van H R. The {CELEX} lexical data base on {CD-ROM}. 1993.
40. Davis CJ. N-Watch: A program for deriving neighborhood size and other psycholinguistic statistics. Behavior Research Methods. 2005;37(1):65–70. pmid:16097345
- View Article
- PubMed/NCBI
- Google Scholar
41. Levenson MR, Kiehl KA, Fitzpatrick CM. Assessing psychopathic attributes in a noninstitutionalized population. Journal of Personality and Social Psychology. 1995;68(1):151–8. pmid:7861311
- View Article
- PubMed/NCBI
- Google Scholar
42. Somma A, Fossati A, Patrick C, Maffei C, Borroni S. The three-factor structure of the Levenson self-report psychopathy scale: fool's gold or true gold? A study in a sample of Italian adult non-clinical participants. Personality and Mental Health. 2014;8(4):337–47. pmid:25132649
- View Article
- PubMed/NCBI
- Google Scholar
43. Ames DR, Rose P, Anderson CP. The NPI-16 as a short measure of narcissism. Journal of Research in Personality. 2006;40(4):440–50.
- View Article
- Google Scholar
44. Christie R, Geis FL. Studies in Machiavellianism. New York: Academic Press. 1970
45. Rauthmann JF. Investigating the MACH-IV with item response theory and proposing the trimmed MACH*. Journal of Personality Assessment. 2013;95(4):388–97. pmid:23186231
- View Article
- PubMed/NCBI
- Google Scholar
46. Saucier G. What Are the Most Important Dimensions of Personality?Evidence from Studies of Descriptors in Diverse Languages. Social and Personality Psychology Compass. 2009;3(4):620–37.
- View Article
- Google Scholar
47. Saucier G. Measures of personality factors found recurrently in human lexicons. In: Boyle GJ, Matthews G, Sakloske D, editors. Handbook of personality theory and testing. 2. London: Sage; 2008.
48. Olderbak S, Sassenrath C, Keller J, Wilhelm O. An emotion-differentiated perspective on empathy with the emotion specific empathy questionnaire. Frontiers in Psychology. 2014;5:653. pmid:25071632
- View Article
- PubMed/NCBI
- Google Scholar
49. Wilhelm O, Schroeders U, Schipolowski S. BEFKI–Berliner Test zur Erfassung fluider und kristalliner Intelligenz (Mittelstufenform) [BEFKI–Berlin Test for the measurement of fluid and crystallized intelligence (middle school form)]. Göttingen:Hogrefe; 2014 (Unpublished Manuscript).
50. Stankov L. The Gf/Gcquickie test battery. Unpublished test battery from the School of Psychology, University of Sydney, Australia. 1997.
51. Jiang Y, Kleitman S. Metacognition and motivation: Links between confidence, self-protection and self-enhancement. Learning and Individual Differences. 2015;37:222–30.
- View Article
- Google Scholar
52. Kleitman S. Metacognition in the rationality debate: Self-confidence and its calibration. VDM Publishing; 2008.
53. Stankov L, Crawford JD. Self-confidence and performance on tests of cognitive abilities. Intelligence. 1997;25(2):93–109.
- View Article
- Google Scholar
54. Kleitman S, Stankov L. Self-confidence and metacognitive processes. Learning and Individual Differences. 2007;17(2):161–73.
- View Article
- Google Scholar
55. Want J, Kleitman S. Imposter phenomenon and self-handicapping: Links with parenting styles and self-confidence. Personality and Individual Differences. 2006;40(5):961–71.
- View Article
- Google Scholar
56. Benjamini Y., & Hochberg Y. (1995). Controlling the false discovery rate: A practical and powerful approach to multiple testing. Journal of the Royal Statistical Society. Series B (Methodological), 57, 289–300.
- View Article
- Google Scholar
57. Westfall P., Tobias R., Rom D., Wolfinger R., & Hochberg Y. (1999). Multiple comparisons and multiple tests using SAS. Cary, NC: SAS Institute.
58. Hartwig M, Granhag P, Strömwall L, Vrij A. Detecting deception via strategic disclosure of evidence. Law and Human Behavior. 2005;29(4):469–484. pmid:16133950
- View Article
- PubMed/NCBI
- Google Scholar
59. Vrij A, Mann S, Leal S, Vernham Z, Vaughan M. Train the Trainers: A First Step towards a Science-Based Cognitive Lie Detection Training Workshop Delivered by a Practitioner. Journal of Investigative Psychology and Offender Profiling. 2015;13(2):110–130.
- View Article
- Google Scholar

[ref1] 1. Levine TR, Park HS, McCornack SA. Accuracy in detecting truths and lies: Documenting the “veracity effect”. Communication Monographs. 1999;66(2):125–44.
View Article
Google Scholar

[2] View Article

[3] Google Scholar

[ref2] 2. Wright GR, Berry CJ, Bird G. "You can't kid a kidder": association between production and detection of deception in an interactive deception task. Frontiers in Human Neuroscience. 2012;6:87. pmid:22529790
View Article
PubMed/NCBI
Google Scholar

[5] View Article

[6] PubMed/NCBI

[7] Google Scholar

[ref3] 3. O'Sullivan M. Unicorns or Tiger Woods: are lie detection experts myths or rarities? A response to on lie detection "wizards" by Bond and Uysal. Law and Human Behavior. 2007;31(1):117–23. pmid:17221308
View Article
PubMed/NCBI
Google Scholar

[9] View Article

[10] PubMed/NCBI

[11] Google Scholar

[ref4] 4. Jupe L, Akehurst L, Vernham Z, Allen J. Teenage Offenders' Ability to Detect Deception in Their Peers. Applied Cognitive Psychology. 2016;30(3):401–408.
View Article
Google Scholar

[13] View Article

[14] Google Scholar

[ref5] 5. O'Sullivan M, Ekman P. The wizards of deception detection. In: Granhag PA, Stromwall LA, editors. Deception detection in forensic contexts. Cambridge, UK: Cambridge Press; 2004. p. 269–86.

[ref6] 6. Ekman P, O'Sullivan M. Who can catch a liar? American Psychologist. 1991;46(9):913–20. pmid:1958011
View Article
PubMed/NCBI
Google Scholar

[17] View Article

[18] PubMed/NCBI

[19] Google Scholar

[ref7] 7. Ekman P, O'Sullivan M, Frank MG. A Few Can Catch a Liar. Psychological Science. 1999;10(3):263–6.
View Article
Google Scholar

[21] View Article

[22] Google Scholar

[ref8] 8. Bond CF Jr., Uysal A. On lie detection "wizards". Law and Human Behavior. 2007;31(1):109–15. pmid:17221309
View Article
PubMed/NCBI
Google Scholar

[24] View Article

[25] PubMed/NCBI

[26] Google Scholar

[ref9] 9. Sporer SL. Deception and Cognitive Load: Expanding Our Horizon with a Working Memory Model. Frontiers in Psychology. 2016;7:420. pmid:27092090
View Article
PubMed/NCBI
Google Scholar

[28] View Article

[29] PubMed/NCBI

[30] Google Scholar

[ref10] 10. Baddeley A. Working memory, thought, and action. Oxford, United Kingdom: Oxford University Press; 2007.

[ref11] 11. Baddeley A. Working memory: theories, models, and controversies. Annual Reviews of Psychology. 2012;63:1–29.
View Article
Google Scholar

[33] View Article

[34] Google Scholar

[ref12] 12. Wegner DM. Ironic processes of mental control. Psychological Review. 1994;101(1):34–52. pmid:8121959
View Article
PubMed/NCBI
Google Scholar

[36] View Article

[37] PubMed/NCBI

[38] Google Scholar

[ref13] 13. Wegner DM. How to think, say, or do precisely the worst thing for any occasion. Science. 2009;325(5936):48–50. pmid:19574380
View Article
PubMed/NCBI
Google Scholar

[40] View Article

[41] PubMed/NCBI

[42] Google Scholar

[ref14] 14. Zuckerman M, DePaulo BM, Rosenthal R. Verbal and nonverbal communication of deception. Advances in Experimental Social Psychology. 1981; 14: 1–59.
View Article
Google Scholar

[44] View Article

[45] Google Scholar

[ref15] 15. Bond CF, Depaulo BM. Individual differences in judging deception: accuracy and bias. Psychological Bulletin. 2008;134(4):477–92. pmid:18605814
View Article
PubMed/NCBI
Google Scholar

[47] View Article

[48] PubMed/NCBI

[49] Google Scholar

[ref16] 16. Bond CF, DePaulo BM. Individual differences in judging deception: Reply to O'Sullivan (2008) and Pigott and Wu (2008). Psychological Bulletin. 2008;134(4):501–3.
View Article
Google Scholar

[51] View Article

[52] Google Scholar

[ref17] 17. Pigott TD, Wu MJ. Methodological issues in meta-analyzing standard deviations: Comment on Bond and DePaulo (2008). Psychological Bulletin. 2008;134(4):498–500; discussion 1–3. pmid:18605816
View Article
PubMed/NCBI
Google Scholar

[54] View Article

[55] PubMed/NCBI

[56] Google Scholar

[ref18] 18. Nunnally JC, Bernstein IH. Psychometric theory. New York: McGraw-Hill; 1994.

[ref19] 19. Hartwig M, Bond C. Lie Detection from Multiple Cues: A Meta-analysis. Applied Cognitive Psychology. 2014;28(5):661–676.
View Article
Google Scholar

[59] View Article

[60] Google Scholar

[ref20] 20. Levine T, Serota K, Shulman H, Clare D, Park H, Shaw A et al. Sender Demeanor: Individual Differences in Sender Believability Have a Powerful Impact on Deception Detection Judgments. Human Communication Research. 2011;37(3):377–403.
View Article
Google Scholar

[62] View Article

[63] Google Scholar

[ref21] 21. Bond CF Jr., DePaulo BM. Accuracy of deception judgments. Personality and Social Psychology Review. 2006;10(3):214–34. pmid:16859438
View Article
PubMed/NCBI
Google Scholar

[65] View Article

[66] PubMed/NCBI

[67] Google Scholar

[ref22] 22. Ruby CL, Brigham JC. The usefulness of the criteria-based content analysis technique in distinguishing between truthful and fabricated allegations: A critical review. Psychology, Public Policy, and Law. 1997;3(4):705–37.
View Article
Google Scholar

[69] View Article

[70] Google Scholar

[ref23] 23. Vrij A, Mann S, Jundi S, Hillman J, Hope L. Detection of Concealment in an Information-Gathering Interview. Applied Cognitive Psychology. 2014;28(6):860–866.
View Article
Google Scholar

[72] View Article

[73] Google Scholar

[ref24] 24. Vrij A. Nonverbal Dominance Versus Verbal Accuracy in Lie Detection: A Plea to Change Police Practice. Criminal Justice and Behavior. 2008;35(10):1323–36.
View Article
Google Scholar

[75] View Article

[76] Google Scholar

[ref25] 25. Hartwig M, Granhag P, Strömwall L, Kronkvist O. Strategic use of evidence during police interviews: When training to detect deception works. Law and Human Behavior. 2006;30(5):603–619. pmid:16977348
View Article
PubMed/NCBI
Google Scholar

[78] View Article

[79] PubMed/NCBI

[80] Google Scholar

[ref26] 26. Granhag P, Rangmar J, Strömwall L. Small Cells of Suspects: Eliciting Cues to Deception by Strategic Interviewing. Journal of Investigative Psychology and Offender Profiling. 2014;12(2):127–141.
View Article
Google Scholar

[82] View Article

[83] Google Scholar

[ref27] 27. Etcoff NL, Ekman P, Magee JJ, Frank MG. Lie detection and language comprehension. Nature. 2000;405(6783):139.
View Article
Google Scholar

[85] View Article

[86] Google Scholar

[ref28] 28. Aamodt MG, Custer H. Who can best catch a liar? A meta-analysis of individual differences in detecting deception. Forensic Examiner. 2006;15:6–11.
View Article
Google Scholar

[88] View Article

[89] Google Scholar

[ref29] 29. Porter S, Campbell MA, Stapleton J, Birt AR. The influence of judge, target, and stimulus characteristics on the accuracy of detecting deceit. Canadian Journal of Behavioural Science/Revue canadienne des Sciences du comportement. 2002;34(3):172–85.
View Article
Google Scholar

[91] View Article

[92] Google Scholar

[ref30] 30. McGrew KS. CHC theory and the human cognitive abilities project: Standing on the shoulders of the giants of psychometric intelligence research. Intelligence. 2009;37(1):1–10.
View Article
Google Scholar

[94] View Article

[95] Google Scholar

[ref31] 31. DePaulo BM, Lindsay JJ, Malone BE, Muhlenbruck L, Charlton K, Cooper H. Cues to deception. Psychological Bulletin. 2003;129(1):74–118. pmid:12555795
View Article
PubMed/NCBI
Google Scholar

[97] View Article

[98] PubMed/NCBI

[99] Google Scholar

[ref32] 32. Ekman P. Expression and the Nature of Emotion. In: Scherer K, Ekman P, editors. Approaches to Emotion. Hillsdale, NJ: Lawrence Erlbaum. 1984. p. 319–43.

[ref33] 33. Mayer JD, Caruso DR, Salovey P. Emotional intelligence meets traditional standards for an intelligence. Intelligence. 1999;27(4):267–98.
View Article
Google Scholar

[102] View Article

[103] Google Scholar

[ref34] 34. Ekman P, O'Sullivan M, Friesen WV, Scherer KR. Invited article: Face, voice, and body in detecting deceit. Journal of Nonverbal Behavior. 1991;15:125–135.
View Article
Google Scholar

[105] View Article

[106] Google Scholar

[ref35] 35. Hill ML, Craig KD. Detecting deception in pain expressions: The structure of genuine and deceptive facial displays. Pain. 2002;98:135–144. pmid:12098625
View Article
PubMed/NCBI
Google Scholar

[108] View Article

[109] PubMed/NCBI

[110] Google Scholar

[ref36] 36. Wilhelm O, Hildebrandt A, Manske K, Schacht A, Sommer W. Test battery for measuring the perception and recognition of facial expressions of emotion. Frontiers in Psychology. 2014;5:404. pmid:24860528
View Article
PubMed/NCBI
Google Scholar

[112] View Article

[113] PubMed/NCBI

[114] Google Scholar

[ref37] 37. Wagner HL. On measuring performance in category judgment studies of nonverbal behavior. Journal of Nonverbal Behavior. 1993;17(1):3–28.
View Article
Google Scholar

[116] View Article

[117] Google Scholar

[ref38] 38. Stankov L, Horn JL. Human abilities revealed through auditory tests. Journal of Educational Psychology. 1980;72(1):21–44. pmid:7372913
View Article
PubMed/NCBI
Google Scholar

[119] View Article

[120] PubMed/NCBI

[121] Google Scholar

[ref39] 39. Baayen RH, Piepenbrock R, van H R. The {CELEX} lexical data base on {CD-ROM}. 1993.

[ref40] 40. Davis CJ. N-Watch: A program for deriving neighborhood size and other psycholinguistic statistics. Behavior Research Methods. 2005;37(1):65–70. pmid:16097345
View Article
PubMed/NCBI
Google Scholar

[124] View Article

[125] PubMed/NCBI

[126] Google Scholar

[ref41] 41. Levenson MR, Kiehl KA, Fitzpatrick CM. Assessing psychopathic attributes in a noninstitutionalized population. Journal of Personality and Social Psychology. 1995;68(1):151–8. pmid:7861311
View Article
PubMed/NCBI
Google Scholar

[128] View Article

[129] PubMed/NCBI

[130] Google Scholar

[ref42] 42. Somma A, Fossati A, Patrick C, Maffei C, Borroni S. The three-factor structure of the Levenson self-report psychopathy scale: fool's gold or true gold? A study in a sample of Italian adult non-clinical participants. Personality and Mental Health. 2014;8(4):337–47. pmid:25132649
View Article
PubMed/NCBI
Google Scholar

[132] View Article

[133] PubMed/NCBI

[134] Google Scholar

[ref43] 43. Ames DR, Rose P, Anderson CP. The NPI-16 as a short measure of narcissism. Journal of Research in Personality. 2006;40(4):440–50.
View Article
Google Scholar

[136] View Article

[137] Google Scholar

[ref44] 44. Christie R, Geis FL. Studies in Machiavellianism. New York: Academic Press. 1970

[ref45] 45. Rauthmann JF. Investigating the MACH-IV with item response theory and proposing the trimmed MACH*. Journal of Personality Assessment. 2013;95(4):388–97. pmid:23186231
View Article
PubMed/NCBI
Google Scholar

[140] View Article

[141] PubMed/NCBI

[142] Google Scholar

[ref46] 46. Saucier G. What Are the Most Important Dimensions of Personality?Evidence from Studies of Descriptors in Diverse Languages. Social and Personality Psychology Compass. 2009;3(4):620–37.
View Article
Google Scholar

[144] View Article

[145] Google Scholar

[ref47] 47. Saucier G. Measures of personality factors found recurrently in human lexicons. In: Boyle GJ, Matthews G, Sakloske D, editors. Handbook of personality theory and testing. 2. London: Sage; 2008.

[ref48] 48. Olderbak S, Sassenrath C, Keller J, Wilhelm O. An emotion-differentiated perspective on empathy with the emotion specific empathy questionnaire. Frontiers in Psychology. 2014;5:653. pmid:25071632
View Article
PubMed/NCBI
Google Scholar

[148] View Article

[149] PubMed/NCBI

[150] Google Scholar

[ref49] 49. Wilhelm O, Schroeders U, Schipolowski S. BEFKI–Berliner Test zur Erfassung fluider und kristalliner Intelligenz (Mittelstufenform) [BEFKI–Berlin Test for the measurement of fluid and crystallized intelligence (middle school form)]. Göttingen:Hogrefe; 2014 (Unpublished Manuscript).

[ref50] 50. Stankov L. The Gf/Gcquickie test battery. Unpublished test battery from the School of Psychology, University of Sydney, Australia. 1997.

[ref51] 51. Jiang Y, Kleitman S. Metacognition and motivation: Links between confidence, self-protection and self-enhancement. Learning and Individual Differences. 2015;37:222–30.
View Article
Google Scholar

[154] View Article

[155] Google Scholar

[ref52] 52. Kleitman S. Metacognition in the rationality debate: Self-confidence and its calibration. VDM Publishing; 2008.

[ref53] 53. Stankov L, Crawford JD. Self-confidence and performance on tests of cognitive abilities. Intelligence. 1997;25(2):93–109.
View Article
Google Scholar

[158] View Article

[159] Google Scholar

[ref54] 54. Kleitman S, Stankov L. Self-confidence and metacognitive processes. Learning and Individual Differences. 2007;17(2):161–73.
View Article
Google Scholar

[161] View Article

[162] Google Scholar

[ref55] 55. Want J, Kleitman S. Imposter phenomenon and self-handicapping: Links with parenting styles and self-confidence. Personality and Individual Differences. 2006;40(5):961–71.
View Article
Google Scholar

[164] View Article

[165] Google Scholar

[ref56] 56. Benjamini Y., & Hochberg Y. (1995). Controlling the false discovery rate: A practical and powerful approach to multiple testing. Journal of the Royal Statistical Society. Series B (Methodological), 57, 289–300.
View Article
Google Scholar

[167] View Article

[168] Google Scholar

[ref57] 57. Westfall P., Tobias R., Rom D., Wolfinger R., & Hochberg Y. (1999). Multiple comparisons and multiple tests using SAS. Cary, NC: SAS Institute.

[ref58] 58. Hartwig M, Granhag P, Strömwall L, Vrij A. Detecting deception via strategic disclosure of evidence. Law and Human Behavior. 2005;29(4):469–484. pmid:16133950
View Article
PubMed/NCBI
Google Scholar

[171] View Article

[172] PubMed/NCBI

[173] Google Scholar

[ref59] 59. Vrij A, Mann S, Leal S, Vernham Z, Vaughan M. Train the Trainers: A First Step towards a Science-Based Cognitive Lie Detection Training Workshop Delivered by a Practitioner. Journal of Investigative Psychology and Offender Profiling. 2015;13(2):110–130.
View Article
Google Scholar

[175] View Article

[176] Google Scholar

Abstract

Figures

1. Introduction

1.1. The lie detection paradigm

1.2. Individual differences in deception detection

1.3. Language comprehension

1.4. Additional individual differences variables

1.5. Aims and hypotheses

2. Study 1

2.1. Methods

2.1.1. Participants.

2.1.2. Measures.

2.1.3. Lie detection paradigm.

2.1.4. Procedure.

2.2. Results

2.2.1. Hypothesis 1.

2.2.2. Hypothesis 2.

2.2.3. Hypothesis 3.

2.2.4. Hypothesis 4.

2.3. Discussion

3. Study 2

3.1. Methods

3.1.1. Participants.

3.1.2. Measures.

3.2. Results

3.2.1. Hypothesis 1.

3.2.2. Hypothesis 2.

3.2.3. Hypothesis 3.

3.2.4. Hypothesis 4.

3.3. Discussion

4. General discussion

4.1. Limitations and future directions

4.2 Conclusion

Supporting information

S1 Appendix.

Acknowledgments

References

Cookie Preference Center

Customize Your Cookie Preference