- Email: [email protected]
Reality Monitoring in the Forensic Context: Digging Deeper into the Speech of Liars
+Model
ARTICLE IN PRESS Journal of Applied Research in Memory and Cognition xxx (2017) xxx–xxx Contents lists available at ScienceDirect
Journal of Applied Research in Memory and Cognition journal homepage: www.elsevier.com/locate/jarmac
Reality Monitoring in the Forensic Context: Digging Deeper into the Speech of Liars Galit Nahari∗ Bar-Ilan University, Israel Reality monitoring (RM) indicates that truthful accounts contain more perceptual and contextual details than false accounts. Considering the tendency of liars to manipulate their accounts by adding false details, I compared truths and lies in terms of the amount and veracity of details provided by suspects across three conditions: a single statement provided immediately; a single statement following a two-week delay; or two statements, the first provided immediately and the second following a two-week delay. Distinguishing truths from lies was possible across conditions, but with varying intensity. Truth-tellers provided only truthful details, whereas liars provided both truthful and false details. While the opportunity to provide truthful details decreased over time for both truthtellers and liars, only the latter compensated for this decrease by adding false details. The current study provides a new empirical approach and significant insight into the application of the RM framework in the forensic context.
General Audience Summary The current study examined the verbal behavior of suspects, who tell the truth or lie when they are interviewed about their involvement in a crime, across three situations: when they provide a single statement immediately after the crime occurred; a single statement following a two-week delay; or two statements, the first provided immediately after the crime occurred and the second following a two-week delay. According to the realitymonitoring approach for lie detection, truth-tellers provide more perceptual (e.g., what they saw, heard, and smelled during the described event) and contextual details (e.g., times and locations) than liars. While truthtellers usually provide truthful details in the interviews, liars, who are motivated to be convincing, manipulate their accounts by adding false details. Results showed that distinguishing truths from lies was possible in all situations, but with varying intensity. Truth-tellers provided only truthful details, whereas liars provided both truthful and false details. While the opportunity to provide truthful details decreased over time for both truth-tellers and liars, only the latter compensated for this decrease by adding false details. The current study provides new insights into the verbal behavior of liars and truth-tellers. Keywords: Reality monitoring, Detection deception, Richness in detail, Memory, Self-manipulated memory
Author Note Correspondence concerning this article should be addressed to Galit Nahari, Department of Criminology, Bar-Ilan University, Ramat Gan 52900, Israel. Contact: [email protected] ∗
Please cite this article in press as: Nahari, G. Reality Monitoring in the Forensic Context: Digging Deeper into the Speech of Liars. Journal of Applied Research in Memory and Cognition (2017), https://doi.org/10.1016/j.jarmac.2018.04.003
+Model
ARTICLE IN PRESS REALITY MONITORING IN THE FORENSIC CONTEXT
Reality monitoring (RM) theory (Johnson & Raye, 1981) describes the process used by an individual to monitor his or her own memories by their content qualities. According to RM, memories for experienced events (externally derived memories) are characterized by perceptual and contextual attributes, while memories for imagined events (internally derived memories) are characterized by the cognitive operational attributes that help to generate them. A few decades ago, the RM framework was adapted to the field of deception detection (Sporer, 1997, 2004), where the process of judging the source of other people’s memories has been labeled interpersonal RM (iRM; Johnson, 2006; Johnson, Bush, & Mitchell, 1998). In this application, true accounts are expected to be richer in perceptual and contextual details than are fabricated accounts (e.g., Vrij, 2008). This extension of RM has been supported by many studies (see DePaulo et al., 2003; Masip, Sporer, Garrido, & Herrero, 2005; Nahari, Vrij, & Fisher, 2012; Vrij, 2005, 2008). Yet, I argue that it occurred without taking into consideration the factors that distinguish fabrication (in iRM) from false memory (in RM). One factor involved in iRM for deception detection and absent in RM, where intended deception and manipulation are not likely, is the tendency of liars to manipulate their accounts to make them seem truthful (Granhag & Hartwig, 2008; Nahari, Vrij, & Fisher, 2014a, 2014b), for example by intentionally adding false perceptual and contextual details to their accounts (Masip & Herrero, 2013; Nahari et al., 2012). Similar to a false memory, a fabrication also entails creation of a memory internally, using thought and imagination. However, in contrast to a false memory, liars operate on the fabricated memory intentionally, in a manipulative way. As the nature of these two memories is different, I argue that the term “internally derived memory” used to describe a “false memory” in RM process is somewhat inaccurate for describing a “fabricated memory.” I therefore suggest to distinguish fabricated memories by terming them “self-manipulated memories” rather than “internal memories.” The current study sought to shed light on the distinctive characteristics of these self-manipulated memories.
Boundaries of Reality Truth-tellers usually believe that sticking with the truth is the best strategy for convincing others of their honesty (Granhag & Hartwig, 2008; Hartwig, Granhag, & Strömwall, 2007; Masip & Herrero, 2013; Nahari et al., 2014b); thus, their accounts generally depend on what actually happened. When asked to repeat their account, truth-tellers still try to be consistent with reality rather than with their previous accounts (Granhag & Strömwall, 2002). Liars, on the other hand, are less limited by reality because they fabricate it. Thus, in their first account liars are relatively free to include many false details. However, their degree of freedom decreases in subsequent statements, as they try to be consistent with their previous accounts (Granhag & Strömwall, 2002). Therefore, from the second time they provide a statement, liars actually recall their first account. As such, for liars, the first account functions as reality does for truth-tellers.
2
Lack of Memory Truth-tellers depend on their memory, which decreases over time (see, Nahari & Ben-Shakhar, 2011), and tend to provide fewer details when interviewed after time has passed. In contrast, fabricated memories do not depend to the same extent on the time when the event occurred. As such, the passage of time does not influence the amount of details that liars can provide (see Harvey, Vrij, Hope, Leal, Mann, 2017; Harvey, Vrij, Leal, Hope, & Mann, 2017). Consequently, the differences between truths and lies may be diminished by a delay—with the truthtellers’ forgetting offsetting their initial advantage over liars, making it more difficult to discriminate between the two groups. The situation is different, though, when liars repeat their original account. As the first account is their reference, the memory factor plays a role for liars when they are required to repeat it. Current Study The current study examined conditions that provide liars with the opportunity to manipulate their accounts by adding false perceptual and contextual details and conditions that hinder them from doing so. Specifically, I examined the effects of statement time (following the event described) and repetition. In Stage 1, participants either committed a mock theft (liars) or left the laboratory and attended to their business for 30 min (truth-tellers). In Stage 2, liars and truth-tellers provided their statements at different times: (a) a single statement immediately after Stage 1, (b) a single statement two weeks after Stage 1, or (c) two statements, the first immediately after Stage 1 and the second two weeks after Stage 1. If I extrapolate from the unadorned RM approach, I would expect truth-tellers to provide more perceptual and contextual details than liars in all conditions (Hypothesis 1a). Yet, I expected that the differences in the amount of details provided by liars and truth-tellers would change across conditions. Since liars create their statement rather than retrieve it from memory the first time they provide it, they are likely to be unaffected by a delay in the provision of their statement, unlike truth-tellers, who are likely to exhibit a decrease in the amount of details they recall with passing time. Thus, I expected that the differences between liars’ and truth-tellers’ statements would be greater when participants provided a single statement immediately than when the initial statement was provided following a delay of two weeks (Hypothesis 1b). Hypothesis 1b assumes that the magnitude of the forgetting effect (i.e., the decline between the immediate and delayed conditions) among truth-tellers is not greater than the magnitude of the veracity effect (i.e., the difference between truth-tellers and liars in the immediate condition). This assumption has some empirical support (Harvey, Vrij, Hope, et al., 2017), and will be discussed further in the Discussion section. However, in the attempt to be consistent with their first statement, liars, like truth-tellers, may exhibit a lack of memory in subsequent repetitions. Therefore, I expected that, just like when providing a single statement immediately, the differences between liars’ and truth-tellers’ statements in the amount of details would be greater when participants provided a second statement after a two-week interval than when only a single
Please cite this article in press as: Nahari, G. Reality Monitoring in the Forensic Context: Digging Deeper into the Speech of Liars. Journal of Applied Research in Memory and Cognition (2017), https://doi.org/10.1016/j.jarmac.2018.04.003
+Model
ARTICLE IN PRESS REALITY MONITORING IN THE FORENSIC CONTEXT
statement was provided following a delay of two weeks (Hypothesis 1c). In addition, the veracity of the perceptual and contextual details in the accounts was examined. I expected that liars would provide more false than truthful details, while truth-tellers would provide more truthful than false details, in all experimental conditions (Hypothesis 2a); that liars would add more false details when providing a single statement (immediately or following a delay) than when providing a second statement after a two-week interval (Hypothesis 2b); and, finally, that truth-tellers would add more truthful details when providing a single statement immediately than when providing a single delayed statement or a second statement after a two-week interval (Hypothesis 2c). Method Participants A total of 142 students and administrative staff members (101 females and 41 males) participated in the experiment for course credit or payment (50 NIS). One female participant who did not follow the instructions (i.e., she confessed in the interview) was eliminated, and thus the final sample size was 141. Their mean age was 23.75 years (SD = 2.74). A 2 × 3 between-participants design was used, with two orthogonal factors: veracity and time of test. The veracity factor had two conditions: (a) participants who performed a mock crime (“liar” condition) and (b) participants who did not perform a mock crime (“truth-teller” condition). The time-of-test factor had three conditions: (a) participants who provided a single statement immediately after the first stage of the experiment, during which the mock crime took place (“immediate” condition), (b) participants who gave a single statement two weeks after the first stage of the experiment (“delayed” condition), and (c) participants who gave two statements, the first immediately and the second two weeks after the first stage of the experiment (“repeating” condition). The participants were randomly assigned to one of the six conditions created by this design (with 23–24 participants in each condition). Table 1 summarizes the experimental procedure for each condition. Procedure Stage 1: innocent or criminal activities. Participants arrived at the laboratory individually at a predetermined time. They were told that the experiment would deal with examining the
3
efficiency of specific lie-detection tools. All participants signed a consent form indicating that participation was voluntary and that they could withdraw from the experiment at any time without penalty. The experimenter then read the instructions aloud. To the truth-tellers, she said, “You are requested to go about your normal business for 30 minutes, such as buying coffee, having a conversation with a friend, visiting the book shop, etc. Make sure that you carry out more than one activity and do not stay in one place for more than ten minutes. Please return to the lab in exactly 30 minutes [the experimenter indicated the exact time] for the second stage of the study.” To the liars, she said, “You are requested to commit a mock crime as part of an experiment about deception. The mock theft involves stealing a chemical formula from Professor Zvi Aharonovitz’s office in the Criminology Department. You should enter the room using a key, log into his computer, find the file with the formula, and copy it. Somebody will pick up the formula from you.” The experimenter then gave the participant a closed envelope, and continued, “In the envelope you will find all the details and instructions regarding stealing the formula. Please read them carefully, carry out the crime according to the instructions, and come back here. Please return to the lab in exactly 30 minutes [the experimenter indicated the exact time] for the second stage of the study. Please make sure that nobody witnesses you committing this crime. Also, do not tell anybody about it.” The liars left the laboratory with the envelope and read the instructions outside. They were asked to open Professor Zvi Aharonovitz’s mail box, using a key that was left for them in the envelope. In the mail box, they found a bundle of keys and had to find out which one opened the professor’s office door. The professor’s computer was password protected. In the envelope, participants found a list of potential passwords and had to try them one by one, until they were able to log into the computer. They looked for the file containing the formula and copied it onto a piece of paper. Once this task was completed, they had to make sure to leave the office in the same condition as they found it, lock the office door, and return the bundle of keys to the professor’s mailbox. They were then told to wait in a specific location for someone who was supposed to pick up the formula from them. They were told to put the envelope containing the formula in a specific location in the event that nobody appeared before they had to return to the lab. In all cases, nobody appeared to pick up the envelope. When the participant returned to the laboratory, he or she was either dismissed and asked to return two weeks later
Table 1 Description of the experimental design Condition Immediate
First meeting 1. Innocent or criminal activity (Stage1) 2. Interview (Stage 2a) 3. Post-interview tasks (Stage 3)
Second meeting (two weeks later) –
Delayed
1. Innocent or criminal activities (Stage 1)
2. Interview (Stage 2a) 3. Post-interview tasks (Stage 3)
Repeating
1. Innocent or criminal activities (Stage 1) 2. Interview (first statement; Stage 2a)
3. Interview (second statement; Stage 2b) 4. Post-interview tasks (Stage 3)
Please cite this article in press as: Nahari, G. Reality Monitoring in the Forensic Context: Digging Deeper into the Speech of Liars. Journal of Applied Research in Memory and Cognition (2017), https://doi.org/10.1016/j.jarmac.2018.04.003
+Model
ARTICLE IN PRESS REALITY MONITORING IN THE FORENSIC CONTEXT
for Stage 2a (truth-tellers and liars in the delayed condition), or immediately continued to Stage 2a (truth-tellers and liars in the immediate and repeating conditions). The participants in the delayed condition were not informed about the purpose of Stage 2a, or that they were going to be interviewed about their activities in Stage 1. Stage 2a: first interview. All participants were interviewed in Stage 2a. In the immediate and repeating conditions, participants were interviewed immediately upon their return to the lab after Stage 1, while in the delayed condition the participants were interviewed two weeks after Stage 1. Before the interview, the experimenter told the participant that someone had broken into Professor Aharonovitz’s computer and was suspected of stealing information from his computer. The participant was informed that he or she was one of the suspects in the theft, and would be interviewed about it immediately. The participant was further told that the interviewer did not know if he or she was guilty and was asked not to admit to committing the crime. The experimenter asked the participants to be as convincing as possible in the interview, and added that if they succeeded in convincing the interviewer that they were innocent, they would participate in a drawing in which four participants would win 150 Israeli shekels each. The experimenter accompanied the participant to the interview room. She asked the participant to wait outside the room until the interviewer called him or her to enter. She suggested that the participant use this time to prepare for the interview, and then she knocked on the interview room door and notified the interviewer that the suspect was waiting outside. The interviewer always waited 5–6 min before calling the participant to come in. A second experimenter served as the interviewer and was blind to the participant’s veracity condition. She said to the participant, “You are suspected of breaking into and stealing confidential information from Professor Aharonovitz’s computer in the Department of Criminology. Please tell me what you were doing during the 30 minutes from the time you left the lab to the time you re-entered the lab. I will turn on the recorder now. When you are ready, please tell me about your activities in as much detail as possible, and do not exclude anything, so I can have an idea of what happened during those 30 minutes. Be sure that you mention all details, activities, people you met, and conversations that took place. Give as much information as you can, including any information that seems irrelevant.” Participant statements were recorded. When the participant finished his or her statement, the interviewer asked if he or she would like to add something, and then turned off the recorder. At this point, participants in the immediate and delayed conditions continued immediately to the third stage of the study. In contrast, the participants in the repeating condition were released and asked to come back to the lab two weeks later, for a second interview (Stage 2b). Stage 2b: second interview. Only participants in the repeating condition provided a second statement. When they returned to the lab two weeks after the first interview, they were asked to provide a second statement regarding the relevant 30 min, before continuing to the third stage of the experiment. The instructions
4
for providing the second statement were identical to those given for providing the first statement. Stage 3: post-interview tasks. At this stage, participants were told that the experiment had ended, that their next responses would not influence the assessment of their statements as true or false, and that they should answer the subsequent questions truthfully. The participants in the immediate and delayed conditions performed the post-interview tasks immediately upon providing their first and only statement (immediately after Stage 2a). The participants in the repeating condition performed the post-interview tasks after they provided their second statement (immediately after Stage 2b). First, each participant listened to the recording of his/her own interview (the participants in the repeating condition listened to the second interview), and identified the false and truthful details that they had provided in their statement. Subsequently, participants completed a post-interview questionnaire, in which they were asked to rate on a 7-point scale from 1 (not at all) to 7 (completely) to what extent they (a) were motivated to appear convincing in their statements; (b) found it difficult to appear convincing in their statements; (c) thought they were convincing in their statements; (d) perceived their likelihood to participate in the drawing; and (e) were interested in receiving the payment bonus. The participants were also asked three yes/no questions to determine whether or not they (a) used a strategy to convince the interviewer of their innocence; (b) provided truthful details; and (c) provided false details. Participants were asked to elaborate on their responses to these questions. Those who reported provision of truthful details were also asked to rate on a 7-point scale (d) to what extent they thought their statement was truthful and (e) to what extent their memory was accurate, and (f) to what extent details were not reported due to forgetfulness. Upon completing the questionnaire, all participants were debriefed and given their course credits. Coding the statements for richness in detail. All statements were coded by two trained, independent coders (blind to the experimental conditions). The coders noted the frequency of occurrence of the relevant RM criteria: (a) perceptual details (information units about what the examinee saw, tasted, heard, felt [by touch], or smelled during their alleged activities); (b) contextual details, including spatial details (information units about locations or the spatial arrangement of people and objects) and temporal details (information units about when the event happened or an explicit description of a sequence of events). The coders initially marked the perceptual and contextual (spatial and temporal) details on the statements and subsequently counted them. For example, the sentence “I entered the Cafeteria at 14:00. A black cat was laying at the entrance” contains six details: I entered the Cafeteria (contextual detail), at 14:00 (contextual detail), black (perceptual detail), cat (perceptual detail), was laying (perceptual detail) at the entrance (contextual detail). Detailed explanation for coding of perceptual and contextual details appears in Nahari (2018). For each statement, the sum of the frequency counts of the perceptual and contextual details was calculated and the resulting variable was labeled richness in detail. One coder coded all the statements and a second one
Please cite this article in press as: Nahari, G. Reality Monitoring in the Forensic Context: Digging Deeper into the Speech of Liars. Journal of Applied Research in Memory and Cognition (2017), https://doi.org/10.1016/j.jarmac.2018.04.003
+Model
ARTICLE IN PRESS REALITY MONITORING IN THE FORENSIC CONTEXT
coded 70% of them. The inter-rater reliability of richness in detail between the coders, measured by the intraclass correlation coefficient (ICC) for consistency, was high and satisfactory (ICC[2,2]= .98, averaged measures). Classifying the type of details. In Stage 3 of the experiment, the participants listened to the recording of their own interview and identified the false and truthful details that they had provided in their statement. Based on their report, the perceptual and contextual details recognized by the coders were classified into truthful and false details. For example, if a participant said in her statement: “I sat on a bench and had a coffee. While drinking it, Zvi approached me.” In this statement there are six details: sat (perceptual detail), on a bench (contextual detail), had a coffee (perceptual detail); while drinking it (contextual detail), Zvi (perceptual detail), approached me (contextual detail). Now, let us consider that the participant reports that this statement is true, except for the fact that Zvi approached her. In such a case, three details should be classified as being true (set, on a bench, had a coffee), and the other three details should be classified as being false (while drinking it, Zvi, approached me).
5
Perception of Difficulty and Success in Being Convincing Liars found it more difficult to appear convincing in their statements (M = 4.20, SD = 1.82) than did truth-tellers (M = 2.47, SD = 1.90), F(1, 135) = 30.36, η2 = .18, p < .001. Liars further assessed their success in being convincing (M = 4.55, SD = 1.40) and their chances of participating in the draw (M = 4.16, SD = 1.70) as lower than truth-tellers (M = 5.78, SD = 1.09 and M = 5.10, SD = 1.60, respectively), F(1, 135) = 34.42, η2 = .20, p < .001 and F(1, 135) = 11.76, η2 = .08, p < .01. There were no differences between time-of-test conditions with respect to any of these measures, and no interaction between veracity and time of test, all F(2, 135)s < 3.02. Perceived Memory In all the time-of-test conditions, truth-tellers perceived their memory as highly accurate (M = 6.47 on 7-point Likert scale, SD = .93) and estimated the number of details that were not reported due to forgetfulness as low (M = 1.93 on 7-point Likert scale, SD = 1.34), F(2, 69) < 1 and F(2, 69) = 2.47, η2 = .07, ns, respectively. The accuracy perceptions of liars are irrelevant, as many of them fabricated the reported memories.
Results Veracity Manipulation Check
Reported Strategy
Significant between-group differences were revealed in the amount of truthful details participants said they reported in their statements, t(36.14) = 7.85, p < .001 (variances of groups [truthful, false] were unequal; t test calculated accordingly). Truth-tellers reported providing more truthful details (M = 6.99 on a 7-point Likert scale, SD = .12) than liars (M = 4.54, SD = 1.89). This indicates that the participants followed the experimenter’s instructions. All truth-tellers said that they provided only truthful details. Fifty-six liars (81.2%) said that they provided false details, with 24 of them providing false details in addition to truthful details. This mimics the real-life behavior of liars, who often embed their lies into truthful statements (Vrij, 2008; Vrij, Granhag, & Porter, 2010). The remaining thirteen liars (18.8%) reported that they did not provide any false details. I checked the statements and responses of these 13 liars, and found that they provided concealed lies (see Nahari et al., 2012). They admitted to being in the department, reported only their non-criminal activities there (e.g., people they saw there, playing with their phone, etc.), and concealed their criminal activities, a strategy that has been found in previous studies (e.g., Nahari et al., 2014a).
More liars (n = 49; 71.0%) than truth-tellers (n = 20; 27.8%) reported using a strategy to appear convincing, χ2 (1) = 26.36, φ = .43, p < .001. This pattern was similar across the three timeof-test conditions.
Participant Motivation The reported motivation of the participants to be convincing in their statements was high (M = 6.02 on a 7-point Likert scale, SD = 1.39). Further, 90.1% of the participants (n = 127) were interested in receiving the payment bonus (ratings of 5, 6, or 7; M = 6.25, SD = 1.44). There was no difference between the experimental conditions (veracity, time of test) with respect to motivation or interest in receiving the payment bonus, all Fs < 2.65.
Testing of Hypotheses Richness in detail. The three predictions of Hypothesis 1 referred to the statement’s richness in detail. To test this hypothesis, I compared the richness in detail of statements provided by truth-tellers and liars in the three time-of-test conditions. Note that participants in the repeating condition provided two statements. However, since the hypothesis refers only to their second statement, their first statement was excluded from the comparison.1 Thus, richness in detail of the second statement in the repeating group was compared with the single statements provided by participants in the immediate condition and the delayed condition. Since number of words can be related to, and therefore confound, richness in detail, I first conducted an ANOVA with number of words as the dependent variable and veracity and time of test as factors. This analysis revealed a statistically significant time-of-test effect, F(2, 135) = 5.38, partial η2 = .07, p < .01. A post hoc test (Scheffe) revealed that participants in
1 I performed a repeated measures ANCOVA to compare the richness in detail of the two statements provided in the repeating condition. The statement (first, second) was the within-subject factor and veracity group (truth-tellers, liars) was the between-subject factor. The number of words of the first statement was entered as a covariate. Neither the effect of statement nor the Statement × Veracity interaction were significant, F(1, 44) = 1.76, partial η2 = .04, ns, and F < 1 respectively. This mirrors the finding that statements in the immediate and repeating conditions did not differ in terms of richness in detail.
Please cite this article in press as: Nahari, G. Reality Monitoring in the Forensic Context: Digging Deeper into the Speech of Liars. Journal of Applied Research in Memory and Cognition (2017), https://doi.org/10.1016/j.jarmac.2018.04.003
+Model
ARTICLE IN PRESS REALITY MONITORING IN THE FORENSIC CONTEXT
6
Table 2 Comparison of richness in detail (estimated marginal Ms and SDs) across veracity and time-of-test conditions Immediate condition
Delayed condition
Repeating condition
Truth-tellers
23.77 (7.01)
16.43 (6.81)
20.68 (6.81)
Liars
11.33 (6.86)
9.99 (6.86)
14.42 (6.91)
Veracity condition comparison
F(1,44) = 33.39 partial η2 = .43 P < .001
F(1,44) = 21.42 partial η2 = .33 P < .001
F(1,44) = 7.87 partial η2 = .15 P < .01
the immediate condition (M = 165.64, SD = 111.66) provided longer statements than participants in the delayed (M = 121.55, SD = 65.30) and repeating (M = 112.79, SD = 63.74) conditions. The main effect of veracity and the Veracity × Time-of-test interaction were not significant, F(1, 135) = 2.10, partial η2 = .02, ns, and F < 1, respectively. I then conducted an ANCOVA with richness in detail as the dependent variable, and veracity and time of test as factors. As noted above, the length of statements differed among the experimental groups. Thus, as in previous studies (e.g., Nahari & Pazuelo, 2015), to ensure that between-group differences in richness in detail were not simply a function of the amount of information provided overall, number of words was inserted as a covariate, which was found to be significant, F(1, 134) = 105.99, partial η2 = .44, p < .001. This analysis revealed a statistically significant time-of-test effect, F(2, 134) = 6.29, partial η2 = .09, p < .01, indicating that participants in the immediate condition (M = 17.49, SD = 6.99) and in the repeating condition (M = 17.55, SD = 6.92) provided more details than did participants in the delayed condition (M = 13.21, SD = 6.79). The participants in the immediate and repeating conditions did not differ in richness in detail. The veracity effect, F(1, 134) = 52.51, partial η2 = .28, p < .001, was also significant. Truth-tellers (M = 20.29, SD = 6.87) provided more details than liars (M = 11.91, SD = 6.81). Finally, a Time-of-test × Veracity interaction was observed, F(2, 134) = 3.12, partial η2 = .04, p < .05. To locate the source of this interaction, tests of simple effects were conducted. As shown in Table 2, truth-tellers provided more details than liars in each of the three time-of-test conditions. However, differences across the time-of-test conditions were observed among truthtellers but not among liars. A post hoc test (Scheffe) revealed that truth-tellers provided more details in the immediate condition than in the delayed condition, and similar amount of details in the immediate and repeating conditions. Liars provided the same amount of details in all three conditions. To summarize, as predicted in Hypothesis 1a, differences between truth-tellers and liars were significant for all time-oftest conditions. In accordance with Hypothesis 1b, truth-tellers provided more details in the immediate condition than in the delayed condition, and thus the differences between truthtellers and liars were greater in the immediate condition (partial η2 = .43) in comparison to the delayed condition (partial η2 = .33). However, contrary to Hypothesis 1c, the amount
Time-of-test condition comparison F(2,68) = 4.85 partial η2 = .13 P < .05 F(2,65) = 2.53 partial η2 = .07 ns
of details provided by truth-tellers and liars did not decrease in the second statement, and differences between truth-tellers and liars were not greater in the repeating condition than in the delay condition. In fact, these differences were smaller in the repeating condition than in the other two conditions (partial η2 = .15). Thus, the first hypothesis was only partially confirmed. Type of details. The three predictions of Hypothesis 2 referred to the type of details (truthful or false) found in statements. To examine these predictions, a repeated measures ANOVA was performed, in which the type of details (truthful, false) was the within-subject factor and veracity and time of test were the between-subject factors. Table 3 presents the means (and standard deviations) for truthful and false details across the experimental conditions. A main effect was found for type of details, F(1, 135) = 98.70, partial η2 = .42, p < .001, showing that participants provided more truthful details (M = 12.64, SD = 13.08) than false details (M = 3.55, SD = 5.65). A Details × Time-of-test interaction, F(2, 135) = 4.50, partial η2 = .06, p < .05, as well as a strong Details × Veracity interaction were found, F(1, 135) = 170.64, partial η2 = .56, p < .001. Finally, the triple interaction was marginally significant, F(2, 135) = 2.96, partial η2 = .04, p = .05. Simple effects showed different patterns for truth-tellers and liars. First, paired-sample t tests revealed that the statements of truth-tellers included more truthful details than false details in all conditions, t(23) = 10.97, d = 2.24, p < .001 in the immediate condition, t(23) = 12.11, d = 2.47, p < .001 in the delayed condition, and t(23) = 7.04, d = 1.44, p < .001 in the repeating condition. The amount of false details in truth-tellers’ statements was negligible (means ranged between .04 and .46 details). In contrast, the statements of liars included more false details than truthful details in the delayed condition, t(22) = −2.16, d = .48, p < .05, but the same amount of false and truthful details in the immediate and repeating conditions, t(22) = −.65, d = .13, ns, and t(22) = −1.87, d = .42, ns respectively. Second, one-way ANOVAs showed that truth-tellers provided different amount of truthful details across conditions, F(2, 69) = 6.69, partial η2 = .16, p < .01, and the same, negligible, amount of false details in all conditions, F(2, 69) = 2.68, partial η2 = .07, ns. A post hoc test (Scheffe) revealed that they provided more truthful details in the immediate condition than in the delayed condition, and the same amount of truthful details in the immediate and repeating conditions. In all conditions, liars provided the same amount of both truthful
Please cite this article in press as: Nahari, G. Reality Monitoring in the Forensic Context: Digging Deeper into the Speech of Liars. Journal of Applied Research in Memory and Cognition (2017), https://doi.org/10.1016/j.jarmac.2018.04.003
+Model
Truth-tellers Truthful details False details
26.96 (11.90) 0.33 ( 0.64)
Liars
Delayed condition Overall
5.35 (7.33 ) 16.38 (14.69) 7.09 (7.98) 3.64 ( 6.51)
Truth-tellers 15.42 (5.69) 0.46 ( 0.88)
Liars 3.00 (3.07 ) 5.83 ( 4.99)
Repeating condition Overall
Truth-tellers
9.34 (7.75) 3.09 ( 4.43)
19.83 (13.81) 0.04 ( 0.20)
Liars
Across conditions Overall
4.21 (5.28 ) 12.19 (13.08) 8.00 ( 5.74) 3.94 ( 5.65)
Truth-tellers 20.74 (11.88) 0.28 ( 0.65)
Liars
Overall
4.19 (5.51 ) 12.64 (12.46) 3.55 ( 5.57) 6.97 ( 6.34)
ARTICLE IN PRESS
Immediate condition
REALITY MONITORING IN THE FORENSIC CONTEXT
Please cite this article in press as: Nahari, G. Reality Monitoring in the Forensic Context: Digging Deeper into the Speech of Liars. Journal of Applied Research in Memory and Cognition (2017), https://doi.org/10.1016/j.jarmac.2018.04.003
Table 3 Means (and standard deviations) of truthful and false details across veracity and time-of-test conditions
7
+Model
ARTICLE IN PRESS REALITY MONITORING IN THE FORENSIC CONTEXT
and false details, F < 1 and F(2, 66) = 1.05, partial η2 = .03, ns, respectively. To summarize, the predictions of Hypothesis 2 were only partially confirmed. In accordance with Hypothesis 2a, truth-tellers provided more truthful than false details, but contrary to this hypothesis, liars provided more false details than truthful details only in the delayed condition. In the immediate and repeating conditions, liars provided the same amount of truthful and false details. In contrast with Hypothesis 2b, liars provided the same amount of false details in all three conditions. Finally, in accordance with Hypothesis 2c, truth-tellers provided more truthful details in the immediate condition than in the delayed condition. However, contrary to this hypothesis, truth-tellers provided the same amount of truthful details in the immediate and repeating conditions. Discussion The current study examined the effect of statement time and repetition on the provision of perceptual and contextual details. In accordance with the RM approach, truth-tellers provided more perceptual and contextual details than liars in all the experimental conditions. I also found that most of the liars added false details to their accounts while truth-tellers stuck with the truth. One may wonder why liars, who apparently could add whatever details they chose, did not reach the same level of detail as did truth-tellers. One suggestion, drawn from the Criteria-Based Content Analysis (CBCA) approach, is that some details are difficult for liars to make up (Köhnken, 1996; Kohnken, 2004). Alternatively, the cognitive load approach proposes that sometimes limitations in cognitive resources may reduce the ability of liars to provide many details (Vrij et al., 2008). A different perspective, however, starts with the observation that liars do not lack imagination (Merckelbach, 2004), and that many liars base their lies on an event that they have actually experienced, albeit at another time (Leins, Fisher, & Ross, 2013), or embed truthful details in their lies (Nahari & Vrij, 2015). Furthermore, there is evidence that liars are able to provide as many details as truth-tellers when encouraged to do so (Nahari & Pazuelo, 2015). Perhaps, therefore, the contrast in detail between truth-tellers and liars is better explained by the verifiability approach (Nahari et al., 2014a, 2014b). According to this strategy-based approach, liars are careful in providing verifiable details, which can be checked by the investigators. That is, even if liars have a rich imagination and even if they discuss previous experiences, they are limited in the number of details that they can provide because of their motivation to keep their accounts unverifiable and difficult to refute. In the current study, differences between liars and truth-tellers were greater when their statements were provided immediately as opposed to following a delay. Truth-tellers, who stuck with the truth, provided more details in the immediate condition than in the delayed condition, presumably due to lack of memory. Liars, who did not suffer from lack of memory, provided the
8
same number of details in both conditions. Interestingly, liars provided a similar number of truthful and false details in the immediate condition, but more false than truthful details in the delayed condition. This implies that liars, as well as truthtellers, had difficulty providing truthful details after a delay. However, in contrast to truth-tellers, liars compensated for this by adding false details. In this way, the difference between liars and truth-tellers decreased when the statement was provided following a delay. The implications of these findings are significant, since in real life, most of the time suspects are interviewed days, weeks, or even years after the crime was committed. In contrast to our expectations, both truth-tellers and liars provided the same number of details in the immediate and the repeating conditions (second statement). That is, the participants in the repeating condition did not show lack of memory when repeating their first statements. One possible explanation is related to the testing effect, which predicts that retrieval practice benefits later memory, even at long delays (see McDermott, Arnold, & Nelson, 2014). Truth-tellers in the repeating condition provided their first statements immediately after the event. This act can be considered retrieval practice or testing, which may have had a beneficial effect on memory of the event when they provided the second statement. However, this explanation is not relevant for liars. Since their memories were fabricated while they provided their first statements, this stage cannot be considered memory practice. It is more likely that liars practiced during the interval between the statements, in order to be consistent. Liars knew that to avoid being discovered, they had to remember what they said in their previous statements (Granhag & Strömwall, 1999). It is thus reasonable to assume that they made an effort to remember what they had said. One point merits attention. I examined the hypotheses on the amount of perceptual and contextual details combined (i.e., richness in detail). Although the reality-monitoring theory does not provide differential predictions for each type of details, one may argue that it is worth further examining the hypotheses for each criterion separately. The sample was too small to allow analyzing perceptual and contextual details separately, and thus a future study is required for this purpose. To conclude, the current study demonstrates that the manipulation factor in verbal behavior of liars is key to the adjustment and improvement of RM as a tool for lie detection, and that for better understanding of this manipulation factor, one should look also at the veracity (false or truthful) of details in different conditions (statement time and repetition). This line of reasoning requires more research. Conflicts of Interest Statement The author declares no conflict of interest. Author Contributions Single author. Responsible for all manuscript components.
Please cite this article in press as: Nahari, G. Reality Monitoring in the Forensic Context: Digging Deeper into the Speech of Liars. Journal of Applied Research in Memory and Cognition (2017), https://doi.org/10.1016/j.jarmac.2018.04.003
+Model
ARTICLE IN PRESS REALITY MONITORING IN THE FORENSIC CONTEXT
Acknowledgment This research was supported by the Israel Science Foundation grant (grant No. 372/14). References DePaulo, B. M., Lindsay, J. J., Malone, B. E., Muhlenbruck, L., Charlton, K., & Cooper, H. (2003). Cues to deception. Psychological Bulletin, 129, 74–118. Granhag, P. A., & Hartwig, M. (2008). A new theoretical perspective on deception detection: On the psychology of instrumental mindreading. Psychology, Crime & Law, 14, 189–200. Granhag, P. A., & Strömwall, L. A. (1999). Repeated interrogations – Stretching the deception detection paradigm. Expert Evidence, 7, 163–174. Granhag, P. A., & Strömwall, L. A. (2002). Repeated interrogations: Verbal and non-verbal cues to deception. Applied Cognitive Psychology, 16, 243–257. Hartwig, M., Granhag, P. A., & Strömwall, L. A. (2007). Guilty and innocent suspects’ strategies during police interrogations. Psychology, Crime & Law, 13, 213–227. Harvey, A. C., Vrij, A., Hope, L., Leal, S., & Mann, S. (2017). A stability bias effect among deceivers. Law and Human Behavior, 41, 519–529. Harvey, A. C., Vrij, A., Leal, S., Hope, L., & Mann, S. (2017). Deception and decay: Verbal lie detection as a function of delay and encoding quality. Journal of Applied Research in Memory and Cognition, 6, 306–318. Johnson, M. K. (2006). Memory and reality. American Psychologist, 61, 760–771. Johnson, M. K., Bush, J. G., & Mitchell, K. J. (1998). Interpersonal reality monitoring: Judging the sources of other people’s memories. Social Cognition, 16, 199–224. Johnson, M. K., & Raye, C. L. (1981). Reality monitoring. Psychological Review, 88, 67–85. Köhnken, G. (1996). Social psychology and the law. In G. R. Semin, & K. Fiedler (Eds.), Applied social psychology (pp. 257–282). London, UK: Sage Publications. Kohnken, G. (2004). Statement validity analysis and the detection of the truth. In P. A. Granhag, & L. A. Stromwall (Eds.), The detection of deception in forensic contexts (pp. 41–63). Cambridge: Cambridge University Press. Leins, D. A., Fisher, P. A., & Ross, S. J. (2013). Exploring liars’ strategies for creating deceptive reports. Legal and Criminological Psychology, 18, 141–151. Masip, J., & Herrero, C. (2013). ‘What would you say if you were guilty?’ Suspects’ strategies during a hypothetical behavior analysis interview concerning a serious crime. Applied Cognitive Psychology, 27, 60–70. Masip, J., Sporer, S. L., Garrido, E., & Herrero, C. (2005). The detection of deception with the reality monitoring approach: A review of the empirical evidence. Psychology, Crime & Law, 11, 99–122.
9
McDermott, K. B., Arnold, K. M., & Nelson, S. M. (2014). The testing effect. In T. J. Perfect, & D. S. Lindsay (Eds.), The SAGE handbook of applied memory (pp. 183–200). Thousand Oaks, CA: SAGE Publications. Merckelbach, H. (2004). Telling a good story: Fantasy proneness and the quality of fabricated memories. Personality and Individual Differences, 37, 1371–1382. Nahari, G. (2018). The applicability of the verifiability approach to the real world. In P. Rosenfeld (Ed.), Detecting concealed information and deception: Recent developments (pp. 329–349). London: Elsevier. Nahari, G., & Ben-Shakhar, G. (2011). Psychophysiological and behavioral measures for detecting concealed information: The role of memory for crime details. Psychophysiology, 48, 733–744. Nahari, G., & Pazuelo, M. (2015). Telling a convincing story: Richness in detail as a function of gender and information. Journal of Applied Research in Memory and Cognition, 4, 363–367. Nahari, G., & Vrij, A. (2015). Can someone fabricate verifiable details when planning in advance? It all depends on the crime scenario. Psychology, Crime & Law, 21, 987–999. Nahari, G., Vrij, A., & Fisher, R. P. (2012). Does the truth come out in the writing? Scan as a lie detection tool. Law and Human Behavior, 36, 68–76. Nahari, G., Vrij, A., & Fisher, R. (2014a). Exploiting liars’ verbal strategies by examining unverifiable details. Legal and Criminological Psychology, 19, 227–239. Nahari, G., Vrij, A., & Fisher, R. (2014b). The verifiability approach: Countermeasures facilitate its ability to discriminate between truths and lies. Applied Cognitive Psychology, 28, 122–128. Sporer, S. L. (1997). The less travelled road to truth: Verbal cues in deception detection in accounts of fabricated and self-experienced events. Applied Cognitive Psychology, 11, 373–397. Sporer, S. L. (2004). Reality monitoring and detection of deception. In P. A. Granhag, & L. A. Stromwall (Eds.), The detection of deception in forensic contexts (pp. 64–102). UK: Cambridge University Press. Vrij, A. (2005). Criteria-based content analysis: A qualitative review of the first 37 studies. Psychology, Public Policy, and Law, 11, 3–41. Vrij, A. (2008). Detecting lies and deceit: Pitfalls and opportunities (2nd ed., pp. 1–502). John Wiley & Sons. Vrij, A., Granhag, P. A., & Porter, S. B. (2010). Pitfalls and opportunities in nonverbal and verbal lie detection. Psychological Science in the Public Interest, 11, 89–121. Vrij, A., Mann, S. A., Fisher, R. P., Leal, S., Milne, R., & Bull, R. (2008). Increasing cognitive load to facilitate lie detection: The benefit of recalling an event in reverse order. Law and Human Behavior, 32, 253–265.
Received 4 July 2017; received in revised form 5 April 2018; accepted 5 April 2018 Available online xxx
Please cite this article in press as: Nahari, G. Reality Monitoring in the Forensic Context: Digging Deeper into the Speech of Liars. Journal of Applied Research in Memory and Cognition (2017), https://doi.org/10.1016/j.jarmac.2018.04.003
ARTICLE IN PRESS Journal of Applied Research in Memory and Cognition xxx (2017) xxx–xxx Contents lists available at ScienceDirect
Journal of Applied Research in Memory and Cognition journal homepage: www.elsevier.com/locate/jarmac
Reality Monitoring in the Forensic Context: Digging Deeper into the Speech of Liars Galit Nahari∗ Bar-Ilan University, Israel Reality monitoring (RM) indicates that truthful accounts contain more perceptual and contextual details than false accounts. Considering the tendency of liars to manipulate their accounts by adding false details, I compared truths and lies in terms of the amount and veracity of details provided by suspects across three conditions: a single statement provided immediately; a single statement following a two-week delay; or two statements, the first provided immediately and the second following a two-week delay. Distinguishing truths from lies was possible across conditions, but with varying intensity. Truth-tellers provided only truthful details, whereas liars provided both truthful and false details. While the opportunity to provide truthful details decreased over time for both truthtellers and liars, only the latter compensated for this decrease by adding false details. The current study provides a new empirical approach and significant insight into the application of the RM framework in the forensic context.
General Audience Summary The current study examined the verbal behavior of suspects, who tell the truth or lie when they are interviewed about their involvement in a crime, across three situations: when they provide a single statement immediately after the crime occurred; a single statement following a two-week delay; or two statements, the first provided immediately after the crime occurred and the second following a two-week delay. According to the realitymonitoring approach for lie detection, truth-tellers provide more perceptual (e.g., what they saw, heard, and smelled during the described event) and contextual details (e.g., times and locations) than liars. While truthtellers usually provide truthful details in the interviews, liars, who are motivated to be convincing, manipulate their accounts by adding false details. Results showed that distinguishing truths from lies was possible in all situations, but with varying intensity. Truth-tellers provided only truthful details, whereas liars provided both truthful and false details. While the opportunity to provide truthful details decreased over time for both truth-tellers and liars, only the latter compensated for this decrease by adding false details. The current study provides new insights into the verbal behavior of liars and truth-tellers. Keywords: Reality monitoring, Detection deception, Richness in detail, Memory, Self-manipulated memory
Author Note Correspondence concerning this article should be addressed to Galit Nahari, Department of Criminology, Bar-Ilan University, Ramat Gan 52900, Israel. Contact: [email protected] ∗
Please cite this article in press as: Nahari, G. Reality Monitoring in the Forensic Context: Digging Deeper into the Speech of Liars. Journal of Applied Research in Memory and Cognition (2017), https://doi.org/10.1016/j.jarmac.2018.04.003
+Model
ARTICLE IN PRESS REALITY MONITORING IN THE FORENSIC CONTEXT
Reality monitoring (RM) theory (Johnson & Raye, 1981) describes the process used by an individual to monitor his or her own memories by their content qualities. According to RM, memories for experienced events (externally derived memories) are characterized by perceptual and contextual attributes, while memories for imagined events (internally derived memories) are characterized by the cognitive operational attributes that help to generate them. A few decades ago, the RM framework was adapted to the field of deception detection (Sporer, 1997, 2004), where the process of judging the source of other people’s memories has been labeled interpersonal RM (iRM; Johnson, 2006; Johnson, Bush, & Mitchell, 1998). In this application, true accounts are expected to be richer in perceptual and contextual details than are fabricated accounts (e.g., Vrij, 2008). This extension of RM has been supported by many studies (see DePaulo et al., 2003; Masip, Sporer, Garrido, & Herrero, 2005; Nahari, Vrij, & Fisher, 2012; Vrij, 2005, 2008). Yet, I argue that it occurred without taking into consideration the factors that distinguish fabrication (in iRM) from false memory (in RM). One factor involved in iRM for deception detection and absent in RM, where intended deception and manipulation are not likely, is the tendency of liars to manipulate their accounts to make them seem truthful (Granhag & Hartwig, 2008; Nahari, Vrij, & Fisher, 2014a, 2014b), for example by intentionally adding false perceptual and contextual details to their accounts (Masip & Herrero, 2013; Nahari et al., 2012). Similar to a false memory, a fabrication also entails creation of a memory internally, using thought and imagination. However, in contrast to a false memory, liars operate on the fabricated memory intentionally, in a manipulative way. As the nature of these two memories is different, I argue that the term “internally derived memory” used to describe a “false memory” in RM process is somewhat inaccurate for describing a “fabricated memory.” I therefore suggest to distinguish fabricated memories by terming them “self-manipulated memories” rather than “internal memories.” The current study sought to shed light on the distinctive characteristics of these self-manipulated memories.
Boundaries of Reality Truth-tellers usually believe that sticking with the truth is the best strategy for convincing others of their honesty (Granhag & Hartwig, 2008; Hartwig, Granhag, & Strömwall, 2007; Masip & Herrero, 2013; Nahari et al., 2014b); thus, their accounts generally depend on what actually happened. When asked to repeat their account, truth-tellers still try to be consistent with reality rather than with their previous accounts (Granhag & Strömwall, 2002). Liars, on the other hand, are less limited by reality because they fabricate it. Thus, in their first account liars are relatively free to include many false details. However, their degree of freedom decreases in subsequent statements, as they try to be consistent with their previous accounts (Granhag & Strömwall, 2002). Therefore, from the second time they provide a statement, liars actually recall their first account. As such, for liars, the first account functions as reality does for truth-tellers.
2
Lack of Memory Truth-tellers depend on their memory, which decreases over time (see, Nahari & Ben-Shakhar, 2011), and tend to provide fewer details when interviewed after time has passed. In contrast, fabricated memories do not depend to the same extent on the time when the event occurred. As such, the passage of time does not influence the amount of details that liars can provide (see Harvey, Vrij, Hope, Leal, Mann, 2017; Harvey, Vrij, Leal, Hope, & Mann, 2017). Consequently, the differences between truths and lies may be diminished by a delay—with the truthtellers’ forgetting offsetting their initial advantage over liars, making it more difficult to discriminate between the two groups. The situation is different, though, when liars repeat their original account. As the first account is their reference, the memory factor plays a role for liars when they are required to repeat it. Current Study The current study examined conditions that provide liars with the opportunity to manipulate their accounts by adding false perceptual and contextual details and conditions that hinder them from doing so. Specifically, I examined the effects of statement time (following the event described) and repetition. In Stage 1, participants either committed a mock theft (liars) or left the laboratory and attended to their business for 30 min (truth-tellers). In Stage 2, liars and truth-tellers provided their statements at different times: (a) a single statement immediately after Stage 1, (b) a single statement two weeks after Stage 1, or (c) two statements, the first immediately after Stage 1 and the second two weeks after Stage 1. If I extrapolate from the unadorned RM approach, I would expect truth-tellers to provide more perceptual and contextual details than liars in all conditions (Hypothesis 1a). Yet, I expected that the differences in the amount of details provided by liars and truth-tellers would change across conditions. Since liars create their statement rather than retrieve it from memory the first time they provide it, they are likely to be unaffected by a delay in the provision of their statement, unlike truth-tellers, who are likely to exhibit a decrease in the amount of details they recall with passing time. Thus, I expected that the differences between liars’ and truth-tellers’ statements would be greater when participants provided a single statement immediately than when the initial statement was provided following a delay of two weeks (Hypothesis 1b). Hypothesis 1b assumes that the magnitude of the forgetting effect (i.e., the decline between the immediate and delayed conditions) among truth-tellers is not greater than the magnitude of the veracity effect (i.e., the difference between truth-tellers and liars in the immediate condition). This assumption has some empirical support (Harvey, Vrij, Hope, et al., 2017), and will be discussed further in the Discussion section. However, in the attempt to be consistent with their first statement, liars, like truth-tellers, may exhibit a lack of memory in subsequent repetitions. Therefore, I expected that, just like when providing a single statement immediately, the differences between liars’ and truth-tellers’ statements in the amount of details would be greater when participants provided a second statement after a two-week interval than when only a single
Please cite this article in press as: Nahari, G. Reality Monitoring in the Forensic Context: Digging Deeper into the Speech of Liars. Journal of Applied Research in Memory and Cognition (2017), https://doi.org/10.1016/j.jarmac.2018.04.003
+Model
ARTICLE IN PRESS REALITY MONITORING IN THE FORENSIC CONTEXT
statement was provided following a delay of two weeks (Hypothesis 1c). In addition, the veracity of the perceptual and contextual details in the accounts was examined. I expected that liars would provide more false than truthful details, while truth-tellers would provide more truthful than false details, in all experimental conditions (Hypothesis 2a); that liars would add more false details when providing a single statement (immediately or following a delay) than when providing a second statement after a two-week interval (Hypothesis 2b); and, finally, that truth-tellers would add more truthful details when providing a single statement immediately than when providing a single delayed statement or a second statement after a two-week interval (Hypothesis 2c). Method Participants A total of 142 students and administrative staff members (101 females and 41 males) participated in the experiment for course credit or payment (50 NIS). One female participant who did not follow the instructions (i.e., she confessed in the interview) was eliminated, and thus the final sample size was 141. Their mean age was 23.75 years (SD = 2.74). A 2 × 3 between-participants design was used, with two orthogonal factors: veracity and time of test. The veracity factor had two conditions: (a) participants who performed a mock crime (“liar” condition) and (b) participants who did not perform a mock crime (“truth-teller” condition). The time-of-test factor had three conditions: (a) participants who provided a single statement immediately after the first stage of the experiment, during which the mock crime took place (“immediate” condition), (b) participants who gave a single statement two weeks after the first stage of the experiment (“delayed” condition), and (c) participants who gave two statements, the first immediately and the second two weeks after the first stage of the experiment (“repeating” condition). The participants were randomly assigned to one of the six conditions created by this design (with 23–24 participants in each condition). Table 1 summarizes the experimental procedure for each condition. Procedure Stage 1: innocent or criminal activities. Participants arrived at the laboratory individually at a predetermined time. They were told that the experiment would deal with examining the
3
efficiency of specific lie-detection tools. All participants signed a consent form indicating that participation was voluntary and that they could withdraw from the experiment at any time without penalty. The experimenter then read the instructions aloud. To the truth-tellers, she said, “You are requested to go about your normal business for 30 minutes, such as buying coffee, having a conversation with a friend, visiting the book shop, etc. Make sure that you carry out more than one activity and do not stay in one place for more than ten minutes. Please return to the lab in exactly 30 minutes [the experimenter indicated the exact time] for the second stage of the study.” To the liars, she said, “You are requested to commit a mock crime as part of an experiment about deception. The mock theft involves stealing a chemical formula from Professor Zvi Aharonovitz’s office in the Criminology Department. You should enter the room using a key, log into his computer, find the file with the formula, and copy it. Somebody will pick up the formula from you.” The experimenter then gave the participant a closed envelope, and continued, “In the envelope you will find all the details and instructions regarding stealing the formula. Please read them carefully, carry out the crime according to the instructions, and come back here. Please return to the lab in exactly 30 minutes [the experimenter indicated the exact time] for the second stage of the study. Please make sure that nobody witnesses you committing this crime. Also, do not tell anybody about it.” The liars left the laboratory with the envelope and read the instructions outside. They were asked to open Professor Zvi Aharonovitz’s mail box, using a key that was left for them in the envelope. In the mail box, they found a bundle of keys and had to find out which one opened the professor’s office door. The professor’s computer was password protected. In the envelope, participants found a list of potential passwords and had to try them one by one, until they were able to log into the computer. They looked for the file containing the formula and copied it onto a piece of paper. Once this task was completed, they had to make sure to leave the office in the same condition as they found it, lock the office door, and return the bundle of keys to the professor’s mailbox. They were then told to wait in a specific location for someone who was supposed to pick up the formula from them. They were told to put the envelope containing the formula in a specific location in the event that nobody appeared before they had to return to the lab. In all cases, nobody appeared to pick up the envelope. When the participant returned to the laboratory, he or she was either dismissed and asked to return two weeks later
Table 1 Description of the experimental design Condition Immediate
First meeting 1. Innocent or criminal activity (Stage1) 2. Interview (Stage 2a) 3. Post-interview tasks (Stage 3)
Second meeting (two weeks later) –
Delayed
1. Innocent or criminal activities (Stage 1)
2. Interview (Stage 2a) 3. Post-interview tasks (Stage 3)
Repeating
1. Innocent or criminal activities (Stage 1) 2. Interview (first statement; Stage 2a)
3. Interview (second statement; Stage 2b) 4. Post-interview tasks (Stage 3)
Please cite this article in press as: Nahari, G. Reality Monitoring in the Forensic Context: Digging Deeper into the Speech of Liars. Journal of Applied Research in Memory and Cognition (2017), https://doi.org/10.1016/j.jarmac.2018.04.003
+Model
ARTICLE IN PRESS REALITY MONITORING IN THE FORENSIC CONTEXT
for Stage 2a (truth-tellers and liars in the delayed condition), or immediately continued to Stage 2a (truth-tellers and liars in the immediate and repeating conditions). The participants in the delayed condition were not informed about the purpose of Stage 2a, or that they were going to be interviewed about their activities in Stage 1. Stage 2a: first interview. All participants were interviewed in Stage 2a. In the immediate and repeating conditions, participants were interviewed immediately upon their return to the lab after Stage 1, while in the delayed condition the participants were interviewed two weeks after Stage 1. Before the interview, the experimenter told the participant that someone had broken into Professor Aharonovitz’s computer and was suspected of stealing information from his computer. The participant was informed that he or she was one of the suspects in the theft, and would be interviewed about it immediately. The participant was further told that the interviewer did not know if he or she was guilty and was asked not to admit to committing the crime. The experimenter asked the participants to be as convincing as possible in the interview, and added that if they succeeded in convincing the interviewer that they were innocent, they would participate in a drawing in which four participants would win 150 Israeli shekels each. The experimenter accompanied the participant to the interview room. She asked the participant to wait outside the room until the interviewer called him or her to enter. She suggested that the participant use this time to prepare for the interview, and then she knocked on the interview room door and notified the interviewer that the suspect was waiting outside. The interviewer always waited 5–6 min before calling the participant to come in. A second experimenter served as the interviewer and was blind to the participant’s veracity condition. She said to the participant, “You are suspected of breaking into and stealing confidential information from Professor Aharonovitz’s computer in the Department of Criminology. Please tell me what you were doing during the 30 minutes from the time you left the lab to the time you re-entered the lab. I will turn on the recorder now. When you are ready, please tell me about your activities in as much detail as possible, and do not exclude anything, so I can have an idea of what happened during those 30 minutes. Be sure that you mention all details, activities, people you met, and conversations that took place. Give as much information as you can, including any information that seems irrelevant.” Participant statements were recorded. When the participant finished his or her statement, the interviewer asked if he or she would like to add something, and then turned off the recorder. At this point, participants in the immediate and delayed conditions continued immediately to the third stage of the study. In contrast, the participants in the repeating condition were released and asked to come back to the lab two weeks later, for a second interview (Stage 2b). Stage 2b: second interview. Only participants in the repeating condition provided a second statement. When they returned to the lab two weeks after the first interview, they were asked to provide a second statement regarding the relevant 30 min, before continuing to the third stage of the experiment. The instructions
4
for providing the second statement were identical to those given for providing the first statement. Stage 3: post-interview tasks. At this stage, participants were told that the experiment had ended, that their next responses would not influence the assessment of their statements as true or false, and that they should answer the subsequent questions truthfully. The participants in the immediate and delayed conditions performed the post-interview tasks immediately upon providing their first and only statement (immediately after Stage 2a). The participants in the repeating condition performed the post-interview tasks after they provided their second statement (immediately after Stage 2b). First, each participant listened to the recording of his/her own interview (the participants in the repeating condition listened to the second interview), and identified the false and truthful details that they had provided in their statement. Subsequently, participants completed a post-interview questionnaire, in which they were asked to rate on a 7-point scale from 1 (not at all) to 7 (completely) to what extent they (a) were motivated to appear convincing in their statements; (b) found it difficult to appear convincing in their statements; (c) thought they were convincing in their statements; (d) perceived their likelihood to participate in the drawing; and (e) were interested in receiving the payment bonus. The participants were also asked three yes/no questions to determine whether or not they (a) used a strategy to convince the interviewer of their innocence; (b) provided truthful details; and (c) provided false details. Participants were asked to elaborate on their responses to these questions. Those who reported provision of truthful details were also asked to rate on a 7-point scale (d) to what extent they thought their statement was truthful and (e) to what extent their memory was accurate, and (f) to what extent details were not reported due to forgetfulness. Upon completing the questionnaire, all participants were debriefed and given their course credits. Coding the statements for richness in detail. All statements were coded by two trained, independent coders (blind to the experimental conditions). The coders noted the frequency of occurrence of the relevant RM criteria: (a) perceptual details (information units about what the examinee saw, tasted, heard, felt [by touch], or smelled during their alleged activities); (b) contextual details, including spatial details (information units about locations or the spatial arrangement of people and objects) and temporal details (information units about when the event happened or an explicit description of a sequence of events). The coders initially marked the perceptual and contextual (spatial and temporal) details on the statements and subsequently counted them. For example, the sentence “I entered the Cafeteria at 14:00. A black cat was laying at the entrance” contains six details: I entered the Cafeteria (contextual detail), at 14:00 (contextual detail), black (perceptual detail), cat (perceptual detail), was laying (perceptual detail) at the entrance (contextual detail). Detailed explanation for coding of perceptual and contextual details appears in Nahari (2018). For each statement, the sum of the frequency counts of the perceptual and contextual details was calculated and the resulting variable was labeled richness in detail. One coder coded all the statements and a second one
Please cite this article in press as: Nahari, G. Reality Monitoring in the Forensic Context: Digging Deeper into the Speech of Liars. Journal of Applied Research in Memory and Cognition (2017), https://doi.org/10.1016/j.jarmac.2018.04.003
+Model
ARTICLE IN PRESS REALITY MONITORING IN THE FORENSIC CONTEXT
coded 70% of them. The inter-rater reliability of richness in detail between the coders, measured by the intraclass correlation coefficient (ICC) for consistency, was high and satisfactory (ICC[2,2]= .98, averaged measures). Classifying the type of details. In Stage 3 of the experiment, the participants listened to the recording of their own interview and identified the false and truthful details that they had provided in their statement. Based on their report, the perceptual and contextual details recognized by the coders were classified into truthful and false details. For example, if a participant said in her statement: “I sat on a bench and had a coffee. While drinking it, Zvi approached me.” In this statement there are six details: sat (perceptual detail), on a bench (contextual detail), had a coffee (perceptual detail); while drinking it (contextual detail), Zvi (perceptual detail), approached me (contextual detail). Now, let us consider that the participant reports that this statement is true, except for the fact that Zvi approached her. In such a case, three details should be classified as being true (set, on a bench, had a coffee), and the other three details should be classified as being false (while drinking it, Zvi, approached me).
5
Perception of Difficulty and Success in Being Convincing Liars found it more difficult to appear convincing in their statements (M = 4.20, SD = 1.82) than did truth-tellers (M = 2.47, SD = 1.90), F(1, 135) = 30.36, η2 = .18, p < .001. Liars further assessed their success in being convincing (M = 4.55, SD = 1.40) and their chances of participating in the draw (M = 4.16, SD = 1.70) as lower than truth-tellers (M = 5.78, SD = 1.09 and M = 5.10, SD = 1.60, respectively), F(1, 135) = 34.42, η2 = .20, p < .001 and F(1, 135) = 11.76, η2 = .08, p < .01. There were no differences between time-of-test conditions with respect to any of these measures, and no interaction between veracity and time of test, all F(2, 135)s < 3.02. Perceived Memory In all the time-of-test conditions, truth-tellers perceived their memory as highly accurate (M = 6.47 on 7-point Likert scale, SD = .93) and estimated the number of details that were not reported due to forgetfulness as low (M = 1.93 on 7-point Likert scale, SD = 1.34), F(2, 69) < 1 and F(2, 69) = 2.47, η2 = .07, ns, respectively. The accuracy perceptions of liars are irrelevant, as many of them fabricated the reported memories.
Results Veracity Manipulation Check
Reported Strategy
Significant between-group differences were revealed in the amount of truthful details participants said they reported in their statements, t(36.14) = 7.85, p < .001 (variances of groups [truthful, false] were unequal; t test calculated accordingly). Truth-tellers reported providing more truthful details (M = 6.99 on a 7-point Likert scale, SD = .12) than liars (M = 4.54, SD = 1.89). This indicates that the participants followed the experimenter’s instructions. All truth-tellers said that they provided only truthful details. Fifty-six liars (81.2%) said that they provided false details, with 24 of them providing false details in addition to truthful details. This mimics the real-life behavior of liars, who often embed their lies into truthful statements (Vrij, 2008; Vrij, Granhag, & Porter, 2010). The remaining thirteen liars (18.8%) reported that they did not provide any false details. I checked the statements and responses of these 13 liars, and found that they provided concealed lies (see Nahari et al., 2012). They admitted to being in the department, reported only their non-criminal activities there (e.g., people they saw there, playing with their phone, etc.), and concealed their criminal activities, a strategy that has been found in previous studies (e.g., Nahari et al., 2014a).
More liars (n = 49; 71.0%) than truth-tellers (n = 20; 27.8%) reported using a strategy to appear convincing, χ2 (1) = 26.36, φ = .43, p < .001. This pattern was similar across the three timeof-test conditions.
Participant Motivation The reported motivation of the participants to be convincing in their statements was high (M = 6.02 on a 7-point Likert scale, SD = 1.39). Further, 90.1% of the participants (n = 127) were interested in receiving the payment bonus (ratings of 5, 6, or 7; M = 6.25, SD = 1.44). There was no difference between the experimental conditions (veracity, time of test) with respect to motivation or interest in receiving the payment bonus, all Fs < 2.65.
Testing of Hypotheses Richness in detail. The three predictions of Hypothesis 1 referred to the statement’s richness in detail. To test this hypothesis, I compared the richness in detail of statements provided by truth-tellers and liars in the three time-of-test conditions. Note that participants in the repeating condition provided two statements. However, since the hypothesis refers only to their second statement, their first statement was excluded from the comparison.1 Thus, richness in detail of the second statement in the repeating group was compared with the single statements provided by participants in the immediate condition and the delayed condition. Since number of words can be related to, and therefore confound, richness in detail, I first conducted an ANOVA with number of words as the dependent variable and veracity and time of test as factors. This analysis revealed a statistically significant time-of-test effect, F(2, 135) = 5.38, partial η2 = .07, p < .01. A post hoc test (Scheffe) revealed that participants in
1 I performed a repeated measures ANCOVA to compare the richness in detail of the two statements provided in the repeating condition. The statement (first, second) was the within-subject factor and veracity group (truth-tellers, liars) was the between-subject factor. The number of words of the first statement was entered as a covariate. Neither the effect of statement nor the Statement × Veracity interaction were significant, F(1, 44) = 1.76, partial η2 = .04, ns, and F < 1 respectively. This mirrors the finding that statements in the immediate and repeating conditions did not differ in terms of richness in detail.
Please cite this article in press as: Nahari, G. Reality Monitoring in the Forensic Context: Digging Deeper into the Speech of Liars. Journal of Applied Research in Memory and Cognition (2017), https://doi.org/10.1016/j.jarmac.2018.04.003
+Model
ARTICLE IN PRESS REALITY MONITORING IN THE FORENSIC CONTEXT
6
Table 2 Comparison of richness in detail (estimated marginal Ms and SDs) across veracity and time-of-test conditions Immediate condition
Delayed condition
Repeating condition
Truth-tellers
23.77 (7.01)
16.43 (6.81)
20.68 (6.81)
Liars
11.33 (6.86)
9.99 (6.86)
14.42 (6.91)
Veracity condition comparison
F(1,44) = 33.39 partial η2 = .43 P < .001
F(1,44) = 21.42 partial η2 = .33 P < .001
F(1,44) = 7.87 partial η2 = .15 P < .01
the immediate condition (M = 165.64, SD = 111.66) provided longer statements than participants in the delayed (M = 121.55, SD = 65.30) and repeating (M = 112.79, SD = 63.74) conditions. The main effect of veracity and the Veracity × Time-of-test interaction were not significant, F(1, 135) = 2.10, partial η2 = .02, ns, and F < 1, respectively. I then conducted an ANCOVA with richness in detail as the dependent variable, and veracity and time of test as factors. As noted above, the length of statements differed among the experimental groups. Thus, as in previous studies (e.g., Nahari & Pazuelo, 2015), to ensure that between-group differences in richness in detail were not simply a function of the amount of information provided overall, number of words was inserted as a covariate, which was found to be significant, F(1, 134) = 105.99, partial η2 = .44, p < .001. This analysis revealed a statistically significant time-of-test effect, F(2, 134) = 6.29, partial η2 = .09, p < .01, indicating that participants in the immediate condition (M = 17.49, SD = 6.99) and in the repeating condition (M = 17.55, SD = 6.92) provided more details than did participants in the delayed condition (M = 13.21, SD = 6.79). The participants in the immediate and repeating conditions did not differ in richness in detail. The veracity effect, F(1, 134) = 52.51, partial η2 = .28, p < .001, was also significant. Truth-tellers (M = 20.29, SD = 6.87) provided more details than liars (M = 11.91, SD = 6.81). Finally, a Time-of-test × Veracity interaction was observed, F(2, 134) = 3.12, partial η2 = .04, p < .05. To locate the source of this interaction, tests of simple effects were conducted. As shown in Table 2, truth-tellers provided more details than liars in each of the three time-of-test conditions. However, differences across the time-of-test conditions were observed among truthtellers but not among liars. A post hoc test (Scheffe) revealed that truth-tellers provided more details in the immediate condition than in the delayed condition, and similar amount of details in the immediate and repeating conditions. Liars provided the same amount of details in all three conditions. To summarize, as predicted in Hypothesis 1a, differences between truth-tellers and liars were significant for all time-oftest conditions. In accordance with Hypothesis 1b, truth-tellers provided more details in the immediate condition than in the delayed condition, and thus the differences between truthtellers and liars were greater in the immediate condition (partial η2 = .43) in comparison to the delayed condition (partial η2 = .33). However, contrary to Hypothesis 1c, the amount
Time-of-test condition comparison F(2,68) = 4.85 partial η2 = .13 P < .05 F(2,65) = 2.53 partial η2 = .07 ns
of details provided by truth-tellers and liars did not decrease in the second statement, and differences between truth-tellers and liars were not greater in the repeating condition than in the delay condition. In fact, these differences were smaller in the repeating condition than in the other two conditions (partial η2 = .15). Thus, the first hypothesis was only partially confirmed. Type of details. The three predictions of Hypothesis 2 referred to the type of details (truthful or false) found in statements. To examine these predictions, a repeated measures ANOVA was performed, in which the type of details (truthful, false) was the within-subject factor and veracity and time of test were the between-subject factors. Table 3 presents the means (and standard deviations) for truthful and false details across the experimental conditions. A main effect was found for type of details, F(1, 135) = 98.70, partial η2 = .42, p < .001, showing that participants provided more truthful details (M = 12.64, SD = 13.08) than false details (M = 3.55, SD = 5.65). A Details × Time-of-test interaction, F(2, 135) = 4.50, partial η2 = .06, p < .05, as well as a strong Details × Veracity interaction were found, F(1, 135) = 170.64, partial η2 = .56, p < .001. Finally, the triple interaction was marginally significant, F(2, 135) = 2.96, partial η2 = .04, p = .05. Simple effects showed different patterns for truth-tellers and liars. First, paired-sample t tests revealed that the statements of truth-tellers included more truthful details than false details in all conditions, t(23) = 10.97, d = 2.24, p < .001 in the immediate condition, t(23) = 12.11, d = 2.47, p < .001 in the delayed condition, and t(23) = 7.04, d = 1.44, p < .001 in the repeating condition. The amount of false details in truth-tellers’ statements was negligible (means ranged between .04 and .46 details). In contrast, the statements of liars included more false details than truthful details in the delayed condition, t(22) = −2.16, d = .48, p < .05, but the same amount of false and truthful details in the immediate and repeating conditions, t(22) = −.65, d = .13, ns, and t(22) = −1.87, d = .42, ns respectively. Second, one-way ANOVAs showed that truth-tellers provided different amount of truthful details across conditions, F(2, 69) = 6.69, partial η2 = .16, p < .01, and the same, negligible, amount of false details in all conditions, F(2, 69) = 2.68, partial η2 = .07, ns. A post hoc test (Scheffe) revealed that they provided more truthful details in the immediate condition than in the delayed condition, and the same amount of truthful details in the immediate and repeating conditions. In all conditions, liars provided the same amount of both truthful
Please cite this article in press as: Nahari, G. Reality Monitoring in the Forensic Context: Digging Deeper into the Speech of Liars. Journal of Applied Research in Memory and Cognition (2017), https://doi.org/10.1016/j.jarmac.2018.04.003
+Model
Truth-tellers Truthful details False details
26.96 (11.90) 0.33 ( 0.64)
Liars
Delayed condition Overall
5.35 (7.33 ) 16.38 (14.69) 7.09 (7.98) 3.64 ( 6.51)
Truth-tellers 15.42 (5.69) 0.46 ( 0.88)
Liars 3.00 (3.07 ) 5.83 ( 4.99)
Repeating condition Overall
Truth-tellers
9.34 (7.75) 3.09 ( 4.43)
19.83 (13.81) 0.04 ( 0.20)
Liars
Across conditions Overall
4.21 (5.28 ) 12.19 (13.08) 8.00 ( 5.74) 3.94 ( 5.65)
Truth-tellers 20.74 (11.88) 0.28 ( 0.65)
Liars
Overall
4.19 (5.51 ) 12.64 (12.46) 3.55 ( 5.57) 6.97 ( 6.34)
ARTICLE IN PRESS
Immediate condition
REALITY MONITORING IN THE FORENSIC CONTEXT
Please cite this article in press as: Nahari, G. Reality Monitoring in the Forensic Context: Digging Deeper into the Speech of Liars. Journal of Applied Research in Memory and Cognition (2017), https://doi.org/10.1016/j.jarmac.2018.04.003
Table 3 Means (and standard deviations) of truthful and false details across veracity and time-of-test conditions
7
+Model
ARTICLE IN PRESS REALITY MONITORING IN THE FORENSIC CONTEXT
and false details, F < 1 and F(2, 66) = 1.05, partial η2 = .03, ns, respectively. To summarize, the predictions of Hypothesis 2 were only partially confirmed. In accordance with Hypothesis 2a, truth-tellers provided more truthful than false details, but contrary to this hypothesis, liars provided more false details than truthful details only in the delayed condition. In the immediate and repeating conditions, liars provided the same amount of truthful and false details. In contrast with Hypothesis 2b, liars provided the same amount of false details in all three conditions. Finally, in accordance with Hypothesis 2c, truth-tellers provided more truthful details in the immediate condition than in the delayed condition. However, contrary to this hypothesis, truth-tellers provided the same amount of truthful details in the immediate and repeating conditions. Discussion The current study examined the effect of statement time and repetition on the provision of perceptual and contextual details. In accordance with the RM approach, truth-tellers provided more perceptual and contextual details than liars in all the experimental conditions. I also found that most of the liars added false details to their accounts while truth-tellers stuck with the truth. One may wonder why liars, who apparently could add whatever details they chose, did not reach the same level of detail as did truth-tellers. One suggestion, drawn from the Criteria-Based Content Analysis (CBCA) approach, is that some details are difficult for liars to make up (Köhnken, 1996; Kohnken, 2004). Alternatively, the cognitive load approach proposes that sometimes limitations in cognitive resources may reduce the ability of liars to provide many details (Vrij et al., 2008). A different perspective, however, starts with the observation that liars do not lack imagination (Merckelbach, 2004), and that many liars base their lies on an event that they have actually experienced, albeit at another time (Leins, Fisher, & Ross, 2013), or embed truthful details in their lies (Nahari & Vrij, 2015). Furthermore, there is evidence that liars are able to provide as many details as truth-tellers when encouraged to do so (Nahari & Pazuelo, 2015). Perhaps, therefore, the contrast in detail between truth-tellers and liars is better explained by the verifiability approach (Nahari et al., 2014a, 2014b). According to this strategy-based approach, liars are careful in providing verifiable details, which can be checked by the investigators. That is, even if liars have a rich imagination and even if they discuss previous experiences, they are limited in the number of details that they can provide because of their motivation to keep their accounts unverifiable and difficult to refute. In the current study, differences between liars and truth-tellers were greater when their statements were provided immediately as opposed to following a delay. Truth-tellers, who stuck with the truth, provided more details in the immediate condition than in the delayed condition, presumably due to lack of memory. Liars, who did not suffer from lack of memory, provided the
8
same number of details in both conditions. Interestingly, liars provided a similar number of truthful and false details in the immediate condition, but more false than truthful details in the delayed condition. This implies that liars, as well as truthtellers, had difficulty providing truthful details after a delay. However, in contrast to truth-tellers, liars compensated for this by adding false details. In this way, the difference between liars and truth-tellers decreased when the statement was provided following a delay. The implications of these findings are significant, since in real life, most of the time suspects are interviewed days, weeks, or even years after the crime was committed. In contrast to our expectations, both truth-tellers and liars provided the same number of details in the immediate and the repeating conditions (second statement). That is, the participants in the repeating condition did not show lack of memory when repeating their first statements. One possible explanation is related to the testing effect, which predicts that retrieval practice benefits later memory, even at long delays (see McDermott, Arnold, & Nelson, 2014). Truth-tellers in the repeating condition provided their first statements immediately after the event. This act can be considered retrieval practice or testing, which may have had a beneficial effect on memory of the event when they provided the second statement. However, this explanation is not relevant for liars. Since their memories were fabricated while they provided their first statements, this stage cannot be considered memory practice. It is more likely that liars practiced during the interval between the statements, in order to be consistent. Liars knew that to avoid being discovered, they had to remember what they said in their previous statements (Granhag & Strömwall, 1999). It is thus reasonable to assume that they made an effort to remember what they had said. One point merits attention. I examined the hypotheses on the amount of perceptual and contextual details combined (i.e., richness in detail). Although the reality-monitoring theory does not provide differential predictions for each type of details, one may argue that it is worth further examining the hypotheses for each criterion separately. The sample was too small to allow analyzing perceptual and contextual details separately, and thus a future study is required for this purpose. To conclude, the current study demonstrates that the manipulation factor in verbal behavior of liars is key to the adjustment and improvement of RM as a tool for lie detection, and that for better understanding of this manipulation factor, one should look also at the veracity (false or truthful) of details in different conditions (statement time and repetition). This line of reasoning requires more research. Conflicts of Interest Statement The author declares no conflict of interest. Author Contributions Single author. Responsible for all manuscript components.
Please cite this article in press as: Nahari, G. Reality Monitoring in the Forensic Context: Digging Deeper into the Speech of Liars. Journal of Applied Research in Memory and Cognition (2017), https://doi.org/10.1016/j.jarmac.2018.04.003
+Model
ARTICLE IN PRESS REALITY MONITORING IN THE FORENSIC CONTEXT
Acknowledgment This research was supported by the Israel Science Foundation grant (grant No. 372/14). References DePaulo, B. M., Lindsay, J. J., Malone, B. E., Muhlenbruck, L., Charlton, K., & Cooper, H. (2003). Cues to deception. Psychological Bulletin, 129, 74–118. Granhag, P. A., & Hartwig, M. (2008). A new theoretical perspective on deception detection: On the psychology of instrumental mindreading. Psychology, Crime & Law, 14, 189–200. Granhag, P. A., & Strömwall, L. A. (1999). Repeated interrogations – Stretching the deception detection paradigm. Expert Evidence, 7, 163–174. Granhag, P. A., & Strömwall, L. A. (2002). Repeated interrogations: Verbal and non-verbal cues to deception. Applied Cognitive Psychology, 16, 243–257. Hartwig, M., Granhag, P. A., & Strömwall, L. A. (2007). Guilty and innocent suspects’ strategies during police interrogations. Psychology, Crime & Law, 13, 213–227. Harvey, A. C., Vrij, A., Hope, L., Leal, S., & Mann, S. (2017). A stability bias effect among deceivers. Law and Human Behavior, 41, 519–529. Harvey, A. C., Vrij, A., Leal, S., Hope, L., & Mann, S. (2017). Deception and decay: Verbal lie detection as a function of delay and encoding quality. Journal of Applied Research in Memory and Cognition, 6, 306–318. Johnson, M. K. (2006). Memory and reality. American Psychologist, 61, 760–771. Johnson, M. K., Bush, J. G., & Mitchell, K. J. (1998). Interpersonal reality monitoring: Judging the sources of other people’s memories. Social Cognition, 16, 199–224. Johnson, M. K., & Raye, C. L. (1981). Reality monitoring. Psychological Review, 88, 67–85. Köhnken, G. (1996). Social psychology and the law. In G. R. Semin, & K. Fiedler (Eds.), Applied social psychology (pp. 257–282). London, UK: Sage Publications. Kohnken, G. (2004). Statement validity analysis and the detection of the truth. In P. A. Granhag, & L. A. Stromwall (Eds.), The detection of deception in forensic contexts (pp. 41–63). Cambridge: Cambridge University Press. Leins, D. A., Fisher, P. A., & Ross, S. J. (2013). Exploring liars’ strategies for creating deceptive reports. Legal and Criminological Psychology, 18, 141–151. Masip, J., & Herrero, C. (2013). ‘What would you say if you were guilty?’ Suspects’ strategies during a hypothetical behavior analysis interview concerning a serious crime. Applied Cognitive Psychology, 27, 60–70. Masip, J., Sporer, S. L., Garrido, E., & Herrero, C. (2005). The detection of deception with the reality monitoring approach: A review of the empirical evidence. Psychology, Crime & Law, 11, 99–122.
9
McDermott, K. B., Arnold, K. M., & Nelson, S. M. (2014). The testing effect. In T. J. Perfect, & D. S. Lindsay (Eds.), The SAGE handbook of applied memory (pp. 183–200). Thousand Oaks, CA: SAGE Publications. Merckelbach, H. (2004). Telling a good story: Fantasy proneness and the quality of fabricated memories. Personality and Individual Differences, 37, 1371–1382. Nahari, G. (2018). The applicability of the verifiability approach to the real world. In P. Rosenfeld (Ed.), Detecting concealed information and deception: Recent developments (pp. 329–349). London: Elsevier. Nahari, G., & Ben-Shakhar, G. (2011). Psychophysiological and behavioral measures for detecting concealed information: The role of memory for crime details. Psychophysiology, 48, 733–744. Nahari, G., & Pazuelo, M. (2015). Telling a convincing story: Richness in detail as a function of gender and information. Journal of Applied Research in Memory and Cognition, 4, 363–367. Nahari, G., & Vrij, A. (2015). Can someone fabricate verifiable details when planning in advance? It all depends on the crime scenario. Psychology, Crime & Law, 21, 987–999. Nahari, G., Vrij, A., & Fisher, R. P. (2012). Does the truth come out in the writing? Scan as a lie detection tool. Law and Human Behavior, 36, 68–76. Nahari, G., Vrij, A., & Fisher, R. (2014a). Exploiting liars’ verbal strategies by examining unverifiable details. Legal and Criminological Psychology, 19, 227–239. Nahari, G., Vrij, A., & Fisher, R. (2014b). The verifiability approach: Countermeasures facilitate its ability to discriminate between truths and lies. Applied Cognitive Psychology, 28, 122–128. Sporer, S. L. (1997). The less travelled road to truth: Verbal cues in deception detection in accounts of fabricated and self-experienced events. Applied Cognitive Psychology, 11, 373–397. Sporer, S. L. (2004). Reality monitoring and detection of deception. In P. A. Granhag, & L. A. Stromwall (Eds.), The detection of deception in forensic contexts (pp. 64–102). UK: Cambridge University Press. Vrij, A. (2005). Criteria-based content analysis: A qualitative review of the first 37 studies. Psychology, Public Policy, and Law, 11, 3–41. Vrij, A. (2008). Detecting lies and deceit: Pitfalls and opportunities (2nd ed., pp. 1–502). John Wiley & Sons. Vrij, A., Granhag, P. A., & Porter, S. B. (2010). Pitfalls and opportunities in nonverbal and verbal lie detection. Psychological Science in the Public Interest, 11, 89–121. Vrij, A., Mann, S. A., Fisher, R. P., Leal, S., Milne, R., & Bull, R. (2008). Increasing cognitive load to facilitate lie detection: The benefit of recalling an event in reverse order. Law and Human Behavior, 32, 253–265.
Received 4 July 2017; received in revised form 5 April 2018; accepted 5 April 2018 Available online xxx
Please cite this article in press as: Nahari, G. Reality Monitoring in the Forensic Context: Digging Deeper into the Speech of Liars. Journal of Applied Research in Memory and Cognition (2017), https://doi.org/10.1016/j.jarmac.2018.04.003