Eliciting Veracity Differences Through a General Memory Recall Test

Method

Ethics

The experiment received ethics approval by the Faculty ethics committee of the university of the first author [C-2022-002].

Participants

A power analysis using G*Power software revealed that at least 131 participants are required to obtain an 95% statistical power, an alpha level of .05, and a medium to large effect size (f2 = .09). A total of 131 participants took part in the experiment of whom 42 were males and 87 were females. Two participants did not say. The average age in the sample was M = 26.77 (SD = 1.35). Most par- ticipants (n = 60) identified themselves as white British, followed by Asian (n = 21), white European (n = 17), African (n = 8), mixed (n = 8), other (n = 7), Arab (n = 5), Black British (n = 4), and Black European (n = 1).

Allocation to the Veracity and Sketch conditions occurred ran- domly: 65 participants were allocated to the truth condition and 66 to the lie condition; 66 to the Sketch-absent condition and 65 to the Sketch-present condition. The sample sizes in each of the four cells ranged from 32 to 33.

Design

Data were analysed utilising a Veracity [truth vs lie] X Sketch [present vs absent] between-subjects design. The following five dependent variables were included in the analysis: Total details, number of tricks accurately reported, number of tricks correctly mentioned in the sequence, number of tricks mentioned in the cor- rect order, and number of self-handicapping comments

Procedure

Participants were recruited via the department’s database [stu- dents, staff, members of the general public interested in taking part in research] and university online platforms. They were informed that they will be watching a video and will tell the truth or lie about it in an interview. The experiment was carried out online via Zoom and each participant was interviewed individually. At least 24 hours prior to their appointment, participants were sent the participant information sheet and consent form via Qualtrics, and the Zoom link to their appointment via email. During their appointment, participants were given the opportunity to ask the experimenter any questions they have, after which they gave their written consent.

First, participants watched a 6.30 min video about a mock secret meeting. In the meeting, a male agent presents to two other female agents the characteristics of a spying device that will be planted in a smoke detector to gather intelligence about a hostile agent’s activities. The presenter then asks the two agents to vote on one of two locations international school or chemical factory] to plant the device. The agents decide to plant it in a lecture room in the international school.

After watching the secret meeting video, the experimenter ran- domly allocated participants to the truth teller or lie teller condi- tion. The participants were led to believe that they would be inter- viewed about the secret meeting video. Truth tellers were told the interviewer could be trusted and were asked to tell the truth about the secret meeting video. Lie tellers were told that the interviewer could not be trusted and could be a double agent. They were there- fore asked to lie about the spying device and the location where it will be planted to protect the mission. The video was also used by [20]. Participants were then given as much time as they needed to prepare for the interview. Truth tellers and lie tellers were never asked to recall this secret meeting video in the interview. Instead, they were asked to recall the tricks video truthfully but were not told that at this stage. In other words, the lie tellers were not asked to lie at all in the interview. Yet, we call them ‘lie tellers’ because they were led to believe that they had to lie in the interview about the secret meeting video. The hypotheses predict that this will im- pact upon their verbal recall of the tricks video. When participants informed the experimenter that they were ready, they were sent the Qualtrics link to a pre-interview questionnaire that measured their measured background characteristics [age, gender, ethnicity]. Af- ter completing the questionnaire, they were given a link to a Zoom room where the interviewer was waiting.

The Interview

The interviewer was a research assistant who was blind to the par- ticipants’ veracity condition and research hypotheses. The inter- viewer instructed the participants as follows: “I understand you saw a video depicting a secret meeting. Before I ask you anything about the secret meeting video, I need to first test your general memory recall. I will show you a tricks video after which you will be asked to report everything that you saw in that tricks video as detailed and as accurately as possible.” Participants were then shown a 10-min video of two physiotherapists describing seven tricks to make an old mattress more comfortable (https://www.youtube.com/watch?v=NJrA7obfZzA&ab_chan- nel=Bob%26Brad). We chose this tricks video because it provides enough detail, un- likely to have been seen by participants before [indeed participants were not familiar with it, see Results] and presents a clear sequence of actions [order in which the seven tricks are presented]. Report- ing the tricks in the order in which they were presented should come natural to participants because the natural way of reporting information is sequentially [6,21]. This gave us the opportunity to introduce two accuracy measures [not only the tricks mentioned but also the order in which they were mentioned].

After watching the tricks video, participants in the Sketch-absent condition received the following instructions: “Now please tell me in as much detail and as accurate as possible everything in the tricks video you just watched. You may take as long as you need to recall the content.” Participants in the Sketch-present condition were instructed as follows: “Now please tell me in as much detail and as accurate as possible everything in the tricks video you just watched. While doing so, I would like you to sketch on a white A4 sheet of paper everything you saw in the video. Thus, you need to sketch and talk at the same time. You may use more than one sheet of paper if needed, and you may take as long as you need to recall the content.” All participants in the Sketch-present condition had blank A4 sheet of paper.

When participants finished recalling the tricks video, the inter- viewer told them that the interview was completed [they were never asked to recall the secret meeting video]. The interviewer redirected the participants to the experimenter who sent them a Qualtrics link to the post-interview questionnaire that they were asked to answer truthfully.

Post-Interview Questionnaire and Debrief

The post-interview questionnaire included statements about par- ticipants paying attention to the tricks video, their motivation to recall details about the tricks video, the extent to which they found themselves detailed when reporting the tricks video and the extent to which they thought they were accurate when reporting the tricks video. All statements could be rated on 7-point Likert scales rang- ing from 1 [not at all] to 7 [very much]. We also asked about their familiarity with the tricks video on a 7-point scale ranging from 1 [not at all] to 7 [very much].

Participants in the sketch condition were further asked to rate the extent to which they found sketching difficult [on a 7-point Likert scale from 1 = not at all to 7 = very dificult], and the extent to which they thought it helped them recall details from the tricks video [on a 7-point Likert scale from 1 = not at all to 7 = very much].

After completing the questionnaire, the experimenter told partici- pants that the experiment was over and explained the true purpose of the research. Participants were then thanked and remunerated with £10.

Coding

The interviews were transcribed and coded. One rater, blind to the Veracity conditions and hypotheses, was taught the coding scheme by the third author who designed most of the coding scheme [for all variables except total details and self-handicapping comments] specifically for this experiment.

Total number of details was the sum of all nouns, verbs adjec- tives and adverbs. Details were counted regardless of accuracy and repetitions were ignored. The word ‘video’ was not counted. The following sentence contains eight details: “On that video I saw two men, two American men, discussing seven tricks that help you to sleep”. The following sentence contains six details: “The second man kept a pillow between his knees”

The rater counted the number of tricks accurately reported. This variable could range from 0 to 7. The tricks in the video were given sequential numbers [trick 1, 2, 3 etc.] representing the order in which they were introduced in the video:

Trick 1: Board or pillow under mattress

Trick 2: Sheet wrapped around lower back

Trick 3: Pillow between legs and/or feet

Trick 4: Arm canal for shoulder pain with two pillows beneath head and a throw pillow beneath chest

Trick 5: Two pillows between arms for upper sore shoulder

Trick 6: Leg wedge for back pain with pillow beneath legs

Trick 7: Throw pillow beneath mid-back

Participants received one point for each trick reported using the correct sequential number. This variable, labelled tricks sequence matched, could range from 0 to 7. We also introduced the tricks sequence general variable, which could also range from 0 to 7. For this variable, the general sequence order is relevant rather than the exact sequence number. For example, participants who say that Trick 1 was a board under the mattress, Trick 2 was a pillow be- tween legs, and Trick 3 was a leg wedge would be given a score of 1 for the tricks sequence matched variable as only Trick 1 matches the sequential number above. However, they would receive a score of 3 for the tricks sequence general variable as the order of tricks was accurately reported.

Finally, every justification given for not being able to provide in- formation was coded as a self-handicapping strategy comment. Examples are “I forgot because there were so many tricks”, “My memory went a bit blank” and “I can’t remember the tricks in or- der, I was too fascinated with the two presenter guys”. A second rater coded a random sample of 60 transcripts. Inter-rater reliability between the two coders, using the two-way random ef- fects model measuring consistency, was very good for total details [Single Measures ICC = .76], total number of tricks accuracy [Sin- gle Measures ICC = .88], tricks sequence matched [Single Mea- sures ICC = .83], tricks sequence general [Single Measures ICC = .82] and self-handicapping comments [Single Measures ICC = .95].

Results

Familiarity with Memory Recall (Tricks) Video

A 2 [Veracity: Truth vs lie] X 2 [Sketch: Present vs absent] ANO- VA was carried out with familiarity with the tricks video as de- pendent variable. The Veracity main effect, F[1, 127] = 0.02, p = .893, d = .03, 95% CI[-0.32,0.37], Sketch main effect, F[1, 127] = 2.73, p = .101, d = .28, 95% CI[-0.07,0.62] and Veracity X Sketch interaction effect F[1, 127] = 2.73, p = .101, η 2 = .02, were not significant. The grand mean score reveals that the tricks video was largely unknown [M = 1.84, SD = 1.57] amongst participants.

Questionnaire Variables

A 2 [Veracity: Truth vs lie] X 2 [Sketch: Present vs absent] MANO- VA was carried out with as dependent variables [all measured on 7-point Likert scales]: Attention to the tricks video, motivation to report details about the tricks video, detailedness in reporting the tricks video and accuracy in reporting the tricks video. At a mul- tivariate level the analysis revealed no significant main effects for Veracity, F [4, 124] = 0.09, p = .985, η _p2 = .003 and Sketch, F [4, 124] = 0.20, p = .938, η_p 2 = .006. Also, the Veracity X Sketch interaction effect was not significant, F [4, 124] = 0.40, p = .809, η_p 2 = .01.

The mean scores for the total sample showed that participants re- ported that they were attentive when watching the tricks video [M = 5.14, SD = 1.84], motivated to report details about the tricks video [M = 5.44, SD = 1.55], and thought to be reasonably detailed [M = 4.63, SD = 1.55], and reasonably accurate [M = 4.93, SD = 1.39] in reporting the tricks video.

Participants in the Sketch-present condition were asked the extent to which they found the sketching task difficult and the extent they thought it helped them in reporting details about the tricks vid- eo. Two ANOVAs with Veracity as the only factor and those two variables as dependent variables did not reveal a significant effect, both Fs < 2.86, both ps > .095. The average means showed that the sketching task was judged to be moderately difficult [M = 4.29, SD = 1.78] and had helped them to a moderate extent to report details [M = 3.87, SD = 1.82].

Hypotheses-Testing

A 2 [Veracity: Truth vs lie] X 2 [Sketch: Present vs absent] MANOVA was carried out with total details reported about the tricks video, number of tricks accurately reported, number of tricks correctly mentioned in the sequence, number of tricks mentioned in the correct order, and number of self-handicapping comments as dependent variables. At a multivariate level, the Veracity main effect, F [5, 123] = 1.86, p = .106, η_p 2 = .07, the Sketch main effect, F [5, 123] = 1.79, p = .120, η_p 2 = .07 and the Veracity X Sketch interaction effect F [5, 123] = 1.18, p = .332, η_p 2 = .05 were not significant.

Despite the absence of multivariate effects, we believe that univar- iate testing is appropriate. First, some of the dependent variables were highly correlated, particularly number of tricks accurately reported with (i) accuracy in tricks sequence matched, r [131] = .728, p < .001 and (ii) accuracy in tricks sequency general, r [131] = .824, p < .001. When dependent variables are correlated the multivariate statistics can become unreliable and may identify effects smaller than regular ANOVAs can find [22]. Second, it is worthwhile to examine the effects of the factors on specific vari- ables. For example, although previous research has shown that the instruction to sketch results in interviewees reporting more details, it is yet unknown whether they also do this more accurately. The univariate main effects for Veracity and Sketch are presented in Tables 1 and 2.

                                   Table 1: Questionnaire Variables and Memory Cues as a Function of Veracity

Note. NHST = Null-hypothesis significance testing.

Table 2: Questionnaire Variables and Memory Cues as a Function of Sketching

Note. NHST = Null-hypothesis significance testing.

Only one significant Veracity effect emerged. Truth tellers report- ed the order of tricks more accurately than lie tellers. Given that all other Veracity effects were not significant, we conclude that Hypothesis 1 was not supported. Two significant Sketch effects emerged. Sketching resulted in interviewees providing more de- tails. It also resulted in more tricks being accurately reported. This supports Hypothesis 2, particularly in terms of the amount of in- formation provided.

No significant Veracity X Sketch interaction effect was obtained at a univariate level, all Fs < 2.38, all ps > .125, but these effects refer to any type of interaction. As we predicted a directional effect with specific group differences, a more informative test of Hypothesis 3 is to statistically test for significant differences between truth tell- ers and lie tellers in each of the two Sketch conditions and to com- pare the d-values in the two sets of analyses [6,23,24] Comparing d-values between the Sketch-present and Sketch-absent conditions is important. A p-value provides information about the statistical relevance but not about the practical importance of an effect [25]. In this article, we are interested in the practical relevance of the effect and d-values are indicators of practical relevance [26].

            Table 3: Memory Cues as a Function of Veracity for Sketch Absent and Sketch Present Conditions

        Note. NHST = Null - Hypothesis Significance Testing

Two MANOVAs were conducted with Veracity as the only factor and the five variables mentioned in Table 3 as dependent variables, one MANOVA for the Sketch-absent condition and one MANO- VA for the Sketch-present condition. The multivariate effect in the MANOVA for the Sketch-absent condition was not significant, F [5, 60] = 0.20, p = .963, η_p 2 = .02. At a univariate level, none of the effects were significant either [all ps > .735] and all effect sizes were very small [all d’s < 0.09]. In contrast, the multivariate effect in the MANOVA for the Sketch-present condition was significant, F[5, 59] = 3.26, p = .012, η_p 2 = .22 and at a univariate level, two significant effects with medium effect sizes emerged [with a com- bined average d = 0.57]. Truth tellers were more accurate than lie tellers in reporting the tricks and the general sequence in which they occurred. This provides support for Hypothesis 3.

Discussion

In this article we reasoned that lie tellers would perform worse than truth tellers on the general memory test [i.e. recalling the tricks video], either due to self-handicapping or due to a lack of focus during the general memory test. Indeed, lie tellers did per- form worse on the general memory test than truth tellers, albeit only in the Sketch-present condition. In that condition, lie tellers were less accurate than truth tellers in reporting the tricks and the general sequence in which they occurred.

Our results did not clarify what caused the effect. Lie tellers did not make more self-handicapping comments [justifications why they could not report information] than truth tellers when recalling the tricks video. The number of self-handicapping comments was low so perhaps this represents a floor effect. It may not be surpris- ing that few self-handicapping comments were made. Interview- ees had watched the tricks video just before the memory recall task and may had difficulty thinking about justifications why they would not be able to recall the video. The information the partic- ipants provided in the post-interview questionnaire [self-reports] did not show evidence of self-handicapping either. That is, truth tellers and lie tellers reported to have been equally detailed and equally accurate when recalling the tricks video. However, self-re- ports are known to be somewhat unreliable [27].

An alternative explanation for lie tellers performing more poorly on the memory test than truth tellers is that lie tellers were more focussed than truth tellers on what they thought would be the task ahead [reporting the secret video], and thus put less effort in recall- ing the tricks video. However, lie tellers and truth tellers self-re- ported to be equally motivated to perform well in the recall of the tricks video and to have been equally attentive in watching the tricks video. Of course, practically it does not matter what exactly causes the effect [self-handicapping, lack of focus, a combination of the two or something else] if the result remains the same [lie tellers perform poorer than truth tellers on the general memory test]. Theoretically it is more relevant. Identifying the cause of the effect provides insight into truth tellers’ and lie tellers’ strategies or mental states and possible differences between them. Interview protocols could subsequently be developed aimed at exploiting these differences [28,29]. Future research could attempt to identify the exact cause of the effect. In such research underlying factors should be examined experimentally rather than through self-re- ports as we did in the current experiment, due to the unreliability of self-reports [27].

We predicted that sketching would enhance the veracity differenc- es in the memory recall test, which is indeed what we found. We found that sketching was a necessary requirement to obtain verac- ity differences because they only occurred in the Sketch-present condition. Introducing instructions such as sketching is called an active interview protocol, in contrast to a passive interview pro- tocol in which an interviewee is just asked to recall the event in as much detail as possible [28]. Our findings suggest that a more active interview protocol is required to elicit veracity differences in memory recall tests. It is a common finding in the verbal decep- tion literature that active interview protocols are required to elicit veracity differences [30-33].

One of the main problems in verbal veracity assessment is that cut- off scores cannot be established. Large individual [e.g. differences in being talkative or memory] and situational differences [e.g. top- ic of investigation] make it impossible to establish unique lie tell- ing and unique truth telling response patterns [34,35,29]. Cut-off scores are helpful when someone must make a veracity decision in an individual case, a situation that practitioners often face. A memory recall test as a method to detect deceit has the potential for establishing cut-off points. That is, data sets could be developed about how much information truth tellers and lie tellers recall in a specific memory recall test. An individual response can then be compared with the distributions of truth tellers and lie tellers in the data set and could give an idea of the likelihood that someone is a truth teller or a lie teller. Or in other words, if someone reports little information when describing their experiences in the investi- gative part of an interview, one possibility is that the interviewee is lying. The memory test could shed light on this. We acknowledge that this train of thought, which is based on one single experiment, is premature. More research should be carried out first: We need to (1) conclusively demonstrate that lie tellers indeed perform worse than truth tellers in memory tests; (2) examine which factors cause the effect and (3) examine the effect of introducing a memory test on the recall of the main event later in an interview. Hopefully this article will encourage researchers to carry out such research.

We further found that the request to sketch resulted in interview- ees reporting more information. This supports the growing body of research that sketching benefits reporting visual experiences in adults [18,19,36,37,]. This finding has been replicated with chil- dren [38-40]. A relevant question is whether the increase in quan- tity of information occurs at the expense of the accuracy of that information [13]. Obtaining more information is more beneficial if it does not negatively affect accuracy. Our results showed that increased quantity did not negatively affect accuracy, demonstrat- ing a real beneficial effect of the sketching instruction to obtain information.

Data Availability Request

The data that support the findings of this study are available from the corresponding author upon reasonable request.

Author Contributions

The first author designed the experiment [together with the sec- ond author], analysed the data [together with the third author] and wrote the initial draft [together with the third author]. The second author came up with the initial idea, designed the experiment [to- gether with the first author] and commented on the initial draft. The third author analysed the data [together with the first author], wrote the initial draft [together with the first author] and set up the experiment. The fourth author commented on the initial draft.

Conflict of interest

On behalf of all authors, the corresponding author states that there is no conflict of interest.

Funding Declaration

There was no external funding for carrying out the experiment.

References

1. Granhag, P. A., & Hartwig, M. (2008). A new theoretical perspective on deception detection: On the psychology of in-strumental mind-reading. Psychology, Crime & Law, 14(3),189-200.
2. Hartwig, M., Granhag, P. A., Stromwall, L. A., & Doering,N. (2010). Impression and information management: On the strategic self-regulation of innocent and guilty suspects. The Open Criminology Journal, 3(1).
3. Nahari, G., Vrij, A., & Fisher, R. P. (2014). Exploiting liars' verbal strategies by examining the verifiability of details. Le- gal and Criminological Psychology, 19(2), 227-239.
4. Granhag, P. A., & Strömwall, L. A. (1999). Repeated interro- gations–stretching the deception detection paradigm. Expert Evidence, 7, 163-174.
5. Strömwall, L. A., & Willén, R. M. (2011). Inside criminal minds: Offenders' strategies when lying. Journal of Investiga- tive Psychology and Offender Profiling, 8(3), 271-281.
6. Deeb, H., Vrij, A., Hope, L., Mann, S., Granhag, P. A., & Lan- caster, G. L. (2017). Suspects' consistency in statements con- cerning two events when different question formats are used. Journal of Investigative Psychology and Offender Profiling, 14(1), 74-87.
7. DePaulo, B. M., Lindsay, J. J., Malone, B. E., Muhlenbruck, L., Charlton, K., & Cooper, H. (2003). Cues to deception. Psy- chological bulletin, 129(1), 74.
8. Kassin, S. M., Appleby, S. C., & Perillo, J. T. (2010). Inter- viewing suspects: Practice, science, and future directions. Le- gal and criminological psychology, 15(1), 39-55.
9. Buller, D. B., & Burgoon, J. K. (1996). Interpersonal decep- tion theory. Communication theory, 6(3), 203-242.
10. Hartwig, M., Anders Granhag, P., & Strömwall, L. A. (2007). Guilty and innocent suspects’ strategies during police interro- gations. Psychology, Crime & Law, 13(2), 213-227.
11. Vrij, A., Palena, N., Leal, S., & Caso, L. (2021). The relation- ship between complications, common knowledge details and self-handicapping strategies and veracity: A meta-analysis. European Journal of Psychology Applied to Legal Context, 13(2), 55-77.
12. Verigin, B. L., Meijer, E. H., Vrij, A., & Zauzig, L. (2020). The interaction of truthful and deceptive information. Psy- chology, Crime & Law, 26(4), 367-383.
13. Koriat, A., & Goldsmith, M. (1996). Monitoring and control processes in the strategic regulation of memory accuracy. Psy- chological review, 103(3), 490.
14. Lancaster, G. L., Vrij, A., Hope, L., & Waller, B. (2013). Sort- ing the liars from the truth tellers: The benefits of asking un- anticipated questions on lie detection. Applied Cognitive Psy- chology, 27(1), 107-114.
15. Suchotzki, K., Verschuere, B., Van Bockstaele, B., Ben- Shakhar, G., & Crombez, G. (2017). Lying takes time: A me- ta-analysis on reaction time measures of deception. Psycho- logical Bulletin, 143(4), 428.
16. Van’t Veer, A. E., Stel, M., & van Beest, I. (2014). Limited capacity to lie: Cognitive load interferes with being dishonest. Judgment and Decision Making, 9(3), 199-206.
17. Wixted, J. T., & Carpenter, S. K. (2007). The Wickelgren power law and the Ebbinghaus savings function. Psychologi-cal Science, 18(2), 133.
18. Dando, C., Wilcock, R., & Milne, R. (2009). The cognitive interview: The efficacy of a modified mental reinstatement of context procedure for frontline police investigators. Applied Cognitive Psychology: The Official Journal of the Society for Applied Research in Memory and Cognition, 23(1), 138-147.
19. Verigin, B. L., Meijer, E. H., Vrij, A., & Zauzig, L. (2020). The interaction of truthful and deceptive information. Psy- chology, Crime & Law, 26(4), 367-383.
20. Ewens, S., Vrij, A., Mann, S., Leal, S., Jo, E., & Houston, K. (2017). The effect of the presence and seating position of an interpreter on eliciting information and cues to deceit. Psy- chology, Crime & Law, 23(2), 180-200.
21. Fisher, R. P., & Geiselman, R. E. (1992). Memory enhancing techniques for investigative interviewing: The cognitive inter- view. Charles C Thomas Publisher.
22. Tabachnick, B. G., Fidell, L. S., & Ullman, J. B. (2013). Us- ing multivariate statistics (Vol. 6, pp. 497-516). Boston, MA: pearson.
23. Nahari, G., & Ben�Shakhar, G. (2011). Psychophysiological and behavioral measures for detecting concealed information: The role of memory for crime details. Psychophysiology, 48(6), 733-744.
24. Vrij, A., Leal, S., Mann, S., Dalton, G., Jo, E., Shaboltas, A.,... & Houston, K. (2017). Using the model statement to elicit information and cues to deceit in interpreter-based interviews. Acta psychologica, 177, 44-53.
25. Du Prel, J-B., Hommel, G., Röhrig, B., & Blettner, M. (2009). Confidence interval or p-value? Deutzches Arzteblatt Interna- tional, 106, 335-339.
26. Fritz, C. O., Morris, P. E., & Richler, J. J. (2012). Effect size estimates: current use, calculations, and interpretation. Jour- nal of experimental psychology: General, 141(1), 2.
27. Nisbett, R. E., & Wilson, T. D. (1977). Telling more than we can know: Verbal reports on mental processes. Psychological review, 84(3), 231.
28. Vrij, A., & Granhag, P. A. (2012). Eliciting cues to deception and truth: What matters are the questions asked. Journal of Applied Research in Memory and Cognition, 1(2), 110-117.
29. Vrij, A., Granhag, P. A., Ashkenazi, T., Ganis, G., Leal, S., & Fisher, R. P. (2022). Verbal Lie Detection: Its Past, Present and Future. Brain Sciences, 12(12), 1644.
30. Colwell, K., Hiscock-Anisman, C., & Fede, J. (2013). Assess- ment criteria indicative of deception: An example of the new paradigm of differential recall enhancement. Applied issues in investigative interviewing, eyewitness memory, and credibili- ty assessment, 259-291.
31. Hartwig, M., Granhag, P. A., & Luke, T. (2014). Strategic use of evidence during investigative interviews: The state of the science. Credibility assessment, 1-36.
32. Palena, N., Caso, L., Vrij, A., & Nahari, G. (2021). The ver- ifiability approach: A meta-analysis. Journal of Applied Re- search in Memory and Cognition, 10(1), 155-166.
33. Vrij, A., Fisher, R. P., & Leal, S. (2022). How researchers can make verbal lie detection more attractive for practitioners. Psychiatry, Psychology and Law, 1-14.
34. Nahari, G., Ashkenazi, T., Fisher, R. P., Granhag, P. A., Hers- hkowitz, I., Masip, J., ... & Vrij, A. (2019). ‘Language of lies’: Urgent issues and prospects in verbal lie detection research. Legal and Criminological Psychology, 24(1), 1-23.
35. Vrij, A. (2016). Baselining as a lie detection method. Applied Cognitive Psychology, 30(6), 1112-1119.
36. Leins, D. A., Fisher, R. P., Pludwinski, L., Rivard, J., & Rob- ertson, B. (2014). Interview protocols to facilitate human in- telligence sources' recollections of meetings. Applied Cogni- tive Psychology, 28(6), 926-935.
37. Mattison, M. L., Dando, C. J., & Ormerod, T. C. (2015). Sketching to remember: Episodic free recall task support for child witnesses and victims with autism spectrum disorder. Journal of Autism and Developmental Disorders, 45, 1751- 1765.
38. Barlow, C. M., Jolley, R. P., & Hallam, J. L. (2011). Drawings as memory aids: Optimising the drawing method to facilitate young children's recall. Applied Cognitive Psychology, 25(3), 480-487.
39. Butler, S., Gross, J., & Hayne, H. (1995). The effect of draw- ing on memory performance in young children. Developmen- tal psychology, 31(4), 597.
40. Wesson, M., & Salmon, K. (2001). Drawing and showing: Helping children to report emotionally laden events. Applied Cognitive Psychology: The Official Journal of the Society for Applied Research in Memory and Cognition, 15(3), 301-319.