{"Affiliation":[{"label":"Affiliation","value":"Arts and Social Sciences, Irving K. Barber Faculty of (Okanagan)","attrs":{"lang":"en","ns":"http:\/\/vivoweb.org\/ontology\/core#departmentOrSchool","classmap":"vivo:EducationalProcess","property":"vivo:departmentOrSchool"},"iri":"http:\/\/vivoweb.org\/ontology\/core#departmentOrSchool","explain":"VIVO-ISF Ontology V1.6 Property; The department or school name within institution; Not intended to be an institution name."}],"AggregatedSourceRepository":[{"label":"Aggregated Source Repository","value":"DSpace","attrs":{"lang":"en","ns":"http:\/\/www.europeana.eu\/schemas\/edm\/dataProvider","classmap":"ore:Aggregation","property":"edm:dataProvider"},"iri":"http:\/\/www.europeana.eu\/schemas\/edm\/dataProvider","explain":"A Europeana Data Model Property; The name or identifier of the organization who contributes data indirectly to an aggregation service (e.g. Europeana)"}],"Campus":[{"label":"Campus","value":"UBCO","attrs":{"lang":"en","ns":"https:\/\/open.library.ubc.ca\/terms#degreeCampus","classmap":"oc:ThesisDescription","property":"oc:degreeCampus"},"iri":"https:\/\/open.library.ubc.ca\/terms#degreeCampus","explain":"UBC Open Collections Metadata Components; Local Field; Identifies the name of the campus from which the graduate completed their degree."}],"Creator":[{"label":"Creator","value":"Wellspring, Ian","attrs":{"lang":"en","ns":"http:\/\/purl.org\/dc\/terms\/creator","classmap":"dpla:SourceResource","property":"dcterms:creator"},"iri":"http:\/\/purl.org\/dc\/terms\/creator","explain":"A Dublin Core Terms Property; An entity primarily responsible for making the resource.; Examples of a Contributor include a person, an organization, or a service."}],"DateAvailable":[{"label":"Date Available","value":"2020-09-01T14:07:02Z","attrs":{"lang":"en","ns":"http:\/\/purl.org\/dc\/terms\/issued","classmap":"edm:WebResource","property":"dcterms:issued"},"iri":"http:\/\/purl.org\/dc\/terms\/issued","explain":"A Dublin Core Terms Property; Date of formal issuance (e.g., publication) of the resource."}],"DateIssued":[{"label":"Date Issued","value":"2020","attrs":{"lang":"en","ns":"http:\/\/purl.org\/dc\/terms\/issued","classmap":"oc:SourceResource","property":"dcterms:issued"},"iri":"http:\/\/purl.org\/dc\/terms\/issued","explain":"A Dublin Core Terms Property; Date of formal issuance (e.g., publication) of the resource."}],"Degree":[{"label":"Degree (Theses)","value":"Master of Arts - MA","attrs":{"lang":"en","ns":"http:\/\/vivoweb.org\/ontology\/core#relatedDegree","classmap":"vivo:ThesisDegree","property":"vivo:relatedDegree"},"iri":"http:\/\/vivoweb.org\/ontology\/core#relatedDegree","explain":"VIVO-ISF Ontology V1.6 Property; The thesis degree; Extended Property specified by UBC, as per https:\/\/wiki.duraspace.org\/display\/VIVO\/Ontology+Editor%27s+Guide"}],"DegreeGrantor":[{"label":"Degree Grantor","value":"University of British Columbia","attrs":{"lang":"en","ns":"https:\/\/open.library.ubc.ca\/terms#degreeGrantor","classmap":"oc:ThesisDescription","property":"oc:degreeGrantor"},"iri":"https:\/\/open.library.ubc.ca\/terms#degreeGrantor","explain":"UBC Open Collections Metadata Components; Local Field; Indicates the institution where thesis was granted."}],"Description":[{"label":"Description","value":"Research on cues for detecting deception have focused on two primary areas, verbal and nonverbal (e.g., body language; Bogaard et al., 2016; DePaulo et al., 2003). The current thesis examined if gender potentially plays a role in the production of language in a deceptive context. The Self-Report Psychopathy Test-IV (Paulhus et al., 2016) assessed the potential presence of psychopathic traits, while also examining the nature of their relationship with both gender and language. University undergraduate students were asked to recall either a truthful or deceptive trip they experienced while being interviewed in either a Be Detailed or Ghostwriter condition. Linguistic output was then analyzed using Linguistic Inquiry and Word Count (LIWC; Pennebaker et al., 2015a) and the Dictionary of Affective Language (DAL; Whissell, 2009). It was found that there was increased verbal output when participants were instructed with the Ghostwriter instructions, suggesting that small alterations in methodology can impact results and may explain previous contradictory findings. Somewhat surprisingly, positive affective language was found to be higher in the deceptive condition in the Be Detailed condition. Gender differences were also found in levels of psychopathic traits, where males tended to score higher than females. Levels of psychopathic traits were a significant covariate for disfluencies in the Be Detailed condition, suggesting that personality influences disfluencies. The current thesis provides additional information regarding linguistic cues to deceit, specifically variables that warrant further investigation (i.e., gender, personality, and context). This study also provides insights into potential methods for obtaining increased information in interview settings.","attrs":{"lang":"en","ns":"http:\/\/purl.org\/dc\/terms\/description","classmap":"dpla:SourceResource","property":"dcterms:description"},"iri":"http:\/\/purl.org\/dc\/terms\/description","explain":"A Dublin Core Terms Property; An account of the resource.; Description may include but is not limited to: an abstract, a table of contents, a graphical representation, or a free-text account of the resource."}],"DigitalResourceOriginalRecord":[{"label":"Digital Resource Original Record","value":"https:\/\/circle.library.ubc.ca\/rest\/handle\/2429\/75804?expand=metadata","attrs":{"lang":"en","ns":"http:\/\/www.europeana.eu\/schemas\/edm\/aggregatedCHO","classmap":"ore:Aggregation","property":"edm:aggregatedCHO"},"iri":"http:\/\/www.europeana.eu\/schemas\/edm\/aggregatedCHO","explain":"A Europeana Data Model Property; The identifier of the source object, e.g. the Mona Lisa itself. This could be a full linked open date URI or an internal identifier"}],"FullText":[{"label":"Full Text","value":"Sounding Authentic: An Examination Of How Individual and Personality Differences in   Language Affect Deception  by Ian Wellspring  B.A. (Hons.), Mount Royal University, 2015  A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF ARTS in THE COLLEGE OF GRADUATE STUDIES (Clinical Psychology)  THE UNIVERSITY OF BRITISH COLUMBIA (Okanagan) August 2020  \u00a9 Ian Wellspring, 2020  ii The following individuals certify that they have read, and recommend to the College of Graduate Studies for acceptance, a thesis\/dissertation entitled: Sounding Authentic: An Examination Of How Individual and Personality Differences in   Language May Affect Deception submitted by Ian Wellspring in partial fulfillment of the requirements of the degree of Master of Arts (M.A.).   Dr. Michael Woodworth, Irving K. Barber School of Arts and Sciences Supervisor Dr. Cynthia Mathieson, Irving K. Barber School of Arts and Sciences Supervisory Committee Member Dr. Harry Miller, Irving K. Barber School of Arts and Sciences Supervisory Committee Member Dr. Shirley Chau, Faculty of Health and Social Development, School of Social Work University Examiner     iii Abstract Research on cues for detecting deception have focused on two primary areas, verbal and non-verbal (e.g., body language; Bogaard et al., 2016; DePaulo et al., 2003). The current thesis examined if gender potentially plays a role in the production of language in a deceptive context. The Self-Report Psychopathy Test-IV (Paulhus et al., 2016) assessed the potential presence of psychopathic traits, while also examining the nature of their relationship with both gender and language. University undergraduate students were asked to recall either a truthful or deceptive trip they experienced while being interviewed in either a Be Detailed or Ghostwriter condition. Linguistic output was then analyzed using Linguistic Inquiry and Word Count (LIWC; Pennebaker et al., 2015a) and the Dictionary of Affective Language (DAL; Whissell, 2009). It was found that there was increased verbal output when participants were instructed with the Ghostwriter instructions, suggesting that small alterations in methodology can impact results and may explain previous contradictory findings. Somewhat surprisingly, positive affective language was found to be higher in the deceptive condition in the Be Detailed condition. Gender differences were also found in levels of psychopathic traits, where males tended to score higher than females. Levels of psychopathic traits were a significant covariate for disfluencies in the Be Detailed condition, suggesting that personality influences disfluencies. The current thesis provides additional information regarding linguistic cues to deceit, specifically variables that warrant further investigation (i.e., gender, personality, and context). This study also provides insights into potential methods for obtaining increased information in interview settings. Keywords: Gender, deception, language, psychopathy   iv Lay Summary The current project investigated how personality and gender may affect the language used when an individual is lying. To investigate the potential impact the type of instructions have on language, participants were asked to tell either an honest or dishonest story of a trip in either a \u201cBe Detailed\u201d or \u201cGhostwriter\u201d condition. Participants produced a longer statement in the Ghostwriter condition, demonstrating the impact that a slight change in instructions can have on results. There was also somewhat surprisingly more positive language in general when a person was lying in the Be Detailed condition, while personality was also related to disfluencies (i.e., \u201cum\u201d). Results further our knowledge of how personality, gender, and instructions influence deceptive language. For example, positivity may be used to distract another person in an attempt to sound more truthful when lying. The current study has implications for how to conduct investigative interviews.    v Preface For the present thesis, I was responsible for managing and coordinating a team of research assistants with data collection. I was also responsible for data transcription, generation of an excel dataset, data analysis, and the written portion herein. This research has been approved by the UBC Okanagan Behavioral Research Ethics Board under Ethics Certificate number H19-02313. This study was also supported by the Canada Graduate Scholarships, Masters (CGSM) \u2013 Social Sciences and Humanities Research Council of Canada (SSHRC). No findings of this study have been published.     vi Table of Contents Abstract ......................................................................................................................................... iii Lay Summary ............................................................................................................................... iv Preface .............................................................................................................................................v Table of Contents ......................................................................................................................... vi Acknowledgements .................................................................................................................... viii Dedication ..................................................................................................................................... ix Chapter 1: Introduction ................................................................................................................1 1.1 Hypotheses .................................................................................................................. 16 Chapter 2: Methodology ..............................................................................................................18 2.1 Participants .................................................................................................................. 18 2.2 Measures ..................................................................................................................... 18 2.2.1 Demographic Questions ........................................................................................ 18     2.2.2   Self-Report Psychopathy Scale .............................................................................. 19 2.2.3 Post Interview Questionnaire ................................................................................ 19 2.3 Linguistic Analysis ..................................................................................................... 20 2.4 Design ......................................................................................................................... 21 Chapter 3: Results ........................................................................................................................24 3.1 Participant Demographics ........................................................................................... 24 3.2 Post Interview Measures ............................................................................................. 24 3.3 SRP-IV ........................................................................................................................ 25 3.3.1 Reliability .............................................................................................................. 25 3.3.2 SRP-IV Demographics & Comparisons ............................................................... 25 3.4 Preparation Time ......................................................................................................... 26 vii 3.5 Ghostwriter vs. Be Detailed ........................................................................................ 26 3.6 Linguistic Cues ........................................................................................................... 27 3.6.1 Binomial Regression ................................................................................................ 28 3.6.2 Analysis of Variance ............................................................................................. 28 3.6.2.1 Main Effects .................................................................................................. 29 3.6.2.2 Interaction Effects ......................................................................................... 30 3.6.2 Analysis of Covariance ......................................................................................... 30 Chapter 4: Discussion ..................................................................................................................32 4.1 Preparation Time ......................................................................................................... 33 4.2 Gender Differences in Language ................................................................................ 35 4.3 Story-telling Conditions .............................................................................................. 37 4.4 Linguistic Cues to Deceit ............................................................................................ 39 4.4.1 Gender Differences & Linguistic Cues to Deceit ................................................. 43 4.5 Levels of Psychopathic Traits ..................................................................................... 44 4.5.1 Psychopathy & Linguistic Cues to Deceit ............................................................ 47 4.6 Implications ................................................................................................................. 51 4.7 Limitations .................................................................................................................. 53 4.8 Future Directions ........................................................................................................ 56 4.9 Conclusions ................................................................................................................. 59 References .....................................................................................................................................62 Appendices ....................................................................................................................................80 Appendix A \u2013 Tables ............................................................................................................ 80         Appendix B \u2013 Figures ............................................................................................................93  viii Acknowledgements I would first like to acknowledge and thank my supervisor, Dr. Michael Woodworth, for your continued support, patience, and guidance. I am so grateful and honoured for all your help, insight, and dependability throughout this process. Your door has always been open and I have felt continuously supported. Words cannot express how much I appreciate all of it. Thank-you for being a fantastic and brilliant role model and teacher.  Thank-you to my committee members \u2013 Dr. Woodworth, Dr. Mathieson, and Dr. Miller \u2013 for devoting your time to this project. Your thoughtfulness, feedback, and insight were immensely appreciated throughout this process and it was a pleasure to work with you all.  I would also like to acknowledge the profound role of my lab mates and peers for their support. Thank-you for the study breaks, last minute peer-reviews, related and unrelated discussions, advice, and humour throughout this process. Thank-you to my family and friends for your encouragement and support.     ix Dedication For Rebecca who has been through it all with me.  1 Chapter 1: Introduction There are a multitude of factors to consider when attempting to discern if an individual is being truthful. While there are a couple of potential indicators to deceit, such as non-verbal cues, there has also been focus and more recent attention on linguistic cues (Bogaard et al., 2016; Bond & DePaulo, 2006; Vrij, 2018). Relatively limited research suggests that a factor to consider that could influence our understanding of deceptive language is gender (Forrest et al., 2004; Lloyd et al., 2018; Porter et al., 2002). Interestingly, research has found gender differences in language in non-deceptive, natural contexts (Chaplin & Aldao, 2013; Eriksson et al., 2012). These gender-based differences may even be developmentally influenced during prenatal periods (Gervain, 2018), and are already noticeable during early developmental stages of a child\u2019s life (Eriksson et al., 2012). For example, some have proposed that genetics play a role (Chaplin & Aldao, 2013; Eriksson et al., 2012), or that hormone exposure during maturation also influences gender differences that manifest in language (Gervain, 2018; Zahn-Waxler et al., 2008). Moreover, these gender differences have been found to persist into adulthood, and have been shown to be consistent across various cultures (e.g., German, English, French; Eriksson et al., 2012). Further, these variations have been illustrated in specific linguistic categories (e.g., hedges; Leaper & Robnett, 2011), as well as broader categories such as language associated with social processes (Newman et al., 2008).  A meta-analysis utilizing adult participants found that there were gender differences in tentative and assertive language (Leaper & Robnett, 2011). Specifically, it was found that females tend to use less assertive and more tentative language such as hedges (e.g., \u201ckind of\u201d, \u201cI guess\u201d), intensifiers (e.g., \u201cvery\u201d, \u201creally\u201d), and expressions of uncertainty (e.g., \u201cthis could be wrong but\u2026\u201d). Newman et al. (2008) utilized a sample of over 14,000 scripts and also found an 2 effect of gender on language. The authors used the Linguistic Inquiry and Word Count software (LIWC; Pennebaker et al., 2015a; Pennebaker et al., 2015b) to examine numerous linguistic dimensions such as overall word count, prepositions, pronouns, and negations across gender. LIWC is also able to differentiate words based on broader categories such as cognitive operations and emotional language (Pennebaker et al., 2015a). Interestingly, LIWC has been shown to be effective in providing potential language cues for the presence of mood and anxiety disorders (Sonnenschein et al., 2018), personality (Hawkins & Boyd, 2017), and other psychopathologies (Junghaenel et al., 2008). Similar to Leaper and Robnett (2011), there was a moderate effect size of gender on language (Newman et al., 2008). Respectively, females tended to use more language associated with social and psychological processes, and used more verbs and pronouns as compared to males. Males were also generally more focused on current concerns and swore more often. Taken together, this research suggests that there are general gender differences in language such as with verbs and hedges, as well as with words associated with psychological processes and current concerns, which could relate to emotional language. One particular area of language that has also demonstrated gender differences in linguistic output is emotional language (Chaplin, 2015; Chaplin & Aldao, 2013; Newman et al., 2008; O\u2019Kearney & Dadds, 2004). For instance, it has been found that females tend to use language that is more associated with both positive emotions (e.g., happiness) and internalized negative emotions (e.g., sadness; Chaplin & Aldao, 2013; O\u2019Kearney & Dadds, 2004). However, males tend to express more linguistic cues of externalizing emotions such as anger and shame (Chaplin, 2015; Chaplin & Aldao, 2013). Consistent with the other language differences considered above, these potential emotional language differences have been found to be present in early childhood (Chaplin & Aldao, 2013), adolescence (O\u2019Kearney & Dadds, 2004), and 3 remain intact during adulthood (Chaplin, 2015). It should be noted that there are societal gender stereotypes in regard to gender differences in emotional language (Chaplin, 2015; Mehl & Pennebaker, 2003; Newman et al., 2008), which may be influenced by societal pressure (Mehl & Pennebaker, 2003; Newman et al., 2008) or developmental upbringing (Chaplin, 2015; Chaplin & Adao, 2013), amongst other explanations. Some research has found results that are similar to these stereotypes, while other research has found contradictory results (Chaplin, 2015; Newman et al., 2008). For instance, Newman et al. (2008) found no gender differences in the use of personal pronouns, which goes against the gender stereotype that males discuss themselves more often (Mehl & Pennebaker, 2003; Mulac et al., 2001). Further, the aforementioned results have found that males have higher rates of externalized negative emotions, which contradicts some gender stereotypes, and females tend to have higher rates of positive affective language which is similar to some other gender stereotypes (Chaplin, 2015; Newman et al., 2008). Nonetheless, gender differences in emotional language have been illustrated in the literature, and various explanations have been proposed that offer some insight into how gender differences in language develop.  Research into gender differences in language have proposed a combination of theories that originate from biological, social-developmental, and social constructivist perspectives (Chaplin & Aldao, 2013; Leaper & Friedman, 2007; Shields, 2002). The biological perspective suggests that gender differences are present at birth or early childhood due to either genetics or maturation (i.e., hormone differences at puberty), which manifests in language (Chaplin & Aldao, 2013; Zahn-Waxler et al., 2008). It has been proposed that these differences are then exacerbated by social development through cognitive learning and experience (Chaplin & Aldao, 2013; Liben & Bigler, 2002). The expression of these gender differences, as theorized by social 4 constructivist theories, may be influenced by contextual factors such speaking with a stranger as opposed to a friend (Shields, 2002). For instance, partially related to genetics, males have less enhanced language abilities in combination with decreased control of negative emotions as compared to females (Zahn-Waxler et al., 2008), which could be exacerbated by an adopted gender role (Liben & Bigler, 2002). These can be more pronounced given situational and contextual factors such as social demands, rapport, and the conversation topic (Shields, 2002). Interestingly, it has also been considered that when there is increased social demand that females attempt to hide negative emotions with positive affect more than males, thereby exacerbating gender differences in language (Chaplin, 2015; Chaplin & Aldao, 2013). Notwithstanding, it has been implied that these gender-based differences in language, specifically emotional language, are potentially an important (and understudied) factor to consider when assessing the credibility of deceptive accounts (Lloyd et al., 2018). Deception detection ability has routinely been found to hover around 54%, regardless of a wide range of factors that have been suggested to influence accuracy such as training, age, and socioeconomic status (Bond & DePaulo, 2006; Hartwig & Bond, 2011; Mann et al., 2004). Researchers and professionals (such as law enforcement and clinicians) have sought to improve deception detection accuracy, from which two essential avenues of progress have arisen: the examination of verbal and non-verbal cues (Bogaard et al., 2016; DePaulo et al., 2003; van Swol et al., 2012; Vrij, 2018). Historically, research on non-verbal cues has received more attention due to the perception that it had the highest potential utility in numerous scenarios ranging from job interviews (Hauch et al., 2015; Jupe et al., 2018) to police investigations (Mann et al., 2004; Mann et al., 2008). Some of the more common non-verbal cues to deception that have been investigated include eye contact (Mann et al., 2013; Walczyk et al., 2013), fidgeting (Siegfried & 5 Schwandt, 2007), and micro-expressions (Ekman & Friesen, 1969; Matsumoto & Hwang, 2018). The previous focus on non-verbal cues has also resulted in more of a reliance on them to give indications of deceit, rather than verbal cues (Gerlach et al., 2019; Mann et al., 2008; Matsumoto et al., 2016). However, research has consistently demonstrated that non-verbal cues to deceit are not as reliable as hoped, nor always accurate (Bogaard et al., 2016; DePaulo et al., 2003; Gerlach et al., 2019; Vrij, 2018). For example, avoiding eye contact has historically been discussed as a non-verbal cue to deceit (Sporer & Schwandt, 2007), but it has also been shown that increased eye contact may also be an indicator of deceit (Mann et al., 2013). Due to mounting questions around the reliability and accuracy of non-verbal cues to deceit, as well as promising results of some verbal indicators, more research is now focusing on linguistic cues to deception.   In the past decade or so the importance of verbal cues to deceit has been increasingly considered and perhaps now is thought to be even more reliable when compared to non-verbal cues (Arciuli et al., 2010; Bogaard et al., 2016; DePaulo et al., 2003; Newman et al., 2003; van Swol et al., 2012; Vrij, 2018). Some research even suggests that verbal cues to deception are able to detect deceit at an accuracy above chance levels (Bogaard et al., 2016; Hauch et al., 2015; McQuaid et al., 2015; Newman et al., 2003). For instance, Newman et al. (2003) found that an automated program used to investigate linguistic cues to deceit was better than human judgements. Studies have found that deceptive individuals use fewer total words as compared to individuals who are being truthful (DePaulo et al., 2003; Pennebaker, 2011). Other research has found a decrease in the number of first-person pronouns (e.g., \u201cI\u201d, \u201cme\u201d; McQuaid et al., 2015; Pennebaker, 2011) and an increase in the number of disfluencies (e.g., \u201cumm\u201d, \u201cjust\u201d) when someone is attempting to be deceptive (McQuaid et al., 2015; Hancock & Woodworth, 2013). It should be noted that these results were within particular research paradigms and arguably need to 6 be considered primarily within these contexts (Gerlach et al., 2019; Nahari et al., 2019). For example, McQuaid et al. (2015) examined press conferences, Newman et al. (2003) looked at opinions of social issues, while Hancock & Woodworth (2013) examined a variety of results (and contexts) specific to computer-mediated-communications.  Various explanations have been proposed in each of the aforementioned contexts to account for observed language differences between liars and truth-tellers. For example, in the missing persons press conference study, it was proposed that psychological distancing could be occurring (McQuaid et al., 2015). Specifically, deceptive individuals in these types of press conferences may be emotionally detaching from the event or topic (arguably either consciously or unconsciously). These individuals who were actually responsible for the disappearance of their loved are attempting to deflect suspicion from themselves, and psychologically distancing could lead to less personal (or emotional) attachment and result in a decrease in the number of personal pronouns. A similar theory (that less first-person pronouns would be present as individuals psychologically distance) has also been proposed in studies examining deceptive language during computer-mediated-communications (i.e., instant messages; Handcock & Woodworth, 2013). Additionally, other researchers investigating expressed opinions of social issues have proposed that constructing a lie is cognitively taxing (Newman et al., 2003). This increased cognitive load can be observed through linguistic cues (Vrij & Granhag, 2012), including a potential decrease in overall detail and increase in disfluencies (Johnson & Raye, 1981; Sporer, 2004; Vrij, 2008). Interestingly, it has been proposed that these indicators of deceit act at a subconscious level, and it is possible that they are beyond the cognitive control of the deceiver (Hancock et al., 2011; McQuaid et al., 2015). Despite varying explanations, there appears to be observable and reliable linguistic indicators of deceit that could serve to increase 7 deception detection accuracy above chance levels (Bond & DePaulo, 2006; Hancock & Woodworth, 2013; Vrij, 2018). Such nuances of language are perhaps more detectable through computerized language analysis programs, which are increasingly being explored for providing indicators that a person is more likely to be providing false information (Hauch et al., 2015; ten Brinke & Porter, 2012). For example, the aforementioned LIWC program has been used as it is able to categorize language (e.g., psychological processes; Pennebaker et al., 2015) in ways that appear to be helpful in uncovering deceptive intent. Programs such as LIWC provide the opportunity for automatic and more accurate coding than can be done by hand across various linguistic categories, which can then be interpreted for linguistic markers of deceit (Bond & Lee, 2005; McQuaid et al., 2015; ten Brinke & Porter, 2012). The use of such tools (e.g., LIWC, Wmatrix) has been illustrated in some of the aforementioned research such as the press conferences study, which would be considered a high-stakes scenario (McQuaid et al., 2015). As discussed above, based on this linguistic output, researchers were able to hypothesize about speakers who were being deceptive about their involvement in the disappearance of their loved one (i.e., were perhaps involved). This finding highlights the prospective benefits (i.e., information gathering) of linguistic analysis in real-world forensic contexts and presupposes that this type of technique could be extended to other high-stakes contexts such as police interviews and testimonies. Consistent with other research, this finding also illustrates how linguistic analysis has yielded more stable and accurate results for providing potential indicators of deception as compared to non-verbal cues, resulting in the increased confidence in considering language as an indicator of deceit (Arciuli et al., 2010; Hauch et al., 2015; McQuaid et al., 2015; Newman et al., 2003; Vrij, 2018). However, given the diversity among suspects who could be inclined to lie in these 8 contexts, there is unfortunately little research examining individual difference factors, such as gender, that may impact deceptive language.  The limited research that has been conducted on gender differences, both for deceptive success and deception detection accuracy, has yielded inconsistent results (Lloyd et al., 2018). For example, in one study participants were asked to judge the veracity of statements about an individual\u2019s attitude towards a particular social topic (Forrest et al., 2004). On average, female participants who were suspected of lying were more correctly identified as being deceptive as compared to male participants, with Lloyd et al. (2018) finding similar results. The authors suggested that gender differences in emotional language offer a greater opportunity for genuine emotions and intentions to be \u201cvisible,\u201d potentially making it easier to detect deception among females. Contrasting this, an earlier study found that females were more successful at deception as compared to males (Porter et al., 2002). Participants were asked to judge the veracity of a statement from a person being deceitful or truthful about an emotional experience with results that indicated that males were more correctly identified as being deceptive. In line with emotional language differences, the authors proposed that the increased use of deceptive emotional language can hide or mask true emotions for females. Interestingly, both these studies suggest that gender differences are a potential explanation, despite diverse results (Lloyd et al., 2018). In summary, research has found differences between males and females in emotional language (Chaplin & Adao 2013; Newman et al., 2008), and provided hypotheses about how this may influence language in a deceptive context (Forrest et al., 2004; Porter et al., 2002). Outside of these contradictory studies, there does not appear to be any extant research considering gender and deceptive success, yet alone focusing on linguistic cues to deceit. It should be noted that these hypotheses were created in response to passive methodologies (i.e., watching or listening), 9 and should arguably only be considered within these types of contexts.  A feasible explanation for the above-mentioned inconsistent results in deception research, is that previous methodologies have varied along a continuum from dynamic-interactive to more non-interactive (passive) detection tasks (Gerlach et al., 2019; Rosenbaum et al., 2014; Vrij, 2018). For example, non-interactive methodologies utilized have participants view or hear videos of potentially deceptive individuals (Bogaard et al., 2016; Rosenbaum et al., 2014). These types of methodologies were utilized in studies such as Porter et al. (2002) and Forrest et al. (2004) and are relatively passive with no interaction between the liar\/truth-teller, and the judging participant. While individuals are sometimes required to make judgements about an individual\u2019s credibility based solely on observation in a natural environment, they more typically would explicitly interact with the person (e.g., law enforcement, airport security) which arguably reduces the ecological validity of passive methodologies that have been used (Bogaard et al., 2016; Burgoon et al., 2001; Gerlach et al., 2019; Vrij, 2018). Interactive methodologies likely prompt participants to become more invested and engaged in the procedure, again potentially better approximating what would occur in a more natural setting (Leal et al., 2019; Vrij, 2018). Indeed, researchers have suggested that deception methodologies should strive to veer away from completely non-interactive tasks (e.g., listening to a recording) and move towards more interactive methods to increase ecological validity (Burgoon, 2015; Rosenbaum et al., 2014; Vrij, 2008; 2018). One interactive context that has received particular attention with regard to deception detection is during interview contexts, such as general (non-forensic) and investigative interviews (Fisher, 2010; Vrij, 2018). The main goal is typically to garner as much information as possible (Brandon, 2014), after which the veracity of the information can be investigated 10 (Fisher, 2010; Vrij, 2018). In order to better approximate the interactive nature of these scenarios, more recent research has developed interactive methodologies that aim to increase generalizability and verbal output from participants (Leal et al., 2018; Leal et al., 2019). For instance, some researchers have developed a series of engaging questions for interviewees (Burgoon et al., 2001), while others have provided exemplar statements for participants (Leal et al., 2018). A more recent and novel methodology developed adjusted the initial instructions given to participants with either a set of \u201cBe Detailed\u201d or \u201cGhostwriter\u201d instructions (Leal et al., 2019). In both conditions the aim was to garner as much information as possible from the participants. However, the Ghostwriter instructions asked participants to actually imagine speaking with a Ghostwriter and also provide as much detail as possible. It was found that the Ghostwriter instructions produced significantly more detail from participants with the researchers hypothesizing that the Ghostwriter instructions provided a clear framework for participants in terms of how much detail to the give the interviewers. Nonetheless, additional research utilizing an interactive methodology is required to determine more precisely how individual characteristics influence deception. For instance, it is also important to consider that particular personality characteristics may further influence language production.   Research into the personality construct known as psychopathy has revealed unique language characteristics (Hancock et al., 2011; Le et al., 2017), as well as considered potential differences between how the disorder manifests between males and females (Guay et al., 2018; Neumann et al., 2012). Observations of psychopathy more formally began with Cleckley (1941) in his clinical descriptions of individuals who appeared \u201cnormal,\u201d but had various problematic characteristics. Cleckley\u2019s clinical documentation of these individuals was subsequently formalized by Dr. Robert Hare, who designed what is still considered by many to be the gold 11 standard for psychopathic assessment, the Psychopathy Checklist \u2013 Revised (PCL-R; Cooke & Selbom, 2019; Hare, 2003; Tonnaer et al., 2013). Psychopathy is a cluster of personality traits that include a propensity for deception, manipulation, superficial charm, lack of guilt and remorse, and antisocial behaviour (Cleckley, 1941; Hare, 2003; Verschuere et al., 2018). Notably, psychopathy is associated with emotional deficits such as shallow affect and lack of empathy (Dawel et al., 2018; Verschuere et al., 2018), which may come across in their language (Le et al., 2018; Porter & Woodworth, 2007).   Levels of psychopathic traits have also been associated with higher levels of aggression and criminal offending (Edwards et al., 2017; Nicholls et al., 2005; Vitacco et al., 2014). While psychopathy is estimated to be found in approximately 1% of the general population (Coid et al., 2009) it is far more prevalent in prison populations, with prevalence rates at approximately 20%-30% (Ogloff, 2006; Vachon et al., 2013). Over the decades the precise relationship between psychopathic traits and aggression has been continuously researched, with one study finding that levels of psychopathic traits were able to predict higher levels of aggression above and beyond general personality characteristics (Dini\u0107 & Wertag, 2018). Meta-analyses have also illustrated the predictive utility of psychopathy for predicting both violent and non-violent crime (Leistico et al., 2009; Vitacco et al., 2014). Another study demonstrated that even if dynamic risk factors (e.g., treatment for substance use, increased social support, etc.) are changed, there is still a high likelihood of psychopaths re-offending and engaging in violence (Mastromanno et al., 2018). Further, other research has found that psychopathy is also related to specific types of violence, particularly instrumental (i.e., for some observable gain) and planned violence, in which case they may be more inclined to deceive others (Vitacco et al., 2014; Walsh et al., 2009; Woodworth & Porter, 2002). Considering the high impact of crime, and the increased levels of 12 potential deception involved, it is important to better comprehend what role gender differences may play in the production of language in a deceptive context, and how levels of psychopathic traits might impact this relationship.   Interestingly, the prevalence rates for psychopathy within the prison system for females is approximately 10-30%, which, similar to males, is substantially higher than the general population of approximately 1% (Beryl et al., 2014; Nicholls et al., 2005). Recent research on psychopathy has uncovered certain differences between males and females (de Vogel & Lancel, 2016; Guay et al., 2018; Schulz et al., 2016). Generally, it has been shown that females tend to score lower on the PCL-R than men, insinuating that psychopathy manifests differently in males and females (Guay et al., 2018). It should be noted that the underlying four-facet structure as proposed by Hare (Interpersonal, Affective, Lifestyle, and Antisocial; Hare, 2003, 2016) has generally remained consistent across studies for both males and females. However, different configurations of the four facets could be present between males and females that have higher levels of psychopathic traits (Guay et al., 2018). For example, it has been found that the Antisocial and Interpersonal facets of the PCL-R are more adversely affected by gender differences (Beryl et al., 2014; Forouzan & Cook, 2005). In terms of the Interpersonal facet, it has been theorized that females might score higher in this area (Cunliffe et al., 2016; Forouzan & Cook, 2005). However, extant research has been mixed, as some research has validated this (Strand & Belfrage, 2005), while other studies have contradicted it (Beryl et al., 2014). Nonetheless, it has been consistently found that males tend to have higher scores in the Antisocial facet (Beryl et al., 2014; Forouzan & Cook, 2005). Whereas, there appears to be more similarities between males and females for the Affective and Impulsive subscales of the PCL-R (Beryl et al., 2014; Guay et al., 2018).  13  De Vogel & Lancel (2016) examined if psychopathy manifests differently across gender and types of crimes. Females having higher levels of psychopathic traits were found to have generally committed more fraud, lower rates of physical violence, and were potentially more manipulative during treatment as compared to males. While psychopathic traits in females were only a moderate predictor of physical violence, they were shown to be a strong predictor in males. However, both male and female psychopathic traits were strong predictors of violence when the definition was expanded to include both physical and verbal types of aggression. Some of the above findings suggest gender differences in psychopathy could play a role in not only types of crimes, but also in interview processes. For example, if females with higher levels of psychopathic traits are more manipulative during treatment, they could also be more difficult to interview for law enforcement officials or may present with separate linguistic cues to deceit.  There has been some intriguing recent research illustrating a link between levels of psychopathic traits and certain linguistic characteristics (Gullhaugen & Saskshaug, 2018; Hancock et al., 2018; Hancock et al., 2011; Le et al., 2017; Porter & Woodworth, 2007). For instance, one study found that individuals who have higher levels of psychopathic traits tend to use more disfluencies (e.g., \u201cumm\u201d, \u201cyou know\u201d), personal pronouns (e.g., \u201cI\u201d, \u201cyou\u201d, \u201chis\u201d), and discussed other people less often (Le et al., 2017). Individuals with higher levels of psychopathic traits were also less emotionally expressive in their language, which was one of the main linguistic predictors of psychopathy. Other studies investigating psychopathic language have also found a dearth of emotional language, suggesting that this particular linguistic characteristic may be an identifiable and integral feature of individuals who have psychopathy (Hancock et al., 2018; Hancock et al., 2011; Porter & Woodworth, 2007), which could influence deception.  14  Research investigating the relationship between levels of psychopathic traits and the propensity for deceit has yielded mixed results (Azizli et al., 2016; Halevy et al., 2014; Hare, 2003; Jones & Paulhus, 2017; Lee et al., 2008; Verschuere & Hout, 2016). For instance, brain imaging research has shown that as levels of psychopathic traits increase there is also an increase in brain activity, mainly in the prefrontal cortex (decision making, cognitive behaviour) and parietal lobe (language, visual processing), when recounting a fabricated crime (Glenn et al., 2017). These neuroanatomical regions are associated with the complex process of formulating a lie, as well as potential increased propensity to engage in deceit, indicating that higher levels of psychopathic traits may be associated with deception. Support for a potential link between psychopathy and deceit has also been found in research investigating the propensity for lying in scenarios in which there is either a high or low risk of being caught (Azizli et al., 2016; Jones & Paulhus, 2017). Higher levels of psychopathic traits were found to be the primary predictor of intentional lying in high-risk scenarios and was indicative of individuals having a higher deceptive success rate (Jones & Paulhus, 2017). This study provides some support for the neuroanatomical studies that suggest a relationship between the propensity for deceptive prowess and psychopathy. In contrast, other research indicates that individuals with high psychopathic traits are not more adept at lying (Klaver et al., 2007; Verschuere & Hout, 2016; Wright et al., 2015). For example, Wright et al. (2015) found that while psychopathy was associated with viewing lying as acceptable, levels of psychopathic traits were not associated with success in deceiving others. Another study found that psychopathic offenders were generally slower to produce their fabricated statement and produced more errors (Verschuere & Hout, 2016).   Despite the somewhat contradictory research illustrating a potential association between 15 psychopathy and deceit, limited research has investigated how language may be affected in these scenarios (Lee et al., 2008; Quayle, 2008). This extant sparse literature has found that psychopathic offenders are less coherent and make more spontaneous corrections when lying (Lee et al., 2008). Interestingly, it was also found that psychopathic offenders included more relevant details when producing a fabricated story as compared to non-psychopathic offenders. However, there is little to no research investigating if linguistic characteristics are consistent across genders when considering the effect of psychopathic traits, yet alone focusing on emotional language or other linguistic dimensions while being deceptive.  In summary, while limited previous research suggests that gender may play a role in deceptive success there has been a dearth of studies considering gender differences specifically pertaining to language use. Further, the research that has been conducted into gender differences and deception have yielded inconsistent results, with emotional language differences being cited as potentially playing a significant role in deception (Forrest et al., 2004; Porter et al., 2002). Extant literature has indicated that linguistic characteristics can be valid cues to deception (Bogaard et al., 2016; DePaulo et al., 2003; van Swol et al., 2012; Vrij, 2018), making this integral research to facilitate our understanding of credibility. Moreover, the literature has also shown that there are gender differences in levels of psychopathic traits, along with particular language characteristics (de Vogel & Lancel, 2016; Guay et al., 2018; Le et al., 2017; Schulz et al., 2016). However, there is little to no research investigating if potential gender differences in language are influenced by psychopathic traits, let alone in a deceptive context.  This research will strive to provide additional insight into individual differences that may impact language when a person is being deceptive. Considering the concerns noted with previous research, the current study will aim to be more generalizable and ecologically sound by 16 attempting to be more relevant to what occurs in a more interactive natural setting. Compared to the more passive methodologies often used in previous research, this study could be extended to more interactive contexts such as general interview settings (e.g., job interview) and arguably be applicable to legal settings such as police interviews and parole adjudications. Another purpose of this study is to investigate the effect that different interviewer instructions would have on participant responses and replicate previous findings in the aforementioned study that used the Be Detailed and Ghostwriter instructions (Leal et al., 2019). Insights obtained about factors that affect verbal deception may influence how professionals conduct their interviews. For example, the potential knowledge gained from this study could guide professionals on the phrasing of interview questions, timing of the presentation of important information, and when to switch into other phases of interviewing. Particularly, it could give better insight for which linguistic cues indicate false information and guide interviewers when to ask more questions or adjust their tactics. The research is also timely as the rate of crimes perpetrated by females has seen a steady increase (Mahony, Jacob, & Hobson, 2017), accounting for approximately 25% of crime in Canada with the majority of offences being assault related (70%; Savage, 2019). The consideration of levels of psychopathic traits can also provide added insight as to how it could potentially impact deceptive verbal ability across gender, while also providing additional understanding into how the personality construct presents across genders. 1.1 Hypotheses It is hypothesized that there will be gender differences present in language for both the truthful and deceptive conditions, with particular differences in the deceptive condition. More specifically, it is hypothesized that potential linguistic cues to deception (e.g., disfluencies, personal pronouns, emotional language) will be more pronounced in females as compared to 17 males in the deceptive condition. As outlined above, some research has found that females use more emotional language in general, potentially leading to an increase in emotional language within a deceptive context. This language is predicted to present linguistic characteristics that reflect how individuals may (consciously or unconsciously) psychologically distance themselves when being deceptive (decrease in personal pronouns), or are indicative of the cognitive load required when lying (increase in disfluencies). Based on the extant research, it is also proposed that levels of psychopathic traits will influence language production in combination with gender in a deceptive context. For example, while there is limited specific research or theory to base a hypothesis on, the literature has previously demonstrated both potential gender differences and language characteristics (e.g., less emotional language, personal pronouns) that are associated with psychopathic traits. However, its respective contribution in combination with gender in a deceptive context has not previously been examined. One of the purposes of this study is to explore its effect to both refine our understanding and develop additional insights, as well as to generate more specific hypotheses for future research. As mentioned, the current study will employ a recent methodology from Leal et al. (2019) which found participants produced more details in the novel \u201cGhostwriter\u201d instructions (as compared to \u201cBe Detailed\u201d). Another purpose of this study is to see if their findings can be replicated.    18 Chapter 2: Methodology 2.1 Participants All psychology student participants volunteered for the study through the University of British Columbia\u2019s (Okanagan) SONA system database. Participants completed the study as part of their course requirements, receiving 1.0% course credit for participation in the study. Initially, the study included a total of 107 participants; however, one was removed as the participant was under the age of 18. This resulted in 106 participants that were further broken down into story-telling conditions (i.e., Ghostwriter and Be Detailed), veracity conditions (i.e., True and False), and gender (see Table 1). All participants provided written informed consent via signature. The study was approved by the Behavioral Research Ethics Board (BREB) of the University. 2.2 Measures 2.2.1 Demographic Questions Participants first completed a questionnaire asking general demographic questions (age, program of study, native language), and assessing levels of psychopathic traits. Included in the demographic questions was a question asking the participants identified gender by asking \u201cI identify as (check whichever applies):\u201d Response options for this question included \u201cmale,\u201d \u201cfemale,\u201d \u201cnon-binary,\u201d \u201cnone of the above,\u201d and \u201cprefer not to answer.\u201d In acknowledgement of current ways of how to identify by gender, the demographics that were asked were adjusted to be more inclusive. While there is little precedent for research in this area, particularly surrounding language and deception, an exploratory analysis would have been conducted if participants identified as \u2018\u201cnon-binary,\u201d \u201cnone of the above,\u201d or \u201cprefer not to answer.\u201d However, all participants responded with either \u201cmale\u201d or \u201cfemale.\u201d It should also be noted that, due to minuscule precedent, the proposed hypotheses and results may not generalize or be 19 straightforward to individuals who do not identify as either \u201cmale\u201d or \u201cfemale\u201d as previous research has focused on this binary conceptualization. Participants were also asked which cities they have visited in the past 12 months (1 year). 2.2.2 Self-Report Psychopathy Scale Levels of psychopathic traits were assessed using the Self-Report Psychopathy Scale (SRP-IV; Paulhus et al., 2016). It contains 64 questions, with 16 questions assessing four-factors (Callous Affect, Interpersonal Manipulation, Erratic Lifestyle, Criminal Tendencies), which is similar to the four-factor structure proposed by Hare (2003). Some of the questions included \u201cI am often rude to other people,\u201d \u201cI find it easy to manipulate people,\u201d \u201cRules are made to be broken,\u201d and \u201cI have broken into a building or vehicle to steal and vandalize.\u201d Participants responded on a Likert scale ranging from 1 (\u201cstrongly disagree\u201d) to 5 (\u201cstrongly agree\u201d). Higher scores on the SRP-IV indicated the potential presence of differing levels of psychopathic traits and primarily tap into constructs such as callous-unemotional traits, empathy, and antisocial behaviour. The SRP-IV reports good internal consistency with overall Cronbach Alpha\u2019s ranging from .86-.90 (Gordts et al., 2017; Paulhus et al., 2016). It has also demonstrated high convergent validity with the four-factor structure of the PCL-R, which is considered the \u201cgold standard\u201d for assessing levels of psychopathic traits (Gordts et al., 2017; Mahmut et al., 2011). It should be noted that for the purposes of the current study, the SRP-IV is being used to provide a potential indication of levels of psychopathic traits and does not indicate the presence of the personality construct, or diagnosis, of psychopathy. 2.2.3 Post Interview Questionnaire Participants completed a post-interview questionnaire assessing levels of motivation, honesty, and their perceptions of the allocated preparation time as per Leal et al. (2019). 20 Participants motivation was assessed through self-report in response to \u201cplease give an indication as to how motivated you were to perform well during the interview.\u201d Participants responded on a Likert scale ranging from 1 (\u201cnot motivated at all\u201d) to 5 (\u201cvery motivated\u201d). Their perceptions of the allocated preparation time were assessed through self-report in response to \u201cplease give an indication as to whether or not you thought you were given enough time to prepare for the interview.\u201d Responses ranged from 1 (\u201cinsufficient\u201d) to 7 (\u201csufficient\u201d). Lastly, participants were asked how truthful they were in their accounts to the interviewer through self-report in response to \u201cplease indicate the extent that you were truthful in the interview.\u201d Participant responses ranged from 0% to 100% in 10% increments (i.e., 0%, 10%, 20%, etc.). The post interview questionnaire was used as a manipulation check to ensure that there were differences in honesty between conditions, as well as to assess participants levels of motivation. 2.3 Linguistic Analysis Responses from participants were transcribed into word documents and analyzed using Linguistic Inquiry and Word Count (LIWC; Pennebaker et al., 2015a). LIWC codes language based upon linguistic dimensions (e.g., pronouns, prepositions, overall word count) as well as broader categories such cognitive operations and emotional language, among other variables (Pennebaker, 2011; Pennebaker et al., 2015a). Using a dictionary of approximately 4,500 words, LIWC counted the amount of words in each category and divided this by the total number to get a percentage for each category. These percentages were then analyzed across gender, veracity conditions, and levels of psychopathic traits (see Hancock et al., 2018; McQuaid et al., 2015; Pennebaker, 2011). For the purposes of this study, LIWC was then used to code for general linguistic dimensions (i.e., personal pronouns, disfluencies), psychological processes, and emotional language. 21 The Whissell Dictionary of Affective Language (DAL) was also used to analyze the linguistic output of participants responses (Whissell, 2009). The DAL is a language analysis program that attempts to quantify and measure the suggested emotionality underlying everyday language by measuring each word on three distinct categories: pleasantness, imagery, and activation (Whissell, 1989; 2009). Pleasantness is how pleasurable a word feels, while activation is suggestive of how active a word feels (Whissell, 1989). In conjunction, imagery is how much the word in question brings an image to mind (Whissell, 2009). Each of these categories were measured on a 3-point Likert scale, in which aggregate values can be calculated based on single words, sentences, paragraphs, or entire texts (e.g., the Bible). In the case of the present study, DAL aggregate values for the entire transcripts were used in the analysis. DAL was then used in conjunction with LIWC as DAL is limited in its ability to code general language categories (e.g., pronouns, verbs, disfluencies) as well as broader linguistic dimensions such as language that is associated with cognitive processes.   2.4 Design As discussed, the generalizability of some previous deception research methodologies is potentially limited. Specifically, many studies employ passive tasks, which can impact the ecological validity and may help to explain inconsistent results (Bogaard et al., 2016; Gerlach et al., 2019; Vrij, 2018). The applied methodology attempted to rectify this issue by using a modified interactive procedure from Leal et al. (2019)\u2019s study (see Figure 1). Participants first completed a questionnaire asking general demographic questions and were asked which cities they have visited in the past 12 months (1 year).   The participants were then randomly assigned to two groups: truth-tellers or deceivers. Truth-tellers were asked to describe a trip to a city they have been to, while deceivers were asked 22 to describe a trip to a city they have not been to. Participants were asked to describe a trip as this would be relatable for individuals partaking in the study, potentially increasing engagement on behalf of the participants (Leal et al., 2018; Vrij et al., 2018). Truth-tellers and deceivers were then left with a computer and given 20 minutes to prepare for the interview, or were able to let the researcher know when they were ready to begin. Participants were assured that they would have enough time to prepare and were told they could take notes during their research time if they would like.   Following this, participants were further placed into two interview conditions, a Be Detailed condition and a Ghostwriter condition. Both conditions had the same introduction, which started with the interviewer stating: \u201cI understand from my colleague that you have visited (the interviewer inserts here the name of the city that the participant had discussed), I am just going to ask you one question, so make sure that you include all the information you want to convey in your answer.\u201d In the Be Detailed condition participants were further prompted to provide as much detail as possible with the interviewer stating, \u201cPlease think of all the details, big and small, and include these in your account, and describe one emotional or interesting event, good or bad, that occurred on your trip.\u201d  In the Ghostwriter condition, participants were instructed to pretend as if they were talking to a ghostwriter. A ghostwriter is an individual who writes books and articles for other individuals and has the task of determining the necessary details for writing a story, to which the interviewees are often told to include every detail, even minor ones. This methodology potentially created elevated expectations about how much information should be provided in a more natural, interview type setting. It also is free-flowing and allowed the interviewees to 23 continue discussing their trip and make associations within their story, potentially increasing linguistic output (Leal et al., 2019). The interviewer introduced this condition by stating: To begin, a Ghostwriter is a person whose job it is to write books or articles for another individual. One example of this type of job would be to write about a celebrity\u2019s life for an article or an autobiography. For example, Albert Einstein has an autobiography that was written by a Ghostwriter. In order to write the book about Albert Einstein the Ghostwriter would have spent many hours listening to accounts of his life and experiences. It would be essential to the writer that Albert included all the details, even the tiny insignificant ones, about each experience, since part of the Ghostwriter\u2019s skill lies in determining what they think are the most interesting parts of a story. What we would like you to do is to imagine that you will be speaking to a Ghostwriter about the trip you recently took. Please think of all the details, big and small and include these in your account, and describe one emotional or interesting event, good or bad, that occurred on your trip. Therefore, if a Ghostwriter were to listen to your account they should, in theory, be able to make a comprehensive story regarding your trip. Do you understand? (Response from participant). Thank you, so while imagining you are talking to a Ghostwriter, could you please tell me in as much detail as possible everything that happened from the moment you started your trip to the moment you left. After describing the trip, responses from the participants were transcribed and analyzed using Linguistic Inquiry and Word Count (LIWC; Pennebaker et al., 2015a) and the Whissell Dictionary of Affective Language (DAL; Whissell, 2009) as described above to code for linguistic properties.     24 Chapter 3: Results 3.1 Participant Demographics There were 106 participants with an average age of 20.19 (SD = 2.39), and an age range of 18 to 30. As shown in Table 1, there were 32 males (Mage = 20.53, SDage = 3.12) and 74 females (Mage= 20.04, SDage = 2.39). Of the 106 participants, 96 indicated English as their primary spoken language, while the remaining participants reported that their primary spoken language was either Bahasa Indonesia (N = 1), Mandarin (N = 4), Farsi (N = 1), Korean (N = 1), or Spanish (N = 2). It should be noted that additional analyses were conducted with only participants that identified their primary spoken language as English, and similar results were found. Table 2 shows the descriptive statistics for the linguistic output of LIWC and DAL, with the average word count across all participants being 586.7 (SD = 485.07). 3.2 Post Interview Measures As shown in Table 3, participants average self-reported motivation was 4.33 (SD = 0.82), with the scale ranging from 1 (\u201cnot motivated at all\u201d) to 5 (\u201cvery motivated\u201d), indicating high levels of motivation. A t-test was conducted to investigate if there were potential gender differences in motivation, which resulted in no significant differences (t = 0.48, df = 54.10, p > .05). Further, participants average perception of the amount of preparation time allotted was 6.33 (SD = 1.18), with the scale ranging from 1 (\u201cinsufficient\u201d) to 7 (\u201csufficient\u201d). There were no significant differences between the truthful (M = 6.29, SD = 1.36) and deceptive (M = 6.33, SD = 0.99) conditions in regard to the perception of preparation time (t = -0.19, df = 92.95, p > .05). Participants were also asked how truthful they were in their responses (percentage rating from 0%-100%), which resulted in a significant difference between the truthful (M = 94.23, SD = 25 13.19) and deceptive (M = 21.11, SD = 25.45) conditions (t = -18.67, df = 80.19, p < .05). This helped to validate the level of truthfulness in each condition.  3.3 SRP-IV 3.3.1 Reliability Cronbach Alpha\u2019s were calculated for the SRP-IV total scores, and for each subsequent subscale (see Table 4). As mentioned above, the SRP-IV has demonstrated good internal consistency with overall Cronbach Alpha\u2019s ranging from .86-.90 in previous validation studies (Gordts et al., 2017; Paulhus et al., 2016). Notably, the Cronbach\u2019s Alpha for the total SRP-IV for the current study was .89, with an average total score across participants of 140.9 (SD = 24.81).  3.3.2 SRP-IV Demographics & Comparisons Multiple t-tests with a Bonferroni correction were conducted to investigate potential gender differences in SRP-IV scores (see Table 5; see Figure 2). Firstly, Shapiro-Wilk tests were conducted to test the assumption of normality. With the exception of the criminal tendencies subscale that violated the assumption of normality (W = 0.88, p < .05), all other subscales met the assumption. Therefore, a Mann-Whitney-U t-test was conducted on the criminal tendencies subscale to correct for violation of the normality assumption given that the number of males was slightly over 30 (Mordkoff, 2016). Notably, the average total SRP-IV score for females was 135.55 (SD = 24.27), and was significantly different from males who had an average score of 153.42 (SD = 21.64), t = -3.71, df = 63.09, p < .05. Further investigation resulted in significant differences between genders in the interpersonal manipulation subscale where males (M = 43.81, SD = 8.92) scored higher than females (M = 38.45, SD = 9.13), t = -2.79, df = 57.51, p < .01. Another t-test was conducted on the callous affect subscale, which also revealed that males (M = 26 42.85, SD = 7.33) scored higher than females (M = 34.62, SD = 7.16), t = -5.11, df = 55.13, p < .01.  3.4 Preparation Time  The amount of preparation time that was used by participants was recorded for 79 participants. After a small portion of participants had completed the study, it was observed that there were potential differences in the usage of preparation time so it was only measured for a smaller portion of the sample. The average amount of preparation time in the deceptive condition was 12.37 minutes (SD = 5.7, range = 0 \u2013 20), while the truth-tellers averaged 2.5 minutes (SD = 3.7, range = 0 \u2013 13.3), which resulted in a statistically significant difference (t = 9.15, df = 67.07, p < .01). Multiple comparisons with a Bonferroni correction were completed to investigate potential differences between males (M = 1.21) and females (M = 2.77) in the truthful condition, which revealed no significant differences (t = 1.49, df = 17.09, p > .05). A subsequent comparison was conducted for the deceptive condition comparing males (M = 13.61) and females (M = 12.06), also resulting in no significant differences (t = -0.78, df = 13.00, p > .05). Lastly, a bivariate correlation was calculated for overall word count and preparation time, resulting in a small and non-significant correlation, r = -.02, p > .05.  3.5 Ghostwriter vs. Be Detailed A comparison as to the average number of words that were produced in the Ghostwriter condition as compared to the number of words in the Be Detailed condition was conducted using a t-test. A Shapiro-Wilk test was conducted to investigate if the assumption of normality was violated for word count, which was significant (W = 0.72, p < .05) indicating that the assumption of normality was violated. However, due to the robustness of t-tests with sample sizes above 30 participants, no corrective measures were deemed necessary (Mordkoff, 2016). The results 27 indicated a statistically significant difference in word count between the Ghostwriter condition (M = 746.55, SD = 609.46) and the Be Detailed condition (M = 457.71, SD = 269.63), t = -3.11, df = 67.73, p < .05 (see Figure 3). 3.6 Linguistic Cues A binomial logistic regression was conducted to investigate the relationship between the high-level descriptive linguistic categories created by LIWC (i.e., analytic, clout, authentic, tone, etc.) for true and false statements. The outcome variable was categorical (true\/false), while the independent\/predictor variables were the aforementioned linguistic categories collapsed across story-telling conditions. The descriptive statistics for the males and females in the deceptive category can be seen in Figure 4, while the truthful statistics can be found in Figure 5. Following this, multiple 2x2 Analysis of Variance\u2019s (ANOVA) were conducted on specified linguistic categories (i.e., positive and negative emotional language, disfluencies, personal pronouns, cognitive processes) with gender (male vs. female) and veracity condition (truthful vs. deceptive) as the categorical variables. These ANOVAs helped to reduce the complexity of creating a regression model given the small number of males in each condition, increasing the potential power in detecting gender differences as the analyses were conducted on a single linguistic category. For instance, including all the linguistic categories in a binomial regression would result in a series of Bonferroni corrections on an already small sample of males for veracity and story-telling conditions. Accordingly, ANOVAs on specified linguistic categories (i.e., affective language, personal pronouns, disfluencies, cognitive processes) were conducted in order to increase power.  28 3.6.1 Binomial Regression Binomial logistic regressions were used to test whether the variables in our model would be able to accurately predict if participants were being truthful or deceptive in their accounts. The regression equations were created for males (see Table 6) and females (see Table 7) separately, and were collapsed across story-telling conditions to ensure that there was an appropriate number of participants per group. A Bonferroni correction was applied using 10 terms in the model, resulting in significance being accepted when p < .005 (Tabachnick & Fidell, 2014). The male model was not better than chance at predicting if participants were being deceptive, \u03c72(10) = 13.19, p > .05. The same was found for the female model where it was not better than chance at predicting if participants were being deceptive, \u03c72(10) = -9.81, p > .05. However, it should be noted that both models were over-saturated due to the model complexity. Within the female model the effect of words associated with affect was considered statistically significant before Bonferroni corrections (see Table 7). However, this effect did not remain once multiple comparisons were accounted for (i.e., not significant).While there may be an effect of affective words in predicting if females are being truthful or deceptive, it appears that other variables also contribute in important ways when included in the overall model. Indeed, a larger sample size is required to elucidate this relationship.  3.6.2 Analysis of Variance  An ANOVA was conducted on specified linguistic categories based upon previous research that included positive and negative affective language, functional language (i.e., personal pronouns, disfluencies), and words associated with cognitive processes (see Table 8). The analysis was a 2x2 between subjects ANOVA with gender (male vs. female) and veracity condition (true vs. false) as the categorical variables (see Figure 6). Following this, another set of 29 ANOVAs were conducted utilizing the Dictionary of Affective Language (DAL) output, with the same categorical variables (see Table 10), which resulted in no significant main or interaction effects for each of the DAL linguistic categories. 3.6.2.1 Main Effects There was a main effect of gender only in the Ghostwriter condition for the use of personal pronouns, F(1, 47) = 7.09, p < .05, h2 = .13. Further post-hoc analysis testing all possible comparisons was conducted utilizing Tukey\u2019s Honestly Significant Difference (HSD). This revealed that there was a higher rate of personal pronouns (i.e., \u201cI\u201d, \u201cme\u201d) for females (M = 11.00, SD = 2.5) than males (M = 9.16, SD = 1.98), resulting in a significant difference between the groups, diff = 1.84 (95% CI [0.45, 3.23]), pAdjusted = .011. There was also a main effect of gender only for disfluencies in the Ghostwriter condition, F(1,47) = 21.20, p < .05, h2 = .08. Further post hoc-analysis revealed that males had a higher rate of disfluencies (M = 4.65, SD = 1.92) as compared to females (M = 3.28, SD = 2.36), resulting in a significant difference, diff = 1.37 (95% CI [0.043, 2.69]), pAdjusted = .043.  In the Be Detailed condition, there was a significant main effect of veracity for positive emotion words, F(1, 51) = 15.20, p < .05, h2 = .16. Post-hoc analysis revealed that there was a higher rate of positive emotion words in the deceptive condition (M = 3.53, SD = 1.4) as compared to the truthful condition (M = 2.48, SD = 0.95), resulting in a significant difference, diff = 1.05 (95% CI [0.39, 1.71]), pAdjusted = .002. It should be noted that there was not a main effect of veracity in the Ghostwriter condition as it is likely that it was captured by a significant interaction effect.  30 3.6.2.2 Interaction Effects There was a significant interaction effect between gender and veracity for positive emotional language in the Ghostwriter condition, F(1, 47) = 6.66, p < .05, h2 = .11 (see Figure 7). Further post-hoc analysis testing all possible comparisons was conducted utilizing Tukey\u2019s HSD (see Table 9), which revealed no significant differences once multiple comparisons were accounted for. However, it should be noted that there was a difference between males in the truthful and deceptive condition that was approaching significance (pAdjusted = .06). Males had a higher rate of positive emotion language in the deceptive condition as compared to the truthful condition, similar to the main effect described above, despite it only being a trend and not quite reaching significance. There was also an interaction between gender and veracity for disfluencies in the Be Detailed that was approaching significance (F(1, 47) = 3.12, p = .083, h2 = .055; see Table 8). 3.6.2 Analysis of Covariance  An analysis of covariance (ANCOVA) was conducted on the aforementioned linguistic categories (i.e., affective words, functional words, cognitive processes words) while including the total SRP-IV scores as covariates (see Table 11). Notably, the effect of psychopathy was a significant covariate for disfluencies in the Be Detailed condition, F(1, 50) = 6.61, p < .05, h2 = .10. Subsequent bivariate correlation analysis illustrated a significant negative relationship between general disfluencies and total SRP-IV score, r = -.32 (p = .019) in the Be Detailed condition. It should be noted that this same negative relationship was found in the overall sample (r = -.18, p > .05) and in the Ghostwriter condition (r = -.04, p > .05). The effect of psychopathy was also able to account for more of the variance for disfluencies, exposing a significant main effect of veracity, F(1, 50) = 4.59, p < .05, h2 = .072. Further post-hoc analysis using Tukey\u2019s 31 HSD for all possible comparisons resulted in a non-significant difference between veracity conditions once multiple comparisons were accounted for, diff = -0.91, (95% CI [-2.1, 0.31]), pAdjusted = .14. Tukey\u2019s HSD is considered a conservative approach to post-hoc comparisons in an effort to reduce Type I error, to which there could be significant differences between veracity conditions if a larger sample size were collected thereby increasing power (Barnett & McLean, 1998). This may partially explain the omnibus ANOVA significant effect, while post-hoc analysis did not reveal any significant differences. It is also possible that the conservativeness of Tukey\u2019s HSD also influenced the non-significant findings in the post-hoc analysis, specifically for the interaction effect with positive affective language in the Ghostwriter condition. Taken together, this suggests that there is a difference between veracity conditions for disfluencies, but that further research is required with a larger sample size to determine the precise effect. It should be noted that the effect of levels of psychopathic traits as a covariate was not significant for other linguistic categories or conditions, with results indicating small effect sizes.   32 Chapter 4: Discussion The current study examined how potential individual differences (i.e., gender and personality) may play a role in the production of verbal indicators of deceit while employing a recent and novel research design (e.g., Leal et al., 2019). Intriguingly, our study demonstrated that individual\u2019s responses can be quite impacted by a shift in methodology. Specifically, there was a significant increase in verbal output for the Ghostwriter instructions as compared to the Be Detailed instructions, replicating previous findings (Leal et al., 2019). The current study also illustrated that positive emotional language was a potential linguistic cue to deceit in one story-telling condition. In regard to gender differences in deceptive language, initial analysis indicated a significant interaction between gender and veracity for the Ghostwriter condition, while follow-up analysis yielded non-significant group differences. The group means did imply that males had a slightly higher rate of positive emotional language as compared to females. However, this was not significant and therefore contradicts previous research and was in a contrary direction the current study\u2019s hypothesis. The present study also revealed gender differences in total SRP-IV scores, and more specifically the callous affect and interpersonal manipulation subscales. It was also found that levels of psychopathic traits had a significant relationship with disfluencies in the Be Detailed condition. As levels of psychopathic traits increased, there was a decrease in the number of general disfluencies (consistent across story-telling and veracity conditions), which contradicts previous research. This difference might be attributable to sample characteristics as previous research (Hancock et al., 2018; Le et al., 2017) used entirely male samples, while the present study was primarily female. It is also possible that there were decreased levels of motivation for individuals who scored higher on psychopathy as there was no \u201creward\u201d for successfully deceiving the interviewer (Azizli et al., 2016). So while 33 there was some evidence of levels of psychopathic traits moderating the effect of disfluencies, it intriguingly was most apparent in one of the instruction conditions. Indeed, the current study clearly demonstrates that context and methodology are important aspects to consider in deception research. In fact, relatively minor alterations in instructions (or other considerations such as sample characteristics) served to facilitate our understanding of the aforementioned inconsistent results across deception studies.  4.1 Preparation Time One of the contextual factors considered in the current study was the use of preparation time between truthful and deceptive accounts. There were no significant differences found in the perception of preparation time between veracity conditions, which is consistent with previous findings (Leal et al., 2019). However, there were differences found in the actual amount of preparation time utilized where individuals in the deceptive condition took substantially more time to research the city they were going to discuss. Burgoon (2015) also found that individuals who are going to produce deceptive accounts engage in increased preparation time if given the opportunity. It has been surmised that preparation time allows for the individual to more effectively deal with (and ideally decrease) the cognitive load necessary to produce a deceptive account (Burgoon, 2015; Sporer & Schwandt, 2006). For example, a study investigating high-stakes fraudulent responses during a company conference call found that prepared responses contained more relevant details as compared to unprepared responses (Burgoon et al., 2015). Indeed, researchers have implied that planned messages may be particularly indistinguishable from unplanned messages for deceptive responses (Bond & DePaulo, 2006; Sporer & Schwandt, 2006). In the current study, deceptive participants appear to have engaged in increased preparation in order to include more relevant and authentic details. It should be noted that the 34 current study did not look at the types of details included, and further research is required to determine if preparation time affects perceived credibility of deceptive accounts, specifically as it pertains to some of the nuances around language. An additional explanation for the observed differences in preparation time is that it is likely that participants in the truthful condition would have previously discussed their trip with a family member, friend, or acquaintance before their involvement in the present study. This would have likely enabled truthful participants an opportunity to previously share details of their trip, perhaps even on numerous occasions. Therefore, less preparation time would be required when asked to discuss their trip for the present study. Indeed, research has found that the amount of times recounting (i.e., \u201crehearsal\u201d) results in an increase in the number of details, aiding in ease of recall at a later time (Himmer et al., 2019; Lindeman et al., 2017). For example, Lindeman et al. (2017) had participants recall a collection of recent memories during an intake interview. Results indicated that in a subsequent interview that as reported memory \u201crehearsal\u201d (such as thinking about the incident) increased so did the vividness of the recount. It was also found that pleasant memories typically did not fade as fast as negative memories, often resulting in increased vividness in detail. The slightly higher than average DAL pleasantness ratings in the current study suggest that participants were more focused on a pleasant event. As proposed by Lindeman et al. (2017), for participants in the truthful condition in the current study this would mean less preparation time was required and could also lead to an increase in detail. In sum, when asked to lie, participants took advantage of more preparation time when given the opportunity. However, further research is required to determine if preparation time influences the perceived authenticity of deceptive statements. Regardless of the amount of preparation time 35 utilized, the language produced by participants could also be influenced by non-contextual factors, such as individual differences.  4.2 Gender Differences in Language There were a handful of statistically significant linguistic patterns noted between genders. Firstly, it was illustrated that females tended to utilize more personal pronouns as compared to males. This is consistent with previous research that has also found significant gender differences for personal pronouns (Leaper & Robnett, 2011; Newman et al., 2008). Specifically, Newman et al. (2008) found that males discussed concrete information more often, leading to a decrease in personal pronouns. It was speculated that males use language as an archive to describe external events or processes, as compared to other conversation topics that could have themselves as the focus. For instance, males may choose to provide more details about the characteristics of the hotel they stayed at (such as whether or not it had a pool) which could take away from the amount of \u201cI\u2019 statements, such as discussing what they are experiencing. These researchers also pointed out that this finding goes against the commonly held stereotype that males discuss themselves more often (which would increase the number of personal pronouns). However, there could be other differences in language (e.g., disfluencies) that are more consistent with perceived gender stereotypes. In the present study males had a higher rate of disfluencies, which is also in line with previous research (Bortfeld et al., 2001; Newman et al., 2008). Past research has found gender differences in cognition (Siedlecki et al., 2019), and more specifically planning where females tend to outperform males (Naglieri & Rojahn, 2001). The gender differences observed for disfluencies in the present study suggests that females are not struggling as much as males when planning or formulating their responses to the interviewer. Interestingly, Bortfeld et al. (2001) 36 also found that disfluencies increase as planning difficulty increases. The notion that disfluencies indicate mental effort is based on the aforementioned studies of cognitive load (e.g., Bortfeld et al., 2001) and the proposition that constructing a lie is cognitively taxing (Newman et al., 2003; Vrij & Granhag, 2012; Vrij, 2008). Nonetheless, it should be mentioned that participants in the present study were admittedly not asked how difficult they found producing an untruthful response was for them. Future research would ideally include a post-interview questionnaire assessing cognitive effort to ensure that the lying paradigm for the study is indeed causing additional cognitive load. In sum, the present study found gender differences in language that could relate to cognitive ability, or the focus of conversation (i.e., concrete information). However, these results could also be further influenced by particular contextual factors. It must also be considered that these variations were found within a particular context where participants were asked to discuss a positive experience (a trip). Perhaps participants experienced slightly elevated levels of social demand and pressure as they were speaking to an interviewer they were unfamiliar with, and in some cases were asked to lie. Given that the social constructivist perspective implies that gender differences may be more noticeable with different situational factors (i.e., social demand), it is possible that there would be different results if some contextual factors were adjusted (e.g., motivation for lying, different topic). In fact, the current study\u2019s gender differences found in disfluencies and personal pronouns were only found in the Ghostwriter condition, again demonstrating that the slight adjustment in instructions between the story-telling conditions does have an effect on verbal output. Specifically, it would appear that when participants are asked to produce their stories utilizing instructions that provide a framework meant to optimize the thoroughness, there was a more pronounced effect of gender differences in language.  37 4.3 Story-telling Conditions The Ghostwriter condition is part of a novel methodology that was implemented in an attempt to not only increase generalizability, but to also replicate previous findings from Leal et al. (2019). As a matter of fact, these findings were replicated revealing that participants who were instructed using the Ghostwriter script produced significantly longer statements about their trip as compared to individuals in the Be Detailed condition. As considered already briefly above, a simple adjustment in initial instructions led to increased verbal output, and may also alter the interpretations as to the veracity of those accounts. Regarding the Ghostwriter instructions, Leal et al. (2019) speculated that the increased output is due to a clear framework (or perhaps, a cognitive schema) that the Ghostwriter method provides. The Ghostwriter\u2019s more precise set of instructions could also serve to clarify and aid in reducing any anxious feelings that participants are experiencing. The Be Detailed instructions only ask participants to \u201cprovide as much detail as possible,\u201d with no clear set minimum or maximum stated. This conceivably could cause some participants to feel anxious as they would like to ensure they are giving the appropriate amount of detail. The Ghostwriter instructions could potentially decrease anxious feelings on behalf of the participants and allow them to more openly discuss their story. Regardless, the Ghostwriter directions resulted in more verbal output which could then impact judgments of veracity as more information would be available to verify (Leal et al., 2019).  The increased verbal output may also be particularly effective in increasing the opportunity for verbal cues of deceit to be observable (Leal et al., 2018; Leal et al., 2019). For instance, it is probable that the increased output provides an opportunity for linguistic cues of deceit to \u201cleak\u201d out, increasing deception detection accuracy. Further, more information also affords an additional opportunity for the details to be verified (Leal et al., 2019; Vrij, 2018). For 38 instance, if an individual states that they stayed at the Four Points Hotel, this can then be validated through records (e.g., financial, communication, receipts). Based on the present findings, the Ghostwriter method demonstrates promise as an effective method to facilitate uncovering a deceptive response, as well as compiling more information.   The Ghostwriter method also shows promise as an effective strategy to employ in some interview settings where the main goal is typically to elicit as much information as possible (Brandon, 2014; Fisher, 2010), particularly during investigative interviews (Vrij et al., 2017). Interestingly, the Canadian RCMP transitioned to a Phase Interview Model (PIM) in 2014, where the goal is to garner as much truthful information as possible (Snook et al., 2020). This is in comparison to previous investigative models that have a more accusatorial stance (e.g., Reid technique) and an emphasis on gaining a confession (Inbau et al., 2011). Further, witness (to which PIM has also been adapted for) and clinical interviews (Vrij et al., 2017) also aim to collect as much information as possible from the interviewee, to which the Ghostwriter method looks promising in facilitating. The contextual findings of the current study (i.e., instructions, preparation time) could also provide an explanation for the aforementioned inconsistent results in deception research. Indeed, even slight adjustments such as the topic of discussion, or the modality of how participants made credibility judgements could also influence the results. For instance, Forrest et al. (2004) had participants view truncated video clips, while Porter et al. (2002) had participants listen or watch entire statements. In each study, there was little discussion as to the instructions that were given to participants for the pre-recorded statements. However, the present findings emphasize that how the statements were recorded can partially explain the contradictory findings. It should be mentioned that the current study did not have interviewers make credibility 39 judgements of the participants. Nonetheless, based on the present findings corroborating Leal et al. (2019), tailoring instructions can facilitate more information, which would ultimately aid the interviewer, help explain inconsistent results in deception research, and influence the observed linguistic cues to deceit.   4.4 Linguistic Cues to Deceit Interestingly, a linguistic cue to deceit that was found in the Be Detailed condition in the present study was for positive affective language. However, somewhat surprisingly, the presence of positive emotion words was found to be a significant predictor of an individual being deceptive, contradicting previous research and theory. In fact, proponents of the theory of reality monitoring would predict that there is a higher rate of overall emotional language in truthful accounts (Bogaard et al., 2016; Sporer, 2004; Vrij et al., 2004). Reality monitoring is based on memory, and conjectures that a real memory will contain more information in general, including emotional information (Bogaard et al., 2016; Sporer, 2004). Interestingly, research investigating changes in language between truthful and deceptive accounts have illustrated that there does not tend to be a change in affective language (DePaulo et al., 2003; Hauch et al., 2015), countering the theory of reality monitoring. This has also been found specifically for positive affective language (DePaulo et al., 2003; Hauch et al., 2015). For instance, while a recent meta-analysis predicted that there would less positive affective language for liars, it was found that there was no difference between veracity conditions (Hauch et al., 2015). Based on these findings the authors suggested that a change in positive affective language is in fact not an indicator of deceit. It should be mentioned that positive emotional language being an indicator of deceit in the current study was only found in the Be Detailed condition.  40 A potential explanation for the unexpected finding is that describing a recent trip tends to be more commonly centered around a positive experience (rather than focusing on some negative aspect). While this could occur regardless of the account being truthful or deceptive, when producing a deceptive account participants might be attempting to conform to a positive schema of a trip. In addition, adopting a positive schema may act as a strategy to distract the listener with a compelling and exciting story that might also sound more authentic. For example, it could be that the increased positivity distracts the interviewer from problematic details, or that it more likely engages the interviewer in a way where they are less inclined to have any concerns with the fabricated story. This is corroborated by the DAL pleasantness rating, which was slightly above average, and the positive valence levels from LIWC as they were substantially higher (than negative valence) for the entire sample. It is also possible that these accounts were enhanced by integrating information from aspects of previous genuinely experienced trips (Vrij, 2010). For instance, a participant may have never gone to Hawaii, but they found their plane trip enjoyable on their last flight to Vancouver and would therefore incorporate that into their deceptive account of a fabricated trip to Hawaii. Indeed, research has found that self-reported good liars employ this integration of genuinely previously experienced content as a tactic in an attempt to increase the believability of their statements (Verigin et al., 2019). Self-reported good liars described employing this as a tactic, regardless of it adding any objective credibility to their statements. The actual effectiveness of this tactic does warrant further investigation as some research has demonstrated that it is effective (Porter & ten Brinke, 2010; Vrij, 2010), while more recent research has found that it does not affect the quality of the statement (Verigin et al., 2020).  Alternatively, the significant difference between truthful and deceptive accounts for positive emotional language in the present study could also be explained by the inclusion of 41 more negative or neutral information in truthful accounts. Research has illustrated that there are different types of details that are included in truthful and deceptive accounts, which includes common knowledge details, complications, and self-handicapping details (Hartwig et al., 2007; Vrij et al., 2018; Vrij, 2018). In particular, it has been found that truth-tellers include more complications (e.g., \u201cthe airline lost our luggage\u201d) in their accounts as compared to deceivers (Leal et al., 2018; Vrij et al., 2018). Complications may be viewed as a more neutral or negative affective event, so in that case truth-teller\u2019s stories could have less positive emotional language. Leal et al. (2017) speculated that deceivers do not include complications in their accounts as it makes their story more complex and difficult to track (and produce more disfluencies due to cognitive load). The increased complexity would also provide more undesired opportunities for interviewers to question if their story was truthful. In the present study it is possible that truthful participants included more of these complicated details leading to a significant decrease in positive emotional language, with more positive emotional language being associated with deceptive stories. The LIWC output did have a slightly higher rate of conjunctions, words indicative of analytic thinking, and prepositions (all thought to be indicative of complexity; Tauszik & Pennebaker, 2010) in the truthful condition. Regarding the Dictionary of Affective Language (DAL; Whissell, 2009), the current study did not find any differences noted between veracity conditions or gender. Again, this could be explained by the nature of the procedure as participants were asked to discuss a recent trip, where the majority appeared to have chosen to focus on a positive experience. Given that all participants discussed a trip, regardless of story-telling or veracity condition, it can be expected that emotionality (in this case positive affect) would remain consistent across the DAL categories. As mentioned, the average pleasantness rating was slightly above average, 42 demonstrating that, while not a significant finding, the general sample was skewed more towards positive affective language. Regardless of veracity condition, they were more inclined to adopt a positive schema. If participants had the opportunity to freely discuss any topic it is likely that there would have been more variation among the DAL categories given that each story would have different emotional valence. It should also be mentioned that there was a miniscule amount of variability for the DAL scores across all three categories, and it is likely that the range was too constricted to detect any meaningful differences.  Interestingly, there were also not any differences between veracity conditions noted for functional language (i.e., personal pronouns, disfluencies), which has previously been found as some of the most salient verbal cues of deception (DePaulo et al., 2003; Hancock & Woodworth, 2013; McQuaid et al., 2015; Vrij, 2018). While it did not reach statistical significance, the present study did note a small decrease in the number of personal pronouns used in the deceptive condition, which is consistent with previous research (McQuaid et al., 2015). Also similar to previous research, there was a small increase in the amount of disfluencies in the deceptive condition, although again it did not reach statistical significance. This lack of foundational language differences might be in part explained by the contextual implications mentioned above, such as the Ghostwriter instructions and preparation time. For instance, as previously considered, it has been speculated that increased preparation time provides an opportunity for deceptive accounts to sound more truthful (Bond & DePaulo, 2006). In this case, the use of preparation time could have decreased the magnitude at which disfluencies and personal pronouns were affected. If participants were given less preparation time, it is probable that there would have been a more noticeable difference in functional language between veracity conditions. This has 43 the implication that functional language as a predictor of deceit may be easily influenced by contextual factors, adversely impacting their reliability.  It should be reiterated that personal pronouns and disfluencies are not direct indicators of deceit per se, but are indicative of cognitive processes (i.e., psychological distancing, cognitive load) that influence language. For instance, as considered briefly earlier, the current study\u2019s task was potentially not cognitively demanding enough (regardless of veracity) to create observable differences in disfluencies. Alternatively, participants who were planning a trip could have strategically embedded more readily available details of a location into their narrative, making it less cognitively taxing, potentially impacting the rate of disfluencies. In sum, the magnitude of change for disfluencies and personal pronouns as indicators of deceit did not reach statistical significance, although it was noted that there was a change in the same direction as previous research (Hancock & Woodworth, 2013; McQuaid et al., 2015). The lack of functional language differences also potentially undermines the reliability of verbal indicators of deceit that has previously been found. Indeed, this might be better understood with a consideration of potential gender differences in affective linguistic cues to deceit. 4.4.1 Gender Differences & Linguistic Cues to Deceit Interestingly, a significant interaction between gender and veracity was found specifically for the Ghostwriter condition, despite non-significant follow-up analysis. Notwithstanding, a comparison of the group means illustrated two group differences that warrant further consideration given the focus of the study and hypotheses. First, it was identified that males in the deceptive condition had a higher rate of positive affective language as compared to males in the truthful condition that was approaching significance. Further, males also had more positive affective language in the deceptive condition as compared to females, although this was less of a 44 pronounced (and non-significant) finding. Indeed, the non-significant post-hoc analysis also contradicts the current study\u2019s hypothesis, which suggests that there were not any significant gender differences. Previous research has shown that there are gender differences in emotion regulation (McRae et al., 2008; Rogier et al., 2019; Zou et al., 2019). Particularly, it has been shown that females potentially engage in higher levels of emotional down-regulation, whereby they suppress their emotions as a management strategy (Rogier et al., 2019; Zou et al., 2019). Due to the novel context of the present study, participants possibly suppressed their emotions for various reasons, such as not knowing the interviewer, ensuring that they wanted to give as much detail as possible, or ensuring that they completed the study procedure correctly. As demonstrated by Rogier et al. (2019), females in the present study might have then engaged in higher levels of this type of emotional management resulting in a decrease of positive affective language. It should be reiterated that these potential explanations are based on trends within the group means that imply gender differences in positive affective language. The follow-up analysis did not reveal significant group differences among these group means and these interpretations should be taken with caution. Indeed, the initial ANOVA (where a significant interaction effect was indicated) may even reflect a spurious finding, with a larger sample size required to better elucidate this relationship. Indeed, personality may further influence affective language, as it was found that there were gender differences in levels of psychopathic traits. 4.5 Levels of Psychopathic Traits Gender differences in levels of psychopathic traits (i.e., SRP-IV) were found in both the overall scores, as well as in the callous affect and interpersonal manipulation subscales. Largely consistent with previous literature, males scored higher on the total SRP-IV as compared to females (Beryl et al., 2014; de Vogel & Lancel, 2016; Guay et al., 2018). Notably, previous 45 research has utilized community or forensic psychiatric samples where there tends to be a broader range of scores for levels of psychopathic traits, including participants scoring on the higher end of the spectrum (e.g., Guay et al., 2018). The present study expands these findings to a Canadian university sample. It should be noted that despite most participants scoring on the lower end of the spectrum, there were still significant gender differences found. Specifically, it was demonstrated that there were gender differences in the callous affect and interpersonal manipulation subscales where males tended to score higher than females. The callous affect subscale reflects characteristics such as being more insensitive and having more disregard for others. Further, the interpersonal manipulation subscale pertains to characteristics such as increased attempts to manipulate or deceive others. This is consistent with previous literature that has found that the lifestyle and antisocial facets of the PCL-R (Hare, 2003) are typically more prone to gender differences, with males tending to score higher (Guay et al., 2018; Richards et al., 2003). However, it should be mentioned that this does oppose some previous research that has theorized (Forouzan & Cooke, 2005) and found (Strand & Belfrage, 2005) that females score higher on measures of interpersonal manipulation. Regardless, the gender differences that were found across two of the facets of the SRP-IV in the current study provides additional support that a different four facet structure should be considered to assess males and females (de Vogel & Lancel,2016; Guay et al., 2018). The higher scores for males on the interpersonal manipulation subscale could reflect an increased propensity for deceit, or that they view lying as more acceptable. As previously considered, research has found that psychopathic offenders include more relevant details in deceptive accounts, leading to them to sound more authentic (Lee et al., 2008). However, there has been research illustrating that there are differences in the types of details included in 46 fabricated stories (e.g., common knowledge details; Leal et al., 2018; Vrij et al., 2018). In an attempt to sound authentic, those who score on the higher end for levels of psychopathic traits likely include more common knowledge details (e.g., Kelowna is located around Okanagan Lake). However, there would be little change in self-handicapping and complication details. Other research has also illustrated that psychopaths view lying as more acceptable, and are potentially more likely to engage in deceit (Jones & Paulhus, 2017; Wright et al., 2015). Callous affect could also influence the propensity for deceit as these characteristics are associated with decreased guilt about lying since they disregard others more readily. Additionally, the higher rates of interpersonal manipulation for males in the present study could relate to the higher rates of positive emotional language. Males who scored on the higher end of this spectrum might have more readily employed positivity as a tactic to deceive the interviewer. In sum, the current findings imply that males may have a more acceptable view of lying, and that they are potentially more likely to engage in deceit by including more relevant details and not feel as much guilt as females about deceiving.  As mentioned, males scoring higher than females on the measure of interpersonal manipulation in the current study is inconsistent with some previous research (Cunliffe et al., 2016; de Vogel & Lancel, 2016; Strand & Belfrage, 2005). While some research has found that females do score higher on measures of manipulation (de Vogel & Lancel, 2016; Strand & Belfrage, 2005), other studies found that males score higher (Beryl et al., 2014). Interestingly, Cunliffe et al. (2016) propose that there are not gender differences in manipulation per se. They suggest that while females may utilize more subtle forms of manipulation, males could be more overt. They also noted that females with higher rates of psychopathy tend to be more concerned with how others perceive them as compared to males. In terms of the present study, it could be 47 the case that there was a higher social desirability effect for females, which would then lead to a decrease in the self-report of manipulative behavior. Indeed, research has found that females are more prone to responding in a desirable and ethical manner (Dalton & Ortegren, 2011). In sum, the current results illustrated gender differences in levels of psychopathic traits, specifically for the interpersonal manipulation and callous affect subscales, providing important implications for future research. 4.5.1 Psychopathy & Linguistic Cues to Deceit Other intriguing findings were uncovered when considering psychopathy\u2019s effect on language, specifically that levels of psychopathic traits were related to disfluencies particularly in the Be Detailed condition. It was also revealed that there was a significant difference between veracity conditions. However, post-hoc analysis did not reveal any statistically significant group differences, which may be in part explained by the aforementioned conservativeness of Tukey\u2019s HSD in determining group differences. As discussed, this suggests that there could be a difference between veracity conditions for disfluencies, but that further research is required with a larger sample size to determine the precise effect. Nonetheless, it is worth mentioning that, despite it not being significant, this increase in disfluencies for deceptive accounts is in the same direction as previous research. Since the effect of levels of psychopathic traits was not significant for other linguistic categories or conditions, the findings did not support the stated hypothesis. It should be mentioned that the significant relationship between levels of psychopathic traits and disfluencies was found only in the Be Detailed condition. Since the story-telling condition did not put as much pressure on participants to provide as much detail as possible, psychopathic traits had a larger impact on disfluencies. For example, as discussed, the Ghostwriter instructions provides a clearer framework for participants to follow. However, these instructions could also 48 put more pressure on the participants to provide more details, whereas it is possible that there was not as much pressure to provide all the details in the Be Detailed condition. Therefore, this decreased pressure might have required less cognitive load and allowed for the effect of levels of psychopathic to be more prominent, attesting to the notion that contextual factors influence deception and may in fact interact with personality. Regardless, further research is required to elucidate its effect given the relationship between levels of psychopathic traits, deception, and in some cases criminal offending that involves conning and manipulation.  Previous research investigating the linguistic markers related to psychopathy have found that there is an increase in the number of disfluencies and personal pronouns for higher levels of psychopathic traits (Le et al., 2017). The current study also found that levels of psychopathic traits had a significant relationship with disfluencies. There was also a relationship between psychopathy and personal pronouns that was approaching significance. Interestingly, additional analysis revealed that as levels of psychopathic traits increased, the rate of disfluencies in general decreased, disaffirming previous research (Hancock et al., 2018; Le et al., 2017). This was consistent across both story-telling and veracity conditions in the present study, but it was found that there was a weaker relationship in the Ghostwriter condition. One explanation for this is differences in sample characteristics in each of the aforementioned studies as compared to the current study. For example, Le et al. (2017) and Hancock et al. (2018) utilized entirely male samples, whereas the majority of the current study\u2019s sample was female. As mentioned, disfluencies are also thought to indicate increased cognitive energy for particular tasks. A possibility is that females with higher levels of psychopathic traits are not as impacted by the cognitive demands of story-telling. Indeed, in the aforementioned studies there were more pauses and disfluencies in the exclusively male samples (Hancock et al., 2018; Le et al., 2017). It is also 49 important to consider that Le et al. (2017)\u2019s study assessed the solely male sample using the PCL-R. The current findings (in addition to other research) begs the question as to whether the PCL-R is appropriate to generalize to females (Beryl et al., 2014; Guay et al., 2018). Some researchers have recommended that there should be lower cut-off scores for the PCL-R for females (Guay et al., 2018), or that the PCL-R should be adjusted to more accurately reflect how psychopathy manifests within females (Wynn et al., 2012). Further research is required assessing the validity and reliability of the PCL-R within females.  Another explanation relates to the \u201cstakes\u201d of the lie itself in the current study. Participants reported high levels of motivation for the study, but it was relatively low-stakes in that if a participant were to be \u201ccaught\u201d being deceptive there was no consequence, nor no particular reward for being successful at deceiving the interviewer. This is clearly different than a genuinely high-stakes scenario, such as the aforementioned press conferences study (McQuaid et al., 2015), or other studies that would have an identifiable gain (e.g., monetary reward) if they were successful. Previous research has found that as levels of psychopathic traits increase, there is an increased likelihood that individuals engage in deceit, particularly if there is an observable gain (Azizli et al., 2016; Jones & Paulhus, 2017). The scenario in the present study might not have been a high enough stakes situation for individuals with higher levels of psychopathic traits to completely engage in the study. This could lead to decreased cognitive effort invested by some participants (particularly those scoring higher on psychopathy). Indeed, the interpersonal manipulation facet of levels of psychopathic traits suggests that the propensity to engage in deception is typically for some identifiable gain. Regarding emotional language, it was also found that levels of psychopathic traits did not influence positive and negative affective language. This is inconsistent with previous literature 50 that has found that a lack of emotional language is an identifiable characteristic of higher psychopathy levels (Hancock et al., 2018; Porter & Woodworth, 2007). This is comparison to individuals with lower levels of psychopathic traits that may have higher rates of emotional language. A potential explanation for these findings could relate to the restricted range of scores for levels of psychopathic traits (specifically at the higher level). Other research has shown that participants who score on the higher end of the spectrum have the most blunted emotional expression (Hare, 2003; Verschuere et al., 2018). The scores in the lower half of the spectrum in the current study could reflect a larger range of emotional variability. Indeed, a wider range of psychopathic traits is important to obtain in future studies. Notwithstanding, the current study still found a moderate effect of SRP-IV scores on disfluencies with follow-up analysis indicating that as levels of psychopathic traits increased, the presence of disfluencies decreased.  The current study also illustrated that there are gender differences across particular facets of psychopathic traits (i.e., interpersonal manipulation, callous affect) that influence deceptive language. One explanation is that in addition to the current contextual findings (i.e., preparation time, instructions), there is also an interaction between personality and the current research context. A recent study illustrated that both situation and personality aspects can influence deception such as the participants engagement and propensity to lie (Markowitz & Levine, 2020). It was found that the effect of personality, including the Dark Triad (Machiavellianism, Psychopathy, Narcissism), fluctuated given the situational context. Specifically, it was found that even with high levels of baseline honesty, participants were more likely to cheat and lie if there was a low probability of being caught (contextual factor), regardless of levels of the Dark Triad. However, higher levels of the Dark Triad indicated a higher likelihood to cheat even when it was more likely that they would have been caught. Regardless, this would propose that given 51 different contexts the effect of personality characteristics, (in the case of current study, a specific facet of level of psychopathic traits), could impact salient linguistic indicators of deceit differently. For instance, the present study\u2019s methodology had participants discuss a recent trip or vacation, to which it seems that participants attempted to portray as a positive event, particularly when they were being deceptive. However, it is possible that levels of psychopathic traits would have a different affect if participants were to discuss a negative event (e.g., funeral) or less emotionally salient topic (e.g., description of how a table is made). For example, it could be more difficult for individuals with higher levels of callous affect (in the case of the present study, males) to utilize positivity to deceive the interviewer if given a different topic. There are also an abundance of situational factors to consider that interact with personality to influence verbal indicators of deceit such as rapport, incentives, and location, amongst others. Taken together, there are several noteworthy findings regarding personality, gender, and contextual factors that impact future deception research.  4.6 Implications There are a number of important considerations that stem from the current results. First and foremost, there are implications for how instructions are utilized in general interview, clinical, or investigative settings with the replication of the Ghostwriter methodology (Leal et al., 2019). As mentioned, previous methodologies have varied along a continuum from dynamic-interactive to non-interactive (passive) tasks (Bogaard et al., 2016; Gerlach et al., 2019; Vrij, 2018). The current study implemented a novel, more interactive methodology that resulted in increased verbal output and provided support for the affect that adjusting initial instructions has on verbal output. While not entirely interactive, the current study did have participants initially interact face to face with the interviewer, increasing generalizability. Interviewers, in either 52 research or practical settings, should consider implementing slightly adjusted instructions to obtain increased verbal output for potentially higher rates of pertinent information. This could be important regardless if the aim is to determine if the person is being deceptive or truthful. Importantly, the current study demonstrates that contradictory results in deception research may be better understood by the influence of differing methodologies. Secondly, the current study provided support for previous research that has demonstrated gender differences in levels of psychopathic traits (de Vogel & Lancel, 2016; Guay et al., 2018). For instance, it was found that females scored lower in levels of psychopathic traits as compared to males, which is mainly consistent with previous research (de Vogel & Lancel; Guay et al., 2018). The current study also expands on previous literature, which has largely found these results in forensic settings (e.g., de Vogel & Lancel, 2016), and found that these gender differences were also present in the current university sample. It also demonstrated that gender differences in levels of psychopathic traits are consistent across the spectrum of psychopathy as gender differences were found despite the restricted range of scores. In comparison to previous studies that utilized exclusively male samples, it was also found that there was a negative relationship between levels of psychopathic traits and disfluencies (e.g., Le et al., 2017). This provides further evidence that psychopathic traits require different cut-off scores for females (de Vogel & Lanvel, 2016; Guay et al., 2018), or that the assessment procedure itself requires adjustment based upon gender (Wynn et al., 2012). In sum, psychopathy appears to manifest differently across genders, and this should be considered when investigating the personality construct in various research and clinical contexts.  53 4.7 Limitations The current study has several limitations. While the overall sample size was adequate for many analyses, there was a smaller number of males once divided across all of the conditions (see Table 1). For some analyses this adversely impacted power, particularly for detecting linguistic characteristics associated with statements for males in each specific story-telling condition. As a general rule of thumb, there should be approximately 15-30 observations per variable when conducting a logistic regression. In this case, the number of observations was well below the ideal number. Additional research is required with larger sample sizes. It should be noted that data collection was ended early due to circumstances beyond the researchers control (i.e., COVID-19). Nonetheless, as illustrated above there were still a number of important findings to consider in the present study. In fact, the majority of analyses, specifically regarding instructions and levels of psychopathic traits, did have a large enough sample to conduct meaningful analyses.  Another potential limitation of the present study is the aforementioned range of scores observed in the SRP-IV scores. While our results were consistent with previous research that has found gender differences on measures of levels of psychopathic traits, our truncated range of scores may have influenced the effect psychopathic traits had as a covariate. Is it possible that there would have been some additional affects if participants were included that scored on the higher end of the spectrum for the SRP-IV. Despite the benefit of considering our results exclusively within a university sample, future research should also aim to obtain participants from a non-university sample (such as a prison population) which will have a greater range of psychopathy scores to consider. A truncated range of psychopathy scores is expected given the university sample. A more generalizable assessment of personality could have also been utilized 54 in the current study (e.g., Big Five Questionnaire). While psychopathy was of particular interest in the current study, this would have allowed for considering a greater breadth of personality constructs, as well as expand the ecological validity of the current study to other interview scenarios (e.g., job interviews) where it would also be expected that levels of psychopathic traits are in the lower range. Indeed, future studies may want to consider including other personality measures to further investigate how other aspects of personality might also influence language in a deceptive context. As discussed, this study utilized a novel research design aimed at increasing the interaction between the interviewer and interviewee as previous research has employed relatively passive methodologies (Gerlach et al., 2019; Vrij, 2018). This increases the generalizability of the current study to a more dynamic scenario that would take place during an interview. Despite aiming to be more interactive than many previous studies, it is acknowledged that the conversations were relatively one directional (i.e., interviewer asking one question to the participant). The type of interaction between individuals in the current study is not really like what would more typically take place during most everyday conversations and is perhaps better suited to inform more formal questioning or interview type contexts.  It should be mentioned that all interviewers utilized in this study were female interviewers, in an attempt to reduce the unknown variance added by having interchanging male and female interviewers. However, this could also be a limitation as the interactivity between male and females could not be measured. Previous research has found that there are gender differences in interpersonal interactions (Hall & Matsumoto, 2004; Hall & Schmid Mast, 2008; Lambert & Hopwood, 2016), with some studies suggesting it might influence deception detection ability and deceptive output (Lloyd et al., 2018).The current results also do help to 55 provide additional insights for the effect gender could have during an interview type setting, specifically in regard to deceptive language. Nonetheless, future studies should also endeavor to consider the interactions of various gender combinations. Another potential limitation of the current study is that we did not ask participants to identify their cultural identity. It has been noted that culture should be considered as a contextual factor when investigating language (Chaplin, 2015), as well as potential gender differences in language (Shields, 2012). The current study\u2019s results do not generalize to individuals whose primary spoken language is not English, which forms a small part of one\u2019s cultural identity. For instance, code switching (or language alteration) occurs when an individual alternates between two or more languages in a single conversation (Beatty-Mart\u00ednez et al., 2020; Hlavac, 2011). Research has demonstrated that when code switching occurs there tends to be more pauses, including disfluencies, which, as discussed, have been implicated as a cue to deceit (Hlavac, 2011). Indeed, language is one aspect of culture to consider as research has also demonstrated that other cultural variations and norms also influence how an individual interacts with others (Thakker et al., 2017). Future research should aim to more carefully consider culture, as well as other individual factors such as socioeconomic status, as potential contextual factors influencing the production of language.    Lastly, the current study\u2019s results should also be contextualized within the current social and political climate. Social movements such as Black Lives Matter (BLM), have increased the amount of attention on the systematic discrimination that has been perpetuated by various police forces (Bleich et al., 2019; Nadal et al., 2017) against those who identify as BIPOC (Black, Indigenous, People of Color). Notably, the results of the current study should not be used to inform public policy that may further discriminate against those who identify as BIPOC or based 56 on gender. Discounting the cultural identity of the individual could lead to further discrimination (either implicitly or explicitly), and lead to a more biased and incorrect veracity judgement. Indeed, the social constructivist perspective considers how language characteristics are based on the in the moment interactions (Shields, 2002). Interviewers (including police officers), also need to consider factors that influence suggestive cues to deceit that are in the moment. For instance, disfluencies should not be used as a \u201cone size fits all\u201d cue to deceit as research has shown that this linguistic domain (as well as others) is influenced by other factors such as culture and cognitive load. As described earlier, if an individual is engaging in code-switching (when English is not their primary language), this could lead to additional disfluencies, which enhances the risk of incorrectly increasing the perception of lying. There should also be careful consideration of one\u2019s own biases and identity (culture, gender, socio-economic status, occupation, etc.), as well as others, and how that influences their interactions with others. Indeed, an emphasis should be put on developing rapport in these situations, to which the PIM model for the RCMP has included as an integral aspect of interviews (Snook et al., 2020). In sum, the results of the current study should not be used to inform public policy, or used as the foundation for the \u201cPinocchio\u2019s nose\u201d to deceit, but rather should be used to expand the breadth of factors to be mindful of for those in the moment interactions when an individual is required to make a veracity judgement. The current results should be used to increase the knowledge base in the deception field as to factors that could warrant further investigation. 4.8 Future Directions As mentioned, the current study utilized undergraduate participants and prospective research should expand into other potential samples and contexts. Specifically, further research is required to investigate verbal cues to deceit in the general population, and specifically the effect 57 of gender and levels of psychopathic traits. Relatedly, previous research has also found that personality and situational factors interact (Markowitz & Levine, 2020). Measures of anxiety, other personality measures (i.e., Big Five), and querying the cognitive load experienced by participants could also be included in future studies as other potential individual difference factors that influences deception. As mentioned, the current study\u2019s results primarily generalize to university undergraduate participants. Therefore, future research should aim to not only expand to other populations and examine culture as a contextual factor, but also consider replicating current and previous findings in other contexts (i.e., changing topic of discussions, instruction alterations, etc.).  Another potential area for expansion is to replicate the current findings with a repeated measures design, which would provide a baseline comparison of verbal deceptive cues. Previous deception research has employed baseline comparisons in a limited fashion that does not allow for increased individualistic comparisons of verbal indicators to deceit (Vrij, 2018). For instance, it could be that personality does play a larger role in verbal output when producing a deceptive statement but that some of its effect is not accounted for without baseline comparisons. This would also provide additional insight into if there are individualistic changes from truthful to deceptive responses. This can also increase generalizability to other scenarios and contexts where there is rapport established between the interviewer and interviewee. Further, as discussed, the statistical analyses used in the current study were ANOVAs. ANOVAs were ultimately determined to be more appropriate than a regression model given the small number of males in each condition, increasing the potential power in detecting gender differences. However, logistical regression would be better suited for future studies with larger sample sizes investigating a combination of personality and contextual factors. This would allow for a more 58 precise examination of the combined effect these factors could have on language in determining if an individual is being deceptive or truthful. The current study also focused solely on the participants linguistic output and did not ask interviewers to make a judgement as to the veracity of their responses. Future studies might consider including multiple participants in the procedure, such as interviewers, after which they are asked if they believe the interviewee is being deceptive or truthful and what cues they utilized in making this assessment. Indeed, a potential extension would be to have participants review transcripts of the stories produced in the current study and make judgements as to their veracity. This would allow for an investigation into how contextual factors (i.e., preparation time, initial instructions) influences the perceived authenticity of the accounts. Another future avenue to explore is including a training opportunity for some participants where there is a brief discussion as to some verbal indicators of deception, and compare participants who receive training against those who do not receive training. While not only providing additional insights into the reliability and generalizability of linguistic cues to deceit, it would also aid in applying this knowledge in an interactive context. Future studies should also consider altering the gender of the interviewer, as well as the interviewee, to determine if there are differences when speaking to an individual of a different gender. Specifically, the linguistic cues to deceit should be considered in this scenario and how they change depending on who is speaking. For instance, it is plausible that verbal cues to deceit fluctuate depending if a male interviewer is speaking with a male interviewee or a female interviewee. This interaction between the participants could be an additional factor to consider as individuals alter their approach depending on the person they are speaking to and requires further investigation. 59 4.9 Conclusions  In summary, there are a number of both intriguing and important findings in the current study. Foremost, it was illustrated that alterations in initial instructions significantly influences the amount of verbal output by interviewees. Importantly, this (along with differences in preparation time used by deceptive and non-deceptive participants) helps to possibly explain previous inconsistent results in deception research. As mentioned, this also has implications for various interview settings (e.g., investigative, clinical, and witness interviews) where the main goal is to garner as much information as possible (Brandon, 2014). Increased verbal output also presents an opportunity for additional linguistic cues of deceit to be observable, and the amount of details that can be verified. The current study also expands upon previous research that has illustrated gender differences in language (see Leaper & Robnett, 2011) as it was illustrated that females utilized personal pronouns more often, while males had a higher rate of disfluencies. Positive affective language was an indicator of an individual being in the deceptive condition, which is inconsistent with extant research. This could reflect a cognitive schema that was adopted by participants, or the fact that positivity was used as a way to distract the interviewer. It should be reiterated that this was found in only one of the story-telling conditions (i.e., Be Detailed). There were also no differences found between veracity conditions for functional language, potentially undermining their utility as a linguistic cue to deceit. Indeed, it is possible that these results are further influenced by personality. In the current study males tended to score higher on measures of psychopathy, particularly in domains of interpersonal manipulation and callous affect. This contributes to the growing body of literature that has found that levels of psychopathic traits manifest and present differently in males and females (Guay et al., 2018). Indeed, in comparison to previous studies 60 that have utilized exclusively male samples, it was found that levels of psychopathic traits had a negative relationship with disfluencies, potentially corroborating gender differences in the presentation of psychopathy. Interestingly, this is contrary to previous research that has found a positive relationship, which is more than likely related to differing samples or the \u201cstakes\u201d of the current study. It was also observed that the effect of levels of psychopathic traits was primarily noticeable in the Be Detailed condition (potentially due to pressure), which demonstrates that individual differences (i.e., personality) and situational factors (type of instruction) impact each other. The above findings provide increasing evidence that gender differences are important to consider in future studies investigating psychopathic traits. These results also substantiate that studies utilizing primarily male samples to investigate levels of psychopathic traits do not generalize to a female sample in either general or prison populations. Above all, these results indicate that individual differences (i.e., gender, personality) and context (i.e., instructions, preparation time), should be considered when reviewing linguistic cues to deceit.   As discussed, deception detection ability has routinely been found to hover around 54% (Bond & DePaulo, 2006; Hartwig & Bond, 2011). This research is timely as more recent research is focusing on verbal cues to deceit, with the current study suggesting that contextual factors influence them. Indeed, applications for this research can be found in the training of individuals in the phrasing of interview questions or avenues to elicit more information. For example, training can include discussions about how to provide clear examples and guidelines to garner more information, rather than just simply stating \u201cprovide as much detail as possible.\u201d Further, it could aid interviewers in attempting to determine if interviewees are being deceptive through training, or these moderators can be added to computer language programs to aid in deception detection. A main takeaway from these results is that the factors that have been 61 theorized to influence language and deception are not as straightforward as they would appear. Indeed, investigations into deception are more routinely finding that the prominent factors to consider vary between each interactional context (e.g., Markowitz & Levine, 2020) and other important aspects such as culture. It appears that a combination of interpretations relating to cognition (cognitive load and schemas), personality (levels of psychopathic traits, Big Five, etc.), identity (gender, cultural, socio-economic status), and context are necessary to most adequately explain the results. Pulling from the social constructivist perspective, these could be factors to be mindful with in the moment interactions where an individual is querying veracity, but a judgement should not be made on a single factor alone. As discussed, these results should also be contextualized within the current social and political climate and should not be used to inform public policy, which could lead to further systematic discrimination. Continued research is required to determine the precise and complex relationship these variables have in the production of verbal cues to deceit, and their applicability in a real-world setting. Indeed, systematic reviews or meta-analyses are potential next steps to consider for deception literature in narrowing down the precise effect these factors have. The current study outlines several of these important variables to consider around individual and contextual differences and in doing so may increase deception detection accuracy. In refining and expanding our knowledge, we could be able to provide more concrete information to interviewers in a variety of contexts.       62 References Abdi, H. & Williams, L. J. (2010). Tukey\u2019s Honestly Significant Different (HSD) test. In Neil  Salkind (Ed.), Encyclopedia of Research Design (pp. 2-5). Thousand Oaks, CA: Sage.  Arciuli, J., Mallard, D., & Villar, G. (2010). \u201cUm, I can tell you\u2019re lying\u201d: Linguistic markers of  deception versus truth-telling in speech. Applied Psycholinguistics, 31, 397-411. doi:10.1017\/S0142716410000044 Azizli, N., Atkinson, B. E., Baughman, H. M., Chin, K., Vernon, P. A., Harris, E., & Veselka, L  (2016). Lies and crimes: Dark Triad, misconduct, and high-stakes deception. Personality and Individual Differences, 89, 34-39. doi:10.1016\/j.paid.2015.09.034 Beatty-Mart\u00ednez, A. L., Navarro-Torres, C. A., Dussias, P. E., Bajo, M. T., Tamargo, R. E. G., &  Kroll, J. F. (2020). Interactional context mediates the consequences of bilingualism for language and cognition. Journal of Experimental Psychology: Learning, Memory, and Cognition, 46(6), 1022-1047. doi:10.1037\/xlm0000770 Beryl, R., Chou, S., & V\u00f6llm, B. (2014). A systematic review of psychopathy in women within  secure settings. Personality and Individual Differences, 71, 185-195. doi:10.1016\/j.paid.2014.07.033 Bleich S. N., Findling, M. G., Casey, L. S., Blendon, R. J., Benson, J. M., SteelFisher, G. K.,  Sayde, J. M., Miller, C. (2019). Discrimination in the United States: Experiences of Black Americans. Health Services Research, 54, 1399-1408. doi:10.1111\/1475-6773.13220 Bogaard, G., Meijer, E. W., Vrij, A., & Merckelbach, H. (2016). Strong, but wrong: Lay  people\u2019s and police officers believes about verbal and nonverbal cues to deception. PLoS ONE, 11(6), 1-19. doi:10.1371\/journal.pone.0156615 Bond, G. D., & Lee, A. Y. (2005). Language of lies in prison: Linguistic classification of  63 prisoners\u2019 truthful and deceptive natural language. Applied Cognitive Psychology, 19, 313\u2013329. doi:10.1002\/acp.1087 Bond, C. F. & DePaulo, B. M. (2006). Accuracy of deception judgments. Personality and Social  Psychology Review, 10(3), 214-234. doi:10.1207\/s15327957pspr1003_2 Bortfeld, H., Leon, S. D., Bloom, J. E., Schober, M. F., & Brennan, S. E. (2001). Disfluency  rates in conversation: Effects of age, relationship, topic, role, and gender. Language and  Speech, 44(2), 123-147. doi:10.1177\/00238309010440020101 Box, G. E. P. & Tidwell, P. W. (1962). Transformation of independent variables. Technometrics,  4, 531-550. doi:10.1080\/00401706.1962.10490038 Brandon, S. (2014). Towards a science of interrogation. Applied Cognitive Psychology, 28, 945- 946. doi:10.1037\/a0024818 Burgoon, J. K., Buller, D. B., & Floyd, K. (2001). Does participant affect deception success? A  test of the interactivity principle. Human Communication Research, 27(4), 503-534. doi:10.1093\/hcr\/27.4.503 Burgoon, J. K. (2015). When is deceptive production more effortful than truth-telling? A baker\u2019s  dozen of moderators. Frontiers in Psychology, 6(1965), 1-9. doi:10.3389\/fpsyg.2015.01965 Burgoon, J. K., Mayew, W. J., Giboney, J. S., Elkins, A. C., Moffitt, K., Newark, N. J., Dorn,  B.,\u2026& Spitzley, L. (2015). Which spoken language markers identify deception in high-stakes settings? Evidence from earnings conference calls. Journal of Language and Social Psychology, 35(2), 123-157. doi:10.1177\/0261927X15586792 Chaplin, T. M. (2015). Gender and emotion expression: A developmental contextual perspective.  Emotion Review, 7(1), 14-21. doi:10.1177\/1754073914544408 64 Chaplin, T. M. & Adao, A. (2013). Gender differences in emotion expression in children: A  meta-analytic review. Psychological Bulletin, 139(4), 735-765. doi:10.1037\/a0030737 Cleckley, H. (1941). The mask of sanity: An attempt to clarify some issues about the so-called  psychopathic personality. St. Louis, MO: Mosby. Coid, J., Yang, M., Ulrich, S., Roberts, A., & Hare, R. D. (2009). Prevalence and correlates of  psychopath traits in the household population of Great Britain. International Journal of Law and Psychiatry, 32(2), 65-73. doi:10.1016\/j.ijlp.2009.01.002 Cook, D. J. & Selbom, M. (2019). An examination of Psychopathy Checklist-Revised latent  factor structure via exploratory structural equation modeling. Psychological Assessment, 31(5), 581-591. doi:10.1037\/pas0000676  Cunliffe, T. B., Gacono, C. B., Smith, J. M., Kivisto, A. J., Meloy, J. R., & Taylor, E. E.  (2016). Assessing psychopathy in women. In C. B. Gacono (Ed.), Personality and clinical psychology series. The clinical and forensic assessment of psychopathy: A practitioner's guide (p. 167\u2013190). Routledge\/Taylor & Francis Group. Dalton, D. & Ortegran, M. (2011). Gender differences in ethics research: The importance of  controlling for the social desirability response bias. Journal of Business Ethics, 103, 73-93. doi:10.1007\/s10551-011-0843-8 Dawel, A., Wright, L., Dumbleton, R., & McKone, E. (2018). All tears are crocodile tears:  Impaired perception of emotion authenticity in psychopathic traits. Personality Disorders: Theory, Research, and Treatment, Advance online publication. doi:10.1037\/per0000301 DePaulo, B. M., Lindsay, J. J., Malone, B. E., Muhlenbruck, L., Charlton, K., & Cooper, H.  65 (2003). Cues to deception. Psychological Bulletin, 129(1), 74-118. doi:10.1037\/0033-2909.129.1.74 de Vogel, V. & Lancel, M. (2016). Gender differences in the assessment and manifestation of  psychopathy: Results from a multicenter study in forensic psychiatric patients. Journal of Forensic Mental Health, 15(1), 97-110. doi:10.1080\/14999013.2016.1138173 Dini\u0107, B. M. & Wertag, A. (2018). Effects of Dark Triad and HEXACO traits on  reactive\/proactive aggression: Exploring the gender differences. Personality and Individual Differences, 123, 44-49. doi:10.1016\/j.paid.2017.11.003 Edwards, B. G., Albertson, E., & Verona, E. (2017). Dark and vulnerable personality trait  correlates of dimensions of criminal behaviour among adult offenders. Journal of Abnormal Psychology, 126(7), 921-927. doi:10.1037\/abn0000281 Ekman, P. & Friesen, W. V. (1969). Nonverbal leakage and cues to deception. Psychiatry, 32(1),  88-106. doi:10.1080\/00332747.1969.11023575 Eriksson, M., Marschik, P. B., Tulviste, T., Almgren, M., Pereira, M. P., Wehberg, S., \u2026 & Gallego, C. (2012). Differences between girls and boys in emerging language skills: Evidence from 10 communities. British Journal of Developmental Psychology, 30, 326-343. doi:10.1111\/j.2044-835X.2011.02042.x Fisher, R. P. (2010). Interviewing cooperative witnesses. Legal and Criminological Psychology,  15, 25-38. doi:10.1348\/135532509X441891 Forrest, J. A., Feldman, R. S., & Tyler, J. M. (2004). When accurate beliefs lead to better lie  detection. Journal of Applied Social Psychology, 34(4), 764-780. doi:10.1111\/j.1559-1816.2004.tb02569.x Forouzan, E., & Cooke, D. J. (2005). Figuring out la femme fatale: Conceptual and assessment  66 issues concerning psychopathy in females. Behavioral Sciences & the Law, 23(6), 765\u2013778. doi:10.1002\/bsl.669 Gerlach, P., Teodorescu, K. & Hertwig, R. (2019). The truth about lies: A meta-analysis on  dishonest behavior. Psychological Bulletin, 145(1), 1-44. doi:10.1037\/bul0000174 Gervain, J. (2018). The role of prenatal experience in language development. Current Opinion in  Behavioral Sciences, 21, 62-67. doi:10.1016\/j.cobeha.2018.02.004 Glenn, A. L., Han, H., Yang, Y., Raine, A., & Schug, R. A. (2017). Associations between  psychopathic traits and brain activity during instructed false responding. Psychiatry Research: Neuroimaging, 266, 123-137. doi:10.1016\/j.pscychresns.2017.06.008 Gordts, S., Uzieblo, K., Neumann, C., den Bussche, E. A., & Rossi, G. (2017). Validity of the  Self-Report Psychopathy Scales (SRP-III and short versions) in a community sample. Assessment, 24(3), 308-325. doi:10.1177\/1073191115606205 Guay, J. P. M, Knight, R. A., Ruscio, J., & Hare, R. D. (2018). A taxometric investigation of  psychopathy in women. Psychiatry Research, 261, 565-573. doi:10.1016\/j.psychres.2018.01.015 Gullhaugen, A. S. & Sakshaug, T. (2018). What can we learn about psychopathic offenders by  studying their communication? A review of the literature. Journal of Psycholinguistic Research. doi:10.1007\/s10936-018-9599-y Halevy, R., Shalvi, S., & Verschuere, B. (2014). Being honest about dishonesty: Correlating self- reports and actual lying. Human Communication Research, 40(1), 54-72. doi:10.1111\/hcre.12019 Hall, J. A. & Matsumoto, D. (2004). Gender differences in judgments of multiple emotions from  facial expressions. Emotion, 4(2), 201-206. doi:10.1037\/1528-3542.4.2.201 67 Hall, J. A. & Schmid Mast, M. (2008). Are women always more interpersonally sensitive than  men? Impact of goals and content domain. Personality and Social Psychology Bulletin, 34(1), 144-155. doi:10.1177\/0146167207309192 Hancock, J., & Woodworth, M. (2013). An \u201cEye\u201d for an \u201cI\u201d: The challenges and opportunities  for spotting credibility in a digital world. In B. S. Cooper, D. Griesel, & M. Ternes (Eds.), Applied Issues in Investigative Interviewing, Eyewitness Memory, and Credibility Assessment (pp. 325\u2013340). New York, NY: Springer. Hancock, J., Woodworth, M. & Boochever, R. (2018). Psychopaths online: The linguistic traces  of psychopathy in email, text messaging, and Facebook. Media and Communication, 6(3), 83-92. doi:10.17645\/mac.v6i3.1499 Hancock, J., Woodworth, M. & Porter, S. (2011). Hungry like the wolf: A word-pattern analysis  of the language of psychopaths. Legal and Criminological Psychology, 18(1), 102-114.  doi:10.1111\/j.2044-8333.2011.02025.x  Hartwig, M., Granhag, P. A., & Str\u00f6mwall, L. (2007). Guilty and innocent suspects' strategies  during police interrogations. Psychology, Crime, & Law, 13, 213\u2013227. doi:10.1080\/10683160600750264 Hare, R. D., 2003. The Psychopathy Checklist-Revised Technical Manual, 2nd ed. Multi- Health Systems, Toronto, ON. Hare, R. D. (2016). Psychopathy, the PCL-R, and criminal justice: Some new findings and  current issues. Canadian Psychology, 57(1), 21-34. doi:10.1037\/cap0000041 Hartwig, M. & Bond, C. F. (2011). Why do lie-catchers fail? A lens model meta-analysis of  human lie judgements. Psychological Bulletin, 137(4), 643-659. doi:10.1037\/a0023589 Hauch, V., Bland\u00f3n-Gitlin, I., Masip, J., & Sporer, S. L. (2015). Are computers effective lie  68 detectors? A meta-analysis of linguistic cues to deception. Personality and Social Psychology Review, 19(4), 307-342. doi:10.1177\/1088868314556539 Hawkins, R. C. & Boyd, R. L. (2017). Such stuff as dreams are made on: Dream language,  LIWC norms, and personality correlates. Dreaming, 27(2), 102-121.  doi:10.1037\/drm0000049 Himmer, L., Sch\u00f6nauer, M., Heib, D. P. J., Schabus, & Gais, S. (2019). Rehearsal initiates  systems memory consolidation, sleep makes it last. Science Advances, 5(4), 1-9. doi:10.1126\/sciadv.aav1695 Hlavac, J. (2011). Hesitation and monitoring phenomena in bilingual speech: A consequence of  code-switching or a strategy to facilitate its incorporation? Journal of Pragmatics, 43, 3793-3806. doi:10.1016\/j.pragma.2011.09.008 Inbau, F. E., Reid, J. E., Buckley, J. P., & Jayne, B. C. (2011). Criminal interrogation and  confessions 5th Edition. Chicago, IL: John E. Reid and Associates, Inc.  Johnson, M. K., & Raye, C. L. (1981). Reality monitoring. Psychological Review, 88, 67-85.  doi:10.1037\/\/0033-295X.88.1.67 Jones, D. N. & Paulhus, D. L. (2017). Duplicity among the Dark Triad: Three faces of deceit.  Journal of Personality and Social Psychology, 113(2), 329-343. doi:10.1037\/pspp0000139 Junghaenel, D. U., Smyth, J. M., & Santner, L. (2008). Linguistic dimensions of  psychopathology: A quantitative analysis. Journal of Social and Clinical Psychology,  27(1), 36-55. doi:10.1521\/jscp.2008.27.1.36  Jupe, L. M., Vrij, A., Leal, S., & Nahari, G. (2018). Are you for real? Exploring language use  69 and unexpected process questions within the detection of identity deception. Applied Cognitive Psychology, 32, 622-634. doi:10.1002\/acp.3446 Klaver, J., Lee, Z., & Hart, S. D. (2007). Psychopathy and nonverbal indicators of deception in  offenders. Law and Human Behavior, 31, 337-351. doi:10.1007\/s10979-006-9063-7 Lambert, J. C. & Hopwood, C. J. (2016). Sex differences in interpersonal sensitivities across  acquaintances, friends, and romantic relationships. Personality and Individual Differences, 89, 162-165. doi:10.1016\/j.paid.2015.10.004 Le, M. T., Woodworth, M., Gillman, L., Hutton, E., & Hare, R. D. (2017). The linguistic output  of psychopathic offenders during a PCL-R interview. Criminal Justice and Behaviour, 44(4), 551-565. doi:10.1177\/0093854816683423 Lee, Z., Klaver, J. R., & Hart, S. D. (2008). Psychopathy and verbal indicators of deception in  offenders. Psychology, Crime and Law 14(1), 73-84. doi:10.1080\/10683160701423738 Leal, S., Vrij, A., Deeb, H., & Jupe, L. (2018). Using the model statement to elicit verbal  differences between truth tellers and liars: The benefit of examining core and peripheral details. Journal of Applied Research in Memory and Cognition, 7, 610-617. doi:10.1016\/j.jarmac.2018.07.001 Leal, S., Vrij, A., Deeb, H., & Kamermans, K. (2019). Encouraging interviewees to say more  and deception: The ghostwriter method. Legal and Criminological Psychology. doi:10.111\/lcrp.12152 Leaper, C., & Friedman, C. K. (2007). The socialization of gender. In J. Grusec & P. Hastings  (Eds.), Handbook of socialization: Theory and research (pp. 561\u2013587). New York, NY: Guilford. Leaper, C., & Robnett, R. D. (2011). Women are more likely than men to use tentative language,  70 aren\u2019t they? A meta-analysis testing for gender differences and moderators. Psychology of Women Quarterly, 35(1), 129-142. doi:10.1177\/0361684310392728 Leistico, A. R., Salekin, R. T., DeCoster, J., & Rogers, R. (2008). A large-scale meta-analysis  relating the Hare measures of psychopathy to antisocial conduct. Law and Human Behavior, 32, 28-45. doi:10.1007\/s10979-007-9096-6 Liben, L. S., & Bigler, R. S. (2002). The developmental course of gender differentiation:  Conceptualizing, measuring, and evaluating constructs and pathways. Monographs of the Society for Research in Child Development, 67, 1\u2013147. Lindeman, M. I. H., Zengel, B., & Skowronski, J. J. (2017). An exploration of the relationship  among valence, fading effect, rehearsel frequency, and memory vividness for past personal events. Memory, 25(6), 724-735. doi:10.1080\/09658211.2016.1210172 Lloyd, E. P., Summer, K. M., Hugenberg, K., & McConnell, A. R. (2018). Revisiting perceiver  and target gender effects in deception detection. Journal of Nonverbal Behaviour, 42, 427-440. doi:10.1007\/s10919-018-0283-6 Mahony, T. H., Jacob, J., & Hobson, H. (2017). Women in Canada: A gender-based statistical  report (Catalogue No. 89-503-X). Statistics Canada. Retrieved from https:\/\/www150.statcan.gc.ca\/n1\/en\/pub\/89-503-x\/2015001\/article\/14785-eng.pdf?st=MHpDoR5u Mahmut, M. K., Menictas, C., Stevenson, R. J., & Homewood, J. (2011). Validating the factor  structure of the Self-Report Psychopathy Scale in a community sample. Psychological Assessment, 23(3), 670-678. doi:10.1037\/a0023090 Mann, S., Vrij, A., & Bull, R. (2004). Detecting true lies: Police officers ability to detect  71 suspects lies. Journal of Applied Psychology, 89(1), 137-149. doi:10.1037\/0021-9010.89.1.137 Mann, S., Ewens, S., Shaw, D., Vrij, A., Leal, S., & Hillman, J. (2013). Lying eyes: Why liars  seek deliberate eye contact. Psychiatry, Psychology and Law, 20(3), 452-461. doi:10.1080\/13218719.2013.791218 Mann, S., Vrij, A., Fisher, R. P., & Robinson, M. (2008). See no lies, hear no lies: Differences in  discrimination accuracy and response bias when watching or listening to police suspect interviews. Applied Cognitive Psychology, 22, 1061-1071. doi:10.1002\/acp.1406 Markowitz, D. M. 7 Levine, T. R. (2020). It\u2019s the situation and your disposition: A test of two  honesty hypotheses. Social Psychological and Personality Science, 1-12. doi:10.1177\/1948550619898976 Mastromanno, B., Brookstein, D. M., Ogloff, J. R. P., Campbell, R., Chu, C. M. & Daffern, M.  (2018). Assessing change in dynamic risk factors in forensic psychiatric inpatients: Relationship with psychopathy and recidivism. The Journal of Forensic Psychiatry and Psychology, 29(2), 323-336. doi:10.1080\/14789949.2017.1377277 Matsumoto, D. & Hwang, H. C. (2018). Microexpressions differentiate truths from lies about  future malicious intent. Frontiers in Psychology, 9, 1-11. doi:10.3389\/fpsyg.2018.02545 Matsumoto, D., Hwang, H. C., Skinner, L. G., & Frank, M. G. (2016). Positive effects in  detecting lies from training to recognize behavioral anomalies. Journal of Police and Criminal Psychology, 29(1), 28-35. doi:10.1007\/s11896-012-9115-5 McQuaid, S. M., Woodworth, M., Hutton, E. L., Porter, S. & ten Brinke, L. (2015). Automated  insights: Verbal cues to deception in real-life high-stakes lies. Psychology, Crime & Law, 21(7), 617-631. doi:10.1080\/1068316X.2015.1008477 72 McRae, K., Ochsner, K. N., Mauss, I. B., Gabrieli, J. J., & Gross, J. J. (2008). Gender differences  in emotion regulation: An FMRI study of cognitive reappraisal. Group Processes & Intergroup Relations, 11, 143-162. doi:10.1177\/1368430207088035 Mehl, M. R., & Pennebaker, J. W. (2003). The sounds of social life: A psychometric analysis of  students\u2019 daily social environments and natural conversations. Journal of Personality &  Social Psychology, 84, 857\u2013870. doi:10.1037\/0022-3514.84.4.857 Mordkoff, J. T. (2016). The assumptions of normality. Retrieved from  http:\/\/www2.psychology.uiowa.edu\/faculty\/mordkoff\/GradStats\/part%201\/I.07%20normal.pdf Mulac, A., Bradac, J. J., & Gibbons, P. (2001). Empirical support for the gender-as-culture  hypothesis: An intercultural analysis of male\/female language differences. Human   Communication Research, 27, 121\u2013152. doi:10.1111\/j.1468-2958.2001.tb00778.x  Nadal, K. L., Davidoff, K. C., Allicock, N., Serpe, C. R., & Erazo, T. (2017). Perceptions of  police, racial profiling, and psychological outcomes: A mixed methodological study. Journal of Social Issues, 73(4), 808-830. doi:10.1111\/josi.12249  Naglieri, J. A. & Rojahn, J. (2001). Gender differences in planning, attention, simultaneous, and  successive (PASS) cognitive processes and achievement. Journal of Educational Psychology, 93(2), 430-437. doi:10.1037\/\/\/0022-0663.93.2.430 Nahari, G., Ashkenazi, T. Fisher, R. P., Granhag, P. A., Hershkowitz, I., Masip, J.,\u2026&Vrij, A.  (2019). \u201cLanguage of lies\u201d: Urgent issues and prospects in verbal lie detection research. Legal and Criminological Psychology, 24, 1-23. doi:10.1111\/lcrp.12148 Neumann, C. S., Schmitt, D. S., Carter, R., Embley, I., & Hare, R. (2012). Psychopathic traits in  73 females and males across the globe. Behavioral Sciences and the Law, 30(5), 557-574. doi:10.1002\/bsl.2038 Newman, M. L., Groom, C. J., Handelman, L. D., & Pennebaker, J. W. (2008). Gender  differences in language use: An analysis of 14,000 text samples. Discourse Processes, 45(3), 21-236. doi:10.1080\/01638530802073712 Newman, M. L., Pennebaker, J. W., Berry, D. S., & Richards, J. M. (2003). Lying words:  Predicting deception from linguistic styles. Personality and Social Psychology Bulletin, 29(5), 665-675. doi:10.1177\/0146167203251529 Nicholls, T. L., Ogloff, J. R. P., Brink, J. B., & Spidel, A. (2005). Psychopathy in women: A  review of its clinical usefulness for assessing risk for aggression and criminality. Behavioral Sciences and the Law, 23, 779-802. doi:10.1002\/bsl.678 Ogloff, J. R. P. (2006). Psychopathy\/antisocial personality disorder conundrum. Australian and  New Zealand Journal of Psychiatry, 40, 519-528. doi:10.1111\/j.1440-1614.2006.01834.x O\u2019Kearney, R. & Dadds, M. (2004). Developmental and gender differences in the language for  emotions across adolescent years. Cognition and Emotion, 18(7), 913-938. doi:10.1080\/02699930341000356 Paulhus, D. L., Neumann, C. S., & Hare, R. D. (2016). Manual for the Self-Report Psychopathy Scale 4th Edition. Toronto: Multi-Health Systems. Pennebaker, J. W. (2011). The secret life of pronouns: What our words say about us. New York,  NY: Bloomsbury Press. Pennebaker, J. W., Boyd, R. L., Jordan, K., & Blackburn, K. (2015a). Linguistic inquiry and  word count. LIWC [Computer software]. Austin, RX: liwc.net Pennebaker, J. W., Boyd, R. L., Jordan, K., & Blackburn, K. (2015b). The development and  74 psychometric properties of LIWC 2015. Austin, TX: University of Texas at Austin.  Porter, S. & ten Brinke, L. (2010). The truth about lies: What works in detecting high-stakes  deception? Legal and Criminological Psychology, 15(1), 57-75. doi:10.1348\/135532509X433151 Porter, S., Campbell, M. A., Stapleton, J., & Birt, A. R. (2002). The influence of judge, target,  and stimulus characteristics on the accuracy of detecting deceit. Canadian Journal of Behavioural Science, 34(3), 172-185. doi:10.1037\/h0087170 Porter, S., & Woodworth, M. (2007). \u201cI\u2019m sorry I did it\u2026but he started it\u201d: A comparison of the  official and self-reported homicide descriptions of psychopaths and non-psychopaths. Law and Human Behavior, 31(1), 91\u2013107. doi:10.1007\/s10979-006-9033-0. Richards, H. J., Casey, J. O., & Lucente, S. W. (2003). Psychopathy and treatment response  in incarcerated female substance abusers. Criminal Justice and Behavior, 30, 251\u2013 276. doi:10.1177\/0093854802251010 Rogier, G., Garofalo, C., & Velotti, P. (2019). Is emotional suppression always bad? A matter of  flexibility and gender differences. Current Psychology: A Journal for Diverse Perspectives on Diverse Psychological Issues, 38(2), 411-420. doi:10.1007\/s12144-017-9623-7 Rosenbaum, S. M., Billinger, S., & Stieglitz, N. (2014). Let\u2019s be honest: A review of  experimental evidence of honest and truth-telling. Journal of Economic Psychology, 45, 181-196. doi:10.1016\/j.joep.2014.10.002 Quayle, J. (2008). Interviewing a psychopathic suspect. Journal of Investigative Psychology and  Offender Profiling, 5(1-2), 79-91. doi:10.1002\/jip.83  Savage, L. (2019). Female offenders in Canada, 2017. (Catalogue No. 85-002-X). Statistics  75 Canada. Retrieved from https:\/\/www150.statcan.gc.ca\/n1\/en\/pub\/85-002- x\/2019001\/article\/00001-eng.pdf?st=zTgQMq8h Schulz., N., Murphy, B., & Verona, E. (2016). Gender differences in psychopathy links to drug  use. Law and Human Behavior, 40(2), 1590-168. doi:10.1037\/lhb0000165 Shields, S. A. (2002). Speaking from the heart: Gender and the social meaning of emotion.  Cambridge, England: Cambridge University Press. Shields, S. A. (2012). Gender and emotion: What we think we know, what we need to know, and  why it matters. Psychology of Women Quarterly, 37(4), 423-435. doi:10.1177\/0361684313502312 Siedlecki, K. L., Falzarano, F., & Salhouse, T. A. (2019). Examining gender differences in  neurocognitive functioning across adulthood. Journal of the International  Neuropsychological Society, 25(10), 1051-1060. doi:10.1017\/S1355617719000821  Siegfried, L. & Schwandt, B. (2007). Moderators of nonverbal indicators of deception.  Psychology, Public Policy, and Law, 13(1), 1-34. doi:10.1037\/1076-8971.13.1.1 Snook, B., Fahmy, W., Fallon, L., Lively, C. J., Luther, K., Meissner, C.,\u2026House, J. C. (2020).  Challenges of a \u201ctoolbox\u201d approach to investigative interviewing: A critical analysis of the RCMP\u2019s Phased Interview Model. Psychology, Public Policy and Law, In Press. Retrieved from https:\/\/eprints.lancs.ac.uk\/id\/eprint\/143508 Sonnenschein, A. R., Hofmann, S. G., Ziegelmayer, T., & Lutz, W. (2018). Linguistic analysis of  patients with mood and anxiety disorders during cognitive behavioral therapy. Cognitive Behavior Therapy, 47(4), 315-327. doi:10.1080\/16506073.2017.1419505 Sporer, S. L. (2004). Reality monitoring and the detection of deception. In P.-A. Granhag & L.  76 Stromwell (Eds.). Deception detection in forensic contexts (pp. 64-102). Cambridge, UK: Cambridge University Press. doi:10.1017\/CBO9780511490071 Sporter, S. L. & Schwandt, B. (2006). Paraverbal indicators of deception: A meta-analytic  synthesis. Applications of Cognitive Psychology, 20, 421-446. doi:10.1002\/acp.1190 Sporer, S. L. & Schwandt, B. (2007). Moderators of nonverbal indicators of deception.  Psychology, Public Policy, and Law 13(1), 1-34. doi: 10.1037\/1076-8971.13.1.1 Strand, S. & Belfrage, H. (2005). Gender differences in psychopathy in a Swedish offender  sample. Behavioral Sciences & the Law, 23(6), 837-850. doi:10.1002\/bsl.674 Tabachnick, B. G., & Fidell, L. S. (2014). Using multivariate statistics. Harlow. Tausczik, Y. R. & Pennebaker, J. W. (2010). The psychological meaning of words: LIWC and  computerized text analysis methods. Journal of Language and Social Psychology, 29(1), 24-54. doi:10.1177\/0261927X09351676 ten Brinke, L. & Porter, S. (2012). Cry me a river: Identifying the behavioral consequences of extremely high-stakes interpersonal deception. Law and Human Behavior, 36, 469\u2013477. doi:10.1037\/h0093929 Thakker, D., Karanasios, S., Blanchard, E., Lau, L., & Dimitrova, V. (2017). Ontology for  cultural variations in interpersonal communication: Building on theoretical models and crowdsourced knowledge. Journal of the Association for Information Science and Technology, 68(6), 1411-1428. doi:10.1002\/asi Tonnaer, F., Cima, M., Sijtsma, K., Uzieblo, K., & Lilienfeld, S. O. (2013). Screening for  psychopathy: Validation of the psychopathic personality inventory short-form with reference scores. Journal of Psychopathology Behavioural Assessment, 35(2), 152-161. doi:10.1007\/s10862-012-9333-2 77 Vachon, D. D., Lynam, D. R., Widiger, T. A., Miller, J. D., McCrae, R. R., & Costa, P. T.  (2013). Basic traits predict the prevalence of personality disorder across the life span: The example of psychopathy. Psychological Science, 24(5), 698-705. doi:10.1177\/0956797612460249 van Swol, L. M., Braun, M. T., & Malhorta, D. (2012). Evidence for the Pinocchio effect:  Linguistic differences between lies, deception by omissions, and truths. Discourse Processes, 49(2), 79-106. doi:10.1080\/0163853X.2011.633331 Verigin, B. L., Meijer, E. H., Bogaard, G., & Vrij, A. (2019). Lie prevalence, lie characteristics  and strategies of self-reported good liars. PLoS ONE, 14(12), 1-16. doi:10.1371\/journal.pone.0225566 Verigin, B. L., Meijer, E. H., & Vrij, A. (2020). Embedding lies into truthful stories does not  affect their quality. Applied Cognitive Psychology, 34, 516-525. doi:10.1002\/acp.3642 Verschuere, B. & Hout, W. (2016). Psychopathic traits and their relationship with the cognitive  costs and compulsive nature of lying offenders. PLoS ONE, 11(7), 1-11. doi:10.1371\/ journal.pone.0158595 Verschuere, B., van Ghesel Grothe, S., Waldorp, L., Watts, A., L., Lilienfeld, S. O., Edens, J.  F.,\u2026& Noordhof, A. (2018). What features of psychopathy might be central? A network analysis of the psychopath checklist-revised (PCL-R) in three large samples. Journal of Abnormal Psychology, 127(1), 51-65. doi:10.1037\/abn0000315 Vitacco, M. J., Newmann, C. S., & Pardini, D. A. (2014). Predicting future criminal offending in  a community-based sample of males used self-reported psychopathy. Criminal Justice and Behaviour, 41(3), 345-363. doi:10.1177\/0093854813500488 Vrij, A. (2008). Detecting lies and deceit: Pitfalls and opportunities. Chichester, UK: Wiley.  78 Vrij, A. (2018). Deception and truth detection when analyzing nonverbal and verbal cues.  Applied Cognitive Psychology, 1-8. doi:10.1002\/acp.3457 Vrij, A., Akehurst, L., Soukara, S., & Bull, R. (2004). Let me inform you how to tell a  convincing story: CBCA and reality monitoring scores as a function of age, coaching and deception. Canadian Journal of Behavioral Science, 36(2), 113-126. doi:10.1037\/h0087222 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, 110-117. doi:10.1016\/j-jamac.2012.02.004.  Vrij, A., Leal, S., Jupe, L., & Harvey, A. (2018). With-in subjects verbal lie detection measures:  A comparison between total detail and proportion of compilations. Legal and Criminological Psychology, 23, 265-279. doi:10.1111\/lcrp.12126 Vrij, A., Meissner, C. A., Fisher, R. P., Kassin, S. M., Morgan, III, A., & Kleinman, S. (2017).  Psychological perspectives on interrogation. Perspectives on Psychological Science, 12, 927-955. doi:10.1177\/1745691617706515 Walczyk, J. J., Griffith, D. A., Yates, R., Visconte, S., & Simoneaux, B. (2013). Eye movements  and other cognitive cues to rehearsed and unrehearsed deception when interrogated about a mock crime. Applied Psychology in Criminal Justice, 9(1), 1-23. Retrieved from https:\/\/www.researchgate.net\/publication\/284635823_Eye_movements_and_other_cognitive_cues_to_rehearsed_and_unrehearsed_deception_when_interrogated_about_a_mock_crime Walsh, Z., Swogger, M. T., & Kosson, D. S. (2009). Psychopathy and instrumental violence:  79 Facet level relationships. Journal of Personality Disorders, 23, 416-424. doi:10.1521\/pedi.2009.23.4.416 Whissell, C. (2009). Using the revised dictionary of affect in language to quantify the emotional  undertones of samples of natural language. Psychological Reports, 105(2), 509-521. doi:10.2466\/pr0.105.2.509-521 Woodworth, M. & Porter, S. (2002). In cold blood: Characteristics of criminal homicides as a  function of psychopathy. Journal of Abnormal Psychology, 111, 436-445. doi:10.1037\/0021-843X.111.3.436 Wright, G. R. T., Berry, C. J., Catmur, C., & Bird, G. (2015). Good liars are neither \u2018dark\u2019 nor  self-deceptive. PLoS ONE, 10(6), 1-11. doi:10.1371\/journal.pone.0127315 Wynn, R., H\u00f8iseth, M. H., & Pettersen, G. (2012). Psychopathy in women: Theoretical and  clinical perspectives. International Journal of Womens Health, 4, 257-263. doi:10.2147\/IJWH.S25518 Zahn-Waxler, C., Shirtcliff, E. A., & Marceau, K. (2008). Disorders of childhood and  adolescence: Gender and psychopathology. Annual Review of Clinical Psychology, 4(11), 1-29. doi:10.1146\/annurev.clinpsy.3.022806.091358 Zou, C., Plaks, J. E., & Petereson, J. B. (2019). Don\u2019t get too excited: Assessing individual  differences in the down-regulation of positive emotions. Journal of Personality Assessment, 101(1), 73-83. doi:10.1080\/00223891.2017.1339711   80 Appendices Appendix A \u2013 Tables Table 1  Participant demographics based on conditions  Condition n Age Word Count   M SD M SD Total 106 20.19 2.39 586.70 485.07 Males 32 20.53 3.12 519.97 278.45 Females 74 20.04 2.39 629.85 549.37 Ghostwriter      Males 17 20.82 3.05 584.24 246.72 Females 34 20.38 2.83 827.71 716.18 Be Detailed      Males 15 20.20 3.28 447.13 302.36 Females 40 19.75 1.93 461.68 260.35          81 Table 2 Descriptive statistics for the high-level LIWC categories and DAL output  Linguistic Category Total Sample Females Males  M SD M SD Range M SD Range LIWC         Word Count 596.7 485.0 629.9 549.4 116-3573 519.9 278.5 161-1086 Analytic 37.1 16.0 36.6 15.2 18.3-91 38.5 18.1 21.1-69.8 Clout 69.5 20.2 72.1 18.4 18.6-98.5 63.7 23.4 11.8-98.5 Authentic 70.4 18.1 71.6 17.3 26.7-99 67.8 20.1 34.5-98.3 Tone 65.1 20.1 65.2 18.6 14.3-84.7 64.9 23.7 20.0-98.1 Function 59.8 3.4 60.1 3.1 52.4-66.8 59.0 4.2 52.5-68.2 Verb 19.6 1.9 16.7 1.9 12.9-21.7 16.9 2.04 11.6-20.2 Affect 3.5 1.4 3.5 1.3 0.9-8.1 3.6 1.4 1.0-6.4 Social 9.7 3.0 10.1 2.7 4.6-18.9 8.7 9.0 2.4-19.4 Cognitive Processes 9.7 3.1 9.5 3.2 1.6-18.1 10.0 2.9 4.4-18.8  Perception 1.7 1.0 1.5 0.9 0-5 1.9 1.2 0-7.2 Informal 5.1 2.6 4.8 2.6 0.2-12.1 5.6 2.6 1.7-9.9 DAL         Imagery 1.5 0.06 1.5 0.06 1.4-1.7 1.5 0.07 1.4-1.6 Pleasant 1.9 0.03 1.9 0.03 1.8-2.0 1.9 0.04 1.8-2.0 Activation 1.7 0.04 1.7 0.04 1.4-1.7 1.7 0.03 1.6-1.8        82 Table 3 Descriptive statistics for participant motivation and perception of allotted preparation time  Condition Motivation Perception of Allotted Time  M SD M SD Ghostwriter     Male 4.06 1.03 6.29 1.05 Female 4.26 0.93 6.21 1.39 Total 4.19 0.96 6.24 1.27 Be Detailed     Male  4.47 0.64 6.27 1.22 Female 4.40 0.67 6.42 1.06 Total 4.42 0.66 6.32 1.09 Both Conditions     Male 4.25 0.88 6.28 1.11 Female 4.34 0.80 6.32 1.22 Total 4.31 0.82 6.31 1.18 Note. Multiple comparisons were conducted with no significant differences found between males and females in each condition for both motivation and perception of allotted preparation time (p > 0.05).        83 Table 4 Psychometric properties for the SRP-IV scale and subscales.   Scale M SD Range Cronbach\u2019s a Interpersonal Manipulation 40.03 9.35 21-67 .87 Callous Affect 36.97 8.05 21-56 .79 Erratic Lifestyle 41.86 8.13 21-63 .76 Criminal Tendencies 22.33 6.09 16-45 .73 Total 140.90 24.81 79-219 .89 Note. The SRP-IV subscale scores can range from 16 to 80. The total SRP-IV can range from 64 to 320.                  84 Table 5 SRP-IV demographics and multiple comparisons results  Scale Male Female t\/U p  M SD M SD   Interpersonal Manipulation 43.81 8.92 38.45 9.13 -2.79 .007* Callous Affect 42.58 7.33 34.62 7.16 -5.11 4.14e-06* Erratic Lifestyle 43.71 7.46 41.08 8.33 -1.59 .12 Criminal Tendencies 23.32 7.02 21.92 5.65 1007.5 .33 Total 153.42 21.64 135.55 24.27 -3.71 .0004*  *p < 0.01 Note. A Mann-Whitney-U t-test was conducted on the criminal tendencies subscale to correct for the violation of normality assumption, resulting in a Wilcoxon rank value as the outcome parameter.                85 Table 6 Binomial regression results for males collapsed across story-telling condition  Variable Estimate Standard Error 95% CI p    LL UL  Intercept 7.73 15.74 1.31e-10 1.992e18 .62 Analytic -0.018 0.037 9.08e-01 1.05e00 .64 Clout -0.0071 0.052 8.98e-01 1.11e00 .89 Authentic -0.051 0.039 8.69e-01 1.03e00 .20 Tone 0.29 0.027 9.77e-01 1.09e00 .27 Function 0.0099 0.19 6.92e-01 1.47e00 .62 Affect 0.23 0.46 5.05e-01 3.27e00 .62 Social -0.16 0.34 3.82e-01 1.56e00 .65 CogProc -0.23 0.26 4.48e-01 1.31e00 .37 Perception 0.084 0.44 4.21e-01 2.74e00 .85 Informal -0.49 0.37 2.52e-01 1.16e00 .19 *p < .05 Note. LL = Lower limit, UP = Upper limit.          86 Table 7 Binomial regression results for females collapsed across story-telling condition  Variable Estimate Standard Error 95% CI p    LL UL  Intercept 2.08 9.35 1.31e-10 1.99e18 .82 Analytic -0.014 0.024 9.08e-01 1.05e00 .55 Clout 0.051 0.029 8.98e-01 1.11e00 .082 Authentic 0.019 0.019 8.69e-01 1.03e00 .34 Tone -0.0067 0.017 9.77e-01 1.09e00 .70 Function -0.16 0.13 6.92e-01 1.47e00 .21 Affect 0.61 0.29 5.05e-01 3.27e00 .034* Social -0.060 0.19 3.82e-01 1.56e00 .75 CogProc .14 0.11 4.48e-01 1.31e00 .22 Perception 0.25 0.34 4.21e-01 2.74e00 .46 Informal -0.0066 0.14 2.52e-01 1.16e00 .96 *p < .05 Note. LL = Lower limit, UP = Upper limit. The p-value for affect was found to be not significant once multiple comparisons were accounted for using a Bonferroni correction.        87 Table 8 ANOVA results for specified linguistic categories based upon story-telling condition  Variable Sum of Squares df Mean Square F p h2 Ghostwriter Condition Positive Emotions       Gender 0.70 1 0.695 0.63 .43 0.012 Veracity 1.05 1 1.053 0.96 .33 0.018 Gender x Veracity 6.66 1 6.659 6.06 .018* 0.11 Negative Emotions       Gender 0.028 1 0.0280 0.12 .73 0.0024 Veracity 0.37 1 0.03722 1.57 .22 0.032 Gender x Veracity 0.16 1 0.157 0.66 .42 0.013 Personal Pronouns       Gender 38.41 1 38.41 7.09 .011* 0.13 Veracity 6.76 1 6.76 1.25 .27 0.022 Gender x Veracity 7.69 1 7.69 1.42 .24 0.025 Disfluencies       Gender 21.2 1 21.199 4.31 .043* 0.080 Veracity 11.31 1 11.311 2.30 .14 0.043 Gender x Veracity 1.19 1 1.187 0.24 .763 0.0044 CogProc       Gender 5.1 1 5.06 0.56 .46 0.011 Veracity 0.1 1 0.14 0.016 .90 0.00033 Gender x Veracity 6.8 1 6.81 0.75 .39 0.016 Be Detailed Condition Positive Emotions       Gender 0.00 1 0.00 0.00 .99 5.6e-07 Veracity 15.20 1 15.197 10.30 .0023* 0.16 Gender x Veracity 0.99 1 0.99 0.67 .42 0.011 Negative Emotions       Gender 0.090 1 0.0900 0.21 .65 0.0038 Veracity 0.98 1 0.980 2.27 .14 0.042 Gender x Veracity 0.32 1 0.323 0.75 .39 0.014 Personal Pronouns       Gender 1.42 1 1.43 0.30 .59 5.8e-3 Veracity 2.08 1 2.078 0.44 .51 8.53e-3 Gender x Veracity 0.00 1 0.001 0.00 .99 3.05e-06 Disfluencies       Gender 3.54 1 3.541 0.71 .40 0.013 Veracity 11.22 1 11.216 2.27 .14 0.039 Gender x Veracity 15.46 1 15.457 3.12 .083 0.055 CogProc       Gender 1.1 1 1.06 0.10 .75 0.0019 88 Veracity 12.9 1 12.94 1.24 .27 0.023 Gender x Veracity 9.3 1 9.33 0.89 .35 0.017 Note. *p < .05; a = see Table 9 for Post Hoc Comparisons; CogProc = Words associated with cognitive processes.                        89 Table 9 Post hoc analysis for the interaction effect of gender and veracity for positive emotion language in the Ghostwriter condition  Comparison Mean Difference 95% CI for Difference p*   LL UL  Male(True) : Female(True) -0.64 -1.88 0.60 .53 Female(False) : Female(True) -0.22 -1.18 0.74 .93 Male(False) : Female(True) 0.69 -0.41 1.79 .35 Female(False) : Male(True) 0.42 -0.85 1.69 .81 Male(False) : Male(True) 1.33 -0.043 2.71 .061 Male(False) : Female(False) 0.91 -0.21 2.04 .15 Note. p-values presented in this table are adjusted for multiple comparisons using Tukey\u2019s Honestly Significant Difference (HSD).                 90 Table 10 ANOVA results for DAL linguistic output based upon story-telling condition  Variable Sum of Squares df Mean Square F p h2 Ghostwriter Condition Pleasantness       Gender 0.0012 1 0.0011 1.30 .25 0.027 Veracity 0.0008 1 0.00084 0.99 .33 0.019 Gender x Veracity 0.0010 1 0.00107 1.21 .27 0.023 Imagery       Gender 0.013 1 0.0127 3.51 .067 0.069 Veracity 0.00011 1 0.000112 0.031 .86 0.0006 Gender x Veracity 0.0019 1 0.00198 0.55 .47 0.011 Activation       Gender 0.0019 1 0.00193 1.04 .32 0.020 Veracity 0.0057 1 0.00573 3.09 .085 0.061 Gender x Veracity 0.00025 1 0.000246 0.13 .71 0.0026 Be Detailed Condition Pleasantness       Gender 0.00001 1 0.00001 0.007 .94 0.00012 Veracity 0.0036 1 0.00355 3.15 .084 0.057 Gender x Veracity 0.0011 1 0.00111 0.98 .33 0.018 Imagery       Gender 0.00001 1 0.00001 0.001 .97 2.59e-05 Veracity 0.0056 1 0.00558 1.46 .23 2.69e-02 Gender x Veracity 0.0069 1 0.00686 1.79 .19 3.31e-02 Activation       Gender 0.0014 1 0.00139 1.76 .19 0.033 Veracity 0.00007 1 0.00007 0.084 .77 0.0016 Gender x Veracity 0.00017 1 0.00017 0.21 .65 0.0041 Note. *p < .05       91 Table 11 ANCOVA results for specified linguistic categories based upon story-telling condition with SRP-IV total scores included as a covariate  Variable Sum of Squares df Mean Square F p h2 Ghostwriter Condition Positive Emotions       SRP Total 0.71 1 0.729 0.66 .41 0.014 Gender 2.41 1 2.407 2.17 .15 0.041 Veracity 0.35 1 0.346 0.31 .58 0.0059 Gender x Veracity 4.93 1 4.926 4.44 .041* 0.085 Negative Emotions       SRP Total 0.003 1 0.0027 0.011 .92 0.00025 Gender 0.033 1 0.0327 0.13 .72 0.0028 Veracity 0.39 1 0.388 1.57 .22 0.033 Gender x Veracity 0.13 1 0.132 0.54 .47 0.011 Personal Pronouns       SRP Total 0.25 1 0.25 0.046 .83 0.0008 Gender 35.68 1 35.68 6.58 .014* 0.12 Veracity 7.35 1 7.35 1.36 .25 0.024 Gender x Veracity 11.08 1 11.08 2.04 .16 0.037 Disfluencies       SRP Total 0.74 1 0.742 0.15 .71 0.0028 Gender 23.64 1 23.643 4.62 .037* 0.089 Veracity 9.12 1 9.117 1.78 .19 0.034 Gender x Veracity 0.66 1 0.662 0.13 .72 0.0025 CogProc       SRP Total 11.1 1 11.084 1.19 .28 0.025 Gender 0.8 1 0.85 0.091 .76 0.0019 Veracity 0.2 1 0.15 0.016 .89 0.00034 Gender x Veracity 6.4 1 6.38 0.69 .41 0.015 Be Detailed Condition Positive Emotions       SRP Total 3.01 1 3.005 2.06 .16 0.034 Gender 0.02 1 0.021 0.014 .90 0.00024 Veracity 12.04 1 12.039 8.24 .006* 0.14 Gender x Veracity 0.55 1 0.554 0.38 .54 0.006 Negative Emotions       SRP Total 0.56 1 0.562 1.31 .26 0.024 Gender 0.013 1 0.013 0.031 .86 0.00057 Veracity 1.01 1 1.009 2.35 .13 0.044 Gender x Veracity 0.37 1 0.3746 0.87 .36 0.016 Personal Pronouns       SRP Total 15.15 1 15.149 3.36 .073 0.062 92 Gender 3.57 1 3.565 0.79 .38 0.015 Veracity 4.00 1 3.998 0.88 .35 0.016 Gender x Veracity 0.24 1 0.243 0.054 .82 0.00099 Disfluencies       SRP Total 28.37 1 28.37 6.61 .013* 0.10 Gender 7.02 1 7.016 1.64 .21 0.026 Veracity 19.74 1 19.74 4.59 .037* 0.072 Gender x Veracity 9.62 1 9.62 2.24 .14 0.035 CogProc       SRP Total 1.6 1 1.62 0.15 .69 0.0029 Gender 0.2 1 0.24 0.022 .88 0.00043 Veracity 11.3 1 11.25 1.055 .31 0.021 Gender x Veracity 10.3 1 10.33 0.97 .33 0.019 Note. *p < .05; CogProc = Words associated with cognitive processes.   93 Appendix B \u2013 Figures    Figure 1. Flow chart depicting the study\u2019s design.     94  Figure 2. SRP-IV scores for males and females with significant differences noted for the Total SRP score, as well as the callous affect and interpersonal manipulation subscales. CA = Callous affective subscale, CT = Criminal Tendencies subscale, ELM = Erratic Lifestyle subscale, IMP = Interpersonal Manipulation subscale.        153.4242.8323.3243.7143.81135.5534.6221.9241.0838.45020406080100120140160180Total SRP CA CT ELM IMPSRP ScoreMale Female95  Figure 3. Average word count for each story-telling and veracity condition.   445.7656.8469.3839.90100200300400500600700800900Be Detailed GhostwriterAverage Word CountTRUE FALSE96   Figure 4. Male and female output for LIWC categories in the deceptive condition. CogProc = Cognitive Processes. 38.459.765.370.959.316.94.037.9210.311.95.3735.674.270.368.459.7173.8210.229.781.695.010 10 20 30 40 50 60 70 80AnalyticCloutAuthenticToneFunctionVerbAffectSocialCogPRocPerceptionInformalAverage PercentageLinguistic CategoryFemaleMale97   Figure 5. Male and female output for LIWC categories in the truthful condition. CogProc = Cognitive Processes. 38.6168.8871.0657.3258.7316.683.159.669.671.865.9737.5370.1172.8862.2860.5416.923.29.969.251.424.590 10 20 30 40 50 60 70 80AnalyticCloutAuthenticToneFunctionVerbAffectSocialCogProcPerceptionInformalAverage PercentageLinguistic CategoryFemaleMale98  0 1 2 3 4 5 6 7 8 9 10 11LIWC OutputFemaleFalseCognitive ProcessesDisfluenciesPersonal PronounsNegative Affective LanguagePositive Affective LanguageTrueCognitive ProcessesDisfluenciesPersonal PronounsNegative Affective LanguagePositive Affective LanguageMaleFalseCognitive ProcessesDisfluenciesPersonal PronounsNegative Affective LanguagePositive Affective LanguageTrueCognitive ProcessesDisfluenciesPersonal PronounsNegative Affective LanguagePositive Affective Language3.140.6711.273.709.782.570.5911.133.499.253.390.659.644.3210.312.400.6510.534.839.67Figure 6. LIWC output for personal pronouns, disfluencies, cognitive processes, negative and positive affective language based upon gender and veracity condition. 99   Figure 7. Line graph depicting the omnibus significant interaction effect in the Ghostwriter condition between gender and veracity. Post-hoc analysis utilizing Tukey\u2019s HSD resulted in no significant group differences.  Legend Male Female ","attrs":{"lang":"en","ns":"http:\/\/www.w3.org\/2009\/08\/skos-reference\/skos.html#note","classmap":"oc:AnnotationContainer"},"iri":"http:\/\/www.w3.org\/2009\/08\/skos-reference\/skos.html#note","explain":"Simple Knowledge Organisation System; Notes are used to provide information relating to SKOS concepts. There is no restriction on the nature of this information, e.g., it could be plain text, hypertext, or an image; it could be a definition, information about the scope of a concept, editorial information, or any other type of information."}],"Genre":[{"label":"Genre","value":"Thesis\/Dissertation","attrs":{"lang":"en","ns":"http:\/\/www.europeana.eu\/schemas\/edm\/hasType","classmap":"dpla:SourceResource","property":"edm:hasType"},"iri":"http:\/\/www.europeana.eu\/schemas\/edm\/hasType","explain":"A Europeana Data Model Property; This property relates a resource with the concepts it belongs to in a suitable type system such as MIME or any thesaurus that captures categories of objects in a given field. It does NOT capture aboutness"}],"GraduationDate":[{"label":"Graduation Date","value":"2020-02","attrs":{"lang":"en","ns":"http:\/\/vivoweb.org\/ontology\/core#dateIssued","classmap":"vivo:DateTimeValue","property":"vivo:dateIssued"},"iri":"http:\/\/vivoweb.org\/ontology\/core#dateIssued","explain":"VIVO-ISF Ontology V1.6 Property; Date Optional Time Value, DateTime+Timezone Preferred "}],"IsShownAt":[{"label":"DOI","value":"10.14288\/1.0394131","attrs":{"lang":"en","ns":"http:\/\/www.europeana.eu\/schemas\/edm\/isShownAt","classmap":"edm:WebResource","property":"edm:isShownAt"},"iri":"http:\/\/www.europeana.eu\/schemas\/edm\/isShownAt","explain":"A Europeana Data Model Property; An unambiguous URL reference to the digital object on the provider\u2019s website in its full information context."}],"Language":[{"label":"Language","value":"eng","attrs":{"lang":"en","ns":"http:\/\/purl.org\/dc\/terms\/language","classmap":"dpla:SourceResource","property":"dcterms:language"},"iri":"http:\/\/purl.org\/dc\/terms\/language","explain":"A Dublin Core Terms Property; A language of the resource.; Recommended best practice is to use a controlled vocabulary such as RFC 4646 [RFC4646]."}],"Program":[{"label":"Program (Theses)","value":"Psychology","attrs":{"lang":"en","ns":"https:\/\/open.library.ubc.ca\/terms#degreeDiscipline","classmap":"oc:ThesisDescription","property":"oc:degreeDiscipline"},"iri":"https:\/\/open.library.ubc.ca\/terms#degreeDiscipline","explain":"UBC Open Collections Metadata Components; Local Field; Indicates the program for which the degree was granted."}],"Provider":[{"label":"Provider","value":"Vancouver : University of British Columbia Library","attrs":{"lang":"en","ns":"http:\/\/www.europeana.eu\/schemas\/edm\/provider","classmap":"ore:Aggregation","property":"edm:provider"},"iri":"http:\/\/www.europeana.eu\/schemas\/edm\/provider","explain":"A Europeana Data Model Property; The name or identifier of the organization who delivers data directly to an aggregation service (e.g. Europeana)"}],"Publisher":[{"label":"Publisher","value":"University of British Columbia","attrs":{"lang":"en","ns":"http:\/\/purl.org\/dc\/terms\/publisher","classmap":"dpla:SourceResource","property":"dcterms:publisher"},"iri":"http:\/\/purl.org\/dc\/terms\/publisher","explain":"A Dublin Core Terms Property; An entity responsible for making the resource available.; Examples of a Publisher include a person, an organization, or a service."}],"Rights":[{"label":"Rights","value":"Attribution-NonCommercial-NoDerivatives 4.0 International","attrs":{"lang":"*","ns":"http:\/\/purl.org\/dc\/terms\/rights","classmap":"edm:WebResource","property":"dcterms:rights"},"iri":"http:\/\/purl.org\/dc\/terms\/rights","explain":"A Dublin Core Terms Property; Information about rights held in and over the resource.; Typically, rights information includes a statement about various property rights associated with the resource, including intellectual property rights."}],"RightsURI":[{"label":"Rights URI","value":"http:\/\/creativecommons.org\/licenses\/by-nc-nd\/4.0\/","attrs":{"lang":"*","ns":"https:\/\/open.library.ubc.ca\/terms#rightsURI","classmap":"oc:PublicationDescription","property":"oc:rightsURI"},"iri":"https:\/\/open.library.ubc.ca\/terms#rightsURI","explain":"UBC Open Collections Metadata Components; Local Field; Indicates the Creative Commons license url."}],"ScholarlyLevel":[{"label":"Scholarly Level","value":"Graduate","attrs":{"lang":"en","ns":"https:\/\/open.library.ubc.ca\/terms#scholarLevel","classmap":"oc:PublicationDescription","property":"oc:scholarLevel"},"iri":"https:\/\/open.library.ubc.ca\/terms#scholarLevel","explain":"UBC Open Collections Metadata Components; Local Field; Identifies the scholarly level of the author(s)\/creator(s)."}],"Title":[{"label":"Title ","value":"Sounding authentic : an examination of how individual and personality differences in language affect deception","attrs":{"lang":"en","ns":"http:\/\/purl.org\/dc\/terms\/title","classmap":"dpla:SourceResource","property":"dcterms:title"},"iri":"http:\/\/purl.org\/dc\/terms\/title","explain":"A Dublin Core Terms Property; The name given to the resource."}],"Type":[{"label":"Type","value":"Text","attrs":{"lang":"en","ns":"http:\/\/purl.org\/dc\/terms\/type","classmap":"dpla:SourceResource","property":"dcterms:type"},"iri":"http:\/\/purl.org\/dc\/terms\/type","explain":"A Dublin Core Terms Property; The nature or genre of the resource.; Recommended best practice is to use a controlled vocabulary such as the DCMI Type Vocabulary [DCMITYPE]. To describe the file format, physical medium, or dimensions of the resource, use the Format element."}],"URI":[{"label":"URI","value":"http:\/\/hdl.handle.net\/2429\/75804","attrs":{"lang":"en","ns":"https:\/\/open.library.ubc.ca\/terms#identifierURI","classmap":"oc:PublicationDescription","property":"oc:identifierURI"},"iri":"https:\/\/open.library.ubc.ca\/terms#identifierURI","explain":"UBC Open Collections Metadata Components; Local Field; Indicates the handle for item record."}],"SortDate":[{"label":"Sort Date","value":"2020-12-31 AD","attrs":{"lang":"en","ns":"http:\/\/purl.org\/dc\/terms\/date","classmap":"oc:InternalResource","property":"dcterms:date"},"iri":"http:\/\/purl.org\/dc\/terms\/date","explain":"A Dublin Core Elements Property; A point or period of time associated with an event in the lifecycle of the resource.; Date may be used to express temporal information at any level of granularity. Recommended best practice is to use an encoding scheme, such as the W3CDTF profile of ISO 8601 [W3CDTF].; A point or period of time associated with an event in the lifecycle of the resource.; Date may be used to express temporal information at any level of granularity. Recommended best practice is to use an encoding scheme, such as the W3CDTF profile of ISO 8601 [W3CDTF]."}]}