Psychopathic-Like Traits Among Detained Female Adolescents

Abstract

This study examined the factor structure, reliability, and validity of the Antisocial Process Screening Device–Self-Report (APSD-SR), the Youth Psychopathic Traits Inventory (YPI), and the YPI–Short Version (YPI-SV) in detained female adolescents aged 12 to 17 years. The proposed three-factor structure of the YPI and YPI-SV was replicated, whereas the proposed three-factor structure of the APSD-SR or alternate models did not yield adequate fit. Overall, reliability indices for the YPI and YPI-SV were higher than those reported for the APSD-SR. APSD-SR and YPI scales were positively related with each other, except the affective dimensions of the instruments. All questionnaires showed good criterion validity but the YPI’s factor structure and reliability was superior to the APSD-SR. This superiority is not because of the larger number of items in the YPI, because we also demonstrated that the factor structure and reliability of the YPI-SV was better than that of the APSD-SR.

Keywords

psychopathy delinquent personality traits self-report adolescence incarcerated

Introduction

Psychopathy is a multidimensional construct often defined as a syndrome comprising a constellation of deficient affective, interpersonal, and behavioral characteristics. To better understand the developmental origins of this syndrome, several instruments have been developed to measure psychopathic-like traits in children and adolescents.¹ Although the Psychopathy Checklist: Youth Version (PCL:YV; Forth, Kosson, & Hare, 2003) is often considered to be the most reliable and valid measure of psychopathic-like traits among institutionalized youth (Andershed, Hodgins, & Tengstrom, 2007), its reliance on file information and the relatively time-consuming assessment procedure makes this instrument difficult to use in settings where file data on the youth is absent and in studies in which large populations need to be assessed (e.g., in community samples). To enable the study of psychopathic-like traits during circumstances where time is an issue and where file documentation is not always present, questionnaires have been developed that rely on parents, teachers, and/or youths themselves. The current study was designed to examine to what extent two of these questionnaires can be used in detained female adolescents: the Antisocial Process Screening Device (APSD; Frick & Hare, 2001) and the Youth Psychopathic Traits Inventory (YPI; Andershed, Kerr, Stattin, & Levander, 2002).

Detained youths are not a random sample of the general population, due to selection factors related to their own behavior and to law enforcement decisions about whom to arrest (Vincent, Grisso, Terry, & Banks, 2008). Therefore, results from normal population studies cannot simply be generalized to these juvenile justice–involved youths. These limits of generalizing results from one setting to another has not only been demonstrated in studies that, for example, addressed the relationship between conduct disorder and recidivism (Colins et al., 2011) but also in studies that focused on psychopathic-like traits in detained youths. For example, although several studies showed that psychopathic-like traits in community children and adolescents were predictive of future serious antisocial behavior (e.g., Dadd, Fraser, Frost, & Hawes, 2005; Lynam, Miller, Vachon, Loeber, & Stouthamer-Loeber, 2009), a prospective relationship with future violence could not (Colins, Vermeiren, De Bolle, & Broekaert, in press), or only to some extent (Cauffman, Kimonis, Dmetrieva, & Monahan, 2009; Salekin, 2008), be replicated among detained youths. To be sure that questionnaires developed and validated in normal population samples are useful in seriously antisocial youths, the factor structure and reliability of these questionnaires and their relation to variables of interest (i.e., validity) must be separately established in detained adolescents. The current study addresses this issue by testing the APSD–Self Report version (APSD-SR) and the YPI, including the short version of the YPI, the YPI-SV.

The APSD-SR and the YPI (and the YPI-SV) were selected for two main reasons. First, these questionnaires rely on self-report. Using self-report assessment of psychopathic-like traits is particularly relevant when parents or teachers are not available, unwilling to cooperate, or when they did not have enough contact with the child to provide useful information on the presence or absence of these personality traits (Loney, Frick, Clements, Ellis, & Kerlin, 2003), which is a reality juvenile detention staff very often have to deal with (Colins, Vermeiren, Schuyten, Broekaert, & Soyez, 2008). In addition, self-report instruments have the advantage that opinions (e.g., crying is a sign of weakness), motivations for actions (e.g., using charm to con others), and the presence or absence of some features (e.g., remorse or guilt) are best known to the individual and may be obscure to other people (see also Raine et al., 2006). Second, the APSD is the most widely used and probably the most well-known questionnaire that taps psychopathic-like traits in children and adolescents. The YPI was developed more recently and is thus less studied than the APSD. Yet several scholars have labeled the YPI as a promising new self-report instrument (Kotler & McMahon, 2005; Skeem & Cauffman, 2003), which may explain the increasing number of YPI studies being published.

Only a small number of studies on psychopathic-like traits in detained female adolescents are available, probably because females commit less serious crimes and are less often detained than males (Miller, Watts, & Jones, 2011; Salekin, 2006). In addition, most studies on psychopathic-like traits in detained females used the PCL:YV (Bauer, Whitman, & Kosson, 2011; Cook, Barese, & Dicataldo, 2010; Kosson et al., 2012; Seagrave & Grisso, 2002). Given that the APSD-SR (Murrie & Cornell, 2002) and the YPI (Skeem & Cauffman, 2003) are merely modestly (r = .30) related with the PCL:YV, it must be tested if findings from PCL:YV studies can be replicated with self-report questionnaires or vice versa. Few studies in detained female adolescents used the APSD-SR and/or the YPI (e.g., Boccaccini et al., 2007; Skeem & Cauffman, 2003). Moreover, studies that included both boys and girls did not present relations between psychopathic-like traits and variables of interest by gender because of a restricted number of detained girls (Pechorro, Maroco, Poiares, & Vieira, in press) or because exploratory analysis revealed no significant differences between boys and girls regarding APSD-SR and YPI scores (e.g., Poythress, Dembo, Wareham, & Greenbaum, 2006; Salekin, Debus, & Barker, 2010). However, the absence of mean-level differences between both gender groups does not necessarily imply that the factor structure and psychometric properties of a given instrument is the same for boys and girls or that the nomological network surrounding “juvenile psychopathy” is similar across detained male and female adolescents.

Previous studies have reported on the three-factor structure of the APSD-SR or the YPI in detained males (Skeem & Cauffman, 2003; Veen et al., 2011) and gender-mixed samples of detained adolescent (Poythress, Dembo, et al., 2006; Vitacco, Rogers, & Neumann, 2003). Interestingly, only a YPI study that exclusively included detained males was able to report good fit indices for the original three factor structure (Veen et al., 2011). Likewise, confirmatory factor analyses (CFA) showed very good fit indices for the three factor structure of the APSD in a predominantly detained male sample (83% males; Vitacco et al., 2003; but see Pechorro et al., in press), but this factor structure could not be replicated in a sample of detained youths with an equal proportion of boys and girls (52% male; Poythress, Dembo, et al., 2006). These findings altogether may raise concerns about the applicability of the APSD and YPI three-factor structure in detained female adolescents, a concern that previously has been supported regarding the use of the PCL:YV (Odgers, Reppucci, & Moretti, 2005). With regard to psychometric properties of the APSD-SR, internal consistency indices for the APSD-SR Callous-Unemotional (CU) scale were consistently poor across 10 studies of juvenile justice–involved youths (Poythress, Douglas, et al., 2006). It has been suggested that this poor CU scale reliability may be because of a defensive response style of juvenile justice–involved youths and that the YPI may be a good alternative, given that its items were written with considerable attention and concern to issues of response style (Poythress, Dembo, et al., 2006). A study in a gender-mixed sample of detained adolescents supported this suggestion, as the reliability indices for the YPI scores were generally better to the APSD-SR (Poythress, Dembo, et al., 2006). With regard to validity, the only study that used the APSD-SR in a pure sample of 94 detained female adolescents showed that the three APSD subscale scores were moderately correlated with self-reported violent offending, theft, and—except for the Narcissism subscale—drug abuse (Vaughn, Newhill, DeLisi, Beaver, & Howard, 2008). To the best of our knowledge, no studies are available that addressed the relationship between YPI scores and variables of interest in terms of validity among detained females.

Aim and Hypotheses of the Present Study

The current study will be the first to examine and compare the factor structure, the reliability, and the validity of the APSD-SR and the YPI (and the YPI-SV) in a relatively large sample of detained female adolescents. First, with regard to factorial validity of the APSD-SR and YPI, CFAs will be performed to examine the factor structure of these self-report questionnaires as presented in the original publications (Andershed et al., 2002; Frick & Hare, 2001). Second, to test whether the reliability of the YPI CU dimension is also superior to the corresponding APSD-SR dimension in detained female adolescents as has been demonstrated in with detained males (Poythress, Dembo, et al., 2006) or community male adolescents (Seals, Sharp, Ha, & Michonski, 2012), reliability indices for the APSD and YPI total scores and factor scores will be presented. Given that Cronbach’s alpha is dependent on the number of items, one may argue that it is not surprising that the lengthiest questionnaire has the best reliability indices. Because the recently developed YPI-SV approximately has the same number of items as the APSD-SR, we will examine how the YPI-SV performs compared with the APSD-SR. Third, to test the convergent validity of both self-report questionnaires, we will examine the associations between the corresponding dimensions of the APSD-SR and YPI. It is expected that each dimension of the APSD-SR (e.g., Narcissism) will be significantly related to the corresponding YPI dimension (e.g., Grandiose-manipulative). Fourth, to test the criterion validity of the APSD-SR and YPI, the current study includes variables that have been considered in previous studies of detained adolescents. Specifically, it is hypothesized that both the APSD-SR and YPI total score would be positively related with self-reported offending (e.g., Poythress, Dembo, et al., 2006), drug use, aggression and attention problems (e.g., Campbell, Porter, & Santor, 2004), and disruptive behavior disorders (DBD; Salekin, Neumann, Leistico, DiCicco, & Duros, 2004) and negatively with the Big Five personality traits agreeableness and conscientiousness (e.g., Salekin et al., 2010).² Fifth, although CU traits have been associated with conduct problems, they appear to be less highly correlated with conduct problems than the other two dimensions of psychopathy (Frick & White, 2008). Therefore, it was hypothesized that the affective dimension will less highly correlated with conduct problems and DBD than the other two dimensions

Method

Participants

There is only one youth detention center (YDC) for girls in Flanders (i.e., the Flemish speaking part of Belgium), which has a capacity of 40 beds. Girls are referred to this YDC by a juvenile judge when having committed a(n) (alleged) criminal offense, when misbehaving (e.g., truancy, running away, aggression, prostitution), and/or living in dangerous environments (e.g., being sexually abused). Placement in this YDC can be considered as the harshest measure a juvenile court judge can take. Given the limited number of beds, only the most severely criminal girls and/or girls with the most severe behavioral and emotional problems end up in this YDC. To recruit a relatively large sample of detained female adolescents (younger than 18), we recruited girls in four different periods: (a) between July 2008 and April 2009, (b) between January 2010 and May 2010, (c) between October 2010 and March 2011, and (d) between August and December 2011. Detained girls were eligible to participate in the study if they had sufficient comprehension of Dutch and had an expected detention duration of a minimum of 1 month. During these four periods, a total of 272 different girls were detained in the participating YDC of which 50 did not meet the inclusion criteria. Of the 222 remaining girls, 14 were not approached and asked whether they would like to participate in the study. In addition, of the 208 remaining girls, 10 girls refused to participate, and 3 girls could not be interviewed because of practical circumstances, resulting in a sample size of 195 detained girls. For the purpose of the current study, four girls were excluded because they were not administered the APSD-SR and/or YPI, resulting in a total sample of 191. The age of the detained girls ranged from 12 to 17 years (M = 15.76, SD = 1.02) and 134 girls (70.2%) were of Belgian origin.

Measures

Youth Psychopathic Traits Inventory

The YPI (Andershed et al., 2002) is a 50-item self-report questionnaire based on the three-factor model of psychopathy and focuses on 10 out of the 13 core traits (the ones applicable to nonincarcerated adolescents) identified by Cooke and Michie (2001). The 50 items of the YPI are organized into 10 subscales (cf. 10 core traits), with five items in each subscale. Each item in the YPI is scored on a 4-point Likert-type scale ranging from “Does not apply at all” to “Applies very well.” Within the YPI interpersonal domain (i.e., Grandiose-Manipulative factor), subscales intend to measure dishonest charm, grandiosity, lying, and manipulation. Within the affective domain (often referred to as the Callous-Unemotional dimension), subscales intend to assess callousness, unemotionality, and remorselessness. In the behavioral domain (i.e., Impulsive-Irresponsible factor), subscales intend to measure impulsiveness, thrill seeking, and irresponsibility. The YPI-SV (van Baardewijk et al., 2010) has been developed through a stepwise selection process of items from the original, 50-item YPI, using a series of exploratory factor analyses and content-related arguments. The final model with 18 items organized into three factors (but without subscales) was then cross-validated in independent samples using CFAs (van Baardewijk et al., 2010). The YPI-SV is scored on the same 4-point Likert-type scale as the original YPI.

Antisocial Process Screening Device–Self-Report Version (ASPD-SR)

The APSD-SR (Frick & Hare, 2001) consists of 20 items that tap psychopathic-like traits; the items are answered on a three-point rating scale: not at all true (0), sometimes true (1), or definitely true (2). Previous factor analyses revealed a three-factor structure, comprising an interpersonal dimension (i.e., Narcissism factor, 7 items), an affective dimension (i.e., Callous-Unemotional factor, 6 items), a behavioral dimension (i.e., Impulsivity factor, 5 items), and two items (i.e., Items 2 and 6) that did not load onto any factor (Frick & Hare, 2001). The sum of these 20 items yields a total score for the APSD-SR.

Youth Self-Report

The YSR (Achenbach & Rescorla, 2001) was used as a dimensional measure of externalizing and internalizing problems. This self-report questionnaire consists of 118 items that are to be answered on a 3-point scale ranging from 0 (not at all) to 2 (often), with higher scores indicating more problems. The YSR distinguishes two broadband scales (internalizing and externalizing) and eight syndrome scales, of which only five will be used in the current study (i.e., Withdrawn/Depressed, Anxious/Depressed, Attention Problems, Delinquent Behavior, and Aggressive Behavior).

Diagnostic Interview Schedule for Children–Fourth Edition (DISC-IV)

The DISC-IV (Shaffer et al., 1996) was used to assess the past-year prevalence of psychiatric disorders. The DISC-IV was designed for interviewing children aged 9 to 17 years and is the instrument most often used to examine mental disorders in detained adolescents (Colins et al., 2010). For our study, conduct disorder (CD), oppositional defiant disorder (ODD), attention-deficit/hyperactivity disorder (ADHD), any substance use disorder (i.e., alcohol, marijuana, and/or other drug use disorder), affective disorder (i.e., depression or dysthymia), and anxiety disorder (i.e., separation anxiety disorder and/or posttraumatic stress disorder) were assessed with the authorized Dutch translation of the DISC-IV (Ferdinand & van der Ende, 2002). In line with previous studies (e.g., Colins, Vermeiren, Schuyten, & Broekaert, 2009; Infante, Slattery, Klein, & Essex, 2007), we also differentiated between three diagnostic categories. The first diagnostic category, “pure externalizing disorder,” refers to having only DISC-IV ADHD, ODD, CD, and/or SUD without comorbid affective and/or anxiety disorders. Likewise, the second diagnostic category, “pure internalizing disorder,” refers to having DISC-IV affective and/or anxiety disorders without comorbid externalizing disorders. The third diagnostic category, “both ex- and internalizing disorders,” finally, refers to the presence of at least one DISC-IV externalizing disorder and at least one DISC-IV internalizing disorder.

Quick Big Five

The personality traits Agreeableness and Conscientiousness were measured using self-ratings on the Quick Big Five (QBF), a well-validated instrument assessing the five domains of the Five-Factor Model of personality (Vermulst & Gerris, 2006). The QBF consists of 30 items that are to be rated on a 7-point Likert-type scale (ranging from 1 = not all to 7 = very well). In the current study, only the Agreeableness and Conscientiousness subscales were used in the context of the convergent validity of the APSD-SR and YPI. Because of a heightened interest to relate psychopathy to Big Five dimensions (Salekin & Lynam, 2010), we also present relations between the APSD-SR and YPI with the three remaining QBF subscales. Each dimension is measured by six items. Items referring to Agreeableness are the following: nice, cooperative, kind, helpful, likeable, and sympathetic. Items referring to Conscientiousness are the following: careless (reversed), organized, orderly, efficient, neat, and systematic. Emotional stability was measured by the items irritable, tense, sensitive, concerned, fearful, and nervous (all items reversed); Extraversion by the items reserved, quiet, closed, withdrawn, shy, and talkative (with only talkative loading not needed to be reversed); and Openness by the items imaginative, investigative, versatile, innovative, artistic, and creative.

Self-Reported Offending

We used the WODC Monitor (van der Laan & Blom, 2005) to investigate whether participants ever committed an offense. For the current study, and based on previous studies on psychopathic-like traits (Colins, Noom, & Vanderplasschen, 2012), we created six continuous offense categories (details available on request from the first author). First, four items referring to violence were classified as “violent offenses” (i.e., using violence or the threat of using violence to steal from someone, hitting or kicking with injuries, hitting or kicking without injuries, trying to have sex while the other refuses). Second, 11 items referring to income-related nonviolent delinquent behavior were classified as “property offenses” (e.g., selling stolen property, burglary, shoplifting, bike/scooter theft, theft in school, using someone else’s credit card without permission). Third, six items referring to deliberately damaging property were classified as “vandalism” (i.e., car damaging, house damaging, public transport damaging, damaging something else, deliberate fire setting, graffiti). Fourth, three items referring to dealing or selling drugs were classified as “drug-related offenses” (i.e., selling party drugs, selling soft drugs, selling hard drugs). Fifth, three items referring to threatening and insulting were classified as “threats and insults” (i.e., scaring someone by email, SMS, or chat; threatening someone at school, home, and/or the street; threatening someone at the street to steal from him/her). Sixth, three items that could not be assigned to one of the five previous types of offending were classified as “other offending” (i.e., possession of weapons to protect oneself, deliberately sending viruses by email or the Internet, hacking into someone’s computer).

Sociodemographics

Standardized information about age and origin was assessed by means of a self-report questionnaire designed by the authors.

Data-Analytical Strategy

The proposed factor structure of the APSD-SR (Frick & Hare, 2001), YPI (Andershed et al., 2002), and YPI-SV (van Baardewijk et al., 2010) was investigated using the Interactive Lisrel (8.80) framework (Jöreskog & Sörbom, 2006). Model fit (Estimator: Maximum Likelihood) was assessed using χ², root mean square error of approximation (RMSEA), and comparative fit index (CFI). RMSEA scores below .05 indicate good fit, whereas scores between .05 and .08 indicate acceptable fit. A CFI score of 095 or above indicates excellent fit, and a CFI score of .90 or more indicates good fit (Hu & Bentler, 1995). With regard to chi square, a good fit is indicated when χ²/df ≤ 2, whereas χ²/df ≤ 3 is indicative of an acceptable fit (Schermelleh-Engel, Moosbrugger, & Müller, 2003). Superiority of one model over another was determined by means of Akaike information criterion (AIC), with a lower AIC indicating a better fit. Previous studies in detained adolescents demonstrated that CFA fit indices for the APSD-SR improve if APSD Items 19 and 20 were removed (Douglas, Epstein, & Poythress, 2008; Poythress, Dembo, et al., 2006) and for the YPI if the “lying subscale” was excluded from the CFA (Poythress, Dembo, et al., 2006). Therefore, we also present CFA fit indices for these alternative models.

To evaluate the internal consistency of the APSD-SR and YPI Cronbach’s alphas are presented. Reliability coefficients <.60 are considered insufficient, .60 to .69 marginal, .70 to .79 acceptable, .80 to .89 good, and .90 excellent (Barker, Pistran, & Elliot, 1994). The mean corrected item-to-total correlation (MCITC) helps identify items that, if excluded from the scale, would improve alpha for that scale, whereas the mean inter-item correlation (MIC) is considered to be a more straightforward indicator of the internal consistency of a scale than alpha (Clarck & Watson, 1995). Besides alphas, and in line with Poythress, Dembo, et al. (2006), we therefore also present MCITC and MIC values. The MCITC should be more than the conventionally recommended value of .30 (Nunnally & Bernstein, 1994), whereas MIC values should be at minimum in the range of .15 to .50 to be considered adequate (Clarck & Watson, 1995).

To examine the convergent and criterion validity of the APSD-SR and YPI, odd ratios were calculated to examine the relation between the APSD-SR and the YPI and categorical dependent variables (i.e., psychiatric disorders). Likewise, regression coefficients (β) were calculated to examine the relationship between APSD-SR and YPI and continuous dependent variables (i.e., YSR scores, personality traits, self-reported offending). For each APSD-SR and YPI dimension (e.g., the interpersonal dimension), criterion-validity analyses were repeated while adjusting for the other two APSD-SR and YPI dimensions (e.g., the affective and behavioral dimensions). This allows us to examine the unique association of each dimension with the validity measures. Because a large number of significance tests were conducted, we set α at < .01 as indicator for statistical significance. Descriptives for all measures used in the convergent validity analyses are shown in the appendix.

Procedure

This study was approved by the institutional review board of the Faculty of Psychology and Educational Sciences, Ghent University. Because screening for emotional problems is a mandatory task of the YDC, the institutional review board waived the requirement of parental consent. Participants were approached and assessed following a standardized protocol. Selected detainees were approached individually and given oral and written information about the aims, the content, and the duration of the study. They were assured that their information would be treated confidentially and that refusal to participate would not affect their judicial status or stay in the YDC. The adolescents could consult their primary caregivers or other adults about participation and written informed consent was given before participation. Participating minors did not receive compensation.

Results

Factorial Validity

For the APSD-SR, four models are compared: (a) a one-factor model, in which the 20 APSD-SR items are indicators of a single latent factor “psychopathy”; (b) an oblique two-factor model, in which 16 of the APSD-SR items load onto a factor Impulsivity and a factor Callous and Unemotional Traits (Frick, Bodin, & Barry, 2000); (c) an oblique three-factor model, in which 18 of the APSD-SR items load onto a factor Impulsivity (Frick & Hare, 2001), a factor Narcissism, and a factor Callous and Unemotional Traits; (d) a variant of the latter model using only 16 of the APSD-SR items (omission of Items 19 and 20; Douglas et al., 2008; Poythress, Dembo, et al., 2006). Table 1 shows that none of the models tested reached an acceptable model fit on all the indices. For the two- and three-factor models, RMSEA and χ²/df meet the criteria for acceptable fit. Table 1 also shows that the two-factor model is better than the three-factor model (AIC = 277.58 and. 363.17, respectively). For several reasons, we decided to use the APSD-SR three-factor structure in all further analysis. First, the developer of the APSD-SR suggested that the three-factor solution is the factor structure of choice, for example, because separating Narcissism from the Callous and Unemotional traits factor may facilitate further examination of their unique associations (Frick et al., 2000). Second, the current literature on the factor structure of psychopathy is predominantly about three or four factors that underlie the psychopathy construct (e.g., Cooke, Michie, & Skeem, 2007; Munoz, Kerr, & Besic, 2008). Because the fit indices of the two- and three-factor models are quite similar, we considered the three-factor model of the APSD-SR as most informative.

Table 1.

Goodness-of-Fit Statistics of All Models Tested.

	χ²	df	χ²/df	RMSEA	CFI	AIC
Antisocial Process Screening Device–Self-Report
One-factor model	438.61	170	2.58	0.091	0.82	518.61
Two-factor model	211.58	103	2.05	0.074	0.85	277.58
Three-factor model (manual)	285.17	132	2.16	0.078	0.85	363.17
Three-factor model (Items 19 and 20 not included)	237.66	101	2.35	0.084	0.86	307.66
Youth Psychopathic Traits Inventory
One-factor model	274.86	35	7.85	0.190	0.85	314.86
Three-factor model	53.70	32	1.68	0.060	0.98	99.70
Three-factor model (Lie scale not included)	47.13	24	1.96	0.071	0.98	89.13
Youth Psychopathic Traits Inventory–Short Version
One-factor model	542.65	135	4.02	0.130	0.76	614.65
Three-factor model	242.61	132	1.84	0.066	0.91	320.61

Note. RMSEA = root mean square error of approximation; CFI = comparative fit index; AIC = Akaike information criterion. Indices meeting the recommended cutoff for acceptable fit are italized; indices meeting the recommended cutoff for good fit are in boldface.

For the YPI original version, three models were compared: (a) a one-factor model, in which the 10-item parcels load onto a single latent factor “psychopathy”; (b) a three-factor model, in which 4 parcels load onto a “Grandiose/Manipulative” factor, 3 parcels load onto a “Callous Unemotional’ factor, and 3 parcels load onto a “Impulsive-Irresponsible” factor (Andershed et al., 2002). (c) a modified three-factor model that did not include the Lie subscale. For the one-factor model, none of the fit indices is acceptable; for the three-factor model, however, RMSEA meets the criterion of acceptable fit and both χ²/df and CFI meet the criteria of good fit.

For the YPI-SV, two models were compared: (a) a one-factor model, in which the 18 items load onto a single latent factor “psychopathy”; (b) a three-factor model, in which 6 items load onto three factors: a “Grandiose/Manipulative” factor, a “Callous/Unemotional” factor, and a “Impulsive-Irresponsible” factor (van Baardewijk et al., 2010). Again, for the one-factor model, none of the fit indices is acceptable, whereas for the three-factor model, however, both RMSEA and CFI meet the criterion of acceptable fit and χ²/df meets the criteria of good fit.

Internal Consistency

Reliability indices, means, and standard deviations of the APSD-SR, YPI, and YPI-SV are shown in Table 2.³ As for the APSD-SR, Cronbach’s alpha was acceptable for the total score, marginal for the interpersonal dimension (ID) and the behavioral dimension (BD), and insufficient for the affective dimension (AD). For the total score and the AD, the MIC values were not in the recommended range (.15 to .50). In contrast to the total score and the other two dimensions, the MCITC for AD was below the recommended value of .30. Regarding the YPI, Cronbach’s alpha for the total score was excellent and good for all three dimensions and all but three subscales. The subscales Grandiosity and Unemotionality had Cronbach’s alphas referring to marginal internal consistency, whereas Cronbach’s alpha for the subscale Callousness suggested that this subscale’s internal consistency was insufficient. However, the MIC values for all dimensions and all subscales were in the recommend range to be considered adequate, although the MIC value for the Callousness subscale was just above the recommended value of .15. Finally, the MCITCs for all dimensions and all but the Callousness subscale (.27) were above the recommended value of .30. As for the YPI-SV, Cronbach’s alpha for the YPI-SV total score was good and acceptable for all three YPI-SV dimensions. Likewise, the total score and the three dimensions’ MICs and MCITCs were in or above the recommended range of values.

Table 2.

Descriptives and Internal Consistency Indices for the APSD-SR, the YPI, and the YPI-SV.

	M	SD	Min, Max	Theoretical range	Alpha	MIC	MCITC
APSD-SR Total score	15.41	5.89	1-29	0-40	0.76	0.14	0.34
APSD-SR Interpersonal dimension	4.19	2.62	0-11	0-14	0.65	0.21	0.37
APSD-SR Affective dimension	4.57	2.13	0-12	0-12	0.43	0.11	0.21
APSD-SR Behavioral dimension	5.16	2.33	0-10	0-10	0.67	0.28	0.42
YPI Total score	99.21	22.27	57-158	50-200	0.92	0.20	0.43
YPI Interpersonal dimension	31.13	9.64	20-64	20-80	0.89	0.29	0.51
YPI Affective dimension	28.99	7.87	16-57	15-60	0.81	0.22	0.42
YPI Behavioral dimension	39.09	9.47	15-60	15-60	0.87	0.30	0.51
YPI Dishonest Charm subscale	8.96	3.55	5-20	5-20	0.81	0.46	0.60
YPI Grandiosity subscale	6.99	2.39	5-17	5-20	0.65	0.31	0.43
YPI Lying subscale	7.04	2.74	5-20	5-20	0.77	0.41	0.55
YPI Manipulation subscale	8.17	3.62	5-20	5-20	0.87	0.58	0.70
YPI Remorselessness subscale	8.93	3.57	5-20	5-20	0.77	0.39	0.53
YPI Unemotionality subscale	9.81	3.11	5-19	5-20	0.62	0.24	0.37
YPI Callousness subscale	10.25	2.90	5-18	5-20	0.47	0.16	0.27
YPI Thrill-seeking subscale	14.15	3.42	5-20	5-20	0.71	0.32	0.47
YPI Impulsiveness subscale	13.05	3.53	5-20	5-20	0.70	0.32	0.45
YPI Irresponsibility subscale	11.88	3.91	5-20	5-20	0.72	0.34	0.49
YPI-SV Total score	35.16	9.49	18-58	18-72	0.82	0.20	0.37
YPI-SV Interpersonal dimension	9.64	3.38	6-19	6-24	0.73	0.30	0.46
YPI-SV Affective dimension	10.07	3.65	6-22	6-24	0.71	0.29	0.45
YPI-SV Behavioral dimension	15.44	3.98	6-24	6-24	0.70	0.28	0.48

Note. APSD-SR = Antisocial Process Screening Device–Self-Report; YPI = Youth Psychopathic Traits Inventory; YPI-SV = Youth Psychopathic Traits Inventory–Short-Version; MIC = mean interitem correlation; MCITC = mean corrected item-to-total correlation. Indices below the recommended value for at least acceptable reliability are in boldface.

The correlations between the APSD-SR dimensions were r .25_ID-AD, r .52_ID-BD, and r .17_AD-BD; and between the total score and each dimension: r .82_total-ID, r. 57_total-AD, and r .78_total-BD. The correlations between the YPI dimensions were r .54_ID-AD, r .53_ID-BD, and r .48_AD-BD and between the YPI Total score and each dimension: r .85_total-ID, r .79_total-AD, and r .83_total-BD. The correlations between the YPI-SV dimensions were r .40_ID-AD, r. 40_ID-BD, and .39_AD-BD and between the YPI-SV total score and each dimension: r .76_total-ID, r .76_total-AD, and r .79_total-BD.

Convergent Validity

The APSD-SR total score, ID, and BD show a strong correlation with the corresponding YPI/YPI-S total score and dimension (rs of .72/.62, .67/.56, and .65/.60, respectively). Yet the APSD-SR AD was poorly correlated to the AD of the YPI/YPI-SV AD (r. 34/.15; see Table 3).

Table 3.

Correlations Between APSD-SR and the YPI and the YPI-SV.

	APSD-SR				YPI
	Total	ID	AD	BD	Total	ID	AD	BD
YPI Total score	0.72***	0.62***	0.31***	0.61***	—	—	—	—
YPI Interpersonal dimension	0.63***	0.67***	0.21**	0.42***	0.85***	—	—	—
YPI Affective dimension	0.56***	0.49***	0.34***	0.42***	0.79***	0.54***	—	—
YPI Behavioral dimension	0.57***	0.38***	0.23**	0.65***	0.83***	0.53***	0.48***	—
YPI Dishonest Charm subscale	0.60***	0.61***	0.26***	0.39***	0.76****	0.88***	0.47***	0.50***
YPI Grandiosity subscale	0.18	0.31**	−0.05	0.07	0.52***	0.56***	0.38***	0.33***
YPI Lying subscale	0.42***	0.48***	0.05	0.33***	0.56***	0.71***	0.28***	0.36***
YPI Manipulation subscale	0.65***	0.62***	0.30***	0.43***	0.76***	0.90***	0.51***	0.44***
YPI Remorselessness subscale	0.60***	0.57***	0.28***	0.47***	0.79***	0.58***	0.86***	0.56***
YPI Unemotionality subscale	0.33***	0.29***	0.15	0.28***	0.62***	0.40***	0.82***	0.38***
YPI Callousness subscale	0.43***	0.31***	0.42***	0.24***	0.50***	0.32***	0.76***	0.21**
YPI Thrill-seeking subscale	0.50***	0.35***	0.16	0.58***	0.75***	0.52***	0.44***	0.87***
YPI Impulsiveness subscale	0.46***	0.30***	0.11	0.60***	0.68***	0.40***	0.38***	0.87***
YPI Irresponsibility subscale	0.53***	0.36***	0.32***	0.52***	0.73***	0.47***	0.43***	0.88***
YPI-SV Total score	0.62***	0.56***	0.22**	0.55***	—	—	—	—
YPI-SV Interpersonal dimension	0.55***	0.56***	0.20**	0.34**	—	—	—	—
YPI-SV Affective dimension	0.38***	0.38***	0.15	0.31***	—	—	—	—
YPI-SV Behavioral dimension	0.50***	0.36***	0.17	0.60***	—	—	—	—

***

p < .001. **p < .01.

With regard to the YPI subscales, the four ID subscales were significantly stronger related to the APSD-SR ID than to the other two APSD-SR dimensions (based on Hotelling’s T2 statistic for tests of difference between dependent correlations). Likewise, the three YPI BD subscales were significantly stronger related to the BD of the APSD-SR. Yet only the YPI Callousness subscale was stronger related to the APSD-SR AD than the other two APSD-R dimensions. In contrast, the YPI Remorselessness and Unemotional subscales were stronger (though not significantly stronger) related to the ID and BD than to the AD of the APSD-SR (Table 3).

Criterion Validity: Psychiatric Disorders (DISC-IV)

As for the APSD-SR, the total score was positively related with ADHD, ODD, CD, and pure externalizing disorders, negatively related with pure internalizing disorders, and unrelated with substance use disorders (SUD), affective disorders, anxiety disorders, and having both externalizing and internalizing disorders. Univariate analyses showed that the ID and the AD were generally not related to psychiatric disorder. In addition, the positive relation between the ID and ODD and CD, and the negative relation between the AD and pure internalizing disorders, remained no longer significant after controlling for the other two APSD-SR dimensions. Univariate and multivariate analyses consistently showed that the BD was positively related with ADHD, ODD, CD, and SUD, negatively related with pure internalizing disorders, and unrelated to all other disorders or category of disorders (Table 4).

Table 4.

Univariate and Multivariate Relations Between APSD-SR, YPI, YPI-SV, and DISC-IV Psychiatric Disorders (Odd Ratios)^a.

	APSD-SR				YPI				YPI-SV
Psychiatric disorder	Total	ID	AD	BD	Total	ID	AD	BD	Total	ID	AD	BD
ADHD	1.58*	1.18 (0.78)	1.26 (1.19)	1.93(2.13)	1.65*	1.27 (0.86)	1.37 (1.05)	2.07** (2.19*)	1.62*	1.24 (0.96)	1.26 (1.00)	1.97** (2.00*)
ODD	2.85**	2.14** (1.39)	1.30 (1.02)	3.46 (2.93)	2.40**	1.67* (0.96)	1.75** (1.15)	2.97 (2.82)	2.18**	1.51* (1.08)	1.42 (1.00)	2.79 (2.70)
CD	4.40**	2.70** (1.62)	1.42 (1.21)	3.59 (2.94)	3.70**	2.78** (1.60)	2.17** (1.21)	3.28 (2.44)	3.04**	2.05**(1.45)	1.63* (1.06)	3.20 (2.75)
Any SUD	2.27**	1.53 (0.93)	1.34 (1.24)	2.40 (2.46)	2.38**	1.68* (0.92)	1.59 (1.03)	2.82 (2.92)	2.10**	1.67* (1.29)	1.37 (0.97)	2.21** (2.02*)
Any affective disorder	0.80	0.79 (0.76)	0.75 (0.78)	1.01 (1.21)	0.99	0.87 (0.75)	0.90 (0.86)	1.21 (1.53)	1.08	0.84 (0.71)	1.09 (1.09)	1.27 (1.41)
Any anxiety disorder^b	0.90	0.98 (1.04)	0.83 (0.82)	0.96 (0.97)	0.87	0.93 (1.06)	0.86 (0.88)	0.87 (0.90)	0.95	1.05 (1.13)	0.93 (0.92)	0.92 (0.90)
Pure externalizing^c	1.62*	1.44 (1.22)	1.38 (1.28)	1.43 (1.24)	1.34	1.37 (1.31)	1.29 (1.15)	1.17 (0.94)	1.20	1.27 (1.27)	1.06 (0.95)	1.14 (1.05)
Pure internalizing^c	0.19**	0.45 (1.12)	0.45* (0.49)	0.25(0.22)	0.11**	0.21* (0.63)	0.62 (2.13)	0.08 (0.07)	0.29*	0.36 (0.46)	1.03 (2.99*)	0.15 (0.10)
Both exernalizing and internalizing^c	1.08	0.95 (0.81)	0.93 (0.92)	1.25 (1.42)	1.23	1.05 (0.84)	1.01 (0.81)	1.55* (1.89*)	1.30	1.08 (0.91)	1.09 (0.95)	1.53* (1.62*)

Note. APSD-SR = Antisocial Process Screening Device–Self-Report; YPI = Youth Psychopathic Traits Inventory; YPI-SV = Youth Psychopathic Traits Inventory–Short-Version; ID = Interpersonal dimension; AD = Affective dimension; BD = Behavioral dimension; ADHD = attention-deficit/hyperactivity disorder; ODD = oppositional defiant disorder; CD = conduct disorder; SUD = substance use disorder.

APSD-SR and YPI scores were standardized before being included as predictor in logistic regression analyses.

Due to missing values: n = 188.

Due to missing values: n = 189.

p < .001. *p < .01.

As for the YPI, Table 4 showed that overall, the relations described above for the APSD-SR are replicated for the YPI and the YPI-SV. There were, however, three remarkable exceptions. First, in contrast to the APSD-SR, the YPI and YPI-SV total score were also positively related with SUD, but not positively related with pure externalizing disorders. Second, the BDs of the YPI and YPI-SV were positively related with having both externalizing and internalizing disorders. Third, whereas the AD of the APSD-S was negatively related to pure internalizing disorders, the AD of the YPI was unrelated, and the AD of the YPI-SV positively related with having pure internalizing disorders (Table 4).

Criterion Validity: Behavioral and Emotional Problems (YSR)

As for the APSD-SR, the total score and BD were positively related with rule-breaking behavior, aggression, and attention problems and unrelated with anxious-depressive and withdrawn-depressive feelings. With regard to the ID, univariate analyses demonstrated the same pattern of relations as for the APSD-SR total score and BD. However, after adjusting for the other two dimensions, only the positive relation between ID and aggression remained significant. Univariate analyses also demonstrated that the AD was positively related to rule-breaking behavior, aggression, and attention problems, negatively to anxious-depressive, and unrelated to withdrawn-depressive feelings anxious-depressive. After controlling for the other two dimensions, the AD remained positively related with aggression and negatively related with anxious-depressive feelings, while a positive relation with withdrawn-depressive feelings emerged.

Using the YPI and YPI-SV (Table 5) revealed the same results as described above, with four exceptions. First, the positive relation between the ID and AD of the YPI/YPI-SV, on the one hand, and aggression, on the other, did not remain significant in the multivariate analyses. Second, in univariate and multivariate analyses, the AD of the YPI and YPI-SV were not negatively related to anxious-depressive feelings. Third, in multivariate analyses, the ID became significantly negatively related with anxious-depressive feelings. Fourth, the AD of the YPI was not positively related with withdrawn-depressive feelings as was the APSD-SR and YPI-SV (Table 5).

Table 5.

Univariate and Multivariate Relations Between APSD-SR, YPI, YPI-SV, and Youth Self-Report Behavioral and Emotional Problems and Quick Big Five Personality Traits (Standardized Betas).

	APSD-SR				YPI				YPI-SV
	Total	ID	AD	BD	Total	ID	AD	BD	Total	ID	AD	BD
Rule breaking^a	0.58**	0.42** (0.14)	0.24* (0.12)	0.58 (0.49)	0.57**	0.40** (0.08)	0.36** (0.05)	0.64 (0.57)	0.48**	0.29** (0.09)	0.25* (0.04)	0.53 (0.48)
Aggression^b	0.66**	0.51 (0.24)	0.34 (0.21)	0.58 (0.43)	0.55**	0.40** (0.10)	0.41** (0.15)	0.55 (0.43)	0.49**	0.33** (0.14)	0.30** (0.10)	0.48 (0.39)
Attention^a	0.43**	0.33** (0.12)	0.21* (0.14)	0.44 (0.36)	0.41**	0.27** (0.02)	0.22* (−0.03)	0.50 (0.51)	0.38**	−0.18 (≥0.01)	0.20* (0.05)	0.45 (0.44)
Anxious-Depressive^b	−0.17	−0.07 (−0.01)	−0.23* (−0.23*)	−0.05 (−0.01)	−0.13	−0.15 (−0.12)	−0.16 (−0.14)	−0.03 (0.10)	−0.07	−0.18 (−0.22*)	0.02 (0.08)	−0.02 (0.04)
Withdrawn Depressive^a	−0.01	−0.02 (−0.04)	0.11 (0.12**)	−0.02 (−0.02)	0.01	−0.07 (−0.17)	0.05 (0.11)	0.04 (0.07)	0.02	−0.10 (−0.18)	0.17 (0.25*)	−0.01 (−0.03)
Agreeableness	−0.28**	−0.14 (0.02)	−0.29 (−0.27)	−0.22* (−0.19)	−0.10	−0.01 (0.13)	−0.18 (−0.24*)	−0.06 (−0.02)	−0.05	0.01 (0.07)	−0.07 (−0.07)	−0.08 (−0.08)
Conscientiousness	−0.38**	−0.19* (0.05)	−0.33 (−0.29)	−0.34 (−0.32)	−0.24*	0.16 (<0.01)	−0.19 (−0.09)	−0.25** (0.21)	−0.19	−0.19* (−0.16)	−0.07 (0.04)	−0.17 (−0.12)
Extraversion	0.07	0.11 (0.08)	0.10 (−0.02)	0.09 (0.05)	0.19	0.22* (0.23)	0.06 (−0.11)	0.16 (0.09)	0.17	0.22* (0.22*)	0.02 (−0.09)	0.15 (0.09)
Emotional stability	0.12	−0.04 (−0.04)	0.10 (0.12)	−0.06 (−0.06)	0.11	0.14 (0.14)	0.20* (0.23*)	−0.04 (−0.23*)	0.08	0.13 (0.14)	0.14 (0.16)	−0.08 (−0.19)
Openness to experience	0.03	0.09 (0.08)	−0.10 (−0.14)	0.10 (0.09)	0.23*	0.23* (0.19)	0.10 (−0.07)	0.22* (0.15)	0.25**	0.23* (0.18)	0.11 (−0.02)	0.23* (0.17)

Due to missing values: n = 182.

Due to missing values: n = 181.

p < .001. *p < .01.

Criterion Validity: Agreeableness and Conscientiousness

As for the APSD-SR, the total score, the AD, and the BD were negatively related with Agreeableness. In multivariate analyses, only the AD remained significantly related to Agreeableness. The APSD-SR total score and all three dimensions were negatively related with Conscientiousness. All these relations remained significant in multivariate analyses, except for the ID (Table 5).

As for the YPI, Table 5 shows that none of the YPI scores were significantly related with Agreeableness, except for the AD, which showed a negative relation with this Big Five personality trait (though only in multivariate analyses). The YPI total score and the BD were negatively related to Conscientiousness (both in univariate and multivariate analyses). With regard to the YPI-SV, the ID was negatively related to Conscientiousness, a finding that became nonsignificant after controlling for the other two dimensions.

Criterion Validity: Self-Reported Offending

As for the APSD-SR, the total score was positively related with all types of offending. In univariate analyses, the ID and BD were positively related with all types of offending (except that the ID was not related with drug offenses; Table 6). The AD was only positively related with violent offending and vandalism. After adjusting for the other two dimensions, the BD remained significantly related with all types of offending, whereas the ID remained significantly related with theft, vandalism, threats, and insults. Finally, the AD only remained significantly related with vandalism (Table 6).

Table 6.

Univariate and Multivariate Relations Between APSD-SR, YPI, YPI-SV, and Type of Offending (Standardized Beta).

	APSD-SR				YPI				YPI-SV
Type of offending	Total	ID	AD	BD	Total	ID	AD	BD	Total	ID	AD	BD
Violent offenses	0.43**	0.32** (0.15)	0.21* (0.13)	0.37 (0.27)	0.31**	0.22* (0.03)	0.24* (0.11)	0.31** (0.24*)	0.29**	0.18** (0.01)	0.16 (0.04)	0.32 (0.28)
Theft	0.48**	0.38** (0.20*)	0.17 (0.07)	0.44 (0.32)	0.49**	0.41** (0.22*)	0.32** (0.06)	0.45 (0.31)	0.46**	0.36** (0.21*)	0.24* (0.03)	0.44 (0.35)
Vandalism	0.57**	0.45 (0.27)	0.28** (0.17*)	0.44 (0.27)	0.49**	0.36** (0.07)	0.43** (0.26*)	0.44** (0.27*)	0.47**	0.31** (0.12)	0.35** (0.20*)	0.41 (0.29)
Threats and insults	0.42**	0.38** (0.25*)	0.17 (0.07)	0.36** (0.22*)	0.32**	0.24* (0.06)	0.25** (0.11)	0.31 (0.22)	0.32**	0.17 (0.01)	0.26** (0.17)	0.30** (0.24*)
Drug offenses	0.30**	0.17 (0.04)	0.08 (0.03)	0.27 (0.24)	0.24**	0.13 (− 0.07)	0.23* (0.18)	0.23* (0.18)	0.22*	0.20* (0.13)	0.15 (0.07)	0.17 (0.09)
Other offenses	0.37**	0.29** (0.14)	0.14 (0.07)	0.33** (0.25*)	0.33**	0.20* (−0.06)	0.33** (0.25*)	0.31** (0.33*)	0.33*	0.19* (0.03)	0.30** (0.22*)	0.28** (0.18)

p < .001. *p < .01.

As for the YPI, Table 6 also shows that the YPI total score was positively related with all types of offending. In univariate analyses, all three dimensions were positively related with all or almost all types of offending. After adjusting for the other two dimensions, the ID only remained significantly positively related to theft, and the AD to vandalism and other offenses. The BD remained significantly positively related to all but one type of offense (i.e., drug offenses). Results for the YPI-SV were similar to the YPI, except that in univariate analyses, the ID was not related to threats and insults, the AD not to violent offenses and drug offenses, and the BD not with drug offenses. Finally, the BD was no longer related with other offenses after controlling for the other two dimensions.

Discussion

The current study examined the factor structure, reliability, and validity of the APSD-SR and YPI in a relative large sample of detained female adolescents. Overall, the current study could only support the proposed three-factor structure of the original 50-item YPI and YPI-SV. Using the often reported “rules of thumb” in psychopathy research, neither the proposed three-factor structure of the APSD-SR nor one of the alternate models yielded at least adequate fit through all fit indices. In addition, reliability indices for the YPI and YPI-SV total score and three dimensions were generally all above the recommended cut-offs and higher to those reported for the APSD-SR. APSD-SR and YPI scales were positively related with each other, supporting the convergent validity of both instruments. However, the AD of the APSD-SR and the YPI/YPI-SV were not significantly related to each other. Overall, all the questionnaires showed good criterion validity.

As cogently argued by others (Hopwood & Donnelan, 2010), asserting internal structure should be regarded as just one element of construct validity among several others. Despite the lack of an adequate factor structure and despite its poor reliability, previous studies as well as the current study showed that the APSD-SR scores in general were related to criterion variables in a way that was expected (e.g., the AD was less highly correlated with CD than the other two dimensions). Therefore, it would not be reasonable to argue that the APSD-SR is not useful and that all evidence gathered thus far should be disregarded. We do suggest that the APSD-SR has room for improvement in terms of how well this self-report questionnaire conforms to its hypothesized structure. Having a stable factor structure that can be replicated in many different samples increases confidence in the usefulness of summary scores (Hopwood & Donnelan, 2010). For example, if the actual items of the AD do not neatly cluster into a single dimension, it is difficult to interpret the meaning of the AD score based on those items and any predictor–criterion relationship involving that AD score (Hopwood & Donnelan, 2010; Kotler & McMahon, 2010). In addition, having a clear factor structure allows researchers to explain differences between adult and youth psychopathy literature. For example, based on adult literature, the ID and AD are expected to be more highly related than the ID and BD (Kotler & McMahon, 2010). However, the current study showed that the reverse was true for the APSD-SR (r_{ID and AD} = .25; r_{ID and BD} = .53), a finding that is in line with previous APSD-SR studies (e.g., Bijttebier & Decoene, 2009). Because the YPI correlations were more closely in line with adult literature (YPI/YPI-SV: _{ID and AD} = .54/.40; _{ID and BD} = .53/.40), it is possible that this difference in factor structure between adult literature and the APSD-SR is a consequence of the psychometric features of this measure and not because of real developmental manifestations.

Although the YPI was developed for use in community youth, an increasing number of studies in justice-involved adolescents provides strong support for its proposed three-factor structure, its reliability (e.g., Colins et al., 2012; Salekin et al., 2010), and its criterion validity (e.g., Skeem & Cauffman, 2003; Veen et al., 2011). The current study substantially adds to this empirical evidence in favor of the YPI, not only because we showed that the three-factor structure of the YPI could be replicated in a sample of detained female adolescents but also because we confirmed the previous suggestion (Poythress, Dembo, et al., 2006) that the factor structure and reliability of the YPI are better than the APSD-SR. Importantly, CFA studies on the YPI typically used the 10 YPI subscale scores rather than the 50 YPI items. This strategy where items are summed to form composite scores prior to FA is known as parceling. Parceling has the advantage that combining items results in composite variables with better psychometric properties (e.g., internal consistency) than item variables and better model fit indices in CFA studies (Cooke et al., 2007). This suggests that the better fit of the YPI compared to the APSD-SR may be a consequence of using parcels (cf. YPI) rather than items (cf. APSD-SR) to test the factor structure of the respective instruments. Interestingly, while the YPI-SV’s CFA and reliability fit indices were not as good as the original 50-item YPI, they were still remarkably better than the APSD-SR. Thus, the YPI also had an acceptable model fit when a similar number of items (and no parcels) as the APSD-SR were used. This also suggests that the poor reliability of the APSD-SR AD (see also Poythress, Douglas, et al., 2006) is not because of its restricted number of items and that extending the number of items is not a necessity to improve the reliability of the assessment of callous-unemotional traits. The reason that the psychometric properties of the APSD-SR are not as good as those of the YPI and YPI-SV may be a consequence of the way the APSD-SR was developed. Because the APSD was not designed with self-report in mind, the APSD-SR items are very straightforward measures of traits that are obviously negative in many cases and are likely to increase response bias (Andershed et al., 2002; Kotler & McMahon, 2010). Consequently, rewording items may be the first strategy of choice to back up the APSD-SR criterion validity with a better internal structure.

Internalizing disorders are often assumed to be negatively related with psychopathic-like traits (Sevecke & Kosson, 2010). However, few studies have examined the relation between internalizing disorders and psychopathic-like traits. After reviewing relevant literature, it has been argued that relationships between psychopathic-like traits and internalizing disorders in youths are remarkably less consistent than relationships between these traits and externalizing disorders (Sevecke & Kosson, 2010). In addition, it has been suggested that, among youths, relationships between internalizing disorders are often quite different than those in adults, in that, most studies among youths reported no relationship or a positive relationship between psychopathic-like traits and affective and anxiety disorders (Sevecke & Kosson, 2010). The current study, indeed, seems to concur with this review of the literature, given that the APSD-SR and YPI total scores were all positively related with ADHD, ODD, and CD but unrelated with anxiety disorders and affective disorders. Interestingly, the theoretically expected negative relationship between psychopathic-like traits and internalizing disorders was confirmed when we took into account that internalizing disorders in detained youths are most often comorbid with externalizing disorders (e.g., Colins et al., 2009). That is, APSD-SR and YPI total score were only negatively related to internalizing disorders that were not comorbid with externalizing disorders and not related to internalizing disorders that co-occur with externalizing disorders. In addition, the APSD-SR total score (but not the YPI/YPI-SV total score) was positively related with pure externalizing disorders. Taken together, these patterns may suggest that the relationship between psychopathic-like traits and internalizing disorder may become more transparent if one focuses on internalizing disorders that do not co-occur with externalizing disorders.

Finally, in line with a previous study (Poythress, Dembo, et al., 2006), we found a weak correlation or no significant correlation at all between the AD of the APSD-SR and the YPI or YPI-SV, respectively. Given that all items refer to the concept of callous-unemotional dimension of the psychopathy construct, it is unclear why the AD of the YPI-SV is not related at all with the corresponding dimension of the APSD-SR. These differences may influence clinical work. For example, if the DSM-IV CU specifier (Frick & Moffit, 2010) is accepted, it is likely that clinicians will use standardized instruments to assess CU (specifier) traits. Having demonstrated that the APSD-SR appears to be effective at assessing narcissism and impulsivity but not callousness, it has recently been strongly recommended that research should continue to attempt to identify the measures that most accurately assess traits related to the psychopathy construct (Dillard, Salekin, Barker, & Grimes, 2012). The current study suggests that the YPI is a more reliable instrument to assess the presence of CU traits in juvenile justice–involved girls than the APSD-SR. However, an intriguing question for future research will be how the YPI can be used to assess the DSM-CU specifier. The four CU specifier criteria can straightforwardly be assessed by means of four corresponding APSD-SR items. Yet the fourth CU specifier criterion (i.e., “unconcerned about performance”) cannot be assessed by the YPI. Specifically, although the YPI does contain items referring to poor performance, it does not ask whether youths are concerned about this performance (see also Colins & Vermeiren, in press). Consequently, it seems that the self-report questionnaire with the best psychometric properties in juvenile justice–involved youths is not entirely able to measure a concept that is very likely to influence research and clinical work in the near future.

The current study has several strengths. First, it included a relatively large sample of seriously antisocial youths who have been understudied, that is, detained female adolescents. Second, we included multiple measures of psychopathology and psychopathic-like traits. By doing so, we answered the call for psychometric studies that includes multiple measures of psychopathic-like traits (Dillard et al., 2012; Sevecke & Kosson, 2010). Third, we used a diagnostic psychiatric interview to assess psychiatric disorders. This enabled us to address previous calls for studies on the relationship between psychopathic-like traits and psychiatric disorders such as anxiety and affective disorders (Salekin et al., 2004; Sevecke & Kosson, 2010).

Despite these strengths, the results of our study must be interpreted in the light of some limitations. First, we entirely relied on self-report information to measure criterion validity. Therefore, correlations between self-report psychopathy measured and criterion variables may partly be explained by shared method variance. Second, the cross-sectional nature of the current study does not allow conclusions about causality, for example, between psychopathic-like traits and offending. We intend to collect official information on criminal recidivism, which will allow us to examine the predictive validity of the APSD-SR and YPI to foretell future crime. Third, we guaranteed anonymity and confidentiality to the participants. Therefore, it is a question whether, for example, the internal structure and validity of the YPI can be replicated when youths administer the YPI as if the information they provide would be available to detention staff. Fourth, all YPI-SV data stem from an administration of the original 50-item YPI. Therefore, it is unclear how the YPI-SV would have performed when participants filled out only the 18 YPI-SV items without the remaining 32 original YPI items.

In conclusion, the current study showed that the YPI’s factor structure and reliability is superior to the APSD-SR in detained female adolescents. This superiority is not because of the larger number of items in the YPI because we also demonstrated that the YPI-SV is better than the APSD-SR when it comes to factor structure and reliability. In line with most previous APSD-SR studies in juvenile justice–involved youths, we replicated the difficulty in gaining a good factorial validity of the APSD-SR as well as a reliable measure of callous-unemotional traits.

Footnotes

Appendix

Descriptives for Variables Used in Convergent Validity Analyses.

Prevalence of psychiatric disorder			YSR problems/QBF personality traits					Self-reported offending
Disorder or disorder cluster	N	%	Problem	M	SD	Min, Max	Theo. R	Type of offense	M	SD	Min, Max	Theo. R
ADHD	45	23.6	YSR Rule Breaking^a	13.71	5.74	1-27	0-30	Violent offenses	1.00	1.02	0-3	0-4
ODD	72	37.7	YSR Aggression^b	12.50	6.78	0-32	0-34	Theft	2.46	2.65	0-10	0-11
CD	118	61.8	YSR Attention ^a	9.08	3.77	1-18	0-18	Vandalism	1.20	1.50	0-6	0-6
Any SUD	136	71.2	YSR Anxious-Depressive^b	9.81	5.96	0-26	0-26	Threats and insults	0.75	0.97	0-3	0-3
Any affective disorder	82	42.9	YSR Withdrawn Depressive^a	5.93	3.25	0-16	0-16	Drug offenses	0.82	1.16	0-3	0-3
Any anxiety disorder^c	72	37.7	QBF Agreeableness	31.77	5.45	15-42	6-42	Other offenses	0.49	0.69	0-3	0-3
Pure Externalizing^d	70	36.6	QBF Conscientiousness	25.86	8.75	7-42	6-42
Pure Internalizing^d	16	8.4	QBF Extraversion	27.44	7.02	9-42	6-42
Both Externalizating and Internalizing^d	93	48.7	QBF Emotional StabilityQBF Openness to experience	20.9026.01	7.227.50	6-426-42	6-426-42

Note. YSR = Youth Self-Report; QBF = Quick Big Five; Theo. R = theoretical range; ADHD = attention-deficit/hyperactivity disorder; ODD = oppositional defiant disorder; CD = conduct disorder; SUD = substance use disorder.

Due to missing values: n = 182.

Due to missing values: n = 181.

Due To missing values: n = 188.

Due to missing values: n = 189.

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

Notes

References

Achenbach

T. M.

Rescorla

L. A.

(2001). Manual for the ASEBA school-age forms & profiles. Burlington: University of Vermont, Research Center for Children, Youth, & Families.

Andershed

(2010). Stability and change of psychopathic-traits. What do we know? In Salekin

R. T.

Lynam

D. R.

(Eds.), Handbook of child and adolescent psychopathy (pp. 233-250). New York, NY: Guilford Press.

Andershed

Hodgins

Tengstrom

(2007). Convergent validity of the Youth Psychopathic Traits Inventory (YPI)—Association with the Psychopathy Checklist: Youth Version (PCL: YV). Assessment, 14, 144-154. doi:10.1177/1073191106298286

Andershed

Kerr

Stattin

Levander

(2002). Psychopathic traits in non-referred youths: Initial test of a new assessment tool. In Blaauw

(Ed.), Psychopaths: Current international perspectives (pp. 131-158). The Hague, Netherlands: Elsevier.

Barker

Pistran

Elliot

(1994). Research methods in clinical and counselling psychology. Chichester, England: Wiley.

Bauer

D. L.

Whitman

L. A.

Kosson

D. S.

(2011). Among incarcerated adolescent girls: Reliability and construct validity of Psychopathy Checklist: Youth Version scores. Criminal Justice and Behavior, 38, 965-987.

Bijttebier

Decoene

(2009). Assessment of psychopathic traits in children and adolescents: Further validation of the Antisocial Process Screening Device and the Childhood Psychopathy Scale. European Journal of Psychological Assessment, 25, 157-163.

Boccaccini

M. T.

Epstein

Poythress

Douglas

K. S.

Campbell

Gardner

Falkenbach

(2007). Self-report measures of child and adolescent psychopathy as predictors of offending in four samples of justice-involved youth. Assessment, 14, 361-374.

Campbell

M. A.

Porter

Santor

(2004). Psychopathic traits in adolescent offenders: An evaluation of criminal history, clinical, and psychosocial correlates. Behavioral Sciences & the Law, 22, 23-47.

10.

Cauffman

Kimonis

E. R.

Dmetrieva

Monahan

(2009). A multimethod assessment of juvenile psychopathy: Comparing the predictive utility of the PCL:YV, YPI, and NEO PRI. Psychological Assessment, 21, 528-542.

11.

Clarck

Watson

(1995). Constructing validity: Basic issues in objective scale development. Psychological Assessment, 7, 309-319.

12.

Colins

Noom

Vanderplasschen

(2012). Youth Psychopathic Traits Inventory-Short Version: A further test of the internal consistency and criterion validity. Journal of Behavioral Assessment and Psychopathology, 34, 476-486. doi:10.1007/s10862-012-9299-0

13.

Colins

Vermeiren

(in press). The usefulness of DSM-IV and DSM-5 conduct disorder subtyping in detained adolescents. Journal of Nervous and Mental Disease.

14.

Colins

Vermeiren

De Bolle

Broekaert

(in press). Self-reported psychopathic-like traits as predictor of recidivism in detained male adolescents. Criminal Justice and Behavior.

15.

Colins

Vermeiren

Schuyten

Broekaert

(2009). Psychiatric disorders in property, violent, and versatile offending detained male adolescents. American Journal of Orthopsychiatry, 79(1), 31-38.

16.

Colins

Vermeiren

Schuyten

Broekaert

Soyez

(2008). Informant agreement in the assessment of disruptive behavior disorders in detained minors in Belgium: A diagnosis-level and symptom-level examination. Journal of Clinical Psychiatry, 69, 141-148.

17.

Colins

Vermeiren

Vahl

Markus

Broekaert

Doreleijers

(2011). Psychiatric disorder in detained male adolescents as risk factor for serious recidivism. Canadian Journal of Psychiatry, 56, 44-50.

18.

Colins

Vermeiren

Vreugdenhil

van den Brink

Doreleijers

Broekaert

(2010). Psychiatric disorders in detained and incarcerated male adolescents: A systematic review of the literature. Canadian Journal of Psychiatry, 55, 255-263.

19.

Cook

N. E.

Barese

T. H.

Dicataldo

(2010). The confluence of mental health and psychopathic traits in adolescent female offenders. Criminal Justice and Behavior, 37, 119-135.

20.

Cooke

D. J.

Michie

(2001). Refining the construct of psychopathy: Towards a hierarchical model. Psychological Assessment, 13, 171-188. doi:10.1037//1040-3590.13.2.171

21.

Cooke

D. J.

Michie

Skeem

(2007). Understanding the structure of the Psychopathy Checklist–Revised: An exploration of methodological confusion. British Journal of Psychiatry, 190, S39-S50. doi:10.1192/bjp.190.5.s39

22.

Dadd

M. R.

Fraser

Frost

Hawes

D. J.

(2005). Disentangling the underlying dimensions of psychopathy and conduct problems in childhood: A community study. Journal of Consulting and Clinical Psychology, 73, 400-410.

23.

Declercq

Markey

Vandist

Verhaeghe

(2009). The Youth Psychopathic Trait Inventory: Factor structure and antisocial behaviour in non-referred 12-17-year-olds. Journal of Forensic Psychiatry & Psychology, 20, 577-594.

24.

Dillard

C. L.

Salekin

R. T.

Barker

E. D.

Grimes

R. D.

(2012). Psychopathy in adolescent offenders: An item response theory study of the Antisocial Process Screening Device–Self Report and the Psychopathy Checklist: Youth Version. Personality Disorders: Theory, Research, and Treatment. Advance online publication. doi:10.1037/a0028439

25.

Douglas

K. S.

Epstein

Poythress

(2008). Criminal recidivism among juvenile offenders: Testing the incremental and predictive validity of three measures of psychopathic features. Law and Human Behavior, 32, 423-438.

26.

Ferdinand

van der Ende

(2002). Diagnostic interview schedule for children. Authorized Dutch translation of the NIMH DISC-IV (Shaffer D, Fisher P, Lucas C and the NIMH Editorial Board). Rotterdam, Netherlands: Sophias Children’s Hospital.

27.

Forth

Kosson

Hare

(2003). The Hare Psychopathy Checklist: Youth Version. North Tonawanda, NY: Multi-Health Systems.

28.

Frick

P. J.

Bodin

S. D.

Barry

C. T.

(2000). Psychopathic traits and conduct problems in community and clinic-referred samples of children: Further development of the psychopathy screening device. Psychological Assessment, 12, 382-393.

29.

Frick

P. J.

Hare

R. D.

(2001). Antisocial Process Screening Device technical manual. Toronto, Ontario, Canada: Multi-Health Systems.

30.

Frick

P. J.

Moffit

T. E.

(2010). A proposal to the DSM-V Childhood Disorders and the ADHD and Disruptive Behavior Disorders Work Groups to include a specifier to the diagnosis of conduct disorder based on the presence of callous-unemotional traits. Washington, DC: American Psychiatric Association.

31.

Frick

P. J.

White

S. F.

(2008). Research review: The importance of callous-unemotional traits for developmental models of aggressive and antisocial behavior. Journal of Child Psychology and Psychiatry, 49, 359-375.

32.

Hopwood

C. J.

Donnelan

B. M.

(2010). How should the internal structure of personality inventories be evaluated? Personality and Social Psychology Review, 14, 332-346.

33.

Bentler

P. M.

(1999). Cutoff criteria for fit indexes in covariance structure analysis: Conventional criteria versus new alternatives. Structural Equation Modeling, 6, 1-55.

34.

Johnstone

Cooke

D. J.

(2004). Psychopathic-like traits in childhood: Conceptual and measurement concerns. Behavioral Sciences & the Law, 22, 103-125.

35.

Infante

Slattery

M. J.

Klein

M. H.

Essex

M. J.

(2007). Association of internalizing disorders and allergies in a child and adolescent psychiatry clinical sample. Journal of Clinical Psychiatry, 68, 1419-1425.

36.

Jöreskog

Sörbom

(2006). LISREL for Windows [Computer software]. Lincolnwood, IL: Scientific Software International.

37.

Kosson

D. S.

Neumann

C. S.

Forth

A. E.

Salekin

R. T.

Hare

R. D.

Krischer

M. K.

Sevecke

(2012). Factor structure of the Hare Psychopathy Checklist: Youth Version (PCL:YV) in adolescent females. Psychological Assessment. Advance online publication. doi:10.1037/a002898

38.

Kotler

J. S.

McMahon

R. J.

(2005). Child psychopathy: Theories, measurement, and relations with the development and persistence of conduct problems. Clinical Child and Family Psychology Review, 8, 291-325.

39.

Kotler

J. S.

McMahon

(2010). Assessment of child and adolescent psychopathy. In Salekin

Lynam

(Eds.), Handbook of child and adolescent psychopathy (pp. 79-112). New York, NY: Guilford Press.

40.

Loney

B. R.

Frick

P. J.

Clements

C. B.

Ellis

M. L.

Kerlin

(2003). Callous-unemotional traits, impulsivity, and emotional processing in adolescents with antisocial behavior problems. Journal of Clinical Child and Adolescent Psychology, 32, 66-80. doi:10.1207/15374420360533077

41.

Lynam

D. R.

Miller

D. J.

Vachon

Loeber

Stouthamer-Loeber

(2009). Psychopathy in adolescence predicts official reports of offending in adulthood. Youth Violence and Juvenile Justice, 7, 189-207.

42.

Miller

D. J.

Watts

Jones

S. E.

(2011). Does psychopathy manifest divergent relations with components of its nomological network depending on gender? Personality and Individual Differences, 50, 264-269.

43.

Munoz

L. C.

Kerr

Besic

(2008). The peer relationships of youths with psychopathic personality traits: A matter of perspective. Criminal Justice and Behavior, 35, 212-227. doi:10.1177/0093854807310159

44.

Murrie

D. C.

Cornell

D. G.

(2002). Psychopathy screening of incarcerated juveniles: A comparison of measures. Psychological Assessment, 14, 390-396.

45.

Nunnally

Bernstein

(1994). Psychometric theory. New York, NY: McGraw-Hill.

46.

Odgers

C. L.

Reppucci

N. D.

Moretti

M. M.

(2005). Nipping psychopathy in the bud: An examination of the convergent, predictive, and theoretical utility of the PCL-YV among adolescent girls. Behavioral Sciences & the Law, 23, 743-763.

47.

Pechorro

Maroco

Poiares

Vieira

R. X.

(in press). Validation of the Portuguese version of the Antisocial Process Screening Device-Self-Report with a focus on delinquent behavior and behavior problems. International Journal of Offender Therapy and Comparative Criminology.

48.

Poythress

N. G.

Dembo

Wareham

Greenbaum

P. E.

(2006). Construct validity of the Youth Psychopathic Traits Inventory (YPI) and the Antisocial Process Screening Device (APSD) with justice-involved adolescents. Criminal Justice and Behavior, 33(1), 26-55. doi:10.1177/0093854808282518

49.

Poythress

N. G.

Douglas

K. S.

Falkenbach

Cruise

Lee

Murrie

D. C.

Vitacco

M. J.

(2006). Internal consistency reliability of the self-report antisocial process screening device. Assessment, 13, 107-113.

50.

Raine

Dodge

Loeber

Gatzke-Kopp

Lynam

D. R.

Reynolds

Liu

(2006). The Reactive–Proactive Aggression Questionnaire: Differential correlates of reactive and proactive aggression in adolescent boys. Aggressive Behavior, 32, 159-172.

51.

Roose

Bijttebier

Decoene

Claes

Frick

P. J.

(2010). Assessing the affective features of psychopathy in adolescence: A further validation of the inventory of callous and unemotional traits. Assessment, 17, 44-57.

52.

Salekin

R. T.

(2006). Psychopathy in children and adolescents. In Patrick

C. J.

(Ed.), Handbook of psychopathy (pp. 389-414). New York, NY: Guilford Press.

53.

Salekin

R. T.

(2008). Psychopathy and recidivism from mid-adolescence to young adulthood: Cumulating legal problems and limiting life opportunities. Journal of Abnormal Psychology, 117, 386-395.

54.

Salekin

R. T.

Neumann

C. S.

Leistico

A. M. R.

DiCicco

T. M.

Duros

R. L.

(2004). Psychopathy and comorbidity in a young offender sample: Taking a closer look at psychopathy’s potential importance over disruptive behavior disorders. Journal of Abnormal Psychology, 113, 416-427. doi:10.1037/0021-843x.113.3.416

55.

Salekin

R. T.

Debus

S.A.

Barker

E. D.

(2010). Adolescent psychopathy and the five factor model: Domain and facet analysis. Journal of Behavioral Assessment and Psychopathy, 32, 501-514.

56.

Salekin

R. T.

Lynam

D. R.

(2010). Handbook of child and adolescent psychopathy. New York, NY: Guilford Press.

57.

Schermelleh-Engel

Moosbrugger

Müller

(2003). Evaluating the goodness of fit of structural equation models: Tests of significance and descriptive goodness-of-fit measures. Methods of Psychological Research Online, 8, 23-74.

58.

Seagrave

Grisso

(2002). Adolescent development and the measurement of juvenile psychopathy. Law and Human Behavior, 26, 219-239.

59.

Seals

W. R.

Sharp

Michonski

J. D.

(2012). The relationship between the Youth Psychopathic Traits Inventory and psychopathology in a U.S. community sample of male youth. Journal of Personality Assessment, 94, 232-243.

60.

Sevecke

Kosson

D. S.

(2010). Relationships of child and adolescent psychopathy to other forms of psychopathology. In Salekin

R. T.

Lynam

D. R.

(Eds.), Handbook of child and adolescent psychopathy (pp. 284-317). New York, NY: Guilford Press.

61.

Sevecke

Lehmkuhl

Krischer

M. K.

(2009). Examining relations between psychopathology and psychopathy dimensions among adolescent female and male offenders. European Child & Adolescent Psychiatry, 18, 85-95. doi:10.1007/s00787-008-0707-7

62.

Shaffer

Fisher

Dulcan

M. K.

Davies

Piacentini

SchwabStone

M. E.

Regier

D. A.

(1996). The NIMH Diagnostic Interview Schedule for Children Version 2.3 (DISC-2.3): Description, acceptability, prevalence rates, and performance in the MECA study. Journal of the American Academy of Child and Adolescent Psychiatry, 35, 865-877.

63.

Skeem

J. L.

Cauffman

(2003). Views of the downward extension: Comparing the youth version of the psychopathy checklist with the youth psychopathic traits inventory. Behavioral Sciences & the Law, 21, 737-770. doi:10.1002/bsl.563

64.

van Baardewijk

Andershed

Stegge

Nilsson

K. W.

Scholte

Vermeiren

(2010). Development and tests of short versions of the Youth Psychopathic Traits Inventory and the Youth Psychopathic Traits Inventory-Child Version. European Journal of Psychological Assessment, 26, 122-128. doi:10.1027/1015-5759/a000017

65.

van der Laan

Blom

(2005). Jeugddelinquentie: Risico’s en bescherming: Bevindingen uit de WODC monitor zelfgerapporteerde jeugdcriminaliteit 2005 [Juvenile delinquency: Risks and protection: Findings from the WODC Monitor self-reported juvenile delinquency in 2005]. The Hague, Netherlands: Boom Juridische Uitgevers.

66.

Vaughn

M. G.

Newhill

C. E.

DeLisi

Beaver

K. M.

Howard

M. O.

(2008). An investigation of psychopathic features among delinquent girls: Violence, theft, and drug abuse. Youth Violence and Juvenile Justice, 6, 240-255.

67.

Veen

Stevens

Andershed

Raaijmakers

Q. A. W.

Doreleijers

Vollebergh

(2011). Cross-ethnic generalizability of the three-factor model of psychopathy: The Youth Psychopathic Traits Inventory in an incarcerated sample of native Dutch and Moroccan immigrant boys. International Journal of Law and Psychiatry, 34, 127-130.

68.

Vermulst

A. A.

Gerris

J. R. M.

(2006). Quick Big Five: Persoonlijkheidsvragenlijst. Een handleiding [Quick Big Five Personality questionnaire. Guideline]. Leeuwarden, Netherlands: LDC.

69.

Vincent

G. M.

Grisso

Terry

Banks

(2008). Sex and race differences in mental health symptoms in juvenile justice: The MAYSI-2 national meta-analysis. Journal of the American Academy of Child and Adolescent Psychiatry, 47, 282-290. doi:10.1097/chi.0b013e318160d516

70.

Vitacco

M. J.

Rogers

Neumann

C. S.

(2003). The antisocial process screening device: An examination of its construct and criterion-related validity. Assessment, 10, 143-150.