Debate Paper

Comparing the Dependability and Associations With Functioning of the DSM–5 Section III Trait Model of Personality Pathology and the DSM–5

Section II Personality Disorder Model

Michael Chmielewski Southern Methodist University

Camilo J. Ruggero University of North Texas

Roman Kotov Stony Brook University

Keke Liu University of North Texas

Robert F. Krueger University of Minnesota

Two competing models of personality psychopathology are included in the fifth edition of the Diagnostic Statistical Manual of Mental Disorders (DSM–5; American Psychiatric Association, 2013); the tradi- tional personality disorder (PD) model included in Section II and an alternative trait-based model included in Section III. Numerous studies have examined the validity of the alternative trait model and its official assessment instrument, the Personality Inventory for DSM–5 (PID-5; Krueger, Derringer, Markon, Watson, & Skodol, 2012). However, few studies have directly compared the trait-based model to the traditional PD model empirically in the same dataset. Moreover, to our knowledge, only a single study (Suzuki, Griffin, & Samuel, 2015) has examined the dependability of the PID-5, which is an essential component of construct validity for traits (Chmielewski & Watson, 2009; McCrae, Kurtz, Yamagata, & Terracciano, 2011). The current study directly compared the dependability of the DSM-5 traits, as assessed by the PID-5, and the traditional PD model, as assessed by the Personality Diagnostic Questionnaire-4 (PDQ-4�), in a large undergraduate sample. In addition, it evaluated and compared their associations with functioning, another essential component of personality pathology. In general, our findings indicate that most DSM–5 traits demonstrate high levels of dependability that are superior to the traditional PD model; however, some of the constructs assessed by the PID-5 may be more state like. The models were roughly equivalent in terms of their associations with functioning. The current results provide additional support for the validity of PID-5 and the DSM–5 Section III personality pathology model.

Keywords: dependability, PID-5, functioning, personality disorders, DSM–5 Section III

The fifth edition the Diagnostic Statistical Manual of Mental Disorders (DSM–5; American Psychiatric Association [APA], 2013) includes two competing models of personality pathology: the traditional categorical personality disorder (PD) model from DSM–IV and an alternative trait-based model in Section III. Prob-

lems with the traditional PD model have been extensively re- viewed (Clark, 2007; Widiger & Samuel, 2005; Widiger & Trull, 2007). They include extreme heterogeneity (Chmielewski & Wat- son, 2008; Johansen, Karterud, Pedersen, Gude, & Falkum, 2004), high rates of diagnostic comorbidity (Oldham et al., 1992), arbi- trary boundaries with normality (Widiger & Samuel, 2005), low interrater reliability (Tyrer et al., 2007), poor convergent/discrim- inant validity (Clark, Livesley, & Morey, 1997), excessive not otherwise specified diagnosis (Verheul & Widiger, 2004), and low diagnostic stability (Shea et al., 2002; Skodol et al., 2005).

Considerable research has been conducted on the DSM–5 alter- native model and the official assessment instrument for the trait aspect of the model, the Personality Inventory for DSM–5 (PID-5; Krueger, Derringer, Markon, Watson, & Skodol, 2012). Numerous studies have demonstrated the structural validity of the PID-5 (De Fruyt et al., 2013; Krueger et al., 2012; Wright et al., 2012; Zimmermann et al., 2014). Moreover, the DSM–5 traits capture the variance within the traditional PD model (Hopwood, Thomas, Markon, Wright, & Krueger, 2012; Miller, Few, Lynam, & MacK-

This article was published Online First September 12, 2016. Michael Chmielewski, Department of Psychology, Southern Methodist

University; Camilo J. Ruggero, Department of Psychology, University of North Texas; Roman Kotov, Department of Psychiatry and Behavioral Sciences, Stony Brook University; Keke Liu, Department of Psychology, University of North Texas; Robert F. Krueger, Department of Psychology, University of Minnesota.

Robert Krueger has served as a paid consultant to preValio LLC, developers of psychological reports based on the Personality Inventory for DSM-5 (PID-5).

Correspondence concerning this article should be addressed to Michael Chmielewski, Department of Psychology, Southern Methodist University, PO Box 75275-0442, Dallas, TX 75275. E-mail: [email protected]

T hi

s do

cu m

en t

is co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

ti on

or on

e of

it s

al li

ed pu

bl is

he rs

. T

hi s

ar ti

cl e

is in

te nd

ed so

le ly

fo r

th e

pe rs

on al

us e

of th

e in

di vi

du al

us er

an d

is no

t to

be di

ss em

in at

ed br

oa dl

y.

Personality Disorders: Theory, Research, and Treatment © 2016 American Psychological Association 2017, Vol. 8, No. 3, 228 –236 1949-2715/17/$12.00 http://dx.doi.org/10.1037/per0000213

228

illop, 2015; Samuel, Hopwood, Krueger, Thomas, & Ruggero, 2013), are strongly associated with other models of personality and personality pathology (Anderson et al., 2012; Ashton, Lee, de Vries, Hendrickse, & Born, 2012; De Fruyt et al., 2013; Gore & Widiger, 2013; Hopwood et al., 2012; Miller et al., 2015; Quilty, Ayearst, Chmielewski, Pollock, & Bagby, 2013; Samuel et al., 2013; Suzuki, Griffin, & Samuel, 2015; Thomas et al., 2012; Zimmermann et al., 2014), and have meaningful ties to other clinical constructs (Hopwood et al., 2013; Zimmermann et al., 2014).

The accumulated research provides support for the validity of the DSM–5 trait model and the PID-5; nevertheless, important gaps remain regarding their construct validity. First, few data exist regarding the dependability of the DSM–5 trait model and the PID-5. Second, only a handful of studies have examined the association between DSM–5 traits and functioning. Third, very few studies have directly compared the DSM–5 traits to the traditional PDs by simultaneously examining the performance of both models in the same dataset.

Dependability, Transient Error, and Stability

Cattell, Eber, and Tatsuoka (1970, p. 30) defined dependability “as the correlation between two administrations of the same test when the lapse of time is insufficient for people themselves to change.” Dependability correlations are the primary method of assessing transient measurement errors, which are systematic mea- surement errors caused by fluctuations in participants’ psycholog- ical state on any particular day (Cattell et al., 1970; Chmielewski & Watson, 2009; Gnambs, 2014; McCrae, Kurtz, Yamagata, & Terracciano, 2011; Schmidt, Le, & Ilies, 2003; Watson, 2004). Because transient errors produce consistent responses during the same assessment session but inconsistent responses across differ- ent assessment sessions they cannot be detected using indices of reliability computed from a single administration (e.g., internal consistency, Cronbach’s �). Importantly, transient errors can mas- querade as true trait change, making it difficult to determine the stability of the construct that a scale assesses (see Anusic, Lucas, & Donnellan, 2012; Chmielewski & Watson, 2009; Gnambs, 2014; McCrae et al., 2011).

As noted in the DSM–5, personality pathology must have “an enduring pattern” and be “stable over time” (APA, 2013, p. 645). One of the many limitations of the traditional DSM PD model is that PDs demonstrate, at best, modest stability (Grilo et al., 2004; Lenzenweger, 1999; McGlashan et al., 2005; Shea et al., 2002; Zanarini, Frankenburg, Reich, & Fitzmaurice, 2012). To our knowledge, only a single study (Wright et al., 2015) has examined the stability of DSM–5 traits. They reported a mean PID-5 domain stability of r � .73 (range: r � .62 to .75) and a median facet stability of r � .68 (range: r � .41 to.78) over a 1.4-year interval in 93 outpatients. Wright et al. (2015) concluded that the DSM–5 traits were “highly stable over the course of the study” (p. 202). However, this broad statement overshadows substantial differ- ences in stability (r � .41 to .78) among the various PID-5 traits. Because a measure’s dependability sets an upper limit on its stability, it is possible that the observed differences are due to differential levels of transient error (i.e., differential dependability) across the PID-5 scales. Moreover, it is possible that transient error is responsible for the poor stability of the traditional DSM PD model. In fact, Zimmerman (1994) concluded that state effects

(i.e., transient error) substantially influence assessments of the traditional PD model.

Transient errors can also distort associations with other con- structs, result in failures to replicate research, and substantially alter study outcomes (Chmielewski, Sala, Tang, & Baldwin, 2016; Chmielewski & Watson, 2009). Moreover, they may lead to an overinclusion of “false positives” and exclusion of “false nega- tives” in clinical samples (Chmielewski & Watson, 2009). Despite the potential influence of transient error, to our knowledge only a single study has examined the dependability of the PID-5. Suzuki et al. (2015) reported a mean domain dependability of .83 (range � .81 to .83) and a mean facet dependability of .78 (range � .66 to .86), which was similar to values for the Revised NEO Personality Inventory (NEO-PI-R) in the sample. Although these findings pro- vide initial evidence for the dependability of the PID-5, some of the facet dependabilities were low, suggesting that transient error may be a concern. Moreover, their sample (N � 266) was below recom- mended cutoffs (N � 300) for examining dependability (Watson, 2004). Given the importance of transient error, replication of these findings in a larger sample is necessary.

Personality Pathology and Functioning

The traditional PD model and the alternative model (Criterion A) both include functional impairment (American Psychiatric As- sociation, 2013). The traditional PD model is associated with functioning across various domains, including work, social rela- tionships, and leisure (Miller, Campbell, & Pilkonis, 2007; Skodol et al., 2005). To our knowledge, only three studies have examined associations between the DSM–5 traits and functioning. Wright et al. (2015) reported that the DSM–5 traits predicted psychosocial functioning 1.4 years later in a patient sample. Ackerman and Corretti (2015) found that higher levels of detachment in college students lead to their roommates feeling less close to them. Keeley, Flanagan, and McCluskey (2014) demonstrated that the DSM–5 traits concurrently explained variance in several functioning do- mains in student and patient samples.

Results from these studies have generally reflected impairments (e.g., interpersonal relationships) aligned with Criterion A of the alternative model. However, several unexpected findings emerged, including significant associations with impairments in mobility and self-care (Keeley et al., 2014). Moreover, Risk Taking, Atten- tion Seeking, and Manipulativeness were not associated with func- tioning in the student sample (Keeley et al., 2014). Given the importance of functioning for personality pathology, replication of these counterintuitive findings is necessary. Finally, is worth not- ing that Keeley et al. (2014) hypothesized transient errors could have influenced their results and called for further research regard- ing this issue.

Current Study

The current study was designed to provide evidence regarding the construct validity of the DSM–5 traits and the PID-5. First, we examined the dependability of the PID-5 in a sample large enough (i.e., minimum N � 300; Watson, 2004) to provide precise de- pendability estimates. Second, we examined the concurrent asso- ciations of the DSM–5 traits with functional impairment across a wide range of domains. Finally, we directly compared the DSM–5

T hi

s do

cu m

en t

is co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

ti on

or on

e of

it s

al li

ed pu

bl is

he rs

. T

hi s

ar ti

cl e

is in

te nd

ed so

le ly

fo r

th e

pe rs

on al

us e

of th

e in

di vi

du al

us er

an d

is no

t to

be di

ss em

in at

ed br

oa dl

y.

229DSM–5 TRAITS AND TRADITIONAL PD MODEL

trait (PID-5) and traditional PD (PDQ-4�) models empirically in the same sample. Although the alternative model was created to address limitations inherent in the traditional PD model, only two studies have directly tested these models against each other in the same sample. They found that the models had similar associations with psychopathology and normal personality (Fossati et al., 2015; Miller et al., 2015). As Chmielewski and Watson (2009, p. 199) note, “consideration of transient error could help determine which PD models are more valid and reliable” making the current anal- yses very relevant in comparing the two models. Finally, we categorically and dimensionally examined traditional PDs because (a) dimensional representations are more valid and reliable than categorical ones (Markon, Chmielewski, & Miller, 2011) and (b) dimensional representations have been suggested as an alternative model of traditional PDs (Oldham & Skodol, 2000).

Method

Participants and Procedures

Participants were undergraduate students (N � 572; 68.3% female, 71.8% Caucasian, 12.4% African American, 18.3% His- panic, mean age � 21.9 years) who completed all measures online. Approximately 2 weeks later all participants were invited to com- plete the PID-5 a second time, 382 of whom did, allowing for precise dependability estimates (see Watson, 2004). The Person- ality Diagnostic Questionnaire-4 (PDQ-4�) was later added to the second assessment, resulting in a subset of participants (N � 202 of the 382) who also completed it twice. Participants who com- pleted both assessments scored higher on PID-5 Separation Inse- curity and lower on Callousness, Manipulativeness, Risk Taking, and Irresponsibility than those completing one assessment. Those who completed both assessments also scored lower on the PDQ-4� Antisocial symptoms and World Health Organization Disability Assessment Schedule 2.0 (WHODAS) Life Activities.

Measures

PID-5. The PID-5 (Krueger et al., 2012) is the official mea- sure of the DSM–5 dimensional trait-based model of personality pathology. It includes 25 lower order traits and 5 higher order domains. Facet scores were calculated as item means; domain scores were calculated based on APA guidelines in which each domain score is the mean of three facet scales. Several studies previously reviewed have documented the validity of the PID-5.

PDQ-4�. The PDQ-4� (Hyler, 1994) maps directly onto the traditional DSM PD model and is one of the most widely used measures of the model (Widiger & Coker, 2002). The PDQ-4� demonstrates high sensitivity and moderate specificity (Davison, Leese, & Taylor, 2001). The PDQ-4� was examined categorically and dimensionally (i.e., by summing symptoms for each PD).

WHODAS. The WHODAS was completed during the first as- sessment. It assesses functioning across six domains: Cognition, Mo- bility, Self-Care, Getting Along (i.e., interpersonal relationships), Life Activities, and Participation in Society.1 Internal consistency esti- mates range from .94 to .96 and dependability estimates range from .93 to .96 (Üstün et al., 2010). The WHODAS is also included in Section III of the DSM–5 as a replacement for the Global Assessment

of Functioning scale and is the same measure used by Keeley et al. (2014), allowing for a direct comparison with their findings.

Results

Cronbach’s �s for the PID-5 (see Table 1) were at least adequate (e.g., �.70; Nunnally, 1978), with 88% being considered good (e.g., �.80; Clark & Watson, 1995). Average interitem correla- tions (AICs) for most PID-5 scales were generally within recom- mendations, suggesting that the scales assess homogenous con- structs (see Clark & Watson, 1995). However, some scales demonstrated higher than optimal AICs, especially considering the breadth of the constructs they assess. This was particularly true for Eccentricity, which demonstrated a very high AIC at both assess- ments (i.e., .65 and .72), suggesting that it contains items that could be considered redundant in this sample. In contrast, none of the �s for the PDQ-4� dimensional symptom counts were above .80. Moreover, 85% were below .70, and the AICs were quite low

1 The Mobility and Self-Care domains were added as the study was in progress and therefore completed by a subset (N � 275) of the sample.

Table 1 PID Internal Consistency and Dependability

Scale

Time 1 Time 2

Dependability� AIC � AIC

Domains Negative Affectivity .94 .41 .95 .45 .91 Detachment .93 .36 .93 .36 .89 Psychoticism .95 .37 .96 .42 .88 Antagonism .91 .33 .92 .35 .86 Disinhibition .92 .34 .93 .38 .86 PID-5 Domain Mean .93 .36 .94 .39 .88

Facets Withdrawal .91 .49 .92 .54 .89 Depressivity .93 .49 .94 .52 .88 Eccentricity .96 .65 .97 .72 .88 Emotional Lability .89 .53 .92 .61 .88 Anxiousness .91 .51 .91 .53 .88 Separation Insecurity .86 .47 .88 .52 .86 Hostility .86 .38 .88 .43 .85 Risk Taking .87 .33 .88 .34 .85 Anhedonia .87 .45 .89 .51 .84 Restricted Affect .83 .40 .85 .45 .84 Deceitfulness .87 .41 .89 .44 .84 Intimacy Avoidance .83 .44 .83 .44 .84 Callousness .86 .31 .88 .33 .84 Distractibility .91 .52 .92 .56 .84 Attention Seeking .89 .50 .91 .55 .84 Manipulativeness .83 .50 .87 .57 .83 Rigid Perfectionism .91 .51 .92 .53 .83 Perceptual Dysregulation .84 .30 .85 .32 .80 Unusual Beliefs .77 .29 .84 .39 .80 Irresponsibility .74 .29 .77 .32 .80 Impulsivity .86 .51 .86 .50 .79 Perseveration .87 .43 .89 .48 .78 Submissiveness .80 .51 .85 .59 .77 Suspiciousness .71 .26 .72 .26 .76 Grandiosity .76 .34 .79 .38 .76 PID facet mean .85 .44 .87 .47 .83

Note. N T � 572, T2 � 382. Italic values represents mean of the scales.

T hi

s do

cu m

en t

is co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

ti on

or on

e of

it s

al li

ed pu

bl is

he rs

. T

hi s

ar ti

cl e

is in

te nd

ed so

le ly

fo r

th e

pe rs

on al

us e

of th

e in

di vi

du al

us er

an d

is no

t to

be di

ss em

in at

ed br

oa dl

y.

230 CHMIELEWSKI, RUGGERO, KOTOV, LIU, AND KRUEGER

(see Table 2). These results reflect the heterogeneity of symptoms included within the traditional PD constructs.

Dependability

It is highly unlikely that true changes in personality pathology would have occurred over the short 2-week dependability interval; therefore, any dependability coefficients (see Table 1) below 1.0 indicate the presence of measurement error. The PID-5 domains demonstrated high levels of dependability that approach 1.0 (e.g., r � .86 to .91, mean domain r � .88).2 However, there was considerable variability across the PID-5 facets (mean dependabil- ity; r � .83), with some approaching 1.0 (e.g., Withdrawal; r � .89) and others being appreciably lower (e.g., Grandiosity and Suspiciousness; r � .76).

In sharp contrast, the dependability of the PDQ-4� symptom count scales (see Table 2) were well below 1.0, ranging from r � .68 (Narcissistic) to r � .82 (Avoidant) with a mean of only r � .75. This low level of dependability indicates a substantial amount of measurement error. The PDQ-4� categorical PDs demonstrated even poorer dependability (� � .36 to .72; mean � � .57), demonstrating that a substantial portion of their variance is due to measurement error. These results are in line with past evidence indicating that categorical representations of psychopathology are less reliable and valid than dimensional ones (Markon et al., 2011). Unexpectedly, strong correlations emerged between dependability and Cronbach’s � (PID-5: domain mean r � .62, facet mean r � .77; PDQ-4�: mean r � .82).

Next, we conducted significance tests (Pearson Filon for non- overlapping variables from the same sample) comparing the PID-5 and PDQ-4� symptom count scales, which were more dependable than the PDQ-4� categories, in the subsample (N � 202) that completed both measures twice. The PID-5 domains were signif- icantly more dependable than the PDQ-4� scales in 94% of the comparisons and the facets were significantly more dependable in 63% of comparisons. The PDQ-4� scales were never more de- pendable than the PID-5 scales.

Associations With Functioning

Because Criterion A of the trait model includes dysfunction in identity, self-direction, empathy, and intimacy, we would expect stronger associations with certain WHODAS domains (e.g., get- ting along with others) and weaker associations with others (e.g., self-care, mobility). Nearly every PID-5 scale was significantly associated with poorer functioning across every WHODAS do- main (see Table 3).3 Many of these associations were medium in size; however, certain traits (e.g., Grandiosity and Manipulative- ness) appear to be less strongly associated with functioning than others. A notable exception was that Risk Taking was not signif- icantly associated with any WHODAS domain, which replicates previous counterintuitive findings (Keeley et al., 2014). In this regard, Crego and Widiger (2014) suggested that reverse-keyed items underperform when assessing psychopathy-related traits. Given the high percentage of reverse-keyed items in the Risk Taking scale (43%), we created two risk-taking composites (one using reverse-keyed items and one using nonreverse-keyed items). These two composites correlated r � .50 with each other; how- ever, the nonreverse-keyed composite demonstrated small positive associations (mean r � .13) and the reverse-keyed composite demonstrated small negative associations (mean r � �.15) with the WHODAS. Next, we conducted separate multiple regression analyses for the PID-5 domains and facets to determine the amount of variance each explained in the WHODAS. The weakest PID-5 associations were with mobility (domain R2 � .12, facet R2 � .17) and self-care (domain R2 � .21, facet R2 � .28), domains hypo- thetically less related to Criterion A, suggesting that the PID-5 demonstrates evidence of specificity. The PDQ-4� symptom count scales (see Table 4) demonstrated significant associations (R2 � .21 to .47) with all WHODAS functioning domains. How- ever, the PDQ-4� scales were more broadly associated with functioning in that the PDQ-4� associations with Mobility (R2 � .22) and Self-Care (R2 � .28) were not as differentiated from associations with the other WHODAS domains.

The incremental ability of each personality pathology measure to predict concurrent functioning was examined using hierarchical regressions analyses (see Table 5). In Step 1, the PDQ-4� symp- tom count scores were entered, then in Step 2 the PID-5 domains (or facets) were entered. The order of entry was then reversed. The PID-5 domains and facets added incremental prediction beyond the PDQ-4� for all areas of functioning except Mobility and Self-Care. For most areas of functioning the additional variance was modest; however, it was more substantial for life activities (domain � 6.6%, facet � 14.8%), cognition (domain � 8.9%, facet � 15.7%), and overall functioning (domain � 3.2%, facet � 11.1%). Likewise, the PDQ-4� symptom counts added incremen- tal validity beyond the PID-5 domains (�4 –13%) and facets (�4 –11%) for all functioning domains, with the greatest incre- mental validity (10 –12% additional variance) for Mobility and Self-Care. Moreover, the PDQ-4� added approximately 10% pre-

2 The dependability the PID-5 Brief Form, scored from the full PID-5, ranged from .78 to .83 with a mean of .80.

3 Associations between the PID-5 Brief Form and functioning were very similar.

Table 2 PDQ-4 Internal Consistency and Dependability

Time 1 Time 2 Dependability

Scale � AIC � AIC Sym. Cat.

Avoidant .77 .32 .78 .34 .82 .63 Borderline .67 .18 .65 .17 .81 .72 Paranoid .65 .21 .70 .25 .78 .63 Schizotypal .62 .15 .68 .19 .78 .44 Antisocial .56 .14 .50 .11 .74 .70 Dependent .68 .21 .66 .20 .73 .52 Schizoid .54 .14 .51 .13 .73 .62 Histrionic .56 .14 .59 .15 .72 .56 Obs. Comp. .47 .10 .50 .11 .69 .53 Narcissistic .55 .12 .52 .11 .68 .36 Mean .61 .17 .61 .18 .75 .57

Note. T1 N � 572, T2 N � 202. Sym. � symptom count; Cat. � categorical scoring; Obs. Comp. � Obsessive Compulsive. Dependability statistic for categorical scoring is �. We also computed dependability for the categorical scoring as intraclass correlations, and results were essen- tially identical. Italic values represents mean of the scales.

T hi

s do

cu m

en t

is co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

ti on

or on

e of

it s

al li

ed pu

bl is

he rs

. T

hi s

ar ti

cl e

is in

te nd

ed so

le ly

fo r

th e

pe rs

on al

us e

of th

e in

di vi

du al

us er

an d

is no

t to

be di

ss em

in at

ed br

oa dl

y.

231DSM–5 TRAITS AND TRADITIONAL PD MODEL

dictive utility for overall functioning, regardless of if the PID-5 domains or facets were included in the model.

Discussion

The current study adds to the evidence supporting the DSM–5 alternative model of personality pathology and the PID-5. To our knowledge, it is only the second study to examine the dependabil- ity of the PID-5 and the first to use a sample large enough to meet recommendations (Watson, 2004). It is also one of a few studies examining the association between DSM–5 pathological personal- ity traits and functioning. Finally, only two other studies (Fossati et al., 2015; Miller et al., 2015) directly tested the alternative model against the traditional PD model in the same dataset, neither of which addressed the current issues.

Dependability

High levels of dependability are essential for the construct validity of trait measures and models (Chmielewski & Watson, 2009; Gnambs, 2014; McCrae et al., 2011; Watson, 2004). Al- though measurement error is present to some degree in all PID-5 scales, the dependability of the PID-5 domains approached 1.0 (mean domain dependability r � .88), indicating relatively low levels of transient error. However, there was considerable variabil- ity among the PID-5 facet scales (mean r � .83), with depend- abilities ranging from high (e.g., r � .89) to relatively low (e.g., r � .76). The current dependabilities appear higher than those reported by Suzuki et al. (2015; domain M � .83, range � .81 to .83; facet M � .78, range � .66 to .86). Despite the differences in magnitude, the correlation of dependability coefficients across the

Table 3 PID-5 Relationship to WHODAS Functioning

Scale Cognition Mobility Self-Care Get Along Life Act Society Overall

NA .49 .26 .36 .42 .39 .49 .52 Anxiousness� .45 .21 .29 .38 .37 .43 .48 Depressivity .48 .33 .40 .50 .42 .45 .56 Emot. Labil.� .44 .29 .36 .36 .34 .43 .51 Perseveration .51 .22 .23 .39 .39 .43 .46 Rigid Perfect. .20 .11 .07 .20 .15 .27 .24 Sep. Insecur.� .35 .16 .26 .33 .27 .37 .34 Submissive. .32 .14 .18 .27 .28 .29 .28 Suspicious. .41 .25 .26 .42 .32 .40 .42

Detachment .40 .20 .21 .50 .33 .35 .42 Anhedonia� .39 .22 .23 .46 .39 .38 .43 Intim. Av.� .23 .14 .12 .29 .17 .16 .28 Restrict Aff .15 .05 .02 .21 .13 .13 .14 Withdrawal� .34 .15 .18 .46 .25 .31 .35

Antagonism .18 .15 .15 .17 .18 .22 .20 Att. Seeking .18 .10 .13 .06 .16 .15 .13 Callousness .16 .15 .17 .26 .14 .18 .20 Deceit.� .27 .20 .19 .24 .20 .27 .24 Hostility .31 .24 .26 .34 .24 .36 .38 Grandiosity� .10 .11 .08 .08 .08 .11 .15 Manipulat.� .10 .09 .11 .11 .16 .18 .12

Disinhibition .51 .32 .36 .37 .42 .38 .53 Distract� .55 .27 .30 .34 .46 .34 .51 Impulsivity� .29 .21 .27 .21 .23 .24 .32 Irresponsib.� .43 .36 .38 .38 .36 .36 .52 Risk Taking �.01 .02 .06 �.04 �.02 �.02 .02

Psychoticism .47 .27 .29 .40 .37 .41 .50 Eccentricity� .37 .18 .20 .36 .31 .32 .41 Perc. Dys.� .52 .32 .37 .41 .36 .44 .55 Unsl. Beliefs� .37 .23 .23 .30 .31 .35 .39

R2

Domains .37 .12 .17 .33 .24 .28 .39 Facets .44 .21 .28 .37 .32 .34 .49 Brief Domains .31 .10 .16 .27 .18 .25 .33

Adjusted R2

Domains .37 .10 .16 .32 .23 .28 .38 Facets .42 .14 .20 .34 .29 .31 .43 Brief Domains .31 .08 .14 .26 .17 .24 .32

Note. N � 561–570 for Cognition, Getting Along with Others, Engaging in Life Activities, and Participation in Society; N � 265–275 for Mobility, Self-Care, and Overall Functioning. Underline � p .05. bold � p .01. � � scales scored in the PID-5 domains. Emot. Labil. � Emotional Lability; Rigid Perfect. � Rigid Perfectionism; NA � Negative Affect; Schizd � Schizoid; STPD � Schizotypal PD; Antisoc � Antisocial; Sep. Ins. � Separation Insecurity; Submissive. � Submissiveness; Suspicious. � Suspiciousness; Intim. Av. � Intimacy Avoidance; Restrict Aff. � Restricted Affect; Att. Seeking � Attention Seeking; Deceit. � Deceitfulness; Manipulat. � Manipulative- ness; Irresponsib. � Irresponsibility; Perc. Dys. � Perceptual Dysregulation; Unsl. Belfs. � Unusual Beliefs; Get Along � Getting Along; Life Act � Life Activities; Society � Participation in Society.

T hi

s do

cu m

en t

is co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

ti on

or on

e of

it s

al li

ed pu

bl is

he rs

. T

hi s

ar ti

cl e

is in

te nd

ed so

le ly

fo r

th e

pe rs

on al

us e

of th

e in

di vi

du al

us er

an d

is no

t to

be di

ss em

in at

ed br

oa dl

y.

232 CHMIELEWSKI, RUGGERO, KOTOV, LIU, AND KRUEGER

two studies was high (r � .66), indicating that the relative depend- ability of these PID-5 scales was similar across the studies. In other words, some PID-5 scales are consistently more dependable than others are.

Overall, the dependability results from the current study and those from Suzuki et al. (2015) compare favorably to dependability estimates for “normal” personality traits in similar samples (Chmielewski et al., 2016; Chmielewski & Watson, 2009; Suzuki et al., 2015). Nevertheless, it should be noted that dependability estimates in the .75 range represent a high level of measurement error (e.g., �25% error variance) and that even fairly low levels of transient error can have substantial effects on study outcomes (see Chmielewski & Watson, 2009). Therefore, we echo previous calls

in the literature for further research into the causes of transient error. This is especially important because previous explanations for transient error, such as item formats, response formats, or instruction sets (see Chmielewski & Watson, 2009; Watson, 2004), cannot explain the differences in dependability across PID-5 scales.

The current results suggest that some of the constructs assessed by the PID-5 may, instead of representing traits, be best concep- tualized as more transient and state like. This interpretation is in line with past suggestions that personality pathology may subsume both traits and acute symptoms (Clark, 2007; McGlashan et al., 2005; Oldham & Skodol, 2000). More recently, Wright et al. (2015) raised this as a possible explanation for the differential

Table 4 PDQ-4 Symptom Count Relationship to Functioning

Scale Cognition Mobility Self-Care Get Along Life Act Society Overall

Paranoid .30 .25 .26 .33 .31 .36 .39 Schizoid .29 .30 .29 .39 .22 .27 .40 Schizotypal .43 .38 .34 .45 .36 .45 .52 Antisocial .21 .18 .31 .19 .20 .28 .29 Borderline .46 .33 .42 .46 .36 .50 .58 Histrionic .32 .34 .36 .23 .26 .32 .40 Narcissistic .25 .20 .19 .17 .24 .24 .27 Avoidant .43 .29 .35 .53 .35 .42 .53 Dependent .46 .31 .38 .43 .38 .42 .53 Obs. Comp .28 .17 .19 .25 .22 .31 .32 R2 .32 .22 .28 .38 .21 .34 .47 Adjusted R2 .31 .19 .25 .37 .20 .32 .45

Note. N � 561–570 for Cognition, Getting Along with Others, Engaging in Life Activities, and Participation in Society; N � 265–275 for Mobility, Self-Care, and Overall Functioning. Underline � p .05. bold � p .01. Obs. Comp � Obsessive Compulsive; Get Along � Getting Along; Life Act � Life Activities; Society � Participation in Society.

Table 5 Hierarchical Regression Analyses Predicting Functioning

PID-5 incremental validity PDQ-4 incremental validity

WHODAS R2 F p R2 F p

Domain level Cognition .089 16.86 .000 .041 3.90 .000 Mobility .022 1.50 .191 .126 4.35 .000 Self-Care .011 .83 .533 .114 4.13 .000 Getting Along .026 4.78 .000 .079 7.33 .000 Life Activities .066 9.94 .000 .038 2.85 .002 Society .023 3.99 .001 .075 6.49 .000 Overall .032 3.18 .008 .109 5.40 .000

Facet level Cognition .157 6.48 .000 .037 3.86 .000 Mobility .100 1.43 .091 .109 3.86 .000 Self-Care .100 1.52 .058 .100 3.82 .000 Getting Along .048 1.80 .011 .058 5.35 .000 Life Activities .148 4.86 .000 .043 3.51 .000 Society .057 1.99 .003 .053 4.66 .000 Overall .111 2.42 .000 .090 4.91 .000

Note. N � 561–570 for Cognition, Getting Along with Others, Engaging in Life Activities, and Participation in Society; N � 265–275 for Mobility, Self-Care, and Overall Functioning; Significant estimates are bolded. Columns 2, 3, and 4 present results of the model where the PDQ-4� was entered in the first block and then the PID-5 (domains or facets) in the second block; Columns 5, 6, and 7 present results of the model where the PID-5 (domains or facets) was entered in the first block and then the PDQ-4� entered in the second block. Society � Participation in Society.

T hi

s do

cu m

en t

is co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

ti on

or on

e of

it s

al li

ed pu

bl is

he rs

. T

hi s

ar ti

cl e

is in

te nd

ed so

le ly

fo r

th e

pe rs

on al

us e

of th

e in

di vi

du al

us er

an d

is no

t to

be di

ss em

in at

ed br

oa dl

y.

233DSM–5 TRAITS AND TRADITIONAL PD MODEL

1.4-year stability of the PID-5 scales, arguing that it was essential for future research to address this issue. The current research, combined with that of Suzuki et al. (2015), provides further evidence for this possibility. Nevertheless, it is worth noting that there was a significant association between the number of items a facet contained and its dependability (r � .49), suggesting that scale length may have influenced the dependability of the PID-5 facets. As such, additional research is necessary to determine which constructs included within the PID-5 are best conceptual- ized as acute states.

Our dependability results suggest that the PID-5 outperforms the PDQ-4� regardless of whether the latter was categorically or dimensionally assessed. Because dependability analyses provide a metric for comparing models of personality pathology (Chmielewski & Watson, 2009), the current results empirically demonstrate the superiority of the alternative DSM–5 model over the traditional PD model, at least as assessed by these commonly used measures. It is worth noting that the PID-5 Brief Form, which contains only 25 items (compared with 99 items in the PDQ-4�), demonstrated as strong if not stronger dependabilities than the PDQ-4�, indicating that the PID-5’s higher dependability is not solely due to it having more items than the PDQ-4�. Our results also highlight that the categorical nature of traditional DSM PDs makes them far more susceptible to transient error, suggesting that transient error is at least partially responsible for the poor stability of the traditional categorical PD model (Chmielewski & Watson, 2009; Clark et al., 1997; Skodol et al., 2005; Zimmerman, 1994). Taken together, these results add to the large body of research indicating that dimensional representations of psychopathology are superior to categorical ones (Markon et al., 2011). It is worth noting that interview assessments of PDs have lower stability than do self-report measures (Samuel et al., 2011). Moreover, past research has documented poor dependability for interviewer- assessed mood and anxiety disorders (Chmielewski, Clark, Bagby, & Watson, 2015), suggesting that dependability of interviewer- based PD assessments would likely be lower than self-report PD assessments.

Functioning

The PID-5 scales demonstrated broad associations with multiple functioning domains. In line with previous research (Keeley et al., 2014), the DSM–5 traits appear to be more strongly associated with functioning domains conceptually aligned with personality pathol- ogy (e.g., getting along with others) than more distal domains (e.g., mobility, self-care). With one notable exception, the current asso- ciations were very similar to those reported by Keeley et al. (2014); the current results align very strongly with results from their student sample for five of the WHODAS domains (mean cross study correlation of the PID-5 associations with functioning was r � .83, range r � .71 to .91). However, a major discrepancy emerged for Participation in Society; in the current study nearly every PID-5 scale was significantly associated with Participation in Society (domain mean r � .28, facet mean r � .34), whereas no PID-5 scales were in the Keeley et al. (2014) student sample (domain mean r � .00, facet mean r � .01). We note that significant correlations did emerge in their clinical sample. Finally, we replicated Keeley et al.’s (2014) finding of no association between PID-5 Risk Taking and functioning. It is interesting to

note that Wright et al. (2015) also found no association between Risk Taking and functioning in their longitudinal study. Keeley et al. (2014) suggested that the self-report nature of the PID-5 might have led to this unexpected finding, and we would add that the high number of reverse-keyed items in this scale may have influ- enced the results (Crego & Widiger, 2014)

Despite demonstrating lower dependability, the PDQ-4� was associated with functioning at least as strongly as the PID-5 was. Although this finding may seem counterintuitive, it is in line with previous studies demonstrating that the traditional PDs and the alternative model have similar associations with psychopathology and normal personality (Fossati et al., 2015; Miller et al., 2015). This equivalence may be due, in part, to the fact that the models share a considerable amount of variance and cover the same basic content organized in a different manner (Hopwood et al., 2012; Miller et al., 2015; Samuel et al., 2013). In addition, the PID-5 specifically assesses the trait component of the model (Criterion B) and not the dysfunction component (Criterion A), whereas the PDQ-4� does not explicitly make such a distinction. It is worth noting that the PID-5 Brief Form demonstrated similar associa- tions to the full PID-5, suggesting that scale length was not a major factor. Finally, the PID-5 and PDQ-4� provided significant incre- mental validity over each other for various functioning domains. The PID-5 provided the most incremental validity for the Cogni- tion domain whereas the PDQ-4� had its greatest incremental validity for Mobility and Self-Care. These results suggest that the two models differ in their specificity in regards to functioning, with the traditional PD model linked to a broader range of func- tioning and the DSM–5 trait model demonstrating more specific links.

Limitations, Future Directions, and Conclusions

Despite its strengths, there are some limitations to the current study. First, although the PID-5 is the APA’s official instrument for assessment of the DSM–5 traits and the PDQ-4� is a widely used measure specifically created to assess traditional PDs, differ- ent self-report measures assessing these models, informant reports, or clinician ratings may lead to different results, which is an important direction for future research. Second, the strong associ- ation between dependability and � is in contrast to previous studies (Chmielewski et al., 2016; Chmielewski & Watson, 2009; Gnambs, 2014; McCrae et al., 2011), including those of the PID-5 (Suzuki et al., 2015). The reasons for the high correlation in the current sample, as well as for the differences in magnitude of PID-5 dependabilities between the current study and that of Suzuki et al. (2015), are unclear, especially considering that past studies have used similar samples and test administration techniques. Third, it is unclear if the current student results would replicate in community or clinical samples. Additional dependability studies of the PID-5 using large samples are required to address these issues. However, previous evidence suggests that the relative level of dependability is consistent across different samples (Chmielewski & Watson, 2009).

In conclusion, the current study provides evidence regarding the dependability, a crucial component of construct validity for trait measures, of the PID-5. In general, the results indicate the DSM–5 traits are highly dependable and outperform the traditional PD model in this regard. However, some constructs assessed by the

T hi

s do

cu m

en t

is co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

ti on

or on

e of

it s

al li

ed pu

bl is

he rs

. T

hi s

ar ti

cl e

is in

te nd

ed so

le ly

fo r

th e

pe rs

on al

us e

of th

e in

di vi

du al

us er

an d

is no

t to

be di

ss em

in at

ed br

oa dl

y.

234 CHMIELEWSKI, RUGGERO, KOTOV, LIU, AND KRUEGER

PID-5 may be best conceptualized as more state like. In addition, the current results replicate previous studies demonstrating impor- tant associations between the DSM–5 traits and functioning. It is worth noting that both the traditional PD and alternative models demonstrated similar associations with functioning, suggesting that the DSM–5 traits capture the important aspects of the tradi- tional PD model.

References

Ackerman, R. A., & Corretti, C. A. (2015). Pathological personality traits and intimacy processes within roommate relationships: Pathological traits and intimacy. European Journal of Personality, 29, 152–172. http://dx.doi.org/10.1002/per.1991

American Psychiatric Association. (2013). Diagnostic and statistical man- ual of mental disorders (5th ed.). Washington, DC: Author.

Anderson, J. L., Sellbom, M., Bagby, R. M., Quilty, L. C., Veltri, C. O. C., Markon, K. E., & Krueger, R. F. (2012). On the convergence between PSY-5 domains and PID-5 domains and facets: Implications for assess- ment of DSM–5 personality traits. Manuscript submitted for publication.

Anusic, I., Lucas, R. E., & Donnellan, M. B. (2012). Dependability of personality, life satisfaction, and affect in short-term longitudinal data. Journal of Personality, 80, 33–58. http://dx.doi.org/10.1111/j.1467- 6494.2011.00714.x

Ashton, M. C., Lee, K., de Vries, R. E., Hendrickse, J., & Born, M. P. (2012). The maladaptive personality traits of the Personality Inventory for DSM–5 (PID-5) in relation to the HEXACO personality factors and schizotypy/dissociation. Journal of Personality Disorders, 26, 641– 659. http://dx.doi.org/10.1521/pedi.2012.26.5.641

Cattell, R. B., Eber, H. W., & Tatsuoka, M. M. (1970). Handbook for the Sixteen Personality Factor Questionnaire (16 PF). Champaign, IL: IPAT.

Chmielewski, M., Clark, L. A., Bagby, R. M., & Watson, D. (2015). Method matters: Understanding diagnostic reliability in DSM–IV and DSM–5. Journal of Abnormal Psychology, 124, 764 –769. http://dx.doi .org/10.1037/abn0000069

Chmielewski, M., Sala, M., Tang, R., & Baldwin, A. S. (2016). Examining the construct validity of affective judgments of physical activity mea- sures. Psychological Assessment, 28, 1128 –1141. http://dx.doi.org/10 .1037/pas0000322

Chmielewski, M., & Watson, D. (2008). The heterogeneous structure of schizotypal personality disorder: Item-level factors of the Schizotypal Personality Questionnaire and their associations with obsessive- compulsive disorder symptoms, dissociative tendencies, and normal personality. Journal of Abnormal Psychology, 117, 364 –376. http://dx .doi.org/10.1037/0021-843X.117.2.364

Chmielewski, M., & Watson, D. (2009). What is being assessed and why it matters: The impact of transient error on trait research. Journal of Personality and Social Psychology, 97, 186 –202. http://dx.doi.org/10 .1037/a0015618

Clark, L. A. (2007). Assessment and diagnosis of personality disorder: Perennial issues and an emerging reconceptualization. Annual Review of Psychology, 58, 227–257. http://dx.doi.org/10.1146/annurev.psych.57 .102904.190200

Clark, L. A., Livesley, W. J., & Morey, L. (1997). Personality disorder assessment: The challenge of construct validity. Journal of Personality Disorders, 11, 205–231. http://dx.doi.org/10.1521/pedi.1997.11.3.205

Clark, L. A., & Watson, D. (1995). Constructing validity: Basic issues in objective scale development. Psychological Assessment, 7, 309 –319. http://dx.doi.org/10.1037/1040-3590.7.3.309

Crego, C., & Widiger, T. A. (2014). Psychopathy, DSM–5, and a caution. Personality Disorders: Theory, Research, and Treatment, 5, 335–347. http://dx.doi.org/10.1037/per0000078

Davison, S., Leese, M., & Taylor, P. J. (2001). Examination of the screening properties of the personality diagnostic questionnaire 4� (PDQ-4�) in a prison population. Journal of Personality Disorders, 15, 180 –194. http://dx.doi.org/10.1521/pedi.15.2.180.19212

De Fruyt, F., De Clercq, B., De Bolle, M., Wille, B., Markon, K., & Krueger, R. F. (2013). General and maladaptive traits in a five-factor framework for DSM–5 in a university student sample. Assessment, 20, 295–307. http://dx.doi.org/10.1177/1073191113475808

Fossati, A., Somma, A., Borroni, S., Maffei, C., Markon, K. E., & Krueger, R. F. (2015). A head-to-head comparison of the Personality Inventory for DSM–5 (PID-5) with the Personality Diagnostic Questionnaire-4 (PDQ-4) in predicting the general level of personality pathology among community dwelling subjects. Journal of Personality Disorders, 30, 82–94. http://dx.doi.org/10.1521/pedi_2015_29_184

Gnambs, T. (2014). A meta-analysis of dependability coefficients (test– retest reliabilities) for measures of the Big Five. Journal of Research in Personality, 52, 20 –28. http://dx.doi.org/10.1016/j.jrp.2014.06.003

Gore, W. L., & Widiger, T. A. (2013). The DSM–5 dimensional trait model and five-factor models of general personality. Journal of Abnormal Psychology, 122, 816 – 821. http://dx.doi.org/10.1037/a0032822

Grilo, C. M., Sanislow, C. A., Gunderson, J. G., Pagano, M. E., Yen, S., Zanarini, M. C., . . . McGlashan, T. H. (2004). Two-year stability and change of schizotypal, borderline, avoidant, and obsessive-compulsive personality disorders. Journal of Consulting and Clinical Psychology, 72, 767–775. http://dx.doi.org/10.1037/0022-006X.72.5.767

Hopwood, C. J., Thomas, K. M., Markon, K. E., Wright, A. G. C., & Krueger, R. F. (2012). DSM–5 personality traits and DSM–IV personal- ity disorders. Journal of Abnormal Psychology, 121, 424 – 432. http:// dx.doi.org/10.1037/a0026656

Hopwood, C. J., Wright, A. G., Krueger, R. F., Schade, N., Markon, K. E., & Morey, L. C. (2013). DSM–5 pathological personality traits and The Personality Assessment Inventory. Assessment, 20, 269 –285. http://dx .doi.org/10.1177/1073191113486286

Hyler, S. E. (1994). Personality Diagnostic Questionnaire-4. New York, NY: New York State Psychiatric Institute.

Johansen, M., Karterud, S., Pedersen, G., Gude, T., & Falkum, E. (2004). An investigation of the prototype validity of the borderline DSM–IV construct. Acta Psychiatrica Scandinavica, 109, 289 –298. http://dx.doi .org/10.1046/j.1600-0447.2003.00268.x

Keeley, J. W., Flanagan, E. H., & McCluskey, D. L. (2014). Functional impairment and the DSM–5 dimensional system for personality disorder. Journal of Personality Disorders, 28, 657– 674. http://dx.doi.org/10 .1521/pedi_2014_28_133

Krueger, R. F., Derringer, J., Markon, K. E., Watson, D., & Skodol, A. E. (2012). Initial construction of a Maladaptive Personality Trait Model and Inventory for DSM–5. Psychological Medicine, 42, 1879 –1890. http:// dx.doi.org/10.1017/S0033291711002674

Lenzenweger, M. F. (1999). Stability and change in personality disorder features: The Longitudinal Study of Personality Disorders. Archives of General Psychiatry, 56, 1009 –1015. http://dx.doi.org/10.1001/archpsyc .56.11.1009

Markon, K. E., Chmielewski, M., & Miller, C. J. (2011). The reliability and validity of discrete and continuous measures of psychopathology: A quantitative review. Psychological Bulletin, 137, 856 – 879. http://dx.doi .org/10.1037/a0023678

McCrae, R. R., Kurtz, J. E., Yamagata, S., & Terracciano, A. (2011). Internal consistency, retest reliability, and their implications for Person- ality Scale validity. Personality and Social Psychology Review, 15, 28 –50. http://dx.doi.org/10.1177/1088868310366253

McGlashan, T. H., Grilo, C. M., Sanislow, C. A., Ralevski, E., Morey, L. C., Gunderson, J. G., . . . Pagano, M. (2005). Two-year prevalence and stability of individual DSM–IV criteria for schizotypal, borderline, avoidant, and obsessive-compulsive personality disorders: Toward a

T hi

s do

cu m

en t

is co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

ti on

or on

e of

it s

al li

ed pu

bl is

he rs

. T

hi s

ar ti

cl e

is in

te nd

ed so

le ly

fo r

th e

pe rs

on al

us e

of th

e in

di vi

du al

us er

an d

is no

t to

be di

ss em

in at

ed br

oa dl

y.

235DSM–5 TRAITS AND TRADITIONAL PD MODEL

hybrid model of axis II disorders. The American Journal of Psychiatry, 162, 883– 889. http://dx.doi.org/10.1176/appi.ajp.162.5.883

Miller, J. D., Campbell, W. K., & Pilkonis, P. A. (2007). Narcissistic personality disorder: Relations with distress and functional impairment. Comprehensive Psychiatry, 48, 170 –177. http://dx.doi.org/10.1016/j .comppsych.2006.10.003

Miller, J. D., Few, L. R., Lynam, D. R., & MacKillop, J. (2015). Patho- logical personality traits can capture DSM–IV personality disorder types. Personality Disorders: Theory, Research, and Treatment, 6, 32– 40. http://dx.doi.org/10.1037/per0000064

Nunnally, J. C. (1978). Psychometric theory (2nd ed.). New York, NY: McGraw-Hill.

Oldham, J. M., & Skodol, A. E. (2000). Charting the future of axis II. Journal of Personality Disorders, 14, 17–29. http://dx.doi.org/10.1521/ pedi.2000.14.1.17

Oldham, J. M., Skodol, A. E., Kellman, H. D., Hyler, S. E., Rosnick, L., & Davies, M. (1992). Diagnosis of DSM–III–R personality disorders by two structured interviews: Patterns of comorbidity. The American Jour- nal of Psychiatry, 149, 213–220. http://dx.doi.org/10.1176/ajp.149.2.213

Quilty, L. C., Ayearst, L., Chmielewski, M., Pollock, B. G., & Bagby, R. M. (2013). The psychometric properties of the Personality Inventory for DSM–5 in an APA DSM–5 field trial sample. Assessment, 20, 362–369. http://dx.doi.org/10.1177/1073191113486183

Samuel, D. B., Hopwood, C. J., Ansell, E. B., Morey, L. C., Sanislow, C. A., Markowitz, J. C., . . . Grilo, C. M. (2011). Comparing the temporal stability of self-report and interview assessed personality dis- order. Journal of Abnormal Psychology, 120, 670 – 680. http://dx.doi .org/10.1037/a0022647

Samuel, D. B., Hopwood, C. J., Krueger, R. F., Thomas, K. M., & Ruggero, C. J. (2013). Comparing methods for scoring personality disorder types using maladaptive traits in. DSM–5. Assessment, 20, 353–361. http://dx.doi.org/10.1177/1073191113486182

Schmidt, F. L., Le, H., & Ilies, R. (2003). Beyond alpha: An empirical examination of the effects of different sources of measurement error on reliability estimates for measures of individual differences constructs. Psychological Methods, 8, 206 –224. http://dx.doi.org/10.1037/1082- 989X.8.2.206

Shea, M. T., Stout, R., Gunderson, J., Morey, L. C., Grilo, C. M., McGlashan, T., . . . Keller, M. B. (2002). Short-term diagnostic stability of schizotypal, borderline, avoidant, and obsessive-compulsive person- ality disorders. The American Journal of Psychiatry, 159, 2036 –2041. http://dx.doi.org/10.1176/appi.ajp.159.12.2036

Skodol, A. E., Gunderson, J. G., Shea, M. T., McGlashan, T. H., Morey, L. C., Sanislow, C. A., . . . Stout, R. L. (2005). The Collaborative Longitudinal Personality Disorders Study (CLPS): Overview and impli- cations. Journal of Personality Disorders, 19, 487–504. http://dx.doi .org/10.1521/pedi.2005.19.5.487

Suzuki, T., Griffin, S. A., & Samuel, D. B. (2016). Capturing the DSM–5 alternative personality disorder model traits in the five-factor model’s nomological net. Journal of Personality. Advance online publication. http://dx.doi.org/10.1111/jopy.12235

Thomas, K. M., Yalch, M. M., Krueger, R. F., Wright, A. G., Markon, K. E., & Hopwood, C. J. (2013). The convergent structure of DSM–5

personality trait facets and five-factor model trait domains. Assessment, 20, 308 –311.

Tyrer, P., Coombs, N., Ibrahimi, F., Mathilakath, A., Bajaj, P., Ranger, M., . . . Din, R. (2007). Critical developments in the assessment of person- ality disorder. The British Journal of Psychiatry, 190, s51–s59. http:// dx.doi.org/10.1192/bjp.190.5.s51

Üstün, T. B., Chatterji, S., Kostanjsek, N., Rehm, J., Kennedy, C., Epping- Jordan, J., . . . the WHO/NIH Joint Project. (2010). Developing the World Health Organization disability assessment schedule 2.0. Bulletin of the World Health Organization, 88, 815– 823. http://dx.doi.org/10 .2471/BLT.09.067231

Verheul, R., & Widiger, T. A. (2004). A meta-analysis of the prevalence and usage of the personality disorder not otherwise specified (PDNOS) diagnosis. Journal of Personality Disorders, 18, 309 –319. http://dx.doi .org/10.1521/pedi.2004.18.4.309

Watson, D. (2004). Stability versus change, dependability versus error: Issues in the assessment of personality over time. Journal of Research in Personality, 38, 319 –350. http://dx.doi.org/10.1016/j.jrp.2004.03.001

Widiger, T. A., & Coker, L. A. (2002). Assessing personality disorders. In J. N. Butcher (Ed.), Clinical personality assessment: Practical ap- proaches (2nd ed., pp. 407– 434). New York, NY: Oxford University Press.

Widiger, T. A., & Samuel, D. B. (2005). Diagnostic categories or dimen- sions? A question for the Diagnostic and Statistical Manual of Mental Disorders–Fifth Edition. Journal of Abnormal Psychology, 114, 494 – 504. http://dx.doi.org/10.1037/0021-843X.114.4.494

Widiger, T. A., & Trull, T. J. (2007). Plate tectonics in the classification of personality disorder: Shifting to a dimensional model. American Psy- chologist, 62, 71– 83. http://dx.doi.org/10.1037/0003-066X.62.2.71

Wright, A. G., Calabrese, W. R., Rudick, M. M., Yam, W. H., Zelazny, K., Williams, T. F., . . . Simms, L. J. (2015). Stability of the DSM–5 Section III pathological personality traits and their longitudinal associations with psychosocial functioning in personality disordered individuals. Journal of Abnormal Psychology, 124, 199 –207. http://dx.doi.org/10.1037/ abn0000018

Wright, A. G. C., Thomas, K. M., Hopwood, C. J., Markon, K. E., Pincus, A. L., & Krueger, R. F. (2012). The hierarchical structure of DSM–5 pathological personality traits. Journal of Abnormal Psychology, 121, 951–957. http://dx.doi.org/10.1037/a0027669

Zanarini, M. C., Frankenburg, F. R., Reich, D. B., & Fitzmaurice, G. (2012). Attainment and stability of sustained symptomatic remission and recovery among patients with borderline personality disorder and axis II comparison subjects: A 16-year prospective follow-up study. The Amer- ican Journal of Psychiatry, 169, 476 – 483. http://dx.doi.org/10.1176/ appi.ajp.2011.11101550

Zimmerman, M. (1994). Diagnosing personality disorders. A review of issues and research methods. Archives of General Psychiatry, 51, 225– 245. http://dx.doi.org/10.1001/archpsyc.1994.03950030061006

Zimmermann, J., Altenstein, D., Krieger, T., Holtforth, M. G., Pretsch, J., Alexopoulos, J., . . . Leising, D. (2014). The structure and correlates of self-reported DSM–5 maladaptive personality traits: Findings from two German-speaking samples. Journal of Personality Disorders, 28, 518 – 540. http://dx.doi.org/10.1521/pedi_2014_28_130

T hi

s do

cu m

en t

is co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

ti on

or on

e of

it s

al li

ed pu

bl is

he rs

. T

hi s

ar ti

cl e

is in

te nd

ed so

le ly

fo r

th e

pe rs

on al

us e

of th

e in

di vi

du al

us er

an d

is no

t to

be di

ss em

in at

ed br

oa dl

y.

236 CHMIELEWSKI, RUGGERO, KOTOV, LIU, AND KRUEGER