56 Graduate Student Journal of Psychology 2018, Vol. 17 Copyright 2018 by the Department of Counseling and Clinical Psychology Teachers College, Columbia University Self-Other Agreement in the Assessment of Personality Pathology: Do We Always Know Ourselves Best? Kimberly Tom, B.S. Teachers College, Columbia University Leonard J. Simms, Ph.D. The State University of New York, University at Buffalo Assessment of personality pathology can vary greatly depending on who is doing the assessment. Agreement between self- and informant-reports can be influenced by deviating patterns of thoughts and behaviors which are characteristic of those with personality pathology. In the current study we investigated whether presence of personality pathology was associated with the level of agreement between self- and informant-reports using a sample of 83 current or recent psychiatric patients, each accompanied by a close other. Target and informant participants reported presence of personality pathology using a measure for personality disorder criteria and a measure for maladaptive personality traits. Personality disorder criteria scores and PID-5 trait scores were correlated with level of agreement on different measures to assess whether personality pathology was associated with self-other agreement. Five-Factor model traits were also included as criteria in the analysis. Results were partially consistent with previous findings. Higher informant-reported personality pathology was associated with lower agreement for a measure of Five-Factor model traits. However, higher target-reported personality pathology was actually associated with increased agreement for Five-Factor model traits. Higher pathology as rated by the self and informant were both associated with increased agreement on PD criteria and maladaptive traits. Higher target-rated personality pathology was also associated with higher agreement on acquaintanceship. In conclusion, insight about one’s own personality pathology is related to self-other agreement, and the strength and direction of this relationship can differ depending on the type of informa- tion that is being judged. Is Personality Pathology Associated with Self-Other Agreement? Personality disorders (PDs) are characterized by enduring and inflexible patterns of thinking and behavior involving interpersonal functioning, cognition, affectivity, and impulse control, that can lead to distress or impairment (American Psychiatric Association, 2013). Such dysfunction may lead others to view someone with a personality disorder differently than how this person would view themselves. These differences may call into question the utility of personality judgments made by the self or an informant. Thus, practitioners who do not get multiple viewpoints may be missing important information about their clients; particularly those with limited insight. Numerous studies have investigated personality assessment from the self and other perspectives (e.g., Carlson, Vazire, & Oltmanns, 2013; Human, Biesanz, Finseth, Pierce, & Le 2014; Vazire, 2010; Vazire & Mehl, 2008) and have found many variables that influ- ence accuracy (e.g., the visibility of different traits, how well the informant knows the person they are judging, the self ’s consistency of behavior). Given the marked interpersonal problems and inconsistency of thoughts and behaviors typically associated with PDs, a plausible yet relatively untested hypothesis is that when judging an individual with more personality pathology, agree- ment between the self and an informant on a particular Override (Hidden running head text): Tom, Simms Self-Other Agreement in Personality Pathology Assessment This study was supported by a research grant to L. J. Simms from the National Institute of Mental Health (No. R01MH080086). Correspondence should be addressed to Kimberly Tom at kt2672@tc.columbia.edu. 57 SELF-OTHER AGREEMENT IN PERSONALITY PATHOLOGY ASSESSMENT measure would be lower than those with less personality pathology. However, few studies have observed whether personality pathology is associated with differences in self-other agreement. If personality pathology does have a relationship to self-other agreement, then learn- ing more about those relationships would be useful in a clinical setting. For example, larger discrepancies in between self- and informant-reports could be a diag- nostic criterion for personality pathology. The current study will investigate whether the presence of person- ality pathology is associated with self-other agreement on measures of personality and personality pathology. Self- and Informant-Reports Differences Between Self- and Informant- Reports. To understand how personality pathology may influence self-other agreement, it is important to first understand how self- and informant-reports inherently differ. Self-report has been considered the standard for personality assessment measures. However, self-reports may be inaccurate for a variety of reasons, including distorted self-views (John & Robins, 1994), self-serving bias (Greenwald, 1980; Holzbach, 1978; John & Robins, 1993; Paulhus & John, 1998; Robins & Beer, 2001), and the fish-and-water effect (i.e. the idea that overt patterns of behavior are highly observable to others, but not the self; Kolar, Funder, & Colvin, 1996). Self-reports demonstrate more accuracy than informant- reports when reporting on inner emotional experience (Spain, Eaton, & Funder 2000). Early research has suggested informant-reports are more accurate than self-reports in predicting behavior (Hofstee, 1994), and the better an informant knows the target, the more accurate their judgments become (Connolly, Kavanagh, & Viswesvaran, 2007; Connelly & Ones 2010; Kurtz & Sherker, 2003; South, Oltmanns, Johnson, & Turkheimer, 2011; Tackett, 2011). Newer research has reached consensus that self- and informant-reports offer similar accuracy for predicting behavior (Ready, Watson, & Clark, 2002; Vazire & Mehl, 2008). Although self- and informant-reports show similar levels of valid- ity, each offers a unique perspective on a given target’s personality (Carlson et al., 2013; Clifton, Turkheimer, & Oltmanns, 2004; Funder, 1995; Oltmanns & Turkheimer, 2009; Vazire, 2010). Self-Other Knowledge Asymmetry Model. According to the Self-Other Knowledge Asymmetry Model (Vazire, 2010), discrepancies between self and informants are mediated by the type of information available to the judge. For example, informants do not have access to inner emotional experiences of the self, and sometimes the self is not consciously aware of behaviors that are visible to informants. For an individual with personality pathology, factors such as identity disturbance or affectivity issues could mediate assessment of personality. Convergence between self- reports and informant-reports might then be a function of these diverging assessments. Realistic Accuracy Model. Unlike the Self-Other Knowledge Asymmetry Model, the Realistic Accuracy Model (Funder, 1995) considers more than just the type of information available to a judge. Differences between self- and informant-reports were described in the “accuracy paradigm,” which denoted a shift of focus from error of personality judgments to accu- racy of personality judgments (Funder, 1995; Funder & Dobroth, 1987). The information available to each judge does not necessarily result in only error, in that each judge is interpreting the given information in a unique way. Each judge is trying to make an accurate judgment, even though what is truly accurate cannot be definitively known.One could reasonably conclude that if a target and informant are agreeing more on some criterion, that those two judgments are likely both more accurate than if they diverged. Such divergence may be more prevalent in those with personality pathology, and creates an interesting question in relation to the study of accuracy. Accuracy of Self- and Informant-Reports. Studies concerned with the accuracy of self- and infor- mant-reports have tried to determine whose judgments were more accurate for different criteria. The self tends to be more accurate in rating traits low in observability, and informants are more accurate on evaluative (e.g., intelligence) traits when predicting behavior (Connelly & Ones, 2010; Vazire, 2010). Carlson et al. (2013) found similar results when evaluating self and informant accu- racy for PDs. They concluded that the self was more accurate for less visible PDs (e.g. Avoidant, Dependent, and Obsessive-compulsive PD), informants were more accurate for more visible/antagonistic PDs (e.g. Antisocial and Histrionic PD), and neither was more accurate for thought disorder and less visible/antago- nistic PDs (e.g. Paranoid and Borderline PD). Carlson 58 TOM, SIMMS et al. used a non-clinical sample, so these findings may not generalize to those with PDs. Additionally, accu- racy analyses were calculated using a composite of self, informant, and clinician judgments. When evaluating the accuracy of people’s judgments by drawing statistical connections between other judgments, conceptualiza- tion of accuracy in this context should be called into question. People with personality pathology tend to report personality traits of others in a different way than those without personality pathology, and are also described by others differently, compared to those with less personal- ity pathology (Tandler, Mosch, Wolf, & Borkenau, 2015). These findings reinforce the idea that people with PDs see the world differently (Human & Biesanz, 2011), because judgments are based on some internal aware- ness of normativity (Biesanz, 2010), which can result in lower accuracy of assessment. Ready, Clark, Watson, and Westerhouse (2000) found that individuals whose own personalities were more associated with their own ratings of others had lower self-other agreement. This implies people sometimes use the self as a guide for judging others. If the target of judgment has a person- ality disorder, they may be judged more inaccurately because an informant may be filling in gaps in knowl- edge with assumptions based on themselves. Judging Undesirable Traits. A number of studies implicate lower self-other agreement for undesirable traits. Clifton et al. (2004) investigated self- and infor- mant-reports on personality pathology traits and found low to moderate agreement. These findings are in line with research suggesting better agreement with more pro-social traits such as extraversion, agreeableness, conscientiousness, and being generally well-adjusted (Colvin, 1993; Connolly et al., 2007). However, South et al. (2011) did find self-spouse agreement on patho- logical personality traits to be better than what has been found with regular self-other agreement samples. The authors conclude that people tend towards positive presentation of the self with people they do not know very well, and so an informant must know someone far better in order to accurately judge socially undesirable traits as opposed normal range personality traits. Individuals with personality pathology are less well-adjusted given the marked interpersonal prob- lems and identity disturbance characteristics associated with personality pathology. Therefore, the self and informants may vary more when reporting personal- ity assessments. Self-reports and informant-reports have unique variance (Connolly et al., 2007), but well adjusted-individuals who are easier to judge show less variance (Furr, Dougherty, Marsh, & Mathias, 2007), perhaps because their actions are more consistent with their personality (Human et al., 2014). Leising, Rehbein, and Sporberg (2006) found that individuals actually downplay their most extreme behaviors, which may be particularly relevant with PDs, which are seen as extreme presentations of normal traits (Lynam & Widiger, 2001; Miller, Lynam, Widiger, & Leukefeld, 2001). Is Personality Pathology Associated with Self-Other Agreement? The literature has suggested that personality pathol- ogy moderates self-other agreement, but such studies are limited. Furr et al. (2007) investigated the role of personality pathology in self-other agreement by look- ing at the difference between two groups: adolescents with conduct disorder and a normal control group. The study concluded that personality pathology moderated self-other agreement. The control group had higher overall self-other agreement on Five-Factor model personality traits than the conduct disorder group. Lower agreement and unique variability in the conduct disorder group’s self- and informant-ratings suggests personality pathology leads to distinct differences in self-views and informant-views. However, Furr et al. also found that presence of conduct disorder had higher agreement for traits more relevant to conduct disorder (e.g. Extraversion and Neuroticism). Generalizing these results to PDs, overall lower agreement in those with PDs could indicate self- and informant-assessments of people with personality pathology is less valid, or flawed, simply because certain traits become harder to assess. One possible explanation would be that even if highly visible traits produce some agreement, less overt traits are obscured by inconsistent thoughts and behaviors of the target. In a study using a clinical sample with multiple infor- mants, Mosterman and Hendriks (2011) concluded that higher self-other disagreement on the Five-Factor Model Inventory was correlated with higher self-rated shyness, introversion, hostility, and depression. They also claimed individuals meeting more DSM-IV Axis II (PD) crite- ria exhibited more self-other disagreement. This last 59 SELF-OTHER AGREEMENT IN PERSONALITY PATHOLOGY ASSESSMENT finding was somewhat unreliable because their conclu- sion was drawn from a simple paired-sample t-test for participants with some personality problems and those without. Only six of those 52 participants with person- ality problems met criteria for a personality disorder. In a round-robin style design, Tandler et al. (2015) assessed interpersonal perception in groups of four acquaintances. Assessment of interpersonal perception extends beyond the scope of the current context, but aspects pertaining to self-other agreement are closely related to the current study. Individuals rated themselves and their group members on Five-Factor model traits. Each participant also completed a self-report on person- ality pathology. Tandler et al. explored multiple types of self-other agreement, but the type synonymous with the current study was expression-based distinctive self-other agreement. This analysis involved predicting agreement coefficients from PD scores. They concluded that personality pathology moderated agreement, where those endorsing more PD symptoms had lower self-other agreement for a Five-Factor model based assessment. The current study will replicate and extend these findings by correlating self-other agreement with both self- and informant-reported PD scores for Five- Factor model traits, maladaptive traits, PD criteria, and personal acquaintanceship. Additionally, participants of the study by Tandler et al. were university students, and results may not generalize to psychiatric populations. Self-other agreement on personality traits does not differ in magnitude or variability when comparing clinical to nonclinical samples (Ready & Clark, 2002), suggesting that other types of pathology do not moder- ate self-other agreement, at least to the degree that personality pathology might. However, Ready and Clark also found no difference in self-other ratings between clinical and nonclinical samples on interpersonal prob- lems, which contradicts research presenting general lower self-other agreement on PD measures (Klonsky, Oltmanns, & Turkheimer, 2002). Summary and Current Study Information that is uniquely available to the self or informants should influence the respective reports in distinctive ways (Carlson et al., 2013; Kenrick & Stringfield, 1980; Spain et al., 2000; Vazire, 2010). The current study aims to find whether agreement on particular traits is associated with self-reported and informant-reported level of personality pathology on a measure of PD. The results of previous research suggest that higher personality pathology is associ- ated with lower agreement on Five-Factor model traits. Previous studies did not test higher personality pathol- ogy as rated by an informant, and also only investigated agreement on Five-Factor model traits. We hypothesized that, consistent with previous find- ings, more personality pathology as reported by the self, and as reported by the informant, would be associated with lower self-other agreement for a Five-Factor model measure. Neuroticism, a salient trait in PDs, was previ- ously found to have increased agreement in a conduct disorder group compared to a non-conduct disorder group (Furr et al., 2007). Individuals with personal- ity pathology may be more likely to display neurotic behaviors or discuss neurotic thoughts with close others. Given this past finding, expectations for the maladap- tive traits and PD criteria and their association with agreement were exploratory. It is possible that more personality pathology would predict some agreement for the maladaptive traits and PD criteria. Method Participants Data for the current study were previously collected as part of the CAT-PD project (Computerized Adaptive Test of Personality Disorder; Simms, Goldberg, Roberts, Watson, Weite, & Rotterman, 2011), designed to develop and validate a comprehensive model of personality disorder traits. The current sample consists of 83 dyads from the original study. 89.2% of the dyads had known each other for at least one year. 44.6% of the dyads had known each other for at least 13–14 years. Ages of the targets ranged from 17 to 65 years. Ages of the informants ranged from 18 to 84 years. Procedure Current or recent clinical patients were recruited via postings in Western New York mental health clin- ics. Participants who responded to the posting were screened for eligibility. Qualified individuals were then asked if they could also bring in a close friend or family member to participate in the study. Only four of the measures from that larger dataset were used in the current study. For each assessment, the target participant 60 TOM, SIMMS (the self) answered questions about themselves on a vari- ety of measures (e.g., I see myself as someone who is talkative.) For the informants, the same measures were administered, except phrasing was slightly modified to address the target participant (e.g., I see the person I came with today as someone who is talkative.), except for one measure, in which an informant form already existed. Each participant received monetary compensa- tion for their time. Measures Structured Clinical Interview for DSM-III-R Personality Disorders (SCID-II). The SCID-II (First, Spitzer, Gibbon, & Williams, 1995) is a semi-structure interview admin- istered by a clinician to diagnose DSM-IV axis II PDs (American Psychiatric Association, 2000). The interview is done after an individual has answered a question- naire based on PD criteria. The clinician follows up on these questionnaire items to determine which criteria for disorders are met. Internal consistencies ranged by disorder (α = .58–.85 for targets, α = .65–.84 for informants). Structured Clinical Interview for DSM-IV Personality Questionnaire (SCID-II PQ). The SCID-II PQ is comprised of 119 yes or no questions (“Do you often worry about being criticized or rejected in social situ- ations?”) based on DSM-IV axis II (PD) (American Psychiatric Association, 2000) criteria. The SCID-II PQ is the questionnaire portion of the SCID-II interview (First et al., 1995). Internal consistency ranged by disor- der (α = .64–.89 for targets, .60–.88 for informants). Big Five Inventory (BFI). The BFI (John, Donahue, & Kentle, 1991) is a 44-item questionnaire that begins with the first half of a statement: “I see myself as someone who . . . ” Items include ends of statements such as “is talkative” or “is depressed, blue” rated on a scale of 1 (Disagree Strongly) to 5 (Agree Strongly). Items corresponded to Five-Factor model personality traits (Extraversion, Agreeableness, Conscientiousness, Neuroticism, and Openness). The BFI traits had adequate internal consistency (α = .74–.75 for targets, α = .72–.89 for informants). Personality Inventory for DSM-5 (PID-5). The PID-5 (Krueger, Derringer, Markon, & Skodol, 2012) consists of 220 items corresponding to pathological personality traits. Statements such as “I do a lot of things that others consider risky” are rated on a four-point scale of 0 (Very False or Often False) to 3 (Very True or Often True). Internal consistencies for almost all 25 traits were high (α = .68–.95 for targets). Personality Inventory for DSM-5-Informant Form (PID-5- IRF). The PID-5-IRF (Markon, Quilty, Bagby, & Krueger, 2013) consists of 218 items corresponding to maladaptive personality traits as rated by an infor- mant, for a target. Statements such as “He or She does a lot of things that others consider risky” are rated on a four-point scale of 0 (Very False or Often False) to 3 (Very True or Often True). Internal consistencies for all traits were good (α = .74–.96) except for Unusual Beliefs and Experiences (α = .52). Results Profile agreement for each measure was calculated as the Pearson correlation between self and informant scale scores within those measures. For the BFI, if an individual had self-rated items of the BFI for Openness, Conscientiousness, Extraversion, Agreeableness, and Neuroticism, summing the items for each trait provided a total score for each trait. This series of numbers was then correlated with the same trait total scores provided by an informant for a given dyad. Profile agreement was calculated using PD criterion counts from the SCID-II PQ, trait scores from the PID-5 and PID-5-IRF, and trait scores from the BFI, so that each measure had an overall inter-dyad agreement correlation coefficient for every target-informant pair. The inter-dyad agreement correlations for each measure then were correlated with self-reported scores on various criteria (Table 3). The SCID-II PD scores were counts of actual PD criteria met by the individual as determined by SCID-II PQ responses, followed-up by a semi-structured interview performed by a clinician. Scores on traits of the PID-5 and the BFI were simply the total of scale scores for those traits on their respec- tive measures. The same calculations were repeated with informant-reported scores and the inter-dyad agreement correlations (Table 4). Though typically only adminis- tered to targets, scores on the SCID-II for informants were determined by semi-structured interviews of the informants about the targets. 61 SELF-OTHER AGREEMENT IN PERSONALITY PATHOLOGY ASSESSMENT Descriptive Statistics Results showed that levels of inter-dyad agreement and self- and informant-ratings had good variability and range. The mean level of inter-dyad agreement correla- tions for all measures were reasonably high, even with some dyads producing negative inter-dyad agreement correlations. The PID-5 had the highest mean level of agreement (M = .48) and the SCID-II PQ had the lowest (M = .35). Readers may email the authors for additional information on means, standard deviations, minimums, maximums of criteria rated by the self and their informants. Intercorrelations Table 2 presents intercorrelations of the four measures utilized for inter-dyad agreement correlations. Only two intercorrelations were significantly associated. Higher inter-dyad agreement on the BFI traits was asso- ciated with higher agreement on PID-5 trait scores. Target-Rated Criteria and Agreement Table 3 displays the correlation matrix for crite- ria rated by targets, correlated with the four inter-dyad agreement variables described above. The general pattern indicated that higher target-rated personal- ity pathology was associated with higher self-other agreement. Surprisingly, no SCID-II rated PD criteria were significantly associated with SCID-II PQ agree- ment. Ratings for several PID-5 traits (Anxiousness, Distractibility, Eccentricity, Emotional Liability, Hostility, Perseveration, and Suspiciousness) were positively correlated with agreement on the SCID-II PQ. One BFI trait (Neuroticism) was positively associated with SCID-II PQ agreement. For agreement on the PID-5, the self-reported rating of only one PD (Obsessive-Compulsive) resulted in a positive association. The PID-5 traits, Anxiousness, Emotional Liability, Hostility, and Risking Taking, were positively associated with PID-5 agreement. Neuroticism was also positively correlated with agree- ment on the PID-5. Increased inter-dyad agreement on the BFI was positively associated with target-rated Dependent PD, Eccentricity, and Unusual Beliefs and Experiences. No other target-rated criteria were significantly associated with BFI agreement. Informant-Rated Criteria and Agreement Results for informant-rated criteria (Table 4) showed a different pattern compared to the results for target-ratings. Informant-ratings for Paranoid, Narcissistic, and Borderline PD criteria had moderate correlations with SCID-II PQ agreement. Almost half of the informant-ratings for PID-5 traits were positively associated with agreement on the SCID-II PQ. For BFI rated traits, Agreeableness and Conscientiousness were negatively correlated with SCID-II PQ agreement, and Neuroticism was positively associated with SCID-II PQ agreement. Only one informant-rating was asso- ciated with PID-5 agreement; the PID-5 trait Rigid Perfectionism. Consistent with previous studies, more personal- ity pathology was associated with lower agreement for Five-Factor model traits. Higher scores of informant- rated Histrionic PD criteria was associated with lower agreement on the BFI. Six informant-rated PID-5 Table 1 Descriptive Statistics M SD Min Max Agreement correlations SCID-II PQ 0.35 0.36 -0.89 0.97 PID-5 0.48 0.26 -0.24 0.89 BFI 0.41 0.47 -0.98 0.96 Notes. N = 57–77. BFI = Big Five Inventory; PID-5 = Personality Inventory for DSM-5; PD = personality disorder; SCID-II PQ = Structured Clinical Interview for DSM-IV, Axis-II, Screening Questionnaire. Table 2 Dyad agreement intercorrelations 1 2 3 1. SCID-II PQ 2. PID-5 .12 3. BFI .12 .38** Notes. *p < 0.05, **p < 0.01, ***p < 0.001. BFI = Big Five Inventory; PAM = Personal Acquaintance Measure; PID-5 = Personality Inventory for DSM-5; SCID-II PQ = Structured Clinical Interview for DSM-IV, Axis-II, Screening Questionnaire. 62 TOM, SIMMS traits were negatively correlated with BFI agree- ment (Callousness, Deceitfulness, Distractibility, Grandiosity, Hostility, and Irresponsibility). BFI traits Agreeableness, Conscientiousness, and Openness were positively associated with BFI agreement. Discussion In previous research, individuals only self- reported personality pathology, and self-other agreement was only calculated for Five-Factor model traits. Overall, the results of these studies found lower agreement for normal range personal- ity traits (Furr et al., 2007; Mosterman and Hendriks, 2011; Tandler et al., 2015). In general, individuals self-reporting more personality pathology tended to agree more with a close other on maladaptive personality traits (PID-5), PD symptoms (SCID-II PQ), , and normal range personality traits (BFI). In the current study, contrary to expectations, higher personality pathology correlated with higher agreement on normal range personality traits. One explanation for this unexpected result with self- ratings of personality pathology is that participants of the current study had received some form of mental health treatment currently or in the past two years. Mental health treatment often focuses on increasing self-reflection and insight. Those indi- viduals who were able to recognize their own higher levels of personality pathology, perhaps were then better at assessing their own normal range personal- ity traits and personal relationships. In general, we had expected to find the same pattern of results with both self- and informant- reports of personality pathology, but this was not the case. The negative relationship between informant-ratings of personality pathology and inter-dyad agreement correlations was similar to previous studies utilizing self-ratings. We specu- late that these findings were more consistent with previous research for a couple of reasons. First, individuals whose informants rated them as higher in personality pathology may not necessarily rate themselves high in personality pathology for reasons such as self-serving biases or lack of insight. These informant ratings may actually be reflecting those individuals who are not aware of their personality Table 3 Relations between dyad agreement and target reports of PD criteria, traits, and acquaintanceship SCID-II PQ agreement PID-5 agreement BFI agreement PD criteria Avoidant .02 .03 .18 Dependent -.16 .05 .27* Obsessive-Compulsive .03 .35** .24 Paranoid .10 .14 .07 Schizotypal -.07 .13 .12 Schizoid -.08 -.07 -.03 Histrionic .10 .04 .16 Narcissistic .17 .18 .22 Borderline .17 .22 .15 Antisocial .10 .04 .10 PID-5 traits Anhedonia .15 .07 .13 Anxiousness .29* .30* .14 Attention Seeking .20 .16 .12 Callousness .14 .10 .12 Deceitfulness .04 .07 .05 Depressivity .18 .19 .05 Distractibility .24* .20 .20 Eccentricity .29* .07 .31* Emotional Lability .33** .29* .22 Grandiosity .16 -.15 .03 Hostility .27* .35* .17 Impulsivity .10 .17 .07 Intimacy .13 -.18 -.10 Irresponsibility .10 -.04 -.01 Manipulativeness .22 .08 .04 Perceptual Dysregulation .21 .00 .20 Perseveration .34** .14 .18 Restricted Affectivity .12 -.05 .12 Rigid Perfectionism .18 .17 .07 Risk Taking .21 .29* .08 Separation Insecurity .15 .23 .15 Submissiveness .14 .06 .01 Suspiciousness .27* .09 .06 Unusual Beliefs and Experiences .14 -.17 .28* Withdrawal .14 -.04 .04 BFI traits Agreeableness -.14 -.12 -.03 Conscientiousness -.07 .11 .09 Extraversion .01 .13 .12 Openness .10 .13 .24 Neuroticism .25* .35* .20 Notes. *p < 0.05, **p < 0.01, ***p < 0.001. BFI = Big Five Inventory; PID-5 = Personality Inventory for DSM-5; PD = personality disorder; SCID-II PQ = Structured Clinical Interview for DSM-IV, Axis-II, Screening Questionnaire. 63 SELF-OTHER AGREEMENT IN PERSONALITY PATHOLOGY ASSESSMENT pathology. In contrast, it seems reasonable that the target individuals who had recognized their own personality pathology would generally have more inter-dyad agreement with their partners. Individuals whose informants reported higher personality pathology also agreed more with their targets about PD symptoms. This was similar to the results for self-rated personality pathology. This again went contrary to expectations, because we tend to predict that individuals with more personality pathology would be oblivious to their maladaptive traits. However, these results appear to indicate personality pathology is quite salient and visible to both targets and informants. Even though maladaptive traits and symptoms are harder to judge and would normally have lower general agreement (Clifton et al., 2004), in the current study, individuals with more personality pathology exhibited higher inter-dyad agreement on these undesirable criteria compared to individuals with less personality pathology. Notably, only one informant rated criterion correlated with higher agreement on maladap- tive traits. This was an odd result, given that many informant-reported criterion for personality pathol- ogy had significant associations. Where higher self-reported personality pathology had resulted in higher agreement on both PD criteria and the maladaptive traits, the informant analysis had more significant associations for the PD criteria measure, but lacked significant correlations for maladaptive traits. We were unable to think of an explanation for this finding. BFI Traits and Self-Other Agreement Although the primary focus of this study was to investigate level of personality pathology and its relationship to self-other agreement, we also included self- and informant-rated BFI traits in the analysis. Similar to previous findings (Furr et al. 2007), Neuroticism had actually correlated positively with inter-dyad agreement correlations of the PD criteria measure for both the target- and informant-analysis. For the target-analysis, higher Neuroticism was also associated with higher agree- ment on the maladaptive trait measure. Table 4 Relations between dyad agreement and informant reports of PD criteria, traits, and acquaintanceship SCID-II PQ agreement PID-5 agreement BFI agreement PD criteria Avoidant .13 .07 -.05 Dependent .19 -.09 -.15 Obsessive-Compulsive .15 .02 -.06 Paranoid .30** .05 -.24 Schizotypal .17 .13 -.16 Schizoid .15 .03 .00 Histrionic .05 -.15 -.29* Narcissistic .34** .01 -.20 Borderline .41*** .07 -.20 Antisocial .13 -.12 -.24 PID-5 traits Anhedonia .25* -.04 -.21 Anxiousness .28* .21 -.19 Attention Seeking .23 .05 -.13 Callousness .29* -.10 -.29* Deceitfulness .16 -.16 -.33* Depressivity .36** .07 -.07 Distractibility .32** .13 -.29* Eccentricity .16 .07 -.20 Emotional Lability .36** -.15 -.20 Grandiosity .18 -.06 -.30* Hostility .29* -.12 -.32* Impulsivity .30* -.10 -.25 Intimacy .14 -.07 -.12 Irresponsibility .19 -.23 -.28* Manipulativeness .23 -.11 -.19 Perceptual Dysregulation .25* .03 -.24 Perseveration .30** .03 -.14 Restricted Affectivity .12 -.10 -.17 Rigid Perfectionism .21 .30* .00 Risk Taking .19 .11 -.17 Separation Insecurity .38** .00 -.09 Submissiveness .21 -.05 .08 Suspiciousness .24* .00 -.23 Unusual Beliefs and Experiences .17 -.04 -.15 Withdrawal .22 -.00 -.20 BFI traits Agreeableness -.29* -.03 .42** Conscientiousness -.36** .10 .30* Extraversion -.13 -.06 -.13 Openness .03 .10 .46** Neuroticism .43** .12 -.18 Notes. *p < 0.05, **p < 0.01, ***p < 0.001. BFI = Big Five Inventory; PID-5 = Personality Inventory for DSM-5; PD = personality disorder; SCID-II PQ = Structured Clinical Interview for DSM-IV, Axis-II, Screening Questionnaire. 64 TOM, SIMMS The other four of the Five-Factor model traits did not have any other significant correlations for the target-analysis. However, for the informant analy- sis, Agreeableness, Conscientiousness, and Openness exhibited some reversed patterns compared to the maladaptive traits. Higher Agreeableness and Openness meant lower inter-dyad agreement on PD criteria. Maladaptive criteria are supposed to be more difficult to judge in those without personality pathol- ogy (Clifton et al., 2004). Individuals rated as higher in Agreeableness and Contentiousness may also be individuals who have less personality pathology, and therefore would have worse self-other agreement on PD criteria. Those whose informants rated them as higher in Agreeableness, Conscientiousness, and Openness, were positively associated with higher inter-dyad agree- ment on the normal range trait measure. These results were in line with previous findings that pro-social traits are easier to judge in well-adjusted individuals (Colvin, 1993; Connolly et al., 2007). Strengths, Limitations, and Future Directions The use of both self- and informant-ratings was a major strength to the study. Self-reports only showed a fraction of the whole picture of the relationship between level of personality pathology and self-other agreement. Informant-reports of personality pathol- ogy yielded more significant correlations and unique patterns of correlations in relation to inter-dyad agree- ment. The Realistic Accuracy Model and the Self-Other Asymmetries Model alluded to differences in reception and interpretation of information by a judge on a given target. The differences in the pattern of results for the correlation matrices of the self- and informant-ratings reveals the importance of questioning accuracy, and that convergence for judgments of others may not be obtainable for someone higher in personality pathology. From the findings of the current study, this statement is of course only true for normal range traits. Indeed, self- and informant-judgments of maladaptive criteria did reach better consensus. One limitation to the current study would be using Pearson’s correlations to calculate self-other agree- ment. The literature has used various other methods for calculating self-other agreement. Pearson’s correlation is considered a simpler, yet valid way of measuring self- other agreement (McCrae, 2008). However, a Pearson’s correlation would not capture self-other agreement well if an informant consistently rated a target as many points above, or below, the target ratings. The same limi- tation has the potential to control for rating tendencies among pairs. For example, if one an informant consis- tently rated one point lower than the target for every item on a measure, the self-other agreement correlation would remain unchanged. The current study had a relatively small sample size that varied depending on different measures. This was a result of the large size of the initial dataset; additionally, the measures used in the current study were often miss- ing data. During data collection, which lasted around four hours, sections were often skipped in the interest of saving time. The missing data was detrimental in the analysis due to lowered power. The sample size for BFI inter-dyad agreement was the lowest, at N = 57. Several correlations had approached significance but did not cross that threshold. Another major limitation to the current study was the use of self- and informant-ratings for calculat- ing inter-dyad agreement correlations, and using the same ratings as criteria to correlate with the inter-dyad agreement correlations. This method was used for the PID-5and BFI. However, ratings of the self and infor- mants for PD criteria used the SCID-II, which involved an interview with a clinician, and scores were strictly PD symptoms from the DSM-IV as determined by the clini- cian. The SCID-II’s inter-dyad agreement counterpart was the SCID-II PQ, which utilized a questionnaire that did not involve an interview. The analysis was done this way presuming that the SCID-II PQ would better reflect the judgments on the self and the informants. Additionally, we inferred that the SCID-II at face value, provided a more valid judgment for the presence of PD symptoms. In the current study, we did not assess inter-dyad agreement of individual traits. We only calculated profile agreement on whole measures. This may have limited the significance of the results for BFI agreement, since this measure included prosocial traits and a maladaptive trait. In general, more significant relationships could be revealed in future studies by expanding the analysis to include agreement on individual traits. This was not done in the current study as to not exponentially multi- ply the volume of analysis. 65 SELF-OTHER AGREEMENT IN PERSONALITY PATHOLOGY ASSESSMENT A study using a non-clinical sample will be the next step in this line of research. Due to the unexpected finding that higher self-ratings of personality pathology resulted in higher agreement on a measure of normal range traits, using a non-clinical sample may produce results more consistent with previous results in the literature. It would be interesting to investigate whether a non-clinical sample would yield self- and informant- results in the same pattern as the current study. Conclusion In this study, we found self- and informant-reports of personality pathology had a positive correlation with self-other agreement of maladaptive criteria. Self-reports of higher personality pathology scores unexpectedly correlated with higher agreement for normal range traits. Informant-reports of personality pathology, consistent with the literature, yielded a nega- tive correlation with self-other agreement on normal range traits. Self- and informant-reports are unique and are likely both picking up on valuable information, as opposed to varying degrees of error depending on circumstances surrounding those judgments. The next steps would be to expand the analysis to include inter- dyad agreement correlations of traits and PD criteria and to investigate this research question in a non-clinical sample using the same methods. The ways in which self-other agreement varies based on the rater and the information being rated reminds us of the subjectivity of self- and informant-reports. These differences highlight the value and importance of obtaining multiple viewpoints when assessing popula- tions that may have limited insight. References Biesanz, J. C. (2010). The social accuracy model of interpersonal perception: Assessing individual differences in perceptive and expressive accuracy. Multivariate Behavioral Research, 45(5), 853–885. doi: 10.1080/00273171.2010.519262 Carlson, E. N., Vazire, S., & Oltmanns, T. F. (2013). Self-other knowledge asymmetries in personality pathology. Journal of Personality, 81(2), 155–170. doi:10.1111/j.1467-6494.2012.00794.x Clifton, A., Turkheimer, E., & Oltmanns, T. F. (2004). Contrasting perspectives on personality problems: descriptions from the self and others. Personality and Individual Differences, 36(7), 1499–1514. doi:10.1016/j. paid.2003.06.002 Colvin, C. R. (1993). ‘Judgable’ people: Personality, behavior, and competing explanations. Journal of Personality and Social Psychology, 64(5), 861–873. doi:10.1037/0022-3514.64.5.861 Connelly, B. S., & Ones, D. S. (2010). Another perspective on personality: Meta-analytic integration of observers’ accuracy and predictive validity. Psychological Bulletin, 136(6), 1092–1122. doi:10.1037/ a002121210.1037/a0021212.supp Connolly, J. J., Kavanagh, E. J., & Viswesvaran, C. (2007). The Convergent Validity between Self and Observer Ratings of Personality: A meta-analytic review. International Journal of Selection and Assessment, 15(1), 110–117. doi:10.1111/j.1468-2389.2007.00371.x Funder, D. C. (1993). Judgments as data for personality and developmental psychology: Error versus accuracy. In D. C. Funder, R. D. Parke, C. Tomlinson-Keasey, K. Widaman, D. C. Funder, R. D. Parke, C. Tomlinson-Keasey, & K. Widaman (Eds.), Studying lives through time: Personality and development. (pp. 121–146). Washington, DC, US: American Psychological Association. Funder, D. C. (1995). On the accuracy of personality judgment: A realistic approach. Psychological Review, 102(4), 652–670. doi:10.1037/0033-295X.102.4.652 Funder, D. C., & Dobroth, K. M. (1987). Differences between traits: Properties associated with interjudge agreement. Journal of Personality and Social Psychology, 52(2), 409–418. doi:10.1037/0022-3514.52.2.409 Furr, R. M., Dougherty, D. M., Marsh, D. M., & Mathias, C. W. (2007). Personality judgment and personality pathology: Self-other agreement in adolescents with conduct disorder. Journal of Personality, 75(3), 629–662. doi:10.1111/j.1467-6494.2007.00451.x First, M. B., Spitzer, R. L., Gibbon, M., & Williams, J. W. (1995). The Structured Clinical Interview for DSM-III-R Personality Disorders (SCID-II): I. Description. Journal of Personality Disorders, 9(2), 83–91. doi:10.1521/pedi.1995.9.2.83 Greenwald, A. G. (1980). The totalitarian ego: Fabrication and revision of personal history. American Psychologist, 35(7), 603–618. doi:10.1037/0003-066X.35.7.603 66 TOM, SIMMS Hofstee, W. K. B. (1994). Who should own the definition of personality? European Journal of Personality, 8(3), 149–162. doi:10.1002/per.2410080302 Holzbach, R. L. (1978). Rater bias in performance ratings: Superior, self-, and peer ratings. Journal of Applied Psychology, 63(5), 579–588. doi:10.1037/0021-9010.63.5.579 Human, L. J., & Biesanz, J. C. (2011). Through the looking glass clearly: Accuracy and assumed similarity in well-adjusted individuals’ first impressions. Journal of Personality and Social Psychology, 100(2), 349–364. doi:10.1037/a0021850 Human, L. J., Biesanz, J. C., Finseth, S. M., Pierce, B., & Le, M. (2014). To thine own self be true: Psychological adjustment promotes judgeability via personality–behavior congruence. Journal of Personality and Social Psychology, 106(2), 286–303. doi:10.1037/a0034860 John, O. P., Donahue, E. M., & Kentle, R. L. (1991). The big five inventory—versions 4a and 54. John, O. P., & Robins, R. W. (1993). Determinants of Interjudge Agreement on Personality Traits: The Big Five Domains, Observability, Evaluativeness, and the Unique Perspective of the Self. Journal of Personality, 61(4), 521–551. doi:10.1111/j.1467-6494.1993. tb00781.x John, O. P., & Robins, R. W. (1994). Accuracy and bias in self-perception: Individual differences in self- enhancement and the role of narcissism. Journal of Personality and Social Psychology, 66(1), 206–219. doi:10.1037/0022-3514.66.1.206 Kenrick, D. T., & Stringfield, D. O. (1980). Personality traits and the eye of the beholder: Crossing some traditional philosophical boundaries in the search for consistency in all of the people. Psychological Review, 87(1), 88–104. doi:10.1037/0033-295X.87.1.88 Klonsky, E. D., Oltmanns, T. F., & Turkheimer, E. (2002). Informant-reports of personality disorder: Relation to self-reports and future research directions. Clinical Psychology: Science and Practice, 9(3), 300–311. doi:10.1093/clipsy/9.3.300 Kolar, D. W., Funder, D. C., & Colvin, C. R. (1996). Comparing the accuracy of personality judgments by the self and knowledgeable others. Journal of Personality, 64(2), 311–337. doi:10.1111/j.1467-6494.1996.tb00513.x 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(9), 1879–1890. doi:10.1017/S0033291711002674 Kurtz, J. E., & Sherker, J. L. (2003). Relationship quality, trait similarity, and self-other agreement on personality ratings in college roommates. Journal of personality, 71(1), 21–48. doi:10.1111/1467-6494. t01-1-00005 Leising, D., Rehbein, D., & Sporberg, D. (2006). Does a fish see the water in which it swims? A study of the ability to correctly judge one’s own interpersonal behavior. Journal of Social and Clinical Psychology, 25(9), 963–974. doi:10.1521/jscp.2006.25.9.963 Lynam, D. R., & Widiger, T. A. (2001). Using the Five- Factor Model to represent the DSM-IV personality disorders: An expert consensus approach. Journal of Abnormal Psychology, 110(3), 401–412. doi:10.1037/0021-843X.110.3.401 McCrae, R. R. (2008). A note on some measures of profile agreement. Journal of Personality Assessment, 90(2), 105–109. doi:10.1080/00223890701845104 Markon, K. E., Quilty, L. C., Bagby, R. M., & Krueger, R. F. (2013). The development and psychometric properties of an informant-report form of the Personality Inventory for DSM-5 (PID-5). Assessment, 20(3), 370–383. doi:10.1177/1073191113486513 Miller, J. D., Lynam, D. R., Widiger, T. A., & Leukefeld, C. (2001). Personality disorders as extreme variants of common personality dimensions: can the Five- Factor Model adequately represent psychopathy?. Journal of personality, 69(2), 253–276. Mosterman, R. M., & Hendriks, A. A. J. (2011). Self– other disagreement in personality assessment: Significance and prognostic value. Clinical Psychology & Psychotherapy, 18(2), 159–171. doi:10.1002/cpp.708 Oltmanns, T. F., & Turkheimer, E. (2009). Person perception and personality pathology. Current Directions in Psychological Science, 18(1), 32–36. doi:10.1111/j.1467-8721.2009.01601.x Paulhus, D. L., & John, O. P. (1998). Egoistic and moralistic biases in self-perception: The interplay of self-deceptive styles with basic traits and motives. Journal of Personality, 66(6), 1025–1060. doi:10.1111/1467-6494.00041 67 SELF-OTHER AGREEMENT IN PERSONALITY PATHOLOGY ASSESSMENT Ready, R. E., & Clark, L. A. (2002). Correspondence of psychiatric patient and informant ratings of personality traits, temperament, and interpersonal problems. Psychological Assessment, 14(1), 39–49. doi:10.1037/1040-3590.14.1.39 Ready, R. E., Clark, L. A., Watson, D., & Westerhouse, K. (2000). Self- and peer-related personality: Agreement, trait ratability, and the ‘self-based heuristic.’ Journal of Research in Personality, 34(2), 208–224. doi:10.1006/ jrpe.1999.2280 Ready, R. E., Watson, D., & Clark, L. A. (2002). Psychiatric patient- and informant-reported personality: Predicting concurrent and future behavior. Assessment, 9(4), 361–372. doi:10.1177/1073191102238157 Robins, R. W., & Beer, J. S. (2001). Positive illusions about the self: short-term benefits and long-term costs. Journal of personality and social psychology, 80(2), 340. doi:10.1037/0022-3514.80.2.340 Simms, L. J., Goldberg, L. R., Roberts, J. E., Watson, D., Welte, J., & Rotterman, J. H. (2011). Computerized adaptive assessment of personality disorder: Introducing the CAT–PD project. Journal of Personality Assessment, 93(4), 380–389. doi:10.1080/ 00223891.2011.577475 South, S. C., Oltmanns, T. F., Johnson, J., & Turkheimer, E. (2011). Level of agreement between self and spouse in the assessment of personality pathology. Assessment, 18(2), 217–226. doi:10.1177/1073191110394772 Spain, J. S., Eaton, L. G., & Funder, D. C. (2000). Perspectives on personality: The relative accuracy of self versus others for the prediction of emotion and behavior. Journal of Personality, 68(5), 837–867. doi:10.1111/1467-6494.00118 Tackett, J. (2011). Parent informants for child personality: Agreement, discrepancies, and clinical utility. Journal of Personality Assessment, 93(6), 539–544. doi:10.108 0/00223891.2011.608763 Tandler, N., Mosch, A., Wolf, A., & Borkenau, P. (2015). Effects of Personality Disorders on Self- Other Agreement and Favorableness in Personality Descriptions. Journal of Personality Disorders, 1–18. doi:10.1521/pedi_2015_29_213 Vazire, S. (2010). Who knows what about a person? The self–other knowledge asymmetry (SOKA) model. Journal of Personality and Social Psychology, 98(2), 281–300. doi:10.1037/a0017908 Vazire, S., & Mehl, M. R. (2008). Knowing me, knowing you: The accuracy and unique predictive validity of self-ratings and other-ratings of daily behavior. Journal of Personality and Social Psychology, 95(5), 1202– 1216. doi:10.1037/a0013314