Where the Narcissistic Mind Wanders: Increased Self-Related Thoughts are More Positive and Future Oriented

Journal of Personality Disorders, Volume 30, 1-24, 2016 © 2016 The Guilford Press KANSKE ET AL. WHERE THE NARCISSISTIC MIND WANDERS BRIEF REPORT WHERE THE NARCISSISTIC MIND WANDERS: INCREASED SELF-RELATED THOUGHTS ARE MORE POSITIVE AND FUTURE ORIENTED Philipp Kanske, PhD, Marjan Sharifi, MSc, Jonathan Smallwood, PhD, Isabel Dziobek, PhD, and Tania Singer, PhD Narcissism is characterized by a preoccupation with fantasies of unlimited success, power, beauty, and similar values, which has been discussed as intra-individual regulation of a grandiose, but vulnerable self-concept. To explore where the narcissistic mind wanders, we used an experience-sampling approach in a sample with large variability in pathological narcissism inventory scores. Multilevel modeling revealed (1) more mind-wandering in participants with higher levels of narcissism and (2) a difference in the content of these thoughts (more self- and other-related, past and future oriented, negative content). Critically, (3) in high levels of narcissism, the selfrelated thoughts were associated with more positive valence and were also more future oriented. The results demonstrate the validity of the assumed grandiose, self-absorbed view of oneself in narcissism, which includes selfindulgent fantasies of future success. We also found additional evidence for negative, past-oriented thoughts in narcissism, a dysfunctional pattern reminiscent of rumination, possibly linked to increased psychopathological vulnerability in narcissism. Narcissism1 is characterized by a pervasive pattern of grandiosity, self-focus, and self-importance (American Psychiatric Association, 2013). Diagnostic criteria for narcissistic personality disorder include a preoccupation with fantasies of unlimited success, power, beauty, and brilliance. Interpersonally, narcissists demand admiration, display a sense of entitlement, are exploit1. Even though narcissism is extensively researched, a consensus regarding an exact definition of the construct has not yet been reached. It is also important to note that, in its extreme form, narcissism can manifest as narcissistic personality disorder, but subclinical variations have also been linked to significant maladjustment, including aggression, delinquency, depression, or addiction. For the present manuscript, we have adopted such a dimensional perspective and use the term narcissism and its variants to refer to (subclinical) variations in both the grandiose and more vulnerable aspects of the construct. From Department of Social Neuroscience, Max Planck Institute for Human Cognitive and Brain Sciences, Leipzig, Germany (P. K., M. S., T. S.); Berlin School of Mind and Brain, Humboldt-Universität zu Berlin, Berlin, Germany (M. S., I. D.); and Department of Psychology, University of York, York, UK (J. S.). The first two authors contributed equally. Address correspondence to Philipp Kanske, Department of Social Neuroscience, Max Planck Institute for Human Cognitive and Brain Sciences, Stephanstraße 1, 04103 Leipzig, Germany. E-mail: kanske@cbs. mpg.de 1 2 KANSKE ET AL. ative, and are perceived as arrogant and aloof. Narcissism has also been shown to increase the risk for mental health problems, including depression, anxiety, and drug addiction (Stinson et al., 2008). Although subclinical narcissism is already higher in Western compared to non-Western societies (Foster, Campbell, & Twenge, 2003), it is still continuously rising among Western youths (Twenge, Konrath, Foster, Campbell, & Bushman, 2008, but also see Trzesniewski, Donnellan, & Robins, 2008). Better understanding of the psychological mechanisms of narcissism is therefore critical. Theoretical accounts have explained narcissism in terms of self-constructive processes (Campbell, 1999; Morf & Rhodewalt, 2001; Robins, Tracy, & Shaver, 2001). The grandiose yet vulnerable self-concept requires constant regulation, and fantasies of success or demanding admiration serve as such intra- and interpersonal self-regulatory strategies. It has been noted that, because the self is so highly inflated and overly positive, it is impossible to confirm it in the long run, which explains why narcissists seem relentlessly driven by their efforts to convince themselves and others that they are special (Morf, Torchetti, & Schürch, 2011). Empirically, there is support for a grandiose, but vulnerable self-concept in narcissism. In an early experience sampling study, for example, subclinical narcissism was related to greater reactivity in self-esteem depending on the valence of real-life social encounters, suggesting increased vulnerability (Rhodewalt, Madrian, & Cheney, 1998). Narcissists also report more favorable views of themselves in questionnaires (Campbell, Rudich, & Sedikides, 2002) and an increased implicit self-esteem (Vater et al., 2013). There is also evidence for the use of self-boosting interpersonal strategies to regulate the grandiose and vulnerable self-concept. For instance in romantic attraction, high-scoring narcissists, in comparison to non-narcissists, prefer admiring to caring others as potential partners (Campbell, 1999). They present themselves in self-enhancing ways, even when this collides with social norms (Collins & Stukas, 2008). And subclinical narcissists also show increased self-handicapping behavior (e.g., choosing to take an exam with distracting music rather than in silence), because this allows them to preserve a self-aggrandizing attribution style when confronted with others’ feedback (Rhodewalt, Tragakis, & Finnerty, 2006). Little is known, however, about intrapersonal regulation strategies such as self-generated thought patterns. Specifically, narcissistic individuals may engage in positive fantasies of themselves, for example, of their future success. A possible reason for the lack of studies may be methodological challenges in accessing these inner mental states. However, recent experimental approaches to induce and probe the amount and content of individuals’ selfgenerated thoughts allow such measurements (often investigated under the term mind-wandering; Smallwood & Schooler, 2015). These studies demonstrated the high prevalence of mind-wandering in the general population, with up to 50% of waking time (Kane et al., 2007; Killingsworth & Gilbert, 2010), and a link to negative mental health outcomes (Killingsworth & Gilbert, 2010; Smallwood & Schooler, 2015). The question about the content of self-generated thoughts, that is, where the mind wanders, has only recently been addressed (Ruby, Smallwood, Engen, & Singer, 2013; Ruby, WHERE THE NARCISSISTIC MIND WANDERS 3 Smallwood, Sackur, & Singer, 2013; Smallwood, Ruby, & Singer, 2013). These studies showed that thoughts can be characterized along a social (selfor other-related), temporal (past or future oriented), and valence dimension (negative or positive). Interestingly, the specific thought patterns have predictive power, for example, in predicting how well people can deal with stressors (Engert, Smallwood, & Singer, 2014). However, the relation of mindwandering to narcissism has not yet been investigated. Therefore, the current study investigated the relation of mind-wandering to narcissism as measured with the Pathological Narcissism Inventory (Pincus et al., 2009). We applied an established paradigm that induces self-generated thought because of its low demands (simple choice reaction time task, CRT) and tests the amount and specific content of self-generated thoughts via probe questions that are asked at random time-points during the task. More specifically, subjects were asked to rate the types of thoughts they had on six dimensions (Ruby, Smallwood, Engen et al., 2013; Smallwood et al., 2013): (1) other-related, (2) self-related (3) negatively valenced, (4) positively valenced, (5) past oriented, and (6) future oriented. Considering the preoccupation with fantasies of success, we expected increased mind-wandering in narcissism, with specific alterations in the content of self-generated thoughts. In particular, we hypothesized to find increased self-related thoughts that should be colored positively and be future oriented in individuals with high compared to low narcissistic traits. If alterations in mind-wandering truly represent currently occurring selfgenerated thoughts, in contrast to general response tendencies, then the differences between individuals with high and low levels of narcissism should vanish with reduced mind-wandering. We therefore tested a second paradigm that has been shown not to induce self-generated thoughts because of increased task demands (working memory task, WM) in a subgroup of the participants. Here, we expected reduced reports of mind-wandering in the probe questions and no relation to narcissism. METHODS PARTICIPANTS Participant recruitment was accomplished via flyers circulated through social media, Berlin University mailing lists, and in the central Berlin region. According to a short screening asking participants whether they had at some point been diagnosed with a mental disorder and whether they had been in treatment for a mental disorder (psychotropic medication, psychotherapy), only participants who reported no previous disorders or treatment were invited into the study. The final sample included 135 native German-speaking participants (89 female, mean age = 30.3 years, SD = 10.7 years), 8 without a high school degree, 3 who had completed an apprenticeship, 66 with a high school degree and currently enrolled in college, and 58 with a college degree. One group of participants completed the CRT task only (n = 75, 50 female, mean age 4 KANSKE ET AL. = 29.4 years, SD =10.3 years); a second group completed both the CRT and the WM task (n = 60, 39 female, mean age = 31.5 years, SD = 11.1 years). The study was approved by the Ethics Commission of the Department of Psychology of the Humboldt University of Berlin. Participants gave written informed consent prior to participation and received 7 euros per hour. GENERAL PROCEDURE AND DATA ACQUISITION Pathological Narcissism Inventory (PNI). All participants completed the Pathological Narcissism Inventory (Pincus et al., 2009; German version, Morf et al., 2016). The PNI is a 54-item self-report measure (52 items in the English version) showing good psychometric properties with internal consistency > .90 for the total score (and > .70 for the subscales; Maxwell, Donnellan, Hopwood, & Ackerman, 2011; Morf et al., 2016; Pincus et al., 2009) and excellent test-retest reliability of .84 for the total score (and > .75 for the subscales; Morf et al., 2016). The factor-analytically validated structure of the PNI comprises narcissistic grandiosity (exploitativeness, grandiose fantasy, self-sacrificing self-enhancement) and narcissistic vulnerability (contingent self-esteem, hiding the self, devaluing, entitlement rage). Thus, it represents a comprehensive measure of narcissism, spanning both grandiose and vulnerable elements. The PNI has been externally validated through its relation to interpersonal problems and in clinical samples (Morf et al., 2016; Pincus et al., 2009). In the present study, the final sample had a mean PNI score of M = 117.4 (SD = 39.1). Internal consistency was good (Cronbach’s alpha of .96 for the total score and .86 and .95 for the grandiosity and vulnerability subscales). The PNI score was normally distributed according to skewness (–0.259, SE = 0.199), kurtosis (0.108, SE = 0.395), and the Shapiro-Wilk test of normality (S-W = 0.984, df = 149, p > .05). The range was relatively large, with individual means of 20 to 219 (out of 270), which includes the means in German community and inpatient clinical samples (99.4 and 122.0, respectively; Morf et al., 2016). The vulnerability and grandiosity subscales were correlated (r = .72, p < 0.001). PNI scores were not different between men and women (t(135) = 0.217, p > 0.1), but were correlated with age (r = –.219, p = 0.015). Choice Reaction Time Task (CRT). The CRT task has been used routinely in mind-wandering studies (e.g., Baird et al., 2012; Ruby, Smallwood, Sackur, et al., 2013; Smallwood et al., 2013). A sequence of frequent black (~86%) and infrequent colored digits (~14%) is presented. The participants’ task is to indicate for each colored digit whether it is odd or even; black digits required no response. Black digits were presented for 1000 ms and colored digits for 2000 ms. Button press responses had to be made while the colored digits were still present on screen, otherwise the trial was considered a miss. Stimuli were separated by a fixation cross of variable duration (2200–4400 ms). Thought probes were presented at randomly selected time-points during the task. Each probe included 9 different questions to be answered on 9-point Likert scales (see Christoff, Gordon, Smallwood, Smith, & Schooler, 2009; WHERE THE NARCISSISTIC MIND WANDERS 5 Mrazek et al., 2012, for previous examples of this method). Participants were asked to rate how much their thoughts were (1) related or unrelated to the task to measure the amount of mind-wandering/off-task thoughts; how much their thoughts were (2) other-related, (3) self-related (4) negatively valenced, (5) positively valenced, (6) past oriented, and (7) future oriented to measure the content of their thoughts; and how much their current mood was (8) positive and (9) negative (see Ruby, Smallwood, Sackur et al., 2013, for examples). The number of probes and their occurrence were randomly determined. Responses were given on a keyboard and question presentation was self-paced. Working Memory Task (WM). The WM task was identical to the CRT, but included the periodic presentation of colored question marks, which then required indication of whether the previous number had been odd or even (question marks were only presented after black digits, because colored digits were already responded to immediately; Smallwood et al., 2013). Like the colored digits, question marks were presented for 2000 ms. Stimuli that required a response (colored digits and question marks) made up ~14% of the trials. The same thought probes used in the CRT were also presented in the WM task. For participants who completed only the CRT task, the task lasted 14 minutes, and the average number of probes was M = 6.39, SE = 0.2, Min = 3, Max = 10. For participants who completed both the CRT and the WM task, each task lasted 7 minutes, and the average number of thought probes was M = 3.18, SE = 0.1, Min = 1, Max = 6 for the CRT task and M = 3.33, SE = .2, Min = 1, Max = 7 for the WM task. The number of probes did not correlate with the PNI scores in any of the groups (all p > .30). Stimuli were presented using E-Prime 2.0 (Schneider, Eschman, & Zuccolotto, 2002). DATA ANALYSIS All statistical analyses were performed using SPSS (IBM SPSS Statistics for Windows, Version 22.0). For analysis of the CRT task data, we included all participants; for analysis of the WM task data, we included those participants who performed this task. For the performance analysis, participants with extreme scores in the CRT or WM task were excluded, in line with previous studies (n = 8 for the participants who did CRT and WM, n = 11 for the complete sample; scores were considered extreme when higher than [Q3 + 1.5 × IQR] or lower than [Q1 – 1.5 × IQR], with Q1 and Q3 the first and third quartiles, and IQR being the interquartile range; see Ruby, Smallwood, Engen, et al., 2013). To test for congruency with previous reports on the contents of mindwandering in healthy participants, we calculated a principal components analysis on the CRT task data (as described in Ruby, Smallwood, Engen, et al., 2013), which yielded a three components solution that conformed to those reported in the literature (see Table S1). For the main analyses, we used multilevel models because they take correlated observations within individuals into account and perform well with 6 KANSKE ET AL. missing data or unequal numbers of data points within individuals (Jahng, Wood, & Trull, 2008). The raw data for all the thought probes and mood ratings were normally distributed and analyzed with linear mixed models. A first set of models (model 1) tested for associations of narcissism with the thought probe ratings. The ratings on each dimension (i.e., how other-related, self-related, negatively valenced, positively valenced, past oriented, future oriented, and off-task the thoughts were, as well as how positive and how negative the mood was) were entered as dependent variable to be predicted in separate models. The main predictor was the PNI total score. Significant effects of PNI would indicate that narcissism is, for instance, positively associated with the level of how self-related the thoughts were. Additional covariates include the number of the particular sampling point within the session (e.g., sample count 5, indicating that this rating value was obtained during the fifth probe for this participant) to control for changes due to the repetition of the rating questions, and order of the specific task (e.g., whether the CRT was done first or second, that is, after the WM task) to control for order effects of the task sequence. A second set of models (model 2) tested for the association of how selfrelated a thought was with how positive, negative, future oriented, or past oriented it was. To check the specificity of the results, we repeated the analysis testing for how other-related thoughts were associated with the remaining ratings. For instance, we tested whether the degree to which thoughts were self-related was associated with how positive the thoughts were. To this end, self and other ratings were entered as dependent variable in separate models. Again, the PNI total score was used as a predictor, but the ratings on how positive, negative, future or past oriented the thoughts were, were entered as additional predictors. Crucially, the interaction of the PNI total score with the additional predictors was included in order to test whether the association of, for example, how self-related and how positive a thought might be was stronger/weaker in participants scoring high in narcissism. All models included a random intercept and random effects for sample count and order. Maximum likelihood (ML) was used as estimation method. Compared to the respective null models (excluding the predictors), all of these models showed significantly increased fit (all χ2(5) > 30, p < .01). To compare the amount of off-task thoughts between the CRT and WM tasks, we included only those participants that did both tasks in a multilevel model, identical to the ones described above, but with task as a predictor. For visualization of the results, we divided the sample according to a median split of the PNI total score (median = 123). RESULTS PERFORMANCE For the complete sample, the average error rate for the CRT task was 0.075% (SE = 0.009) and the average reaction time was 813.7 ms (SE = 20.3). For participants who did both the CRT and WM tasks, the average error rate for the CRT task was higher than that of the WM task, CRT: 0.070% (SE = WHERE THE NARCISSISTIC MIND WANDERS 7 0.011), WM: 0.028% (SE = 0.008); paired sample t-test, t = 3.237, p < .01. Numerically, the average reaction time for the CRT was also higher than for the WM task. However, this difference did not reach significance, CRT: 803.5 ms (SE = 20.93); WM: 766.4 ms (SE = 35.85); t = 1.253, p = 0.2. In line with previous reports (Baird et al., 2012; Ruby, Smallwood, Sackur, et al., 2013), the performance data suggest that participants were more focused during the WM than during the CRT task, which is substantiated by higher off-task thought ratings during the CRT than during the WM task (b = 12.9, SE = 2.3, p < 0.001). There was no correlation between narcissism and RTs or error rates in the CRT or WM task (both p > 0.1). CHOICE REACTION TIME TASK Thought Probes. Multilevel models revealed a positive relation of narcissism to off-task thoughts (b = 0.137, SE = 0.053, p < 0.01), indicating more mindwandering in narcissistic individuals (Figure 1; Table S2). Interestingly, narcissists also differed in the content of their self-generated thoughts. Higher narcissism scores were associated with more self-related (b = 00.125, SE = 00.046, p < 0.01) and other-related thoughts (b = 0.114, SE = 0.046, p < 0.05), as well as more past-oriented (b = 0.129, SE = 00.041, p < 0.01) and future-oriented thoughts (b = 00.114, SE = 00.048, p < 0.005). Regarding thought valence, narcissism was associated with higher levels of negative thoughts (b = 00.124, SE = 0.040, p < 0.01) and marginally with lower levels of positive thoughts (b = –0.062, SE = 0.038, p < 0.1). Characterizing Self- and Other-Related Thoughts. Given the hypothesis that narcissists’ preoccupation with fantasies of success should color their self-related thoughts, we tested whether the degree to which thoughts were self-related covaried with the valence of the thoughts (negative, positive) and their temporal dimension (past, future) (Figure 2; Table S3). The results revealed significant interactions of narcissism with how negative (b = –0.003, SE = 0.001, p < 0.05), positive (b = 0.005, SE = 0.002, p < 0.01), past-oriented (b = 0.004, SE = 0.001, p < 0.001), and future-oriented (b = 0.003, SE = 0.001, p < 0.01) thoughts were in predicting self-related thoughts. The results show that narcissists’ self-related thoughts were more positive and less negative, as well as less past and more future oriented than in individuals with low levels of narcissism. Testing for the same relations with regard to other-related thoughts yielded no significant interactions with narcissism (Table S4). Mood Probes. Narcissism also had an effect on mood ratings, with a more negative mood (b = 0.142, SE = 0.042, p < 0.001) and a less positive mood in higher levels of narcissism (b = –0.127, SE = 0.037, p < 0.001) (Figure S1; Table S2). WORKING MEMORY TASK Thought Probes. Repeating the analyses described for the CRT for the WM task, we again found a positive association of narcissism and off-task 8 KANSKE ET AL. FIGURE 1. Level of reported off-task thoughts, other- and selfrelated, positive and negative, and future- and past-oriented thoughts in the Choice reaction time (A) and the Working memory task (B), separately for individuals scoring high and low in narcissism (median split) (rating scale numbers without unit varying between 0 and 100). *p < .05. #p < .10. thoughts (b = 0.118, SE = 0.052, p < .05), indicating more mind-wandering in narcissists, but no further significant effects (Figure 1; Table S5). Mood Probes. No significant effects were found (Figure S1 and Table S5). VULNERABILITY VERSUS GRANDIOSITY To further explore whether vulnerable and grandiose aspects of narcissism show distinct patterns of self-generated thoughts, we repeated all analyses using the PNI subscales on vulnerability and grandiosity instead of the PNI total score. The results showed the same patterns for the two subscales that also conform to the pattern observed for the PNI total score (see Tables S6–11). WHERE THE NARCISSISTIC MIND WANDERS 9 FIGURE 2. Scatterplot depicting the relation of the level of self-related thoughts to how positive (A), negative (B), future-oriented (C), and past-oriented (D) thoughts were, separately for individuals scoring high and low in narcissism (median split). Intra-individual standard errors are displayed in grey and model predictions from the multilevel model as lines in the respective shape. DISCUSSION The present study set out to test how narcissistic personality traits are associated with specific patterns of self-generated thoughts by applying an experience sampling method with multiple thought probes. The results yielded several new insights into the relation of narcissism and mind-wandering. First, we found increased mind-wandering in narcissism, which is in line with the posited preoccupation with fantasies (of future success, for example). Second, and also in line with the hypothesis that narcissists engage more in fantasies 10 KANSKE ET AL. of grandiosity, thoughts were indeed more self-related, and these self-related thoughts were more positive and future oriented in people with high than in people with low levels of narcissism. Third, the degree of narcissism was related to overall more negative, other-related, and past-oriented thoughts, as well as to increased off-task thoughts also in the high-demanding working memory task—both dysfunctional patterns that may tentatively be linked to the increased psychopathological vulnerability in narcissism. The present results conform with theoretical accounts of narcissism that propose a grandiose, but vulnerable self-concept to require constant regulation through intra- and interpersonal regulation strategies (Campbell, 1999; Morf & Rhodewalt, 2001; Robins et al., 2001). While there is ample evidence for interpersonal regulation through favorable self-presentation and demanding admiration (Collins & Stukas, 2008; Rhodewalt et al., 2006), our results corroborate the assumption of narcissists’ preoccupation with overly positive self-related thoughts that may serve as intrapersonal regulation strategy (Morf, Torchetti, et al., 2011). Further, self-related thoughts were more future oriented in narcissism. While the specific content of these thoughts is not clear, the pattern would be consistent with the interpretation of these thoughts as fantasies about future success, power, or beauty (cf. American Psychiatric Association, 2013). As we did not measure self- and other-related thoughts on a bipolar scale, but rather on separate unipolar scales, we could document some specificity of the association of self-related to positive and future-oriented thoughts. This pattern was not present with other-related thoughts. A remarkable feature of the present results is that higher degrees of narcissism are also associated with overall more negative, other-related and past- and future-oriented thoughts. Thus, the association to increased positively valenced and future-oriented thoughts in narcissism is specific to those thoughts that are about oneself. In addition, narcissism was also related to more off-task thoughts, both in the easy choice reaction time task and the more demanding working memory task. We did not observe a relationship between the degree of narcissism and performance in the two tasks, but previous studies have linked task-unrelated thinking to performance impairment (Smallwood, Baracaia, Lowe, & Obonsawin, 2003; Smallwood, Fishman, & Schooler, 2007), and excessive mind-wandering has also been linked to negative mental health outcome (Killingsworth & Gilbert, 2010; Smallwood & Schooler, 2015). The increased mind-wandering in narcissism, together with the more negative, past- or future-oriented thoughts, reminiscent of rumination and worry (Nolen-Hoeksema, Wisco, & Lyubomirsky, 2008), may therefore indicate a dysfunctional pattern, which may help to explain the high risk for psychopathology—even in subclinical narcissism (Thomaes & Brummelman, 2015). It is interesting that the grandiose and vulnerable aspects of narcissism as measured with the PNI do not show different self-generated thought patterns, but conform to the results of the PNI total score. While this may be due to the high correlation of the two subscales in our sample (r = 0.72 compared to previously reported r = 0.39–0.63; Bresin & Gordon, 2011; Glover, Miller, Lynam, Crego, & Widiger, 2012), it may also be an expression of the WHERE THE NARCISSISTIC MIND WANDERS 11 intricate nature of vulnerable and grandiose aspects in pathological narcissism. Hence, in line with self-constructive process models of narcissism, it may be the grandiose, yet vulnerable self that gives rise to ruminative and overly positive self-related thought patterns (Morf & Rhodewalt, 2001). The results allow a first characterization of self-generated thoughts in narcissism. Considering the high prevalence of mind-wandering between 25 and 50% of the waking time (Kane et al., 2007; Killingsworth & Gilbert, 2010), the question arises, however, what the consequences of these altered thought patterns might be. So far, little is known about the effects of specific thoughts, but a recent study could show that particularly negative, past- and other-oriented thoughts are associated with increased cortisol responses to psychosocial stress (Engert et al., 2014). What we occupy our minds with is thus consequential, and future studies should elucidate the concomitant behavioral and health-related outcomes of the specific narcissism-related thought patterns. In contrast to questionnaire studies, probing online thoughts during mind-wandering, as we did in the present study, gives a more realistic and ecologically valid characterization of narcissists’ thought patterns. It does not require participants to integrate previous experience themselves, which can induce biases, for instance due to differences in meta-cognitive capabilities. For some patient groups, it could already be shown that sampling from the actual momentary experience of participants characterizes affective dysregulation more realistically (Ebner-Priemer et al., 2007, 2015). The present study demonstrates the validity of using experience-sampling tasks as a tool to investigate associations with personality traits and possibly clinical groups. A limitation of the present study could be that we investigated narcissism in a subclinical sample, thus not as pathological narcissism as defined in DSM-5, for example (American Psychiatric Association, 2013). This raises the question of how well subclinical and clinical narcissism map onto each other. Interestingly, even subclinically, narcissism predicts significant maladjustment, including aggression, violence, delinquency, and psychopathology such as anxiety, depression, and addiction (Thomaes & Brummelman, 2015). Further, subclinical narcissism correlates substantially with interview assessments of narcissistic personality disorder; both have similar external correlations. There is no qualitative jump from normal to extreme narcissism, and dimensional approaches to the construct have therefore been advocated (Brummelman et al., 2015; Miller & Campbell, 2010). Second, the experience sampling method allows probing the actual current experience of the participants, which is however still subjective. Applying the proposed combinations of experience sampling with (neuro-)physiological measures (Ottaviani, Shapiro, Davydov, Goldstein, & Mills, 2009; Tusche, Smallwood, Bernhardt, & Singer, 2014) to the study of narcissism could provide additional validation of the results. Third, the rating procedure applied in the present study gives a general overview of the critical thought-space dimensions (Ruby, Smallwood, Engen, et al., 2013), but an interesting next step would be to ask for the specific content of current thoughts, such as future success, for instance. A fourth limitation concerns the ecologic valid- 12 KANSKE ET AL. ity of the present results. While measuring self-generated thoughts in a lab setting standardizes the environment, thereby eliminating different outside influences that may affect mind-wandering, future studies should directly relate lab data to probes that were sampled throughout the day in more natural settings. Last, we did not test for the specificity of the observed thought patterns to narcissism. In particular, as the vulnerability subscale of the PNI correlates with neuroticism (Miller et al., 2011), future studies should compare different (clinical) conditions to explore if they are associated with distinct thought patterns. As we assessed narcissism subclinically, any therapeutic implications are of course preliminary. Suggestions for future clinical research derived from the present results could relate to two aspects: First, because increased levels of mind-wandering, as observed in participants with higher narcissism scores, have been characterized as dysfunctional (Killingsworth & Gilbert, 2010), future studies may elucidate the effects of mindfulness interventions that have been shown to reduce mind-wandering (Mrazek, Franklin, Phillips, Baird, & Schooler, 2013) also in narcissism psychotherapy. Second, the increased self-related, more positive and future-oriented thoughts associated with narcissism could be interpreted as an intrapersonal regulation strategy that may be dysfunctional in the long run (e.g., it might lead to resistance to others’ critical feedback; Morf, Horvath, & Torchetti, 2011). It could, thus, be an interesting target for intervention. Taken together, the present study elucidated the amount and specific content of self-generated thoughts in narcissism, critically showing increased mind-wandering with negative, past-oriented thoughts, reminiscent of rumination, and with particularly positive and future-oriented thoughts about themselves. The results are in line with theoretical accounts that explain narcissism in terms of self-constructive processes (Morf & Rhodewalt, 2001), suggesting that engaging in overly positive self-related thoughts may be an intra-individual self-regulatory mechanism in narcissism. REFERENCES American Psychiatric Association. (2013). Diagnostic and statistical manual of mental disorders (5th ed.). Washington, DC: American Psychiatric Publishing. Baird, B., Smallwood, J., Mrazek, M. D., Kam, J. W., Franklin, M. S., & Schooler, J. W. (2012). Inspired by distraction: Mind wandering facilitates creative incubation. Psychological Science, 23(10), 1117–1122. Bresin, K., & Gordon, K. H. (2011). Characterizing pathological narcissism in terms of the HEXACO model of personality. Journal of Psychopathology and Behavioral Assessment, 33(2), 228–235. Brummelman, E., Thomaes, S., Nelemans, S. A., de Castro, B. O., Overbeek, G., & Bushman, B. J. (2015). Reply to Kealy et al.: Theoretical precision in the study of narcissism and its origins. Proceedings of the National Academy of Sciences of the United States of America, 112(23), E2987–E2987. Campbell, W. K. (1999). Narcissism and romantic attraction. Journal of Personality and Social Psychology, 77(6), 1254–1270. Campbell, W. K., Rudich, E. A., & Sedikides, C. (2002). Narcissism, self-esteem, and the positivity of self-views: Two portraits of self-love. Personality and Social Psychology Bulletin, 28(3), 358–368. Christoff, K., Gordon, A. M., Smallwood, J., Smith, R., & Schooler, J. W. (2009). Experience sampling during fMRI reveals default network and executive system contributions to mind wandering. Proceedings of the National Academy of Sciences of the United States of America, 106(21), 8719–8724. WHERE THE NARCISSISTIC MIND WANDERS Collins, D. R., & Stukas, A. A. (2008). Narcissism and self-presentation: The moderating effects of accountability and contingencies of self-worth. Journal of Research in Personality, 42(6), 1629–1634. Ebner-Priemer, U. W., Houben, M., Santangelo, P., Kleindienst, N., Tuerlinckx, F., Oravecz, Z., . . . Kuppens, P. (2015). Unraveling affective dysregulation in borderline personality disorder: A theoretical model and empirical evidence. Journal of Abnormal Psychology, 124(1), 186–198. Ebner-Priemer, U. W., Kuo, J., Kleindienst, N., Welch, S. S., Reisch, T., Reinhard, I., . . . Bohus, M. (2007). State affective instability in borderline personality disorder assessed by ambulatory monitoring. Psychological Medicine, 37(7), 961–970. Engert, V., Smallwood, J., & Singer, T. (2014). Mind your thoughts: Associations between self-generated thoughts and stress-induced and baseline levels of cortisol and alphaamylase. Biological Psychology, 103, 283– 291. Foster, J. D., Campbell, W. K., & Twenge, J. M. (2003). Individual differences in narcissism: Inflated self-views across the lifespan and around the world. Journal of Research in Personality, 37(6), 469–486. Glover, N., Miller, J. D., Lynam, D. R., Crego, C., & Widiger, T. A. (2012). The Five-Factor Narcissism Inventory: A five-factor measure of narcissistic personality traits. Journal of Personality Assessment, 94(5), 500–512. Jahng, S., Wood, P. K., & Trull, T. J. (2008). Analysis of affective instability in ecological momentary assessment: Indices using successive difference and group comparison via multilevel modeling. Psychological Methods, 13(4), 354–375. Kane, M. J., Brown, L. H., McVay, J. C., Silvia, P. J., Myin-Germeys, I., & Kwapil, T. R. (2007). For whom the mind wanders, and when: An experience-sampling study of working memory and executive control in daily life. Psychological Science, 18(7), 614–621. Killingsworth, M. A., & Gilbert, D. T. (2010). A wandering mind is an unhappy mind. Science, 330(6006), 932. Maxwell, K., Donnellan, M. B., Hopwood, C. J., & Ackerman, R. A. (2011). The two faces of Narcissus? An empirical comparison of the Narcissistic Personality Inventory and the Pathological Narcissism Inventory. Personality and Individual Differences, 50(5), 577–582. Miller, J. D., & Campbell, W. K. (2010). The case for using research on trait narcissism as a building block for understanding narcissistic personality disorder. Personality Disor- 13 ders: Theory Research and Treatment, 1(3), 180–191. Miller, J. D., Hoffman, B. J., Gaughan, E. T., Gentile, B., Maples, J., & Campbell, W. K. (2011). Grandiose and vulnerable narcissism: A nomological network analysis. Journal of Personality, 79(5), 1013–1042. Morf, C. C., Horvath, S., & Torchetti, L. (2011). Narcissistic self-enhancement: Tales of (successful?) self-portrayal. In M. D. Alicke & C. Sedikides (Eds.), Handbook of selfenhancement and self-protection (pp. 399– 424). New York, NY: Guilford. Morf, C. C., & Rhodewalt, F. (2001). Unraveling the paradoxes of narcissism: A dynamic self-regulatory processing model. Psychological Inquiry, 12(4), 177–196. Morf, C. C., Schürch, E., Küfner, A., Siegrist, P., Vater, A., Back, M., . . . Schröder-Abé, M. (2016). Expanding the nomological net of the Pathological Narcissism Inventory: German validation and extension in a clinical inpatient sample. Assessment. Advance online publication. doi:10.1177/1073191115627010 Morf, C. C., Torchetti, L., & Schürch, E. (2011). Narcissism from the perspective of the dynamic self-regulatory processing model. In W. K. Campbell & J. Miller (Eds.), The handbook of narcissism and narcissistic personality disorder: Theoretical approaches, empirical findings, and treatment (pp. 56–70). Hoboken, NJ: Wiley. Mrazek, M. D., Franklin, M. S., Phillips, D. T., Baird, B., & Schooler, J. W. (2013). Mindfulness training improves working memory capacity and GRE performance while reducing mind wandering. Psychological Science, 24(5), 776–781. Mrazek, M. D., Smallwood, J., Franklin, M. S., Chin, J. M., Baird, B., & Schooler, J. W. (2012). The role of mind-wandering in measurements of general aptitude. Journal of Experimental Psychology: General, 141(4), 788–798. Nolen-Hoeksema, S., Wisco, B. E., & Lyubomirsky, S. (2008). Rethinking rumination. Perspectives on Psychological Science, 3(5), 400–424. Ottaviani, C., Shapiro, D., Davydov, D. M., Goldstein, I. B., & Mills, P. J. (2009). The autonomic phenotype of rumination. International Journal of Psychophysiology, 72(3), 267–275. Pincus, A. L., Ansell, E. B., Pimentel, C. A., Cain, N. M., Wright, A. G. C., & Levy, K. N. (2009). Initial construction and validation of the Pathological Narcissism Inventory. Psychological Assessment, 21(3), 365–379. Rhodewalt, F., Madrian, J. C., & Cheney, S. (1998). Narcissism, self-knowledge organization, and emotional reactivity: The effect 14 of daily experiences on self-esteem and affect. Personality and Social Psychology Bulletin, 24(1), 75–87. Rhodewalt, F., Tragakis, M. W., & Finnerty, J. (2006). Narcissism and self-handicapping: Linking self-aggrandizement to behavior. Journal of Research in Personality, 40(5), 573–597. Robins, R. W., Tracy, J. L., & Shaver, P. R. (2001). Shamed into self-love: Dynamics, roots, and functions of narcissism. Psychological Inquiry, 12(4), 230–236. Ruby, F. J. M., Smallwood, J., Engen, H., & Singer, T. (2013). How self-generated thought shapes mood—The relation between mindwandering and mood depends on the sociotemporal content of thoughts. PLoS One, 8(10), e77554. Ruby, F. J. M., Smallwood, J., Sackur, J., & Singer, T. (2013). Is self-generated thought a means of social problem solving? Frontiers in Psychology, 4, 962. Schneider, W., Eschman, A., & Zuccolotto, A. (2002). E-Prime: User’s guide. Sharpsburg, PA: Psychology Software Inc. Smallwood, J., Baracaia, S. F., Lowe, M., & Obonsawin, M. (2003). Task unrelated thought whilst encoding information. Consciousness and Cognition, 12(3), 452–484. Smallwood, J., Fishman, D. J., & Schooler, J. W. (2007). Counting the cost of an absent mind: Mind wandering as an underrecognized influence on educational performance. Psychonomic Bulletin & Review, 14(2), 230–236. Smallwood, J., Ruby, F. J. M., & Singer, T. (2013). Letting go of the present: Mind-wandering is associated with reduced delay discount- KANSKE ET AL. ing. Consciousness and Cognition, 22(1), 1-7. Smallwood, J., & Schooler, J. W. (2015). The science of mind wandering: Empirically navigating the stream of consciousness. Annual Review of Psychology, 66, 487–518. Stinson, F. S., Dawson, D. A., Goldstein, R. B., Chou, S. P., Huang, B., Smith, S. M., . . . Grant, B. F. (2008). Prevalence, correlates, disability, and comorbidity of DSM-IV narcissistic personality disorder: Results from the wave 2 national epidemiologic survey on alcohol and related conditions. Journal of Clinical Psychiatry, 69(7), 1033–1045. Thomaes, S., & Brummelman, E. (2015). Narcissism. In D. Cicchetti (Ed.), Developmental psychopathology (3rd ed., vol. 3, pp. 679– 725). Hoboken, NJ: Wiley. Trzesniewski, K. H., Donnellan, M. B., & Robins, R. W. (2008). Is “Generation Me” really more narcissistic than previous generations? Journal of Personality, 76(4), 903–917. Tusche, A., Smallwood, J., Bernhardt, B. C., & Singer, T. (2014). Classifying the wandering mind: Revealing the affective content of thoughts during task-free rest periods. Neuroimage, 97, 107–116. Twenge, J. M., Konrath, S., Foster, J. D., Campbell, W. K., & Bushman, B. J. (2008). Egos inflating over time: A cross-temporal meta-analysis of the narcissistic personality inventory. Journal of Personality, 76(4), 875–901. Vater, A., Ritter, K., Schroder-Abe, M., Schutz, A., Lammers, C. H., Bosson, J. K., & Roepke, S. (2013). When grandiosity and vulnerability collide: Implicit and explicit self-esteem in patients with narcissistic personality dis- WHERE THE NARCISSISTIC MIND WANDERS 15 TABLE S1. Principal Components Analysis Showing the Component Loadings for Each Rating Question Off-Task Other Self Negative Positive Past Future F1 F2 F3 affect 0.299 –0.050 0.108 0.845 –0.913 –0.157 –0.006 past-other 0.593 0.798 0.170 0.192 –0.038 0.834 0.073 future-self 0.340 0.274 0.720 0.180 0.048 –0.072 0.863 Note. Three components were observed: the affect component positively weighted on positive ratings and negatively on negative ratings, the socio-temporal past-other component weighted positively on other and past ratings, and the socio-temporal future-self component weighted positively on self and future. TABLE S2. Relation of Narcissism to Ratings Levels in the Choice Reaction Time (CRT) Task as Estimated With Multilevel Modeling CRT Model 1 Parameters Off-Task Other Self Negative Positive Past Future Mood Negative Mood Positive Predictors Estimate SE p value Intercept PNI Total Sample Count Order Intercept PNI Total Sample Count Order Intercept PNI Total Sample Count Order Intercept PNI Total Sample Count Order Intercept PNI Total Sample Count Order Intercept PNI Total Sample Count Order Intercept PNI Total Sample Count Order Intercept PNI Total Sample Count Order Intercept PNI Total Sample Count Order 36.566 0.137 4.199 –10.866 43.945 0.114 1.487 –20.091 42.231 0.125 1.948 –13.030 31.5a39 0.124 1.492 –12.730 66.292 –0.062 –0.831 0.766 28.738 0.129 0.608 –15.683 41.821 0.114 1.226 –11.355 31.913 0.142 1.371 –11.948 75.311 –0.127 –1.330 –1.911 7.885 0.053 0.586 5.290 6.939 0.046 0.590 4.685 6.948 0.046 0.572 4.683 6.007 0.040 0.440 4.028 5.606 0.038 0.388 3.759 6.166 0.041 0.503 4.173 7.175 0.048 0.668 4.843 6.271 0.042 0.453 4.194 5.499 0.037 0.374 3.677 < 0.001 0.010 < 0.001 0.042 < 0.001 0.015 0.014 < 0.001 < 0.001 0.008 0.001 0.006 < 0.001 0.003 0.001 0.002 < 0.001 0.099 0.035 0.839 < 0.001 0.002 0.229 < 0.001 < 0.001 0.017 0.071 0.020 < 0.001 0.001 0.003 0.005 < 0.001 0.001 0.001 0.604 PNI: Pathological Narcissism Inventory. 16 KANSKE ET AL. TABLE S3. Association of Narcissism With the Relation of Self-Related Ratings to the Positivity, Negativity, Past Orientation, and Future Orientation of Thoughts as Estimated as Interactions in Multilevel Models in the CRT Task CRT Model 2 Parameters Self Predictors Estimate SE p value Intercept PNI Total Sample Count Order Negative PNI × Negative Intercept PNI Total Sample Count Order Positive PNI × Positive Intercept PNI Total Sample Count Order Past PNI × Past Intercept PNI Total Sample Count Order Future PNI × Future 20.317 0.206 1.558 –9.783 0.676 –0.003 78.191 –0.160 1.956 –13.432 –0.589 0.005 22.289 0.223 1.785 –9.249 0.647 –0.004 42.101 –0.034 1.523 –9.276 0.011 0.003 8.986 0.068 0.556 4.730 0.170 0.001 13.810 0.104 0.571 4.700 0.195 0.002 7.833 0.055 0.567 4.549 0.132 0.001 8.060 0.059 0.555 4.383 0.112 0.001 0.025 0.006 0.006 0.040 < 0.001 0.017 < 0.001 0.127 0.001 0.005 0.003 0.003 0.005 < 0.001 0.002 0.044 < 0.001 < 0.001 < 0.001 0.571 0.007 0.036 0.925 0.003 TABLE S4. Association of Narcissism With the Relation of Other-Related Ratings to the Positivity, Negativity, Past Orientation, and Future Orientation of Thoughts as Estimated as Interactions in Multilevel Models in the CRT Task CRT Model 2 Parameters Other Predictors Estimate SE p value Intercept PNI Total Sample Count Order Negative PNI × Negative Intercept PNI Total Sample Count Order Positive PNI × Positive Intercept PNI Total Sample Count Order Past PNI × Past Intercept PNI Total Sample Count Order Future PNI × Future 44.529 0.069 1.340 –18.732 –0.013 0.001 19.201 0.255 1.598 –20.010 0.397 –0.002 37.334 0.030 1.266 –14.555 0.230 0.001 32.473 0.095 1.185 –17.267 0.270 0.000 9.074 0.069 0.591 4.675 0.179 0.001 14.192 0.107 0.594 4.735 0.202 0.002 7.313 0.052 0.585 4.186 0.130 0.001 8.266 0.565 0.565 4.406 0.120 0.001 < 0.001 0.316 0.026 < 0.001 0.941 0.529 0.177 0.018 0.009 < 0.001 0.050 0.166 < 0.001 0.556 0.033 0.001 0.078 0.238 < 0.001 0.039 0.039 < 0.001 0.024 0.844 WHERE THE NARCISSISTIC MIND WANDERS 17 TABLE S5. Relation of Narcissism to Ratings Levels in the Working Memory (WM) Task as Estimated With Multilevel Modeling WM Model 1 Parameters Off-Task Other Self Negative Positive Past Future Mood Negative Mood Positive Predictors Estimate SE p value Intercept PNI Total Sample Count Order Intercept PNI Total Sample Count Order Intercept PNI Total Sample Count Order Intercept PNI Total Sample Count Order Intercept PNI Total Sample Count Order Intercept PNI Total Sample Count Order Intercept PNI Total Sample Count Order Intercept PNI Total Sample Count Order Intercept PNI Total Sample Count Order 18.252 0.118 5.345 2.071 52.214 –0.064 2.239 –1.976 44.069 0.012 1.977 0.978 31.198 0.089 1.936 –0.080 72.355 –0.083 –1.801 –1.783 45.892 –0.044 2.284 –3.458 36.247 0.073 2.405 2.474 39.339 0.049 0.528 –2.184 70.112 –0.074 –1.110 1.904476 6.872 0.052 1.001 4.038 8.587 0.065 1.318 5.073 8.110 0.061 1.183 4.784 7.067 0.053 0.851 4.169 6.843 0.052 0.672 4.050 7.871 0.060 1.210 4.648 8.240 0.062 1.418 4.877 7.369 0.056 0.809 4.361 5.759 0.044 0.564 3.41 0.010 0.027 < 0.001 0.610 < 0.001 0.327 0.094 0.698 < 0.001 0.850 0.105 0.839 < 0.001 0.102 0.024 0.985 < 0.001 0.114 0.008 0.661 < 0.001 0.462 0.065 0.460 < 0.001 0.251 0.095 0.614 < 0.001 0.382 0.518 0.618 < 0.001 0.095 0.051 0.578365 18 KANSKE ET AL. TABLE S6. Relation of the Vulnerability Subscale of the PNI to Ratings Levels in the CRT Task as Estimated With Multilevel Modeling CRT Model 1 Parameters Off-Task Other Self Negative Positive Past Future Mood Negative Mood Positive Predictors Estimate SE p value Intercept PNI Vuln Sample Count Order Intercept PNI Vuln Sample Count Order Intercept PNI Vuln Sample Count Order Intercept PNI Vuln Sample Count Order Intercept PNI Vuln Sample Count Order Intercept PNI Vuln Sample Count Order Intercept PNI Vuln Sample Count Order Intercept PNI Vuln Sample Count Order Intercept PNI Vuln Sample Count Order 38.476 0.195 4.198 –11.147 45.814 0.158 1.484 –20.326 43.656 0.182 1.944 –13.293 31.882 0.195 1.488 –13.004 66.756 –0.107 –0.832 0.908 30.050 0.191 0.604 –15.966 43.255 0.165 1.225 –11.596 32.750 0.218 1.367 –12.256 74.447 –0.193 –1.330 –1.644 7.195 0.073 0.586 5.283 6.342 0.064 0.589 4.685 6.335 0.064 0.571 4.670 5.435 0.055 0.439 3.988 5.097 0.052 0.388 3.740 5.605 0.056 0.500 4.150 6.550 0.066 0.668 4.835 5.677 0.058 0.453 4.156 4.982 0.051 0.374 3.646 < 0.001 0.008 < 0.001 0.036 < 0.001 0.014 0.014 < 0.001 < 0.001 0.005 0.001 0.005 < 0.001 0.001 0.001 0.001 < 0.001 0.041 0.035 0.808 < 0.001 0.001 0.230 < 0.001 < 0.001 0.013 0.071 0.018 < 0.001 < 0.001 0.003 0.004 < 0.001 < 0.001 0.001 0.653 WHERE THE NARCISSISTIC MIND WANDERS 19 TABLE S7. Relation of the Grandiosity Subscale of the PNI to Ratings Levels in the CRT Task as Estimated With Multilevel Modeling CRT Model 1 Parameters Off-Task Other Self Negative Positive Past Future Mood Negative Mood Positive Predictors Estimate SE p value Intercept PNI Grand Sample Count Order Intercept PNI Grand Sample Count Order Intercept PNI Grand Sample Count Order Intercept PNI Grand Sample Count Order Intercept PNI Grand Sample Count Order Intercept PNI Grand Sample Count Order Intercept PNI Grand Sample Count Order Intercept PNI Grand Sample Count Order Intercept PNI Grand Sample Count Order 38.865 0.310 4.201 –10.467 45.175 0.272 1.492 –19.729 45.327 0.261 1.958 –12.701 37.415 0.194 1.498 –12.492 61.549 –0.059 –0.830 0.718 32.356 0.261 0.618 –15.342 44.312 0.246 1.227 –11.048 37.417 0.252 1.379 –11.625 70.759 –0.232 –1.329 –2.202 8.589 0.158 0.587 5.346 7.551 0.138 0.589 4.732 7.602 0.139 0.574 4.753 6.637 0.122 0.442 4.127 6.103 0.112 0.389 3.796 6.787 0.124 0.506 4.261 7.815 0.142 0.668 4.896 6.944 0.128 0.454 4.307 6.085 0.112 0.374 3.773 < 0.001 0.051 < 0.001 0.052 < 0.001 0.050 0.013 < 0.001 < 0.001 0.062 0.001 0.008 < 0.001 0.113 0.001 0.003 < 0.001 0.602 0.036 0.850 < 0.001 0.037 0.225 < 0.001 < 0.001 0.086 0.070 0.025 < 0.001 0.051 0.003 0.008 < 0.001 0.041 0.001 0.560 20 KANSKE ET AL. TABLE S8. Association of the Vulnerability Subscale of the PNI With the Relation of Self- and OtherRelated Ratings to the Positivity, Negativity, Past Orientation, and Future Orientation of Thoughts as Estimated as Interactions in Multilevel Models in the CRT Task CRT Model 2 Parameters Self Other Predictors Estimate SE p value Intercept PNI Vuln Sample Count Order Negative PNI Vuln × Negative Intercept PNI Vuln Sample Count Order Positive PNI Vuln × Positive Intercept PNI Vuln Sample Count Order Past PNI Vuln × Past Intercept PNI Vuln Sample Count Order Future PNI Vuln × Future Intercept PNI Vuln Sample Count Order Negative PNI Vuln × Negative Intercept PNI Vuln Sample Count Order Positive PNI Vuln × Positive Intercept PNI Vuln Sample Count Order Past PNI Vuln × Past Intercept PNI Vuln Sample Count Order Future PNI Vuln × Future 24.737 0.266 1.527 –9.757 0.581 –0.004 67.506 –0.115 1.948 –13.545 –0.394 0.005 26.450 0.296 1.798 –9.372 0.565 –0.005 40.701 –0.023 1.578 –9.612 0.066 0.003 43.920 0.119 1.332 –18.822 0.063 0.000 23.252 0.358 1.582 –20.089 0.364 –0.003 37.418 0.049 1.211 –14.479 0.272 0.001 31.885 0.159 1.225 –17.290 0.322 –0.001 8.035 0.093 0.477 4.704 0.148 0.002 12.340 0.144 0.486 4.647 0.174 0.002 7.016 0.075 0.477 4.509 0.118 0.001 7.187 0.081 0.455 4.334 0.097 0.001 8.112 0.095 0.514 4.657 0.155 0.002 12.678 0.148 0.507 4.696 0.180 0.002 6.571 0.071 0.484 4.166 0.117 0.001 7.390 0.084 0.494 4.387 0.104 0.001 0.002 0.005 0.001 0.040 < 0.001 0.035 < 0.001 0.423 < 0.001 0.004 0.024 0.022 < 0.001 < 0.001 < 0.001 0.039 < 0.001 0.001 < 0.001 0.773 0.001 0.028 0.495 0.005 < 0.001 0.210 0.010 < 0.001 0.686 0.811 0.067 0.016 0.002 < 0.001 0.044 0.154 < 0.001 0.489 0.013 0.001 0.021 0.372 < 0.001 0.058 0.013 < 0.001 0.002 0.461 WHERE THE NARCISSISTIC MIND WANDERS 21 order. of Behavior Therapy and ExTABLE S9. Journal Association of the Grandiosity Subscale of the PNI With the Relation of Self- and Otherperimental Psychiatry, 44(1), 37–47. Related Ratings to the Positivity, Negativity, Past Orientation, and Future Orientation of Thoughts as Estimated as Interactions in Multilevel Models in the CRT Task CRT Model 2 Parameters Self Other Predictors Estimate SE p value Intercept PNI Grand Sample Count Order Negative PNI Grand × Negative Intercept PNI Grand Sample Count Order Positive PNI Grand × Positive Intercept PNI Grand Sample Count Order Past PNI Grand × Past Intercept PNI Grand Sample Count Order Future PNI Grand × Future Intercept PNI Grand Sample Count Order Negative PNI Grand × Negative Intercept PNI Grand Sample Count Order Positive PNI Grand × Positive Intercept PNI Grand Sample Count Order Past PNI Grand × Past Intercept PNI Grand Sample Count Order Future PNI Grand × Future 21.733 0.511 1.436 –9.602 0.691 –0.009 95.682 –0.842 1.875 –13.435 –0.803 0.018 26.957 0.490 1.665 –9.355 0.561 –0.009 45.915 –0.174 1.552 –9.229 –0.020 0.008 47.773 0.081 1.347 –17.811 –0.098 0.005 26.462 0.547 1.615 –19.463 0.305 –0.004 41.836 –0.031 1.311 –14.140 0.099 0.006 40.480 0.054 1.204 –16.835 0.102 0.003 9.544 0.196 0.478 4.699 0.184 0.004 14.713 0.308 0.483 4.703 0.203 0.004 8.522 0.166 0.480 4.570 0.137 0.003 8.699 0.174 0.454 4.403 0.121 0.003 9.671 0.200 0.514 4.675 0.193 0.004 15.178 0.319 0.507 4.762 0.211 0.005 7.923 0.154 0.483 4.189 0.134 0.003 8.895 0.179 0.494 4.416 0.129 0.003 0.024 0.010 0.003 0.043 < 0.001 0.026 < 0.001 0.007 < 0.001 0.005 < 0.001 < 0.001 0.002 0.003 0.001 0.042 < 0.001 0.005 < 0.001 0.319 0.001 0.038 0.866 0.003 < 0.001 0.684 0.009 < 0.001 0.613 0.264 0.082 0.087 0.002 < 0.001 0.149 0.338 < 0.001 0.841 0.007 0.001 0.462 0.034 < 0.001 0.763 0.015 < 0.001 0.432 0.226 22 KANSKE ET AL. TABLE S10. Relation of the Vulnerability Subscale of the PNI to Ratings Levels in the WM Task as Estimated With Multilevel Modeling WM Model 1 Parameters Off-Task Other Self Negative Positive Past Future Mood Negative Mood Positive Predictors Estimate SE p value Intercept PNI Vuln Sample Count Order Intercept PNI Vuln Sample Count Order Intercept PNI Vuln Sample Count Order Intercept PNI Vuln Sample Count Order Intercept PNI Vuln Sample Count Order Intercept PNI Vuln Sample Count Order Intercept PNI Vuln Sample Count Order Intercept PNI Vuln Sample Count Order Intercept PNI Vuln Sample Count Order 20.204 0.162 5.315 2.020 53.455 –0.120 2.258 –1.897 45.559 –0.002 1.970 1.013 32.373 0.126 1.924 –0.127 71.752 –0.126 –1.793 –1.719 45.705 –0.068 2.297 –3.432 37.774 0.095 2.378 2.481 39.248 0.080 0.523 –2.236 69.882 –0.116 –1.103 1.968 6.492 0.077 1.000 4.068 8.008 0.095 1.319 5.044 7.607 0.090 1.184 4.786 6.626 0.079 0.851 4.176 6.389 0.076 0.672 4.046 7.378 0.088 1.212 4.645 7.760 0.092 1.419 4.892 6.881 0.082 0.808 4.355 5.368 0.064 0.564 3.399 0.003 0.040 < 0.001 0.621 < 0.001 0.214 0.092 0.708 < 0.001 0.984 0.107 0.833 < 0.001 0.115 0.025 0.976 < 0.001 0.105 0.008 0.673 < 0.001 0.442 0.064 0.463 < 0.001 0.308 0.099 0.614 < 0.001 0.332 0.521 0.610 < 0.001 0.076 0.052 0.565 WHERE THE NARCISSISTIC MIND WANDERS 23 TABLE S11. Relation of the Grandiosity Subscale of the PNI to Ratings Levels in the WM Task as Estimated With Multilevel Modeling WM Model 1 Parameters Off-Task Other Self Negative Positive Past Future Mood Negative Mood Positive Predictors Estimate SE p value Intercept PNI Grand Sample Count Order Intercept PNI Grand Sample Count Order Intercept PNI Grand Sample Count Order Intercept PNI Grand Sample Count Order Intercept PNI Grand Sample Count Order Intercept PNI Grand Sample Count Order Intercept PNI Grand Sample Count Order Intercept PNI Grand Sample Count Order Intercept PNI Grand Sample Count Order 17.801 0.323 5.421 2.243 47.740 –0.069 2.249 –2.096 41.318 0.093 1.996 0.973 31.476 0.229 1.954 0.050 71.043 –0.191 –1.809 –1.927 44.945 –0.096 2.273 –3.519 35.286 0.216 2.449 2.507 41.054 0.092 0.530 –2.098 68.302 –0.155 –1.116 1.777 7.009 0.141 1.010 4.028 8.845 0.179 1.320 5.112 8.268 0.167 1.185 4.768 7.278 0.147 0.851 4.186 7.079 0.143 0.672 4.075 8.060 0.163 1.210 4.656 8.358 0.170 1.418 4.850 7.592 0.153 0.809 4.375 5.976 0.121 0.565 3.440 0.014 0.026 < 0.001 0.610 < 0.001 0.702 0.093 0.683 < 0.001 0.579 0.103 0.839 < 0.001 0.124 0.023 0.990 < 0.001 0.187 0.008 0.638 < 0.001 0.560 0.066 0.453 < 0.001 0.210 0.089 0.607 < 0.001 0.553 0.516 0.633 < 0.001 0.203 0.050 0.607 24 KANSKE ET AL. FIGURE S1. Mood level in the CRT (A) and WM (B) tasks.