Replicable Facets of Positive Emotionality and Their Relations to Psychopathology

Abstract

General individual differences in positive emotionality are negatively related to depression, social anxiety, and schizotypy/schizophrenia, and positively related to mania. However, the structure of positive emotionality remains unclear at the facet level, as there are significant disparities in the types of content assessed across emotionality measures. This study analyzed the lower order structure of positive emotionality in two samples, finding evidence for a replicable two-factor structure of Joviality and Experience Seeking. These factors demonstrated a markedly different pattern of relations in both direction and magnitude with internalizing, externalizing, and schizotypal symptoms. Joviality seems to represent an adaptive variant of positive emotionality, as it showed strong positive relations with well-being and moderate negative relations with measures of depression, social anxiety, and social anhedonia. In contrast, Experience Seeking appears to be somewhat maladaptive. It generally related positively to psychopathology, correlating most strongly with indicators of manic and externalizing symptoms.

Keywords

positive emotionality psychopathology depressive disorders mania anxiety disorders schizotypy positive affect

Trait affect reflects enduring individual differences in the tendency to experience positive and negative moods and feelings (Watson, 2000) and is composed of two broad dimensions: positive emotionality and negative emotionality. Positive emotionality consists of positively valenced mood states and correlates with Five-Factor model extraversion, whereas negative emotionality, which is characterized by subjective distress, correlates with Five-Factor model neuroticism (Clark & Watson, 1999a; Watson, Wiese, Vaidya, & Tellegen, 1999). These associations between (a) extraversion and positive emotionality and (b) neuroticism and negative emotionality are well-supported in the personality literature (Costa & McCrae, 1980; Emmons & Diener, 1985; Lucas, Diener, Grob, Suh, & Shao, 2000); in fact, the associations are so strong that positive emotionality is regarded by some as the core of extraversion, and negative emotionality is viewed as synonymous with neuroticism (Diener, Suh, Lucas, & Smith, 1999; Watson & Clark, 1984, 1997).

A large body of research indicates that negative emotionality and neuroticism are broadly related to a wide range of psychopathology (Watson, 2000). Measures of negative affect correlate strongly with both depression and anxiety (Mineka, Watson, & Clark, 1998; Watson, Clark, & Carey, 1988), and one recent meta-analysis found large affect sizes for neuroticism with every analyzed disorder, including major depression, social phobia, panic disorder, agoraphobia, generalized anxiety disorder, obsessive–compulsive disorder (OCD), posttraumatic stress disorder (PTSD), and specific phobia (Kotov, Gamez, Schmidt, & Watson, 2010).

In contrast, positive emotionality displays greater specificity in its associations, as it generally shows stronger negative relations with depression than anxiety (Mineka et al., 1998; Watson & Naragon-Gainey, 2010). Social anxiety represents one exception, however. Depression and social anxiety both are characterized by positive affect deficits, meaning that people with these disorders report low levels of positive emotionality (Kashdan, 2002, 2007; Watson & Naragon-Gainey, 2010); however, the available evidence indicates that this positive emotionality deficit is somewhat larger in depression than in social anxiety. Both depression and social anxiety are also negatively related to extraversion. Some research indicates that social anxiety is broadly related to all facets defining the trait, and especially to interpersonal facets such as dominance and sociability, whereas depression is strongly related only to the low positive emotionality facet of extraversion (Naragon-Gainey, Watson, & Markon, 2009).

Other forms of psychopathology also demonstrate interesting associations with positive emotionality. For example, low levels of positive emotionality are a core feature of schizophrenia and schizotypy (Horan, Blanchard, Clark, & Green, 2008). Some evidence indicates that individuals with schizophrenia have the ability to fully experience positive emotions, but that they are affected by an anticipatory pleasure deficit (Cohen & Minor, 2010; Gard, Kring, Gard, Horan, & Green, 2007; Kring & Moran, 2008). Put differently, even though these individuals experience pleasure after fulfilling events, they still do not expect to experience high levels of positive affect when engaging in such events in the future. It is noteworthy, moreover, that positive emotionality is negatively correlated with the negative symptoms of schizophrenia/schizotypy (e.g., constricted affect, social anhedonia) but is only weakly related to the positive symptoms (e.g., magical thinking, perceptual aberrations, suspiciousness).

Finally, mania is a particularly striking case, in that, unlike most disorders, it is characterized by elevated positive emotionality. Whereas goal-directed behavior is lacking in disorders associated with strong positive emotionality deficits, mania often includes aspects of excessive or unrealistic reward-seeking (Gruber, 2011; Watson & Naragon-Gainey, 2010).

Hierarchical Structure of Emotionality

Affective experience is hierarchically structured, such that these general dimensions can be further broken down into specific facets of positive emotionality and negative emotionality. Although they are distinguishable, these facets are related to one another due to their common relations with either the general positive emotionality or negative emotionality factor (Tellegen, Watson, & Clark, 1999; Watson et al., 1999; Watson & Tellegen, 1985). For example, potential facets of positive emotionality include joy, excitement, and interest, whereas negative affect includes the specific components of sadness, fear, and anger. These facets model specific content variance and lie at the lower order level in a hierarchical model of affect.

Although there is a general consensus that negative emotionality includes facets of sadness, fear, and anger, there is less agreement on the specific facets of positive emotionality (Watson & Clark, 1992). Unfortunately, mood researchers have made little progress in defining the lower order structure of affect in the last 30 to 40 years, and the issue appears to have been largely neglected (Watson & Vaidya, 2013). Most measures of affect include at least a general positive emotionality scale, and many do include specific facet level scales. However, these instruments show poor convergence at the facet level, which we examine in greater detail in the next section.

Overview of Existing Self-Report Measures of Positive Emotionality

Positive emotionality also can be assessed as a transient, short-term state; here the term positive affect is commonly used. Examining existing positive affect measures provides a starting point for improved assessment at the facet level. We will focus here on popular inventories that were specifically designed to measure multiple specific affects; these instruments include the Mood Adjective Checklist (MACL; Nowlis, 1965), the Profile of Mood States–Second Edition (POMS 2; Heuchert & McNair, 2012), the Differential Emotions Scale (DES; Izard, Libero, Putnam, & Haynes, 1993), the Multiple Affect Adjective Checklist–Revised (MAACL-R; Zuckerman & Lubin, 1985), and the Positive and Negative Affect Schedule–Expanded Form (PANAS-X; Watson & Clark, 1999; for a more detailed review of these instruments and others, see Gray & Watson, 2007). Although these measures often are used to assess short-term positive affect, they can be used to assess trait emotionality by asking participants to indicate how strongly they feel a particular type of affect in general or on average.

All these measures assess positive affect, but as stated previously, there are significant discrepancies in the specific types of positive affect that they tap, indicating a lack of agreement regarding the facets that comprise positive emotionality. In fact, as seen in Table 1, there is little consensus on the specific facets, as these sets of measures vary widely in both (a) their number of positive affect scales and (b) the content they assess. For example, the PANAS-X and MACL assess a wider range of content using a greater number of facet scales than the MAACL-R and POMS 2, which each include only two facet scales. The MAACL-R, POMS 2, and DES all contain relatively few scales to assess positive affect, but there is relatively little correspondence across these measures as each includes an idiosyncratic facet scale (e.g., MAACL-R Sensation Seeking, POMS 2 Friendliness, DES Interest). Emotional states of happiness, excitement, interest, and enthusiasm all are positive in nature, but which specific individual affects actually encompass positive mood experience?

Table 1.

Positive Affect Scales Included in Multi-Affect Mood Measures.

Measure	Positive affect scales
MACL	Surgency, Elation, Concentration, Social Affection, Vigor
POMS-2	Vigor-Activity, Friendliness
DES	Interest, Joy, Surprise
MAACL-R	Positive Affect, Sensation Seeking
PANAS-X	Joviality, Self-Assurance, Attentiveness, Serenity

Note. MACL = Mood Adjective Checklist; POMS 2 = Profile of Mood States–Second Edition; DES = Differential Emotions Scale; MAACL-R = Multiple Affect Adjective Checklist–Revised; PANAS-X = Positive and Negative Affect Schedule–Expanded Form.

Importance of Understanding Specific Positive Emotionality–Psychopathology Relations

The lack of a consensus on the specific facets of positive emotionality has important implications for psychopathology research. If researchers only measure positive emotionality at a general level or are using facet level scales that are not fully representative of positive mood experience, then it is difficult to understand fully how positive emotionality relates to various forms of psychopathology. A number of articles have already called for additional research examining how specific affects and personality traits relate to psychopathology (Kashdan, 2007; Paunonen, 2003; Sellbom, Ben-Porath, & Bagby, 2008; Stanton & Watson, 2014). However, little research has examined the relations between specific aspects of positive emotionality and psychopathology. Without a fine-grained understanding of positive mood experience, it is difficult to conceptualize positive mood dysfunction in disorders beyond a broad and general level.

Limited evidence suggests that considering such lowerorder relations provides additional information and explanatory power that is not available by examining only general positive emotionality. For example, Watson, Clark, and Stasik (2011) found that PANAS-X Joviality was more strongly related to diagnoses of depression (mean r = −.51) than either PANAS-X Attentiveness (mean r = −.42) or PANAS-X Self-Assurance (mean r = −.35); moreover, it was the only positive affect scale to contribute significantly to the prediction of these diagnoses in a series of logistic regression analyses. Other personality research suggests some specificity for the relations between psychopathology and facets of extraversion and neuroticism (Bienvenu et al., 2004; Naragon-Gainey et al., 2009; Rector, Bagby, Huta, & Ayearst, 2012; Rector, Hood, Richter, & Bagby, 2002); overall, however, there still are little data available to indicate how the specific facets of positive emotionality relate to psychopathology.

Goals of the Current Study

The current study has two basic goals. Its first goal is to delineate the lower order structure of trait positive emotionality through an analysis of existing scales in the mood and personality literature. We sought to achieve this goal by (a) analyzing the structure of the PANAS-X and MACL at the item level, (b) creating scales from this initial factor analysis, and then (c) embedding them into a more comprehensive analysis using a larger battery of positive emotionality measures. We began with a structural analysis of the PANAS-X and MACL because these instruments share a common format and contain overlapping items. Specifically, lively is included in both PANAS-X Joviality and MACL Surgency, attentive and concentrating are contained in both MACL Concentration and PANAS-X Attentiveness, and energetic is scored in both PANAS-X Joviality and MACL Vigor. These shared items would artifactually inflate the correlations between these MACL and PANAS-X scales, rendering subsequent structural analyses problematic. Consequently, we conducted initial factor analyses to create a new set of scales with non-overlapping items for use in subsequent analyses.

Only markers that represented affective states or included significant affective content were retained for analysis, as we sought to include items and scales that would load onto a general positive emotionality factor at both phases of our structural analyses. Thus, for our initial item-level analysis, the adjectives from some positively valenced MACL scales (e.g., Concentration, Social Affection) were dropped, as they loaded weakly on the general positive emotionality factor and were deemed to represent behavioral and cognitive aspects of personality, rather than positive emotionality per se. In this regard, although the MACL scales were derived factor-analytically, they were created more than 50 years ago (see Nowlis, 1965). Mood and affectivity research has progressed significantly—and the domain has been defined more clearly (and narrowly)—since that time (Watson & Vaidya, 2013). The expanded battery for the scale level analyses consisted of scales from several well-validated positive emotionality measures in addition to those newly created from the PANAS-X and MACL item-level analyses.

The second goal of this study is to determine how the specific, lower order facets of positive emotionality that emerge in these analyses relate to symptoms of several types of psychopathology, including depression, social anxiety, schizotypy, and mania. As discussed previously, these disorders are characterized by positive mood dysfunction, with affected individuals reporting either significant deficits (depression, social anxiety, schizotypy) or elevations (mania) in positive emotionality. We also will examine relations with other indicators of psychopathology (e.g., substance use) on a more exploratory basis.

Analysis of the lower order structure of positive emotionality in this study will (a) provide a structural framework to guide future research and (b) allow for a more detailed examination of the relations between specific facets of positive emotionality and psychopathology. Doing so will provide for a more precise understanding of positive emotional dysfunction in these forms of psychopathology and add to a more general body of literature focusing on the interplay between mood and psychopathology.

Method

Overview

Participants from both an online community sample and student sample completed the same battery of self-report positive emotionality items. The community participants also completed several psychopathology scales from the Expanded Version of the Inventory of Depression and Anxiety Symptoms (IDAS-II; Watson et al., 2012). In contrast, the student participants completed a wider range of psychopathology measures including the entire IDAS-II (except for Suicidality); the Alcohol Use Disorders Identification Test (AUDIT; Saunders, Aasland, Babor, de la Fuente, & Grant, 1993); the Drug Use Survey (DUS; Clark & Watson, 1999b); the Hypomanic Personality Scale (HPS; Eckblad & Chapman, 1986); the Social fear scale from the Phobic Stimuli Response Scales (PSRS; Cutshall & Watson, 2004); the Schizotypal Personality Questionnaire (SPQ; Raine, 1991); and the Short Michigan Alcoholism Screening Test (SMAST; Selzer, Vinokur, & Van Rooijen, 1975). The characteristics of each sample and the measures used for this study are described below.

Participants

Online Community Sample

Participants (N = 451) were recruited online through Amazon Mechanical Turk and were paid $5 for completing the study. Most of the participants reported that they were living in the United States (78%), but there was a substantial portion who indicated they were living in India (20%). This sample was 60.5% White, 28.2% Asian, 6.0% African American, 3.1% Multiracial, 1.8% American Indian, and 0.2% Native Hawaiian or other Pacific Islander; 6.4% of participants in this sample identified as Hispanic. The participants had a mean age of 32.94 years (SD = 10.19); 54.8% of the sample was male. Complete positive emotionality data are available on 450 participants (99.8%), and data from these participants will be used in the structural, correlation, and regression analyses.

Student Sample

This sample (N = 436) consisted of University of Notre Dame undergraduates who completed the study for credit toward their required research participation hours. The students were 79.6% White, 10.7% Asian, 7.0% Multiracial, 2.1% African American, and 0.7% American Indian. Additionally, 10.6% of the undergraduates identified as Hispanic. The participants in this sample had a mean age of 19.36 years (SD = 1.29); 58.7% of the sample was female. Complete positive emotionality data are available on 422 participants (96.8%), and data from these participants will be used in the structural, correlation, and regression analyses.

Affect Measures

PANAS-X and MACL

The PANAS-X (Watson & Clark, 1999) is a 60-item measure of self-rated affect. It is comprised of 13 scales that include the two general factors of Positive Affect and Negative Affect, as well as 11 specific affects (Fear, Sadness, Guilt, Hostility, Shyness, Fatigue, Surprise, Joviality, Self-Assurance, Attentiveness, and Serenity). Participants completed only the items from the Joviality (8 items; e.g., happy, enthusiastic), Self-Assurance (6 items; e.g., confident, daring), Attentiveness (4 items; e.g., alert, determined), and Serenity (3 items; e.g., calm, relaxed) scales.

The MACL (Nowlis, 1965) includes both positively and negatively valenced affect scales assessed using a series of adjectives. Of its 12 scales, only the items from the Surgency (5 items; e.g., lively, carefree), Elation (4 items; e.g., elated, overjoyed), Concentration (6 items; e.g., attentive, serious), Social Affection (4 items; e.g., forgiving, kindly), and Vigor (3 items; active, vigorous) scales were administered. As discussed earlier, however, certain Concentration items and all the Social Affection items were excluded from subsequent analyses due to their nonaffective content. In addition, items that overlapped between the PANAS-X and MACL (e.g., lively, attentive) were not repeated but were administered only once. Thus, the battery of PANAS-X and MACL items consisted of 31 items. Participants indicated the extent to which they have experienced these adjectives “in general” or “on average,” using a 5-point scale ranging from very slightly to extremely.

DES

The DES (Izard et al., 1993) is a self-report instrument designed to measure 12 discrete emotions: Interest, Joy, Surprise, Sadness, Anger, Disgust, Contempt, Fear, Shame, Shyness, Guilt, and Hostility. The DES can be modified to assess trait affect using directions asking participants to indicate how often statements are true of themselves in their daily lives. Participants responded to DES items using a 5-point scale ranging from rarely or never to very often. Only the positively valenced Joy (3 items; e.g., “feel glad”) and Interest (3 items; e.g., “alert and curious about something”) scales were administered. Surprise also is classified as a specific facet of positive affect in the DES, but it is uncertain whether it truly represents a positive affective state or trait. As a result, this scale was not administered.

International Personality Item Pool Scales

The International Personality Item Pool (IPIP; Goldberg et al., 2006) is an online repository of public domain scales assessing a wide range of individual differences. Many of the IPIP scales were developed to measure constructs from popular personality inventories. A 10-item Excitement-Seeking scale (e.g., “I love excitement”) modeled after a similarly named facet from the Revised NEO Personality Inventory (NEO PI-R; Costa & McCrae, 1992) was included in this study. Participants rated themselves using a 5-point Likert-type scale ranging from strongly disagree to strongly agree.

Temperament and Affectivity Inventory

The Temperament and Affectivity Inventory (TAI; Watson, Stasik, Chmielewski, & Naragon-Gainey, 2014) is 93-item self-report inventory. It is a multidimensional measure of trait affect intended to provide more stable assessment of emotionality than previous measures, such as the PANAS-X (see Chmielewski & Watson, 2009 for a more detailed discussion of this topic). The inventory contains 12 scales assessing a range of positively and negatively valenced affects. However, only the positively valenced scales were used in this study; these include Attentiveness (6 items; e.g., “easy for me to focus my attention”), Vigor (5 items; e.g., “enjoy having a lot going on”), Experience Seeking (9 items; e.g., “can be pretty fearless”), and Geniality (10 items; e.g., “easy for me to laugh”). Participants answered the items on a 5-point Likert-type scale ranging from strongly disagree to 5 strongly agree.

Psychopathology Measures

Alcohol Use Disorders Identification Test

The AUDIT (Saunders et al., 1993) is a 10-item measure assessing alcohol use. It includes items assessing both drinking frequency and the consequences of drinking excessively. The AUDIT can be administered as either an interview or as a self-report measure; the latter format was used in this study. Participants responded to each item using a 5-point scale.

Drug Use Survey

The DUS (Clark & Watson, 1999b) assesses how frequently participants use a wide range of drugs, including marijuana, cocaine, amphetamines, diet pills, tranquilizers, psychedelics, narcotics, amyl/butyl nitrates, inhalants, and ecstasy. It is a 10-item self-report measure; participants indicated how often they used each type of drug on a 7-point scale ranging from never to 40 times or more.

Hypomanic Personality Scale

The HPS (Eckblad & Chapman, 1986) is a 48-item self-report scale assessing hypomanic personality, which is characterized by hyperactivity and grandiosity. Schalet, Durbin, and Revelle (2011) analyzed the HPS items and found evidence for a 3-factor structure consisting of Social Vitality (19 items; e.g., “uncommon ability to persuade others”), Mood Volatility (13 items; e.g., “feel an emotion with extreme intensity”), and Excitement (7 items; e.g., “excited and happy for no reason”; Schalet et al., 2011). We report results on these three subscales in this study. Participants completed the HPS items using a true or false response format.

Inventory of Depression and Anxiety Symptoms–II

The Expanded Version of the IDAS-II (Watson et al., 2012) is a 99-item measure designed to assess specific symptoms of the emotional disorders (i.e., major depression, bipolar disorder, OCD, PTSD, and the anxiety disorders; see Watson, 2005). It is a revised version of the original IDAS (Watson et al., 2007), which contained 11 symptom scales. The IDAS Dysphoria scale—which contains items assessing depressed mood, anhedonia, worry, worthlessness, guilt, hopelessness, psychomotor disturbance, and cognitive difficulties—was created to capture the core affective and cognitive symptoms of depression and anxiety (see Watson et al., 2007). Five scales represent specific symptoms of major depression: Insomnia, Lassitude (i.e., fatigue, lack of energy, and hypersomnia), Suicidality, Appetite Loss, and Appetite Gain. In addition, Ill Temper assesses feelings of anger and hostility, whereas Well-Being taps content reflecting high energy and positive affect that is specifically related to depression (see Watson et al., 2007; Watson & Naragon-Gainey, 2010). The final three scales assess symptoms related to PTSD (Traumatic Intrusions), social phobia (Social Anxiety), and panic disorder/somatic anxiety (Panic).

The expanded IDAS-II features seven new scales that broaden the coverage of the instrument; these new scales assess symptoms related to PTSD (Traumatic Avoidance), OCD Checking, Ordering, Cleaning), agoraphobia/specific phobia (Claustrophobia), and bipolar disorder (Mania, Euphoria). The IDAS-II also contains an expanded, 6-item version of Social Anxiety to yield a total of 18 scales. The IDAS-II scales can be used individually, as each scale measures a separate symptom dimension.

The Well-being (e.g., “I felt optimistic”), Dysphoria (e.g., “I felt depressed”), Social Anxiety (e.g., “I felt self-conscious knowing that others were watching me”), Mania (e.g., “My thoughts jumped rapidly from one idea to another”), and Euphoria (e.g., “I felt elated for no special reason”) scales were administered to participants in the community sample. As stated earlier, participants in the student sample completed the full IDAS-II, with the exception of Suicidality. Participants answered the items using a 5-point scale ranging from not at all to extremely, indicating how much they had experienced each symptom in the past 2 weeks.

Phobic Stimuli Response Scales

The PSRS (Cutshall & Watson, 2004) consist of five separate scales; however, the student participants completed only the Social Scale. The PSRS Social Scale assesses fear of social situations using a 4-point scale ranging from strongly disagree to strongly agree.

Schizotypal Personality Questionnaire

The SPQ (Raine, 1991) is a 74-item self-report scale intended to model the Diagnostic and Statistical Manual of Mental Disorders (3rd ed., rev.; American Psychiatric Association, 1987) criteria for Schizotypal Personality Disorder; participants responded to the items use a yes/no response format. Although the SPQ is divided into nine subscales (Ideas of Reference, Excessive Social Anxiety, Odd Beliefs/Magical Thinking, Unusual Perceptual Experiences, Odd or Eccentric Behavior, No Close Friends, Odd Speech, Constricted Affect, and Suspiciousness), an item-level structural analysis revealed a five-factor structure of Social Anhedonia (e.g., “others find me elusive in conversation”), Unusual Beliefs and Experiences (e.g., “seen things invisible to others”), Social Anxiety (e.g., “often nervous around unfamiliar people”), Mistrust (e.g., “others are talking about me behind my back”), and Eccentricity/Oddity (e.g., “others think I am bizarre”; Chmielewski & Watson, 2008). Results are reported based on this five-factor model.

Short Form of the Michigan Alcoholism Screening Test

Similar to the AUDIT, the SMAST (Selzer et al., 1975) assesses alcohol use and its consequences. The SMAST consists of 13 self-report items, and it is a short form of the 25-item Michigan Alcoholism Screening Test (Selzer, 1971). Participants responded to the items using a dichotomous yes or no format.

Results

Preliminary Analyses

Table 2 presents the scale means, standard deviations, and coefficient alphas for both the online community and student samples (note that the table includes data for the newly created PANAS-X/MACL scales to be described later). Coefficient alphas for the positive emotionality scales ranged from .69 to .96 in the community sample and from .57 to .94 in the student sample. In both samples, the DES Interest scale (α = .69 community, .57 student), had the lowest alphas, and was the only positive emotionality scale with an alpha below .75. The low DES Interest coefficient alphas reflect, in part, the brevity of this scale. Coefficient alphas for the psychopathology scales ranged from .62 to .93 across samples. The alphas for the HPS Excitement subscale (α = .70) and the SMAST (α = .62) were at the bottom end of this range, as they were the only two scales with alphas below .75. Although these values are not ideal, these scales provided markers of specific content areas wherein few options are otherwise available; they therefore were retained in subsequent analyses.

Table 2.

Scale Descriptive Statistics and Alphas.

	Community sample			Student sample
Scale	Mean	SD	α	Mean	SD	α
DES Interest (3)**	10.76	2.00	.69	9.99	1.73	.57
DES Joy (3)	10.54	2.49	.86	10.77	1.97	.80
IPIP Exc. Seeking (10)***	28.52	7.35	.85	32.68	6.07	.82
Attentiveness (5)**	18.46	3.89	.86	16.98	3.27	.81
Joviality (13)	39.89	11.92	.96	40.96	9.14	.94
Self-Assurance (5)*	13.79	4.35	.83	13.05	3.72	.79
Serenity (3)**	10.51	2.66	.84	9.37	2.32	.80
TAI Attentiveness (6)***	22.02	4.72	.87	18.87	4.52	.84
TAI Exp. Seeking (9)	28.20	6.67	.86	28.04	5.08	.77
TAI Geniality (10)	40.53	4.97	.82	41.01	4.43	.83
TAI Vigor (5)***	16.40	4.49	.89	18.70	3.26	.81
IDAS-II Dysphoria (10)	21.31	8.92	.93	21.00	7.13	.87
IDAS-II Euphoria (5)**	10.62	4.61	.85	9.56	3.70	.80
IDAS-II Mania (5)	10.10	4.67	.87	9.66	3.98	.81
IDAS-II Social Anxiety (6)**	13.04	6.26	.91	11.40	4.62	.85
IDAS-II Well-being (8)	24.76	7.28	.92	24.33	5.97	.88
IDAS-II Appetite Gain (3)	—	—	—	6.46	2.72	.76
IDAS-II Appetite Loss (3)	—	—	—	5.30	2.61	.84
IDAS-II Claustrophobia (5)	—	—	—	6.47	2.97	.89
IDAS-II Checking (3)	—	—	—	5.92	2.77	.84
IDAS-II Cleaning (7)	—	—	—	10.17	4.69	.91
IDAS-II Ill Temper (5)	—	—	—	7.61	3.53	.87
IDAS-II Insomnia (6)	—	—	—	11.78	4.45	.80
IDAS-II Lassitude (6)	—	—	—	14.44	4.89	.80
IDAS-II Ordering (5)	—	—	—	8.97	3.73	.81
IDAS-II Panic (8)	—	—	—	12.21	5.34	.90
IDAS-II Traum. Avoid. (4)	—	—	—	7.37	3.70	.91
IDAS-II Traum. Intrus. (4)	—	—	—	6.86	3.39	.86
AUDIT (10)	—	—	—	16.91	6.34	.86
DUS (10)	—	—	—	11.56	4.17	.78
HPS Excitement (7)	—	—	—	1.56	1.70	.70
HPS Mood Volatility (13)	—	—	—	3.85	2.97	.76
HPS Social Vitality (19)	—	—	—	5.87	3.80	.77
PSRS Social (11)	—	—	—	27.67	5.30	.86
SMAST (13)	—	—	—	1.28	1.56	.62
SPQ Eccent./Odd. (14)	—	—	—	3.82	3.65	.86
SPQ Mistrust (13)	—	—	—	2.82	3.02	.83
SPQ Social Anhedonia (15)	—	—	—	1.63	2.40	.81
SPQ Social Anxiety (17)	—	—	—	4.91	4.32	.87
SPQ Unus. Beliefs and Exp. (15)	—	—	—	2.22	2.49	.78

Note. N = 450-451 for the community sample and N = 422-436 for the student sample. The number of items for each scale are in parentheses following the scale name. Scales that are starred (*) are significantly different between samples, scales double-starred (**) are significantly different with a small effect size, and scales that are triple-starred (***) are significantly different with a medium effect size. DES = Differential Emotions Scale; DIS-I = Disinhibition Inventory; IPIP = International Pool of Personality Items; MACL = Mood Adjective Checklist; PANAS-X = Positive and Negative Affect Schedule–Expanded Form; TAI = Temperament and Affectivity Inventory; IDAS-II = Inventory for Depression and Anxiety Symptoms-II; AUDIT = Alcohol Use Disorders Identification Test; DUS = Drug Use Survey; HPS = Hypomanic Personality Questionnaire; PSRS = Phobic Stimuli Response Scales; SMAST = Short Michigan Alcoholism Screening Test; SPQ = Schizotypal Personality Questionnaire.

The mean scores across samples were significantly different from each other in many instances (see Table 2), but these differences are considered “small” or inconsequential in magnitude in most cases. However, the Cohen’s d values (Cohen, 1988) were of “medium” magnitude for three scales: the students reported substantially higher scores on IPIP Excitement-Seeking (d = 0.62) and TAI Vigor (d = 0.58), whereas the community patients were higher on TAI Attentiveness (d = 0.68).

To address the first goal of the study (viz., delineating the structure of positive emotionality at the facet level), we first conducted factor analyses of the PANAS-X and MACL items in both the student and community samples. These analyses are presented first, followed by a description of the subsequent scale level analyses of the larger battery of positive emotionality measures in both samples.

Factor Analysis of the PANAS-X and MACL items

Determining the Number of Factors

An initial principal components analysis was conducted on the 31 PANAS-X and MACL adjectives in both samples to determine the number of factors to be extracted. First, a parallel analysis (O’Connor, 2000) was conducted in both samples. In parallel analysis, the observed eigenvalues from the principal components analyses are compared with eigenvalues of random data sets containing the same numbers of variables with the same sample sizes. O’Connor’s (2000) SAS program was used to conduct this analysis (Ncases = 450, Nvars = 31, Ndatasets = 1,000, percent = 95 for the community sample; Ncases = 421, Nvars = 31, Ndatasets = 1,000, percent = 95 for the student sample). In the community sample, results indicated that the third eigenvalue (1.64) from the principal components analysis was greater than its random counterpart (1.40), but the fourth value was not (1.24 vs. 1.35). Results from the student sample indicated that the fourth eigenvalue (1.64) exceeded its random counterpart (1.37), but the fifth value did not (1.07 vs. 1.32).

Second, Velicer’s (1976) minimum average partial (MAP) test was used in both samples to provide additional information in determining the optimal number of factors to extract. The MAP test is based on analysis of residual correlation matrices, as it involves computing the averaged squared partial correlation for a range of factor solutions. An increasing number of factors are extracted, and the optimal number of factors is indicated by the factor solution that yields the lowest mean value. The MAP test on the community data indicated that the mean squared partial correlation decreased from the two-factor (.0174) to the three-factor (.0149) to the four-factor solution (.0117), but increased when five (.0130) and six factors (.0145) were extracted. A similar pattern was observed in the student data, as the mean squared partial correlation decreased from the two-factor (.0184) to three-factor (.0152) to four-factor solutions (.0119), but again rose when a fifth (.0123) and sixth factor (.0131) were extracted.

Thus, the MAP test converged with the parallel analysis to indicate that four factors should be extracted in the student data. The MAP test indicated that four factors also should be extracted in the community data, whereas the parallel analysis indicated that a maximum of three should be extracted in this sample. Although it was less clear which factor solution was optimal in the community data, we eventually adopted the four-factor structure. As will be discussed later, extracting four rather than three factors produces results that are more comparable with those for a four-factor solution in the student data. Furthermore, the preponderance of evidence supports a four-factor solution when considering the results from the parallel analyses and MAP tests across these two samples, as three of the four analyses indicated that four factors should be extracted.

Description of the Factors

Next, a series of principal factor analyses were conducted. All solutions up to a maximum of four factors were examined using an oblique promax rotation (power = 3). The one-factor solution yielded evidence of a strong positive emotionality factor, as all 31 items loaded ≥.40 on this general factor in both samples.

The two-factor solution reflected a general distinction between high and low arousal positive emotionality in both samples, as such adjectives as calm and at ease were strong markers for the low arousal factor and items such as bold and vigorous defined the high arousal factor. Although these factors differ in arousal, they were highly correlated in both samples (r = .73 and .65 in the community and student data, respectively). The PANAS-X Attentiveness adjectives (e.g., attentive, determined) were strong markers of the high arousal factor in the student sample, but loaded more strongly on the low arousal factor in the community sample.

The three-factor structure was highly similar across samples: In addition to the same basic high arousal and low arousal factors, the PANAS-X Attentiveness adjectives now provided strong markers of a third Attentiveness dimension.

Last, the four-factor solution provided a very clear and replicable factor structure across samples. Promax loadings from this solution are provided in Tables 3 (community sample) and 4 (student sample). Most of the MACL Surgency and Elation adjectives loaded strongly with the PANAS-X Joviality items (e.g., excited, joyful) on the first factor, although the MACL adjectives vigorous and witty were markers of the third factor, along with several of the PANAS-X Self-Assurance items (e.g., daring, fearless). The PANAS-X Attentiveness adjectives and the adjective confident were clear markers of the second factor in both samples, and the three PANAS-X Serenity items were strong markers of the fourth factor. It is noteworthy that this structural model essentially recreated the four PANAS-X positive emotionality facets, with the MACL adjectives providing additional markers for Factors I and III; as a result, following the PANAS-X terminology, these factors were labeled Joviality, Attentiveness, Self-Assurance, and Serenity, respectively. Scales based on this replicable structural model were created and used in subsequent analyses.

Table 3.

Promax Factor Loadings of the Positive Emotionality Adjectives in a Community Sample.

Positive emotionality adjective	I	II	III	IV
Excited	.84	−.01	.09	−.08
Joyful	.84	.03	−.07	.09
Cheerful	.72	.09	−.06	.18
Delighted	.72	.11	.07	.03
Happy	.71	.08	−.10	.22
Elated	.69	−.06	.24	−.01
Playful	.69	−.05	.15	.01
Enthusiastic	.69	.23	.12	−.14
Overjoyed	.66	−.05	.21	.03
Lively	.61	.22	.01	.11
Pleased	.59	.11	−.01	.22
Talkative	.53	−.01	.22	−.09
Refreshed	.40	.21	.07	.26
Attentive	.02	.81	−.11	.02
Determined	.02	.70	.03	.03
Alert	.03	.69	−.03	−.04
Concentrating	−.03	.69	.06	.10
Energetic	.49	.46	.00	−.03
Confident	.16	.46	.22	.18
Active	.41	.44	.11	−.05
Strong	.16	.41	.37	−.05
Daring	−.02	−.03	.79	.08
Fearless	−.05	.15	.61	.18
Bold	.20	.18	.60	−.05
Witty	.08	−.02	.54	−.08
Vigorous	.26	.05	.50	−.07
Carefree	.16	−.28	.45	.40
Proud	.37	.16	.32	.05
Calm	−.13	.21	−.02	.75
At ease	.13	.01	−.03	.73
Relaxed	.08	−.04	.05	.72

Note. N = 450. Loadings ≥ |.40| are shown in boldface.

Table 4.

Promax Factor Loadings of the Positive Emotionality Adjectives in a Student Sample.

Positive emotionality adjective	I	II	III	IV
Delighted	.85	−.03	−.04	.04
Overjoyed	.82	−.17	.09	−.03
Excited	.80	.04	−.00	−.03
Joyful	.78	.01	−.00	.05
Cheerful	.76	.14	−.18	.06
Elated	.74	−.03	.12	−.00
Enthusiastic	.68	.09	.11	−.01
Pleased	.67	.11	−.08	.14
Lively	.66	.07	.16	.02
Happy	.60	.28	−.28	.13
Talkative	.55	.01	.21	−.11
Proud	.45	.22	.23	−.10
Playful	.43	.04	.22	.07
Refreshed	.38	.12	.09	.21
Attentive	−.11	.85	−.02	.08
Determined	.02	.71	−.05	−.04
Concentrating	.06	.63	.05	−.00
Alert	.05	.60	.02	−.05
Confident	.07	.51	.21	.11
Energetic	.32	.48	.12	−.00
Active	.18	.44	.15	−.06
Daring	−.12	−.03	.82	.06
Bold	.13	.05	.69	−.05
Fearless	−.09	.10	.66	.17
Witty	.17	−.05	.42	.07
Vigorous	.27	.11	.38	−.14
Strong	.16	.27	.38	−.02
Relaxed	−.10	.05	.08	.76
Calm	.00	.08	−.04	.70
At ease	.16	.00	−.02	.69
Carefree	.27	−.26	.23	.52

Note. N = 422. Loadings ≥ |.40| are shown in boldface.

Analysis of Factor Similarity

To establish the replicability of this four-factor solution more formally, we computed comparability coefficients (Finn, 1986)—which involve deriving regression-based factor scores for each solution (Everett & Entrekin, 1980; Finn, 1986; Gorsuch, 1983; Harman, 1976)—to quantify the level of similarity across samples. Our two four-factor solutions each generated a set of regression-based factor scoring weights, which then were applied to participants’ actual scale scores in each sample, yielding a total of eight factor scores (four representing community factors and four representing the student factors) in each data set. If the two solutions are highly similar, then the corresponding weights for each factor (e.g., the Joviality weights in the student sample vs. the Joviality weights in the community data) will produce scores that are very highly correlated; specifically, correlations ≥.90 indicate that the same factor emerged across samples (see Everett, 1983). The values for the comparability coefficients ranged from .98 to .99 (overall M value = .99); this very strong level of convergence indicates that the two samples yielded the same four-factor solution.

Scale Creation

As noted, most items were clear markers of the same factor across samples; however, five of the adjectives were more problematic and, therefore, were not used to create scales for further analyses. The adjectives energetic, strong, and carefree had their strongest loadings on different factors across samples. Active was a clear marker of Attentiveness in the student data, but it had nearly equivalent loadings on Joviality (.41) and Attentiveness (.44) in the community data. Likewise, proud was a clear Joviality marker in the student data, but had similar loadings on Joviality (.37) and Self-Assurance (.32) in the community data. After these five items were dropped, the remaining 26 items were summed together based on their factor loadings to create scales to represent the Joviality (13 items), Attentiveness (five items), Self-Assurance (five items), and Serenity (three items) factors.

Factor Analysis of the Positive Emotionality Scales

Determining the Number of Factors

Next, an initial principal components analysis was conducted on the 11 positive emotionality scales, which included the newly created Joviality, Attentiveness, Self-Assurance, and Serenity scales from the item-level analyses, as well as the DES, IPIP, and TAI scales described earlier. As with the analysis of the positive emotionality adjectives, both parallel analyses and MAP tests were used to determine the number of factors to extract. The parallel analysis in the community data (Ncases = 450, Nvars = 11, Ndatasets = 1,000, percent = 95) indicated that a maximum of two factors should be extracted, as the second eigenvalue (1.53) was larger than its random counterpart (1.18), but the third eigenvalue (.79 vs. 1.13) was not. Likewise, in the parallel analysis of the student data (Ncases = 421, Nvars = 11, Ndatasets = 1,000, percent = 95), the second value (1.99) exceeded its random counterpart (1.19) but the third eigenvalue (1.11 vs. 1.13) did not.

The MAP test on the community data indicated that mean squared partial correlation was lowest in the two-factor solution (.0455), as it increased when three (.0662) and four (.0906) factors were extracted. Similarly, the MAP test on the student data indicated that the value from the two-factor solution (.0522) was smaller than the subsequent three- (.0635) and four-factor solutions (.0760). Thus, both the parallel analyses and MAP tests in the community and student data converged to indicate that two factors should be extracted in both samples.

Description of the Factors

Next, we conducted a series of principal factor analyses using an oblique promax rotation (power = 3); based on the results of the parallel analyses and MAP tests, only one- and two-factor solutions were examined. The one-factor solution in the community data yielded evidence of a strong general positive emotionality factor, as every scale loaded ≥.40 on this dimension. The one-factor solution in the student data yielded similar results; however, three scales had loadings smaller than .40: IPIP Excitement Seeking (loading = .39), TAI Attentiveness (loading = .33), and TAI Experience Seeking (loading = .28). Nonetheless, 21 of the 22 scales (95.5%) still loaded >.30 across the two analyses, indicating that this set of scales all defined a general positive emotionality factor.

In both samples, the two-factor structure represented a distinction between a fairly general first factor consisting of happiness, energy, and attentiveness markers, and a more specific second factor consisting of reward-seeking positive emotionality content (loadings in the community and student samples are presented in Tables 5 and 6, respectively). The Joviality, Attentiveness, and DES Joy scales all loaded ≥.70 on this more general first factor across samples; the TAI Attentiveness, Serenity, TAI Geniality, TAI Vigor, and DES Interest scales also helped to define this factor. The TAI Experiencing Seeking, IPIP Excitement-Seeking, and Self-Assurance (e.g., daring, bold, fearless) scales were the strongest markers of the second factor in both samples.

Table 5.

Promax Factor Loadings of the Positive Emotionality Scales in a Community Sample.

Positive emotionality scale	I	II
Attentiveness	.84	−.04
Joviality	.77	.23
DES Joy	.76	.07
TAI Attentiveness	.68	−.26
Serenity	.58	.09
TAI Geniality	.56	.05
TAI Vigor	.49	.31
DES Interest	.48	.15
IPIP Excitement-Seeking	−.15	.91
TAI Experience Seeking	−.03	.89
Self-Assurance	.35	.58

Note. N = 450. Loadings ≥ |.40| are shown in boldface. DES = Differential Emotions Scale; IPIP = International Personality Item Pool; TAI = Temperament and Affectivity Inventory.

Table 6.

Promax Factor Loadings of the Positive Emotionality Scales in a Student Sample.

Positive emotionality scale	I	II
Joviality	.79	.15
DES Joy	.78	−.11
Attentiveness	.72	.05
TAI Geniality	.59	−.04
DES Interest	.53	.02
TAI Vigor	.48	−.01
TAI Attentiveness	.45	−.19
Serenity	.44	.17
TAI Experience Seeking	−.18	.83
IPIP Excitement-Seeking	−.04	.78
Self-Assurance	.26	.67

Note. N = 421. Loadings ≥ |.40| are shown in boldface. DES = Differential Emotions Scale; IPIP = International Personality Item Pool; TAI = Temperament and Affectivity Inventory.

It is noteworthy that these factors were strongly related in the community data (r = .51) and moderately correlated in the student data (.28), suggesting that there is considerable communality between them (average r = .40). These two factors were labeled Joviality (marked by such scales as Joviality, Attentiveness, and DES Joy) and Experience Seeking (marked by TAI Experience Seeking, IPIP Excitement-Seeking, and Self-Assurance). Regression-based factor scores were created to model these two factors in subsequent analyses. The correlations between the factor scores were slightly larger than the interfactor correlations (r = .58 in the community data; r = .34 in the student data; average r = .46), which again indicates that these dimensions are moderately to strongly correlated with one another.

Analysis of Factor Similarity

Comparability coefficients again were computed to determine the similarity of these factors across samples. The comparability coefficients were .99 for both factors when scored in the community data and also were .99 for both factors when scored in the student data. Thus, the two samples yielded a highly similar set of factors, which is unsurprising given that all 11 scales had their strongest loadings on the same factor in both solutions.

Associations Between Factors of Positive Emotionality and Psychopathology

Overview

To address the second goal of the study (viz., to explicate the associations between positive emotionality and psychopathology), we related the facet level factor scores to the psychopathology measures in each sample. As stated earlier, the community participants were assessed on only five IDAS-II scales (Dysphoria, Social Anxiety, Mania, Euphoria, and Well-Being), whereas the student respondents completed a much broader range of psychopathology measures. Correlations between the facet factor scores and the psychopathology scales are shown in Tables 7 (IDAS-II scales) and 8 (all other psychopathology measures). Because the factor scores were correlated, we also conducted multiple regression analyses to identify the unique, incremental predictive power of each facet; standardized beta weights from these analyses are presented in Tables 9 (IDAS-II scales) and 10 (all other psychopathology measures).

Table 7.

Correlations Between the Positive Emotionality Factors and IDAS-II Scales.

Scale	Joviality	Experience seeking
Community
IDAS-II Well-being	.80 *	.53
IDAS-II Euphoria	.59	.59
IDAS-II Dysphoria	−.38 *	.02
IDAS-II Social Anxiety	−.17*	−.01
IDAS-II Mania	.15*	.40
Student
IDAS-II Well-being	.70 *	.25
IDAS-II Dysphoria	−.42 *	.01
IDAS-II Social Anxiety	−.38 *	−.11
IDAS-II Euphoria	.33	.35
IDAS-II Panic	−.30 *	.13
IDAS-II Appetite Loss	−.29 *	.03
IDAS-II Lassitude	−.29 *	.08
IDAS-II Traumatic Intrusions	−.27 *	.03
IDAS-II Insomnia	−.24*	.09
IDAS-II Traumatic Avoidance	−.23*	−.07
IDAS-II Claustrophobia	−.20*	.07
IDAS-II Ill Temper	−.19*	.17
IDAS-II Cleaning	−.16*	−.02
IDAS-II Mania	−.12*	.13
IDAS-II Appetite Gain	−.05	.04
IDAS-II Ordering	−.04	−.06
IDAS-II Checking	−.03	−.04

Note. N = 449 in the community sample and N = 421 in the student sample. IDAS-II = Expanded version of the Inventory of Depression and Anxiety Symptoms. Correlations with an asterisk (*) are significantly different from the other value in their row, and correlations ≥ |.25| are shown in boldface.

Table 8.

Correlations between the Positive Emotionality Factors and the Additional Psychopathology Scales in a Student Sample.

Scale	Joviality	Experience seeking
SPQ Social Anxiety	−.43 *	−.26
SPQ Social Anhedonia	−.42 *	.00
PSRS Social	−.38	−.33
SPQ Eccentricity/Oddity	−.24 *	.09
SMAST	−.22*	.15
SPQ Mistrust	−.20*	.07
HPS Social Vitality	.25 *	.56
AUDIT	−.14*	.27
HPS Excitement	.17	.24
HPS Mood Volatility	−.18*	.22
Drug Use Survey	−.17*	.21
SPQ Unusual Beliefs and Experiences	−.03*	.16

Note. N = 419-421. HPS = Hypomanic Personality Scale; SMAST = Short Michigan Alcoholism Screening Test; AUDIT = Alcohol Use Disorders Identification Test; PSRS = Phobic Stimuli Response Scales; SPQ = Schizotypal Personality Questionnaire. Correlations with an asterisk (*) are significantly different from the other value in their row, and correlations ≥ |.25| are shown in boldface.

Table 9.

Strength of the Relations between the Positive Emotionality Factors and IDAS-II Scales in the Regression Analyses.

Scale	Joviality	Experience seeking
Community
IDAS-II Well-being	.74	.11
IDAS-II Dysphoria	−.58	.35
IDAS-II Social Anxiety	−.25	.14
IDAS-II Mania	−.13	.47
IDAS-II Euphoria	.37	.37
Student
IDAS-II Well-being	.69	.02
IDAS-II Dysphoria	−.48	.17
IDAS-II Social Anxiety	−.39	.02
IDAS-II Panic	−.39	.26
IDAS-II Lassitude	−.36	.20
IDAS-II Appetite Loss	−.34	.15
IDAS-II Traumatic Intrusions	−.31	.13
IDAS-II Insomnia	−.31	.19
IDAS-II Ill Temper	−.28	.26
IDAS-II Claustrophobia	−.25	.15
IDAS-II Traumatic Avoidance	−.23	.00
IDAS-II Cleaning	−.17	.04
IDAS-II Euphoria	.24	.27
IDAS-II Mania	−.19	.20
IDAS-II Appetite Gain	−.07	.07
IDAS-II Ordering	−.03	−.05
IDAS-II Checking	−.01	−.04

Note. N = 449 for the community sample and N = 421 for the student sample. The values shown are standardized beta weights. Values ≥ |.25| are shown in boldface, and values ≥ |.10| are significant at p < .05. IDAS-II = Inventory of Depression and Anxiety Symptoms-II.

Table 10.

Strength of the Relations between the Positive Emotionality Factors and the Additional Psychopathology Scales in the Regression Analyses.

Scale	Joviality	Experience seeking
SPQ Social Anhedonia	−.48	.16
SPQ Social Anxiety	−.39	−.13
SPQ Eccentricity/Oddity	−.31	.20
SMAST	−.30	.25
PSRS Social	−.30	−.23
SPQ Mistrust	−.25	.16
HPS Social Vitality	.06	.54
AUDIT	−.26	.36
HPS Mood Volatility	−.29	.31
Drug Use Survey	−.27	.30
HPS Excitement	.10	.20
SPQ Unusual Beliefs and Experiences	−.09	.18

Note. N = 419-421. The values shown are standardized beta weights. Values ≥ |.25| are shown in boldface, and values ≥ |.10| are significant at p < .05. HPS = Hypomanic Personality Scale; SMAST = Short Michigan Alcoholism Screening Test; AUDIT = Alcohol Use Disorders Identification Test; PSRS = Phobic Stimuli Response Scales; SPQ = Schizotypal Personality Questionnaire.

It is noteworthy that these associations display considerable specificity/variability, both within and across factors. To help make sense of these data, we offer summary descriptions for each type of symptom measure.

Well-Being

Joviality demonstrated particularly strong relations with IDAS-II Well-being, and the regression analyses indicated that it was a unique predictor of this construct in both samples (βs = .74 and .69). Although Experience Seeking correlated strongly with Well-being in the community sample (r = .53), the strength of this relation dropped substantially in the regression analyses (β = .11). Its relation in the student sample was relatively weaker (r = .25), and it showed no incremental predictive power in the regression analyses (β = .02).

Depressive Symptoms

The IDAS-II models a range of depressive symptoms (e.g., Dysphoria, Lassitude, Appetite Loss/Gain, Insomnia), and Joviality showed comparatively stronger relations with them than did Experience Seeking. Moreover, the associations for Joviality and Experience Seeking were generally in the opposite direction, as Joviality related negatively to these symptoms whereas Experience Seeking actually showed weak to moderate positive relations with them in the regression analyses. For example, Joviality was a strong negative predictor of IDAS-II Dysphoria in the community sample and a moderate negative predictor in the student sample (βs = −.58 and −.48, respectively); in contrast, Experience Seeking was a positive predictor in both samples (βs = .35 and .17, respectively). More generally, Joviality demonstrated moderate negative relations with these symptoms, whereas Experience Seeking exhibited very weak relations in the correlational analyses (rs ranged from .01 to .09) and slightly stronger positive relations in the regression analyses (βs ranged from .07 to .35).

Anxiety

The IDAS-II also includes a number of scales modeling anxiety symptoms (e.g., Social Anxiety, Claustrophobia, Panic), and the PSRS Social scale provided another marker of social anxiety in the student sample. Joviality was also a stronger predictor of these internalizing symptoms than was Experience Seeking. Joviality had moderate negative relations with social anxiety as measured by the IDAS-II (βs = −.25 and −.39) and PSRS (r = −.38; β = −.30), and was also a negative predictor of claustrophobia and panic (βs = −.25 and −.39, respectively). As with the depressive symptoms, Experience Seeking showed weak, but mostly positive relations with these symptoms (βs ranged from −.23 to .26).

Manic Symptoms

Joviality and Experience Seeking demonstrated comparable relations with the IDAS-II Euphoria scale across samples, but the latter had a significantly stronger relation with IDAS-II Mania in the community sample (r = .40; β = .47). Experience Seeking related positively to all three HPS scales (Social Vitality, Mood Volatility, and Excitement), and had a particular affinity with HPS Social Vitality (r = .56; β = .54). In contrast, Joviality related weakly to the HPS scales (rs ranged from −.18 to .25, βs ranged from −.29 to .10).

Schizotypal Symptoms

Chmielewski and Watson (2008) found that positive emotionality related more strongly to negative schizotypal symptoms such as social withdrawal and social anhedonia than to positive symptoms such as oddity and hallucinations. Joviality mirrored those results in the analyses with the SPQ, as it was comparatively stronger predictor of social anxiety (r = −.43; β = −.39) and social anhedonia (r = −.42; β = −.48) than of eccentricity (r = −.24; β = −.31), mistrust (r = −.20; β = −.25), and unusual beliefs and experiences (r = −.03; β = −.09). Experience Seeking was a much weaker predictor, showing both positive and negative relations with the SPQ scales (rs range from −.16 to .16, βs ranged from −.13 to .20).

Substance Use

Joviality and Experience Seeking also were differential predictors of substance use as measured by the DUS (drug use), AUDIT, and SMAST (both alcohol use). Whereas Joviality related negatively to all three measures in both the correlation and regression analyses (rs ranged from −.14 to −.22, βs ranged from −.26 to −.30), Experience Seeking showed positive relations with all three, and was even a moderate predictor of scores on the AUDIT (β = .36) and DUS (β = .30) in the regression analyses.

OCD and PTSD

This study included limited coverage of OCD and PTSD symptoms; as a result, the relations for both are summarized together. Both factors related weakly to OCD symptoms (Checking, Cleaning, Ordering). Neither factor showed particularly strong relations with PTSD symptoms as measured by the IDAS-II Traumatic Intrusions and Traumatic Avoidance scales, although Joviality was a moderate negative predictor of the former in the regression analyses (β = −.31).

Discussion

Summary of Results

The primary goals of this study were (a) to determine the lower order structure of trait positive emotionality and (b) to determine how the specific facets of positive emotionality that emerged from our structural analyses relate to psychopathology. Adjectives from the PANAS-X and MACL were factor analyzed prior to factor analyses with other scales because they contain overlapping items that would complicate subsequent analyses. Strong evidence for a replicable four-factor structure of Joviality, Attentiveness, Self-Assurance, and Serenity was obtained across samples in these analyses, with the resulting factors closely resembling the facet level structure of the PANAS-X positive emotionality scales.

Scales were created to model each of these four factors in subsequent structural analyses at the scale level, which also included positive emotionality scales from the DES, IPIP, and TAI. A replicable two-factor structure of Joviality and Experience Seeking was obtained across samples in these analyses. These two factors demonstrated considerable specificity in their relations with the psychopathology measures. As shown in Tables 7 through 10, the strength of these relations varied considerably not only within each factor, but also across factors.

Joviality—which consisted of markers representing a blend of happiness, attentiveness, and cheerfulness content—demonstrated a strong positive relation with IDAS-II Well-Being and also correlated negatively with 28 of the 32 (87.5%) psychopathology scales across samples. Experience Seeking emerged as a narrower factor marked by the TAI Experience Seeking, IPIP Excitement-Seeking, and Self-Assurance scales in both samples. Although Experience Seeking and Joviality had moderate to strong positive correlations with one another, the former demonstrated markedly different relations with the psychopathology scales, as it correlated positively in 24 of 32 (75.0%) instances. The correlations for these factors differed in both direction and magnitude, as there were significant differences in their correlations with the well-being and psychopathology scales in 27 of 34 analyses (79.4%). Joviality tended to show stronger relations with internalizing and schizotypal symptoms, whereas Experience Seeking correlated more strongly with mania.

The results from the regression analyses closely mirrored the bivariate associations for each factor: Joviality again related negatively to 28 of the 32 psychopathology scales, whereas Experience Seeking related positively to an even greater number of symptoms (29 of 32; 90.6%); moreover, two of its three negative relations were very weak in magnitude. Thus, an interesting pattern emerges, as these two positive emotionality factors demonstrated strikingly different relations with psychopathology despite showing moderate to strong positive correlations with each other. Summaries of the psychopathology relations for the Joviality and Experience Seeking facets are provided below.

Joviality

As stated, Joviality was a negative predictor of a broad range of symptoms. Although it showed moderate to strong positive relations with the IDAS-II Euphoria scale (e.g., “felt elated for no reason”; βs = .37 and .24 in the community and student data, respectively), it was negatively associated with most internalizing, externalizing, and schizotypal symptoms. Joviality had particularly strong negative relations with measures of depressive and, in some instances, social anxiety symptoms. Thus, its relations seem to mirror those for general positive emotionality that have been reported in previous research (Watson & Naragon-Gainey, 2010). Furthermore, it was moderately related to negative symptoms of schizotypy (e.g., social anxiety and social anhedonia), but showed comparatively weaker relations with positive symptoms (i.e., eccentricity, mistrust, unusual experiences). Lastly, its strong positive relations with Well-Being in both samples (rs = .80 and .70 in the community and student data, respectively) provide additional evidence that high scores on this dimension are healthy and adaptive.

Experience Seeking

Experience Seeking emerged as a generally maladaptive facet of positive emotionality in our analyses, despite its moderate to strong positive correlations with Joviality. Unlike Joviality, Experience Seeking generally had weak associations with internalizing and schizotypy symptoms, but related more strongly to manic and externalizing symptoms, such as drug and alcohol use. As discussed earlier, nearly all of its associations with psychopathology were positive, and it related particularly strongly to the HPS Social Vitality subscale (e.g., “usually the ‘life of the party’”; β = .54). These findings are in accord with other research suggesting that mania shows stronger links to reward seeking positive emotions than to other types of positive affect (Gruber, 2011; Gruber & Johnson, 2009). Although, Experience Seeking had a strong positive correlation with Well-Being in the community data (r = .53), the strength of this relation was noticeably weaker in the regression analyses (β = .11). Taken together, these data suggest that high levels of Experience Seeking are somewhat maladaptive, as this factor relates positively to manic and externalizing symptoms and only weakly to Well-Being.

Future Directions and Limitations

This study provides a solid foundation for future research examining the structure of positive emotionality and how its facets relate to psychopathology. It has some notable strengths. In particular, we were able to establish the replicability of our two-factor structure across two large samples and to make a distinction between the more adaptive and less adaptive aspects of positive emotionality.

At the same time, however, our study has several limitations that are worth noting. First, both of our samples were generally healthy and nonclinical in nature, so that future research with more clinically oriented samples would be useful. This is especially important in the study of the bipolar and psychotic disorders, which have low base rates in the general population. Second, this study included measures from a fairly small number of positive emotionality instruments, as only items and scales from the PANAS-X, MACL, DES, IPIP, and TAI were included in the factor analyses. The small number of scales may have limited the number of factors that could have emerged in the scale level analysis, which may help to explain why the Joviality factor was somewhat broad and nonspecific. Research incorporating a larger number of scales would be useful to provide a more comprehensive analysis of positive emotionality.

Third, we had limited coverage of psychotic and externalizing psychopathology in this study: the former was assessed using just the SPQ scales, whereas the latter was assessed by several short substance abuse scales. Determining the strength and direction of the relations between positive emotionality facets and personality pathology offers a potentially valuable avenue for future research, given that there already are data to indicate that personality facets containing positive emotionality content relate positively to narcissistic, histrionic, and antisocial personality disorders (Ruiz, Pincus, & Schinka, 2008; Samuel & Widiger, 2008). Recently developed instruments, such as the Personality Inventory for DSM-5 (Krueger, Derringer, Markon, Watson, & Skodol, 2012) and the Computerized Adaptive Test of Personality Disorder (Simms et al., 2011), offer particularly promising opportunities for explicating the broader psychopathological correlates of positive emotionality facets.

Last, our data were based entirely on self-report; it will be important in future research to extend these findings using informant ratings of personality and/or interview measures of psychopathology. Related to this, psychopathology was assessed only at the symptom level, and further research examining similar relations at the diagnostic level is needed as well.

Summary

Data from two large samples (N = 451 and N = 436) provide preliminary evidence for a two-factor structure of Joviality and Experience Seeking, with the former representing adaptive aspects of positive emotionality and the latter reflecting a more maladaptive variant. Although this is a promising start, future research including a larger number of positive emotionality measures is needed to determine if this structure replicates in other samples or if the facet level structure of positive emotionality is better represented by a larger number of more specific facets. Joviality and Experience Seeking demonstrated specificity both within and across factors when a wide range of symptoms were considered; our data indicate that (a) there are different, specific types of positive emotionality deficits depending on the symptom being examined and, furthermore, that (b) certain types of positive emotionality are not necessarily adaptive. Ideally, further investigation of the role of positive emotionality at the facet level will build from these results and from others (Watson et al., 2011) to clarify the nature of positive emotion dysfunction across a range of psychopathology.

Footnotes

Declaration of Conflicting Interests

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

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research was supported by a grant from the University of Notre Dame Institute for Scholarship in the Liberal Arts awarded to Kasey Stanton.

References

American Psychiatric Association. (1987). Diagnostic and statistical manual of mental disorders (3rd ed., text revision). Washington, DC: Author.

Bienvenu

O. J.

Samuels

J. F.

Costa

P. T.

Jr. Reti

I. M.

Eaton

W. W.

Nestadt

(2004). Anxiety and depressive disorders and the five-factor model of personality: A higher- and lower-order personality trait investigation in a community sample. Depression and Anxiety, 20, 92-97.

Chmielewski

Watson

(2008). The heterogeneous structure of schizotypal personality disorder: Item-level factors of the Schizotypal Personality Questionnaire and their associations with OCD symptoms, dissociative tendencies, and normal personality. Journal of Abnormal Psychology, 117, 364-376.

Chmielewski

Watson

(2009). What is being assessed and why it matters: The impact of transient error on trait research. Journal of Personality and Social Psychology, 97, 186-202.

Clark

L. A.

Watson

(1999a). Temperament: A new paradigm for trait psychology. In Pervin

John

(Eds.), Handbook of personality: Theory and research (2nd ed., pp. 399-423). New York, NY: Guilford Press.

Clark

L. A.

Watson

(1999b). The drug use survey (Unpublished manuscript). University of Iowa, Iowa City.

Cohen

A. S.

Minor

K. S.

(2010). Emotional experience in patients with schizophrenia revisited: Meta-analysis of laboratory studies. Schizophrenia Bulletin, 36, 143-150.

Cohen

(1988). Statistical power analysis for the behavioral sciences (2nd ed.). Hillsdale, NJ: Erlbaum.

Costa

P. T.

McCrae

R. R.

(1980). Influence of extraversion and neuroticism on subjective well-being: Happy and unhappy people. Journal of Personality and Social Psychology, 38, 668-678.

10.

Costa

P. T.

Jr. McCrae

R. R.

(1992). Revised NEO Personality Inventory (NEO PI-R) and NEO Five-Factor Inventory (NEO-FFI) professional manual. Odessa, FL: Psychological Assessment Resources.

11.

Cutshall

Watson

(2004). The phobic stimuli response scales: A new self-report measure of fear. Behaviour Research and Therapy, 42, 1193-1201.

12.

Diener

Suh

E. M.

Lucas

R. E.

Smith

H. L.

(1999). Subjective well-being: Three decades of progress. Psychological Bulletin, 125, 276-302.

13.

Eckblad

Chapman

L. J.

(1986). Development and validation of a scale for hypomanic personality. Journal of Abnormal Psychology, 95, 214-222.

14.

Emmons

R. A.

Diener

(1985). Personality correlates of subjective well-being. Personality and Social Psychology Bulletin, 11, 89-97.

15.

Everett

J. E.

(1983). Factor comparability as a means of determining the number of factors and their rotation. Multivariate Behavioral Research, 18, 197-218.

16.

Everett

J. E.

Entrekin

L. V.

(1980). Factor comparability and the advantages of multiple group factor analysis. Multivariate Behavioral Research, 2, 165-180.

17.

Finn

S. E.

(1986). Structural stability of the MMPI in adult males. Journal of Consulting and Clinical Psychology, 54, 703-707.

18.

Gard

D. E.

Kring

A. M.

Gard

M. G.

Horan

W. P.

Green

M. F.

(2007). Anhedonia in schizophrenia: Distinctions between anticipatory and consummatory pleasure. Schizophrenia Research, 93, 253-260.

19.

Goldberg

L. R.

Johnson

J. A.

Eber

H. W.

Hogan

Ashton

M. C.

Cloninger

C. R.

Gough

H. G.

(2006). The international personality item pool and the future of public-domain personality measures. Journal of Research in Personality, 40, 84-96.

20.

Gorsuch

R. L.

(1983). Factor analysis (2nd ed.). Hillsdale, NJ: Erlbaum.

21.

Gray

E. K.

Watson

(2007). Assessing positive and negative affect via self-report. In Coan

J. A.

Allen

J. J.

(Eds.), Handbook of emotion elicitation and assessment (pp. 171-183). New York, NY: Oxford University Press.

22.

Gruber

(2011). Can feeling too good be bad? Positive emotion persistence (PEP) in bipolar disorder. Current Directions in Psychological Science, 20, 217-221.

23.

Gruber

Johnson

S. L.

(2009). Positive emotional traits and ambitious goals among people at risk for mania: The need for specificity. International Journal of Cognitive Therapy, 2, 176-187.

24.

Harman

H. H.

(1976). Modern factor analysis (3rd ed.). Chicago, IL: University of Chicago Press.

25.

Heuchert

J. P.

McNair

D. M.

(2012). POMS 2. North Tonawanda, NY: Multi-Health Systems Incorporated.

26.

Horan

W. P.

Blanchard

J. J.

Clark

L. A.

Green

M. F.

(2008). Affective traits in schizophrenia and schizotypy. Schizophrenia Bulletin, 34, 856-874.

27.

Izard

C. E.

Libero

D. Z.

Putnam

Haynes

O. M.

(1993). Stability of emotion expression experiences and their relations to traits of personality. Journal of Personality and Social Psychology, 64, 847-860.

28.

Kashdan

T. B.

(2002). Social anxiety dimensions, neuroticism, and the contours of positive psychological functioning. Cognitive Therapy and Research, 26, 789-810.

29.

Kashdan

T. B.

(2007). Social anxiety spectrum and diminished positive experiences: Theoretical synthesis and meta-analysis. Clinical Psychology Review, 27, 348-365.

30.

Kotov

Gamez

Schmidt

Watson

(2010). Linking “big” personality traits to anxiety, depressive, and substance use disorders: A meta-analysis. Psychological Bulletin, 136, 768-821.

31.

Kring

A. M.

Moran

E. K.

(2008). Emotional response deficits in schizophrenia: Insights from affective science. Schizophrenia Bulletin, 34, 819-834.

32.

Krueger

R. F.

Derringer

Markon

K. E.

Watson

Skodol

A. E.

(2012). Initial construction of a maladaptive personality trait model and inventory for DSM-5. Psychological Medicine, 42, 1879-1890.

33.

Lucas

R. E.

Diener

Grob

Suh

E. M.

Shao

(2000). Cross-cultural evidence for the fundamental features of extraversion. Journal of Personality and Social Psychology, 79, 452-468.

34.

Mineka

Watson

Clark

L. A.

(1998). Comorbidity of anxiety and unipolar mood disorders. Annual Review of Psychology, 49, 377-412.

35.

Naragon-Gainey

Watson

Markon

K. E.

(2009). Differential relations of depression and social anxiety symptoms to the facets of extraversion/positive emotionality. Journal of Abnormal Psychology, 118, 299-310.

36.

Nowlis

(1965). Research with the Mood Adjective Checklist. In Tompkins

S. S.

Izard

C. E.

(Eds.), Affect, cognition, and personality: Empirical studies (pp. 352-389). New York, NY: Spring.

37.

O’Connor

B. P.

(2000). SPSS and SAS programs for determining the number of components using parallel analysis and Velicer’s MAP test. Behavior Research Methods, Instruments, & Computers, 32, 396-402.

38.

Paunonen

S. V.

(2003). Big Five factors of personality and replicated predictions of behavior. Journal of Personality and Social Psychology, 84, 411-424.

39.

Raine

(1991). The SPQ: A scale for the assessment of schizotypal personality based on DSM-III-R criteria. Schizophrenia Bulletin, 17, 555-564.

40.

Rector

N. A.

Bagby

R. M.

Huta

Ayearst

L. E.

(2012). Examination of the trait facets of the five-factor model in discriminating specific mood and anxiety disorders. Psychiatry Research, 199, 131-139.

41.

Rector

N. A.

Hood

Richter

M. A.

Bagby

R. M.

(2002). Obsessive-compulsive disorder and the five-factor model of personality: Distinction and overlap with major depressive disorder. Behaviour Research and Therapy, 40, 1205-1219.

42.

Ruiz

M. A.

Pincus

A. L.

Schinka

J. A.

(2008). Externalizing pathology and the five-factor model: A meta-analysis of personality traits associated with antisocial personality disorder, substance use disorder, and their co-occurrence. Journal of Personality Disorders, 22, 365-388.

43.

Samuel

D. B.

Widiger

T. A.

(2008). A meta-analytic review of the relationships between the five-factor model and DSM-IV-TR personality disorders: A facet level analysis. Clinical Psychology Review, 28, 1326-1342.

44.

Saunders

J. B.

Aasland

O. G.

Babor

T. F.

Fuente

J. R.

Grant

(1993). Development of the Alcohol Use Disorders Identification Test (AUDIT): WHO collaborative project on early detection of persons with harmful alcohol consumption-II. Addiction, 88, 791-804.

45.

Schalet

B. D.

Durbin

C. E.

Revelle

(2011). Multidimensional structure of the Hypomanic Personality Scale. Psychological Assessment, 23, 504-522.

46.

Sellbom

Ben-Porath

Y. S.

Bagby

R. M.

(2008). On the hierarchical structure of mood and anxiety disorders: Confirmatory evidence and elaboration of a model of temperament markers. Journal of Abnormal Psychology, 117, 576-590.

47.

Selzer

M. L.

(1971). The Michigan Alcoholism Screening Test: The quest for a new diagnostic instrument. American Journal of Psychiatry, 127, 1653-1658.

48.

Selzer

M. L

Vinokur

Van Rooijen

(1975). A self-administered Short Michigan Alcoholism Screening Test (SMAST). Journal of Studies on Alcohol, 36, 117-126.

49.

Simms

L. J.

Goldberg

L. R.

Roberts

J. E.

Watson

Welte

Rotterman

J. H.

(2011). Computerized adaptive assessment of personality disorder: Introducing the CAT-PD project. Journal of Personality Assessment, 93, 380-389.

50.

Stanton

Watson

(2014). Positive and negative affective dysfunction in psychopathology. Social and Personality Psychology Compass, 8(9), 555-567.

51.

Tellegen

Watson

Clark

L. A.

(1999). On the dimensional and hierarchical structure of affect. Psychological Science, 10, 297-303.

52.

Velicer

W. F.

(1976). Determining the number of components from the matrix of partial correlations. Psychometrika, 41, 321-327.

53.

Watson

(2000). Mood and temperament. New York, NY: Guilford Press.

54.

Watson

(2005). Rethinking the mood and anxiety disorders: A quantitative hierarchical model for DSM-V. Journal of Abnormal Psychology, 114, 522-536.

55.

Watson

Clark

L. A.

(1984). Negative affectivity: The disposition to experience negative affective states. Psychological Bulletin, 96, 465-490.

56.

Watson

Clark

L. A.

(1992). Affects separable and inseparable: On the hierarchical arrangement of the negative affects. Journal of Personality and Social Psychology, 62, 489-505.

57.

Watson

Clark

L. A.

(1997). Extraversion and its positive emotional core. In Hogan

Johnson

Briggs

(Eds.), Handbook of personality psychology (pp. 767-793). San Diego, CA: Academic Press.

58.

Watson

Clark

L. A.

(1999). The PANAS-X: Manual for the Positive and Negative Affect Schedule-Expanded Form. Retrieved from http://ir.uiowa.edu/psychology_pubs/11/

59.

Watson

Clark

L. A.

Carey

(1988). Positive and negative affectivity and their relation to anxiety and depressive disorders. Journal of Abnormal Psychology, 97, 346-353.

60.

Watson

Clark

L. A.

Stasik

S. M.

(2011). Emotions and the emotional disorders: A quantitative hierarchical perspective. International Journal of Clinical and Health Psychology, 11, 429-442.

61.

Watson

O’Hara

M. W.

Naragon-Gainey

Koffel

Chmielewski

Kotov

. . .Ruggero

C. J.

(2012). Development and validation of new anxiety and bipolar symptom scales for an expanded version of the IDAS (the IDAS-II). Assessment, 19, 399-420.

62.

Watson

O’Hara

M. W.

Simms

L. J.

Kotov

Chmielewski

McDade-Montez

E. A.

. . .Stuart

(2007). Development and validation of the Inventory of Depression and Anxiety Symptoms (IDAS). Psychological Assessment, 19, 253-268.

63.

Watson

Naragon-Gainey

(2010). On the specificity of positive emotional dysfunction in psychopathology: Evidence from the mood and anxiety disorders and schizophrenia/schizotypy. Clinical Psychology Review, 30, 839-848.

64.

Watson

Stasik

M. R.

Chmielewski

Naragon-Gainey

(2014). Development and validation of the Temperament and Affectivity Inventory (TAI). Manuscript submitted for publication.

65.

Watson

Tellegen

(1985). Toward a consensual structure of mood. Psychological Bulletin, 98, 219-235.

66.

Watson

Vaidya

J. G.

(2013). Mood measurement: Current status and future directions. In Schinka

J. A.

Velicer

W. F.

(Eds.), Handbook of psychology, Volume 2: Research methods in psychology (pp. 369-394). Hoboken, NJ: Wiley.

67.

Watson

Wiese

Vaidya

Tellegen

(1999). The two general activation systems of affect: Structural findings, evolutionary considerations, and psychobiological evidence. Journal of Personality and Social Psychology, 76, 820-838.

68.

Zuckerman

Lubin

(1985). Manual for the MAACL-R: The Multiple Adjective Check List-Revised. San Diego, CA: Educational and Industrial Testing Service.