Optimism and Pessimism in Social Context: An Interpersonal Perspective on Resilience and Risk

Abstract

Using the interpersonal perspective, we examined social correlates of dispositional optimism. In Study 1, optimism and pessimism were associated with warm-dominant and hostile-submissive interpersonal styles, respectively, across four samples, and had expected associations with social support and interpersonal stressors. In 300 married couples, Study 2 replicated these findings regarding interpersonal styles, using self-reports and spouse ratings. Optimism-pessimism also had significant actor and partner associations with marital quality. In Study 3 (120 couples), husbands’ and wives’ optimism predicted increases in their own marital adjustment over time, and husbands’ optimism predicted increases in wives’ marital adjustment. Thus, the interpersonal perspective is a useful integrative framework for examining social processes that could contribute to associations of optimism-pessimism with physical health and emotional adjustment.

1. Introduction

The tendency to hold optimistic expectations about future events is associated with better emotional adjustment, physical health, and general well-being (for reviews, see Boehm & Kubzansky, 2012; Carver, Scheier, & Segerstrom, 2010; Rasmussen, Scheier, & Greenhouse, 2009). Across multiple frameworks and measures dispositional optimism is conceptualized as a personality characteristic (Carver et al., 2010; Norem & Chang, 2001) that influences health and well-being through primarily intra-individual processes, such as adaptive self-regulation and coping. However, dispositional optimism is also related to important inter-personal processes that may contribute to its association with subsequent health and well-being (Carver et al., 2010).

Research on psychosocial risk and resilience factors such as dispositional optimism typically distinguishes between characteristics of people (e.g., personality traits) and aspects of their social circumstances (e.g., social support, quality of close relationships), and these two classes of risk factors are most commonly studied separately (Smith & Cundiff, 2013; Smith, Glazer, Ruiz, & Gallo, 2004). Separation of the person and the social context potentially impedes the development of a more integrative perspective, in the specific case of dispositional optimism and more generally in the study of risk and resilience. Here we use dispositional optimism to illustrate a general approach to the integration of personality-based risk and resilience factors with risk and resilience factors typically seen as aspects of the social environment. Our approach utilizes the interpersonal tradition in personality and clinical psychology (Horowitz & Strack, 2011; Kiesler, 1996; Pincus & Ansell, 2003), which describes personality traits and important features of the social environment as closely interconnected. We have previously used this perspective to integrate the role of individual differences in anger and hostility with social environmental risk factors, such as social support and conflict (Smith et al., 2004). Here, we present the overall approach, and apply it to dispositional optimism in a series of three studies.

1.1 Dispositional Optimism as Resilience and Risk

There are two predominant models of dispositional optimism. The first describes a bipolar dimension anchored by optimism (i.e., the general expectation that good things will happen in the future) and pessimism (i.e., the expectation of negative outcomes), primarily measured with the self-report Life Orientation Test (LOT; Scheier, Carver, & Bridges, 1994) though other measures are used (e.g., Giltay et al., 2004). In the second, dispositional optimism refers to the tendency to make internal, global, and stable causal attributions for positive events, and external, specific, and unstable attributions for negative events; pessimism is the opposite attributional pattern (Peterson, 2009). A variety of self-report measures and rating methods have been developed to assess these explanatory styles.

Dispositional optimism is a well-established correlate and longitudinal predictor of emotional adjustment (e.g., Andersson, 1996; Giltay, Zitman, & Kromhout, 2006); optimists are less likely to experience and develop symptoms of depression and other types of distress (for a review, see Carver et al., 2010). Optimism also predicts physical health outcomes, including longevity, all-cause mortality, and the development and course of cardiovascular disease (for reviews, see Boehm & Kubzansky, 2012; Rasmussen et al., 2009).

To date, most studies of mechanisms linking dispositional optimism with emotional adjustment and physical health emphasize self-regulation and goal-directed behavior (Carver & Scheier, 1998). When facing difficult but potentially controllable circumstances, optimists persist in effortful engagement and goal-striving, and cope through problem-solving and planning. In contrast, pessimists are more likely to disengage, withdraw effort, and cope through denial and wishful thinking (for reviews, see Carver et al., 2010; Nes & Segerstrom, 2006). Compared to pessimists, optimists are more likely to match their coping strategies with characteristics of the situation, and balance conflicting demands and goals (Segerstrom & Nes, 2006). They are also more likely to use problem-solving when facing controllable stressors, and to use adaptive emotion-focused coping (e.g., reappraisal, acceptance) and pursue alternative goals when facing uncontrollable stressors (Aspinwall & Richter, 1999).

However, optimism is also associated with seeking social support when facing difficulties (Nes & Segerstrom, 2006), and Carver et al. (2010) point out that dispositional optimism could influence health and well-being through this coping strategy and other interpersonal processes. For example, optimism is associated with the development of more social support among undergraduates moving away to college (Brissette, Scheier, & Carver, 2002), and compared to pessimists, optimists are generally better liked (Carver, Cus, & Scheier, 1994). Day to day social interactions are more positive for optimists (Raikkonen, Matthews, Flory, Owens, & Gump, 1999), and they typically report greater social support (e.g., Vollman, Antoniw, Hartung, & Renner, 2011) and less loneliness (Rius-Ottenheim et al., 2012).

In a study of dating couples, optimism was associated with greater relationship satisfaction - for individuals and their partners, both concurrently and over time (Srivastava, McGonigal, Richards, Butler, & Gross, 2006). These actor and partner effects (Kenny, Kashy, & Cook, 2006) of optimism were mediated by greater perceived support from the partner. Also, men's optimism, but not women's, predicted relationship continuation (Srivastava et al., 2006). In a sample of couples who were predominantly married and cohabitating, optimism had similar concurrent and longitudinal actor and partner effects on relationship quality, and this association was due in part to more cooperative relationship problem solving among optimists (Assad, Donnellen, & Conger, 2007). In this case women's optimism, but not men's, predicted continuation of the relationship over time (Assad et al., 2007). However, it is important to note that neither Assad et al. (2007) nor Srivastava et al. (2006) examined the possible overlapping role of depression in these prospective associations with relationship outcomes.

These interpersonal correlates of dispositional optimism have clear relevance for health and well-being. Social support is a strong predictor of emotional adjustment, serious disease, and longevity (Marroquin, 2011; Holt-Lunstad, Smith, & Layton, 2010). Also, conflict and disruption (i.e., separation, divorce) in marriage and other close relationships similarly predicts emotional distress (Whisman & Schonbrun, 2010) and serious physical illness (e.g., Smith, Uchino, Berg, & Florsheim, 2012; for a review, see Smith, Baron, & Caska, in press). Hence, the association of dispositional optimism with social support and relationship quality could contribute to the association of this personality characteristic with multiple health outcomes.

1.2 An Interpersonal Perspective on Risk and Resilience

In pursuing this possibility, a systematic theory-driven approach to the integration of individual differences in dispositional optimism with health-relevant aspects of the social environment would be valuable. We have used the interpersonal perspective in personality and clinical psychology in this general endeavor (Smith et al., 2004; Smith & Cundiff, 2011; Smith, Cundiff, & Uchino, 2012; Smith et al., 2003; Smith et al., 2010). This perspective can be summarized through: (a) a general assumption about personality and social processes, (b) a model of the basic dimensions or structure of interpersonal behavior, and (c) a model of the process of social transactions. The general assumption is evident in Sullivan's (1953) definition of personality as, “the relatively enduring pattern of interpersonal situations which characterize a human life” (p. 111). Hence, in the interpersonal perspective individual differences such as dispositional optimism are closely related to recurring patterns of social behavior, as optimists and pessimists influence and are influenced by their social circumstances in distinct ways.

The structural model at the heart of this perspective is the interpersonal circumplex (IPC) (Kiesler, 1996; Leary, 1957). The two broad IPC dimensions – affiliation (i.e., warmth vs. hostility) and control (i.e., dominance vs. submissiveness) – describe momentary interpersonal events and actions, individual differences in social behavior (i.e., personality traits), and relatively stable features of interpersonal contexts, such as social support (Trobst, 2000) and relationship conflict (Smith et al., 2010). Given this range of application, the IPC can be used to identify similarities and differences among measures of a wide variety of personality traits (Wiggins & Broughton, 1991) and other psychosocial risk factors (Smith et al., 2010). The multiple correlation of a given characteristic with the two IPC dimensions of affiliation and control indicates the degree of its association with social behavior (i.e., “interpersonalness”); associations with the specific IPC dimensions describe the related interpersonal style (Gurtman, 1991, Smith et al., 2010). If an individual difference such as optimism-pessimism reflects substantial interpersonal content, it is plausible that interpersonal processes might be involved in its association with health and well-being (Smith & Cundiff, 2011; Smith et al., 2010).

The interpersonal perspective contains two primary elements regarding interactional processes. The principle of complementarity (Kiesler, 1996; Sadler, Ethier, & Woody, 2011) suggests that one actor's behavior tends to, “pull, invite, or evoke restricted classes of responses” from interaction partners (Pincus & Ansell, 2003, p.215). When described using the IPC, “invited” or evoked complementary behavior is similar in affiliation and opposite in control. That is, warmth invites warmth in return, whereas cold or hostile actions invite hostility. Dominant or controlling behavior invites submissiveness or deference; submissive actions invite dominant or controlling responses. However, research on complementarity suggests it is generally more apparent for affiliation than control (Sadler et al., 2011).

The processes underlying complementarity and social interactions generally are described in the transactional cycle. Aspects of an actor's internal experience (e.g., goals, motives, expectations, representations of others, interaction scripts) guide his or her overt behavior, as when mistrusting persons are outwardly guarded, cold, and even quarrelsome. This behavior restricts the experiences and reactions of interaction partners, as when they appraise the actor as rude or arrogant, leading that partner to behave in ways that are complementary to the actor's original negative expectations (i.e., cold, hostile). These transactional processes can shape the quality of social interactions across a variety of contexts, including close relationships.

Through these transactional processes, complementarity contributes to the aggregation of risk or resilience factors that are typically seen as characteristics of individuals (e.g., personality traits) and features of the social contexts they inhabit (e.g., support, conflict) (Gallo & Smith, 1999). For example, warm and trusting individuals generally evoke greater social support from others and lower levels of interpersonal conflict. In contrast, antagonistic persons are likely to experience less support and more frequent, severe, and prolonged conflict with others. In this way, risk factors for poor health and reduced well-being that are traditionally seen as reflecting intrapersonal characteristics may be more accurately seen as involving interpersonal processes - recurring patterns of person-environment transaction. The frequency, severity, and duration of exposure to interpersonal stressors resulting from these processes over time, as well as levels of social support versus isolation, could be principal mechanisms linking characteristics of persons – such as dispositional optimism – to subsequent health and well-being (Smith & Cundiff, 2011; Smith et al., 2004). Similar models describe the reciprocal association between negative affect and interpersonal events (Eberhart & Hammen, 2009; Liu & Alloy, 2010).

1.3 Overview of Present Studies

The present studies illustrate the application of this perspective to the integration of risk and resilience factors conceptualized as intrapersonal processes –such as dispositional optimism - with interpersonal factors that influence health and well-being. First, we determine interpersonal styles associated with dispositional optimism. We also test the resulting predictions based on the complementarity principle regarding associations of dispositional optimism with risk and resilience factors typically seen as features of the social environment (i.e., social support, aversive interpersonal experiences). Finally, we examine associations of dispositional optimism with marital adjustment - an important outcome in its own right, but one which also predicts emotional adjustment (Whisman & Schonbrun, 2010) and physical health (Smith et al., in press).

Rather than a single continuum, dispositional optimism may be more accurately seen as two separate albeit correlated optimism and pessimism dimensions (Marshall, Wortman, Kusulas, Hervig, & Vickers, 1992; Chang, Maydeu-Olivares, & D'Zurilla, 1997). Measurement modeling studies support this view (e.g., Herzberg, Glaesmer & Hoyer, 2006; Kubzansky, Kubzansky, & Maselko, 2004), although the two-dimensional model might not improve predictive utility (Segerstrom, Evans & Eusen-Moul, 2011). A correlated, two-dimensional structure raises questions about the extent to which optimism confers resilience, pessimism confers risk, or both. Studies of health outcomes have reported differing effects of the two dimensions (e.g., Milan, Richardson, Marks, Kemper, & McCutchan, 2004; Raikkonen et al., 1989; Schulz, Bookwala, Knapp, Scheier, & Williamson, 1996), although the literature as a whole suggests both dimensions are important (Rasmussen et al., 2009). In the present studies, we examined the interpersonal correlates of bipolar optimism, unipolar optimism, and unipolar pessimism.

2. Study 1: Circumplex Descriptions, Social Support, and Negative Social Experiences

The first step in this application of the interpersonal perspective is the identification of the interpersonal style associated with dispositional optimism. Using established procedures (Wiggins & Broughton, 1991) Terrill and colleagues (2010) regressed undergraduates’ LOT scores for bipolar optimism, unipolar optimism, and unipolar pessimism on IPC-based affiliation and control scores (Trapnell & Wiggins, 1990). The multiple R values ranged from .39 (uniploar pessimism) to .47 (bipolar optimism), suggesting moderate levels of interpersonal content (i.e., “interpersonalness”; Gurtman, 1991). Bipolar optimism and unipolar optimism were associated with higher control and affiliation, indicating a warm and dominant interpersonal style; unipolar pessimism was associated with low control and low affiliation, indicating a hostile and submissive interpersonal style. The aims of the present study were to replicate these IPC analyses of dispositional optimism across multiple samples, and to test related predictions based on the complementarity principle that bipolar and unipolar optimism would be associated with high levels of social support and low levels of aversive interpersonal experiences (e.g., conflict).

2.1 Method

2.1.1 Participants

Four samples of University of Utah undergraduates participated, receiving partial course credit (Sample 1: 150 women, 136 men; Sample 2: 146 women, 98 men; Sample 3: 141 women, 104 men; Sample 4: 77 women, 84 men). Mean age was 22 years, and approximately 84% were Caucasian, with Asian/pacific Islander (10%) and Hispanics (4%) the next largest groups.

2.1.2 Measures and Procedures

Participants completed measures in small groups. Dispositional optimism was assessed the 10-item LOT-R (Scheier et al., 1994). Three items reflect unipolar optimism (e.g., “In uncertain times, I usually expect the best”), and three reflect unipolar pessimism (e.g., “If something can go wrong for me, it will”); four filler items are not scored. The pessimism items are reversed scores in calculating bipolar optimism (i.e., total LOT scores). An extensive body of research supports the construct validity of the scale (Carver et al., 2010).

The IPC was measured with the Interpersonal Adjectives Scales –Big Five (IASR-B5; Trapnell & Wiggins, 1990). This 64-ittem, adjective-based measure generates scores for the IPC affiliation and control dimensions. Both the octant and dimension scores scales have demonstrated high levels of internal consistency, expected circumplex structure, and construct validity (Trapnell & Wiggins, 1990; Wiggins & Broughton, 1991).

The Test of Negative Social Exchanges (TENSE) measures exposure to aversive social experiences (Ruehlman & Karoly, 1991), with four subscales: Insensitivity, Hostility/Impatience, Interference, and Ridicule. The total score and subscale scores have good internal consistency and evidence of construct validity (e.g., Gallo, Smith, & Ruiz, 2003). The Interpersonal Support Evaluation List—12 (ISEL-12) (Cohen & Hoberman, 1983) assesses three aspects of social support with Likert-type items: Appraisal Support, Belonging Support, and Tangible Support. Internal consistency of the subscale scores ranges from .80 to .90, and an extensive body of research supports the construct validity of the subscales and total social support scale (Cohen, 2008). Participants in all four samples completed the LOT-R and IAS-R. Participants in sample 1 completed the TENSE and ISEL. Degrees of freedom vary slightly due to incomplete questionnaire responses. Analyses testing gender differences in associations reported below were all non-significant. Therefore, results are reported for men and women together.

2.2 Results and Discussion

Multiple regression results for total LOT-R scores, and uniploar optimism and pessimism scores, are presented in Table 1 and depicted in Figure 1. As in prior research (Terrill et al., 2010), total LOT-R scores and unipolar optimism scores were associated with a friendly-dominant interpersonal style, whereas unipolar pessimism scores were associated with a hostile-submissive style. The multiple R values indicate a moderate degree of interpersonal content.

Table 1.

Multiple Regression Results for Associations of Optimism-Pessimism Scales with IAS-R Affiliation and Control Dimensions in Study 1

	Mean (SD)	α	R	R2	F	β CON	β AFF
Sample 1					df (2,239)
Total LOT	18.4 (4.2)	.80	.48	.24	36.6***	.21***	.23***
Optimism	10.8 (2.4)	.78	.46	.21	32.2***	.25***	.18***
Pessimism	7.4 (2.4)	.65	.40	.16	22.9***	−.12*	−.21***
Sample 2					df (2,224)
Total LOT	19.6 (4.7)	.81	.43	.18	25.1***	.13*	.26***
Optimism	11.2 (2.3)	.80	46	21	30.0***	.27***	.24***
Pessimism	8.3 (2.7)	.62	.39	.15	20.5***	−.01	−.27***
Sample 3					df (2,228)
Total LOT	22.5 (5.7)	.83	.55	.31	50.3***	.29***	.35***
Optimism	13.8 (3.3) .78		.50	.25	38.2***	.36***	.32***
Pessimism	8.6 (3.4)	.82	.46	.21	30.5***	−.17***	−.28***
Sample 4					df (2,156)
Total LOT	22.2 (4.0)	.73	.52	.27	28.1***	.39***	.29***
Optimism	11.2 (2.3)	.60	.48	.23	22.8***	.38***	.25***
Pessimism	7.0 (2.4)	.69	.34	.12	10.1***	−.27***	−.17*

^*p<.05

**p<.01

^***p<.001 (two-tailed)

Figure 1.

Circumplex locations of Total LOT, optimism, and pessimism for Study 1, from multiple regression analyses predicting optimism-pessimism from IAS-R control and affiliation scales. Outer radius represents a multiple R of .6; inner radius represents multiple R of .3.

As expected, higher total LOT and unipolar optimism scores were associated with greater social support and lower levels of negative interpersonal experiences, whereas higher unipolar pessimism scores were associated with lower support and greater levels of negative interpersonal experiences (see Table 2). When examined simultaneously unipolar optimism and pessimism were independently associated with total ISEL scores and total TENSE scores (see Table 3).

Table 2.

Correlations of Total LOT, Optimism, and Pessimism with Social Support (ISEL) and Aversive Interpersonal Experiences (TENSE) in Study 1

	Total LOT	Optimism	Pessimism
ISEL – Social Support
Appraisal	.37***	.35***	−.30***
Belonging	.43***	.38***	−.38***
Tangible	.36***	.32***	−.31***
Total Support	.47***	.44***	−.40***
TENSE - Aversive Experience
Hostile/Impatience	−.24***	−.23***	.19**
Insensitivity	−.37***	−.32***	.34***
Interference	−.21***	−.13*	.24***
Ridicule	−.21***	−.22***	.16*
Total Aversive Experience	−.32***	−.32***	.29***

N = 237

^*p<.05

^**p<.01

^***p<.001 (two-tailed)

Table 3.

Multiple Regression Results for Associations of Social Support and Aversive Interpersonal Experiences with Optimism and Pessimism in Study 1

	R	R2	F	β Optimism	β Pessimism
ISEL			df (2, 240)
Total Support	.42	.18	25.8***	.27***	−.21**
Appraisal	.37	.14	19.2***	.27***	−.15*
Belonging	.43	.19	26.8***	.24***	−.25***
Tangible	.36	.13	17.1***	.22**	−.18*
TENSE
Total	.32	.10	13.9***	−.18*	.19*
Hostility	.24	.06	7.3***	−.17*	.09
Insensitivity	.37	.14	19.2***	−.19**	.23**
Interference	.24	.06	7.1***	.00	.24**
Ridicule	.22	.05	6.22**	−.18*	.06

^*p<.05

^**p<.01

^***p<.001 (two-tailed)

These results provide consistent evidence that higher bipolar and unipolar optimism are associated with a warm and dominant interpersonal style, whereas unipolar pessimism reflects a hostile and submissive style. Consistent with the complementarity principle, higher bipolar and unipolar optimism were associated with greater reports of social support and lower reports of negative interpersonal experiences, and higher unipolar pessimism was associated with less social support and greater negative interpersonal experiences. Importantly, unipolar optimism and pessimism generally had independent associations with these interpersonal outcomes.

There are potentially important limitations of Study 1. The undergraduate samples raise concerns about generalizability, and associations of these traits with interpersonal style could be inflated by the common method variance among self-report scales. Nonetheless, these results provide important initial support for the interpersonal perspective on dispositional optimism.

3. Study2: Interpersonal Style and Concurrent Relationship Quality in Married Couples

Marriage provides an important opportunity to address limitations of Study 1, while replicating and extending those findings. In Study 2 we replicated associations of dispositional optimism with self-reported interpersonal style, and extended them to the perspective of spouses, in middle-aged and older married couples. Associations of dispositional optimism with spouse ratings of affiliation and control avoid the potential inflation of these effects by common method variance, and provide important evidence regarding the interpersonal impact of optimism.

We also examined associations of dispositional optimism with marital functioning. We included measures of general marital adjustment and specific positive and negative aspects of this relationship (i.e., support from spouses vs. marital conflict). Dyadic analyses tested associations of dispositional optimism with the individual's own interpersonal experiences, and with their partner's, again providing evidence regarding the interpersonal impact of these traits. Finally, because optimism is associated with depression, which in turn is a well-established correlate of marital dysfunction (Whisman & Schonbrun, 2010), we also examined the actor and partner effects of optimism-pessimism on marital functioning while controlling depression.

3.1 Method

3.1.1 Participants

Participants were 147 middle aged (Wives, M = 43.9 years old; Husbands, M = 45.8) and 154 older couples (Wives, M = 62.2 years old; Husbands, M = 64.7) from the Salt Lake City, Utah area (96% Caucasian). Eligibility included: 1) married for at least 5 years, 2) one member either between 40 and 50 years old or between 60 and 70 years old, and 3) no more than a 10-year age difference. Couples were married for an average of 28 years.

3.1.2 Measures and Procedure

Participants completed the LOT-R. To measure the IPC, participants completed self-report and spouse-rating versions of the NEO-PI-R (Costa & McCrae, 1992), from which 48 items are used to measure IPC octants (i.e., 6 items each). In this sample we confirmed the internal consistency and circumplex structure of the self-report and spouse rating versions of these scales, as well as the convergent and discriminant validity of octant scales and the control and affiliation dimension scores derived from the octant scales (Traupman et al., 2009).

Participants also completed the Locke-Wallace Marital Adjustment Test (MAT; Locke & Wallace, 1959), which has good internal consistency (alpha = .90) and extensive evidence of construct validity (e.g., Smith et al., 2010), the Support and Conflict subscales of the Quality of Relationship Inventory (QRI; Pierce, Sarason, & Sarason, 1991) (alphas = .80, .89, respectively), and the widely-used Center for Epidemiologic Studies Depression Scale (CESD; Radloff, 1977).

Participants were recruited through a polling firm, advertisements, and presentations at community programs. Husbands and wives received separate survey packets that included a consent form, a demographic questionnaire, the LOT-R, QRI and MAT. Participants later completed a second packet of questionnaires, including the two versions of the NEO PI-R.

3.2 Results

3.2.1 Dispositional Optimism and Interpersonal Style

Results of regressions of husbands’ and wives’ total LOT, unipolar optimism, and unipolar pessimism scores on self-reported affiliation and control scores are presented in Table 4 and depicted in Figure 2 (panel A). Higher total LOT scores and unipolar optimism scores were associated with a warm and dominant interpersonal style, whereas higher unipolar pessimism was associated with a hostile and somewhat submissive style. The multiple R values again suggested a moderate degree of interpersonal content. Results of regressions of husbands’ and wives’ dispositional optimism on spouse ratings of the participants’ affiliation and control are also presented in Table 4 and depicted in Figure 2 (panel B). Compared to the associations with self-reported interpersonal style, the associations with spouse ratings are somewhat smaller, but significant and describe a similar interpersonal style in each case. Hence, common method variance may contribute to - but clearly does not account for - the association of dispositional optimism with interpersonal style. Further, the associations of optimism and pessimism with spouse ratings of affiliation and control indicate a definite interpersonal impact of these traits.

Table 4.

Multiple Regression Results for Associations of Wives' and Husbands' Total LOT, Optimism, and Pessimism Scales with Self-Reported and Spouse Rated NEO-IPC Affiliation and Control Scores in Study 2.

Self-Report	R	R2	F (2,297)	β Control	β Affiliation
Wife
Total LOT	.39	.16	27.3***	.25***	.26***
Optimism	.37	.14	23.5***	.23***	.26***
Pessimism	.33	.11	17.9***	−.21***	−.21***
Husband
Total LOT	.53	.28	58.6***	.33***	.37***
Optimism	.51	.26	51.0***	.29***	.37***
Pessimism	.41	.17	30.1***	−.27***	−.27***
Spouse Rating
Wife
Total LOT	.24	.06	9.4***	.14*	.20***
Optimism	.25	.06	9.8***	.16**	.19***
Pessimism	.19	.04	5.4**	−.09	−.16**
Husband
Total LOT	.40	.16	29.0***	.29***	.26***
Optimism	.53	.13	21.2***	.24***	.24***
Pessimism	.34	.12	19.4***	−.26***	−.20***

^*p<.05

^**p<01

^***p<001 (two-tailed)

Figure 2.

Circumplex locations of Total LOT, optimism, and pessimism for self-reports (Panel A) and spouse ratings (Panel B) of NEO-PI-IPC trait affiliation and control scales for Study 2. Outer radius represents a multiple R of .6; inner radius represents multiple R of .3.

3.2.2 Associations with Marital Functioning

Correlations of husbands’ and wives’ dispositional optimism with MAT, QRI-support, and QRI-conflict scores are presented in Table 5. For both husbands and wives, total LOT, and unipolar optimism, and pessimism scores were significantly associated with their own reports of all three marital outcomes. Husbands’ total LOT, optimism, and pessimism scores were also significantly associated with their wives’ reports of each of the three marital outcomes, with the exception of wives’ QRI-support. Wives’ total LOT, optimism, and pessimism scores were each significantly associated with their husbands’ reports of marital conflict. Wives’ total LOT and optimism scores were significantly associated with husbands’ MAT scores, but wives’ pessimism scores were not. Correlations of wives’ total LOT, optimism, and pessimism scores with husbands’ reports of support from their spouses were not significant.

Table 5.

Correlations of Total LOT, Optimism, and Pessimism with Marital Adjustment (MAT), Perceived Support from Spouse, and Conflict with Spouse in Study 2.

		WIVES			HUSBANDS
	Mean(SD)	T-LOT	OPT	PESS	T-LOT	OPT	PESS
WIVES
MAT	112.8 (26.2)	.33***	.29***	−.30***	.22***	.19***	−.18**
Support	24.0 (3.8)	.29***	.24***	−.26***	.15**	.09	−.16**
Conflict	26.3 (7.2)	−.34***	−.29***	.31***	−.20***	−.18**	.17**
HUSBANDS
MAT	114.9 (24.1)	.15**	.20***	−.08	.38***	.33***	−.32***
Support	24.9 (2.9)	.04	.06	−.01	.30***	.28***	−.24***
Conflict	26.1 (6.1)	−.20***	−.22***	.13*	−.28***	−.18***	.29***
Mean (SD)		16.6(4.2)	8.3 (2.2)	3.7 (2.6)	16.3(4.0)	8.3 (2.2)	4.0 (2.5)

N = 300

^*p<.05

^**p<.01

^***p<.001

We conducted dyadic SEM analyses (Kenny et al., 2006) of the actor and partner effects of dispositional optimism on each of the three marital functioning measures, considered separately. Specifically, we performed path analyses in Amos 7.0, and tested actor and partner effects of unipolar optimism and unipolar pessimism on the three self-reported measures of marital quality (MAT, QRI-Conflict, QRI-Support). We considered unipolar optimism and unipolar pessimism simultaneously to address the issue of their independent effects. Total LOT scores (i.e., bipolar optimism) were not included, given its high degree of association with the unipolar optimism and pessimism scales. The results are depicted in Figure 3. For overall marital adjustment (i.e., MAT scores; see panel A) there were significant actor effects of unipolar optimism for both husbands and wives, and unipolar pessimism for both husbands and wives. The partner effects of optimism were also significant for both husbands and wives, but neither partner effect of pessimism was significant. These effects did not change when controlling for depression, with the exception that the actor effect of wives’ optimism was no longer significant. Thus, for both husbands and wives higher optimism was independently associated with greater self-reported marital satisfaction, both within (i.e., actor effects) and between spouses (i.e., partner effects). Higher pessimism was independently associated with lower marital satisfaction for both husbands and wives, but only for the individual's own reports of marital quality.

Figure 3.

Associations of wives’ and husbands’ optimism and pessimism with marital adjustment (Panel A), perceived support from spouse (Panel B), and marital conflict (panel C). (*p<.05; **p<.01;***p<.001, all two-tailed)

The analyses of perceived support from the spouse (Panel B) revealed significant actor effects of unipolar optimism for both husbands and wives, and unipolar pessimism again for both husbands and wives. Higher optimism was associated with greater perceived support from the spouse, and higher pessimism was independently associated with lower perceived support. No partner effects were significant. Most of these effects became non-significant when controlling for depression; only the actor effect of husbands’ optimism remained significant (B = .18, p < .01). Finally, analyses of reported marital conflict (Panel C) revealed a significant actor effect for wives’ optimism but not husbands’ optimism. The actor effect of pessimism was significant for both husbands and wives. Further, the partner effect of optimism was significant for both husbands and wives, but neither partner effect of pessimism was significant. These effects did not change when analyses controlled for depression, with the exception that the actor effect of wives’ optimism was no longer significant.

3.3 Discussion

These results replicate and extend Study 1. Using a sample of middle-aged and older adults and a different IPC measure, we replicated the association of total LOT scores, unipolar optimism, and unipolar pessimism with self-reported interpersonal style; higher LOT scores and unipolar optimism scores were associated with a warm and dominant style, whereas unipolar pessimism was associated with a hostile and submissive style. Further, analyses using spouse ratings of affiliation and control provided evidence that this interpersonal style does not simply reflect common method variance, but instead involves behavior that is apparent to spouses.

Given these interpersonal styles, the complementarity principle predicts that optimism should be associated with generally warm marital interactions, whereas pessimism should be associated with cold or quarrelsome interactions with the spouse. In both actor and partner associations, these predictions were largely supported. For overall marital adjustment, unipolar optimism and unipolar pessimism both had significant independent actor effects for both husbands and wives. Only optimism had significant partner effects, again for both husbands and wives. Also for both husbands and wives optimism and pessimism had significant actor effects on perceived support from the spouse. However, there were no significant partner effects on perceived support. In analyses of reported conflict with the spouse, the actor effect of optimism was significant for wives but not husbands, whereas the actor effects of pessimism were significant for both spouses. The independent partner effect of optimism on conflict with the spouse was significant for both spouses, but neither spouse effect of pessimism was significant.

Several aspects of these results are noteworthy. First, the significant partner effects of optimism on overall marital adjustment and conflict with the spouse provide additional evidence that correlates of this trait in close relationships do not simply involve common method variance or the individual's own perceptions, but reflect interpersonal processes. Also, the independent associations involving optimism and pessimism suggest that they have distinct associations with marital quality, consistent with the view that they can be examined separately (e.g., Kubzansky et al., 2004; Terrill et al., 2010). The partner associations for optimism but not pessimism also support this distinction and suggest that the positive interpersonal impact of optimism may be greater than the negative impact of pessimism. Finally, many of these associations – notably all of the partner effects of optimism – remained significant when controlling depressive symptoms. However, the overlap with depressive symptoms is clearly an important concern, as several of the actor effects were eliminated when depressive symptoms were controlled.

These results extend prior studies of dispositional optimism and close relationships by providing a more specific depiction of the associations of these traits with multiple aspects of relationship quality. These findings point to a potential interpersonal mechanism linking dispositional optimism with physical health, emotional adjustment, and general well-being. However, the cross-sectional design creates the possibility that associations observed here are due to effects of relationship experiences on optimistic versus pessimistic expectations, rather than the interpersonal consequences of these traits.

4. Study 3: Concurrent and Prospective Association with Marital Adjustment

To address the primary limitation of Study 2, we examined the longitudinal association of dispositional optimism and pessimism with change in marital adjustment over an 18-month period. Specifically, we examined the longitudinal actor and partner effects of optimism and pessimism on change in marital adjustment, and we also tested these associations when controlling the potentially overlapping effects of depression.

4.1 Method

4.1.1 Participants

A total of 122 married couples from the Salt Lake City area participated in the initial assessment of personality and relationship quality. Couples were required to have been married for a minimum of nine months, and most (52%) were married between 1 and 3 years. Average age was 30.1 years for men, and 28.5 years for women. Nearly all couples (92%) were in their first marriage, and consistent with the demographics of the surrounding community most participants were Caucasian (88%), with a small number of Hispanic (5.3%) and Asian (3.5%) participants. Most participants were employed full or part time (81% of men, 64% of women), and many were currently enrolled full or part time at the University (64% men; 44% women).

4.1.2 Measures and Procedure

Participants individually completed the LOT-R, MAT, and the CESD. A follow-up MAT was mailed to participants an average of 18 months later. Participants who did not respond were contacted by phone, and completed a telephone version (Krokoff, 1989). There were no baseline or follow-up differences on any measure between participants who completed the second MAT by mail versus the telephone. At follow-up 78% of the couples provided MAT scores for both husbands and wives. Completers and non-completers did not differ on baseline measures or demographic variables. For women, the initial (115.8; SD = 22.6) and follow-up (117.3 SD = 25.5) MAT scores did not differ, and were similarly unchanged for men (112.0 vs 117.3; SDs = 22.1, 25.4), respectively. Hence, on average participants reported stable levels of marital adjustment within the range expected for non-distressed couples (Locke & Wallace, 1959).

4.2 Results and Discussion

Concurrent correlations of dispositional optimism with MAT scores are presented in Table 6. With one exception (i.e., correlation of husbands’ unipolar optimism with wives’ MAT), the within and between spouse associations were significant and consistent with the results of Study 2; higher total LOT scores and unipolar optimism were associated with greater marital adjustment, whereas higher unipolar pessimism scores were associated with lower levels.

Table 6.

Correlations of Total LOT, Optimism, and Pessimism with Marital Adjustment (MAT) at Time 1 in Study 3.

Wives	Means (SD)	Wives' MAT	Husbands' MAT
Total LOT	23.5 (4.5)	.32***	.34***
Optimism	11.7 (2.3)	.34***	.33***
Pessimism	6.1 (2.6)	−.25**	−.29***
Husbands
Total LOT	23.4 (4.2)	.22*	.28**
Optimism	12.0 (2.2)	.10	.21*
Pessimism	6.3 (2.2)	−.27**	−.34***
Means (SD)		1115.8 (22.6)	112.0 (22.1)

N = 300

^*p<.05

^**p<.01

^***p<.001

In dyadic longitudinal SEM analyses testing the independent associations of unipolar optimism and pessimism at Time 1 with MAT scores at Time 2 (controlling Time 1 MAT) (see Figure 4), the actor effects of both husbands’ and wives’ optimism on their own change in marital adjustment (i.e., time 2 MAT, controlling time 1 MAT) were significant; higher optimism scores were associated with higher subsequent reports marital adjustment. There was also a significant partner effect of husbands’ optimism on wives’ change in marital adjustment, such that the more optimistic the husband reported being at baseline, the higher the marital adjustment reported by the wife 18 months later. A comparison of a model that held the partner effects constant for husbands and wives with one that allowed them to vary was significant, X² (1) = 3.8, p = .05, indicating that the partner effect of husbands’ optimism on wives’ marital adjustment was larger than the partner effect of wives’ optimism. No actor or partner effects of pessimism were significant. These effects did not change after controlling for depression.

Figure 4.

Associations of wives’ and husbands’ optimism and pessimism with 18-month change in marital adjustment in Study 3. (*p<.05; **p<.01;***p<.001, all two-tailed)

These results indicate that husbands’ and wives’ optimism is associated with changes in their own marital satisfaction over time, whereas pessimism had no independent effects. In terms of interpersonal impacts, the partner effect of husbands’ optimism on wives’ change in marital satisfaction was significant. Hence, our specific longitudinal findings were more consistent with those of Srivastava et al. (2006) than those of Assad et al. (2007). These longitudinal effects indicate that associations of optimism with more positive marital adjustment seen in Study 2 cannot be explained readily as simply reflecting the effects of experiences in close relationships on optimistic versus pessimistic expectations. Rather, both within and between spouses, optimism predicts beneficial changes in the quality of this important relationship.

5. General Discussion

Across multiple samples and IPC measures, higher bipolar and unipolar optimism were associated with a warm and dominant interpersonal style, whereas higher unipolar pessimism was associated with a cold, aloof or hostile and somewhat submissive style. This pattern was demonstrated using both self-reports of social behavior where common method variance might contribute to the associations, and using spouse ratings where common method variance is minimized. In interpersonal theory the warm and dominant style of optimists is predicted to “pull, invite, or evoke” (Pincus & Ansell, 2003, p.215) warm and accommodating actions (e.g., cooperation, trust) in return. In contrast, the cold, hostile, and submissive style of the pessimist is predicted to invite distance, but also quarrelsomeness and criticism.

Consistent with these predictions, higher bipolar and unipolar optimism was associated with increased social support and decreased aversive interpersonal experiences (e.g., conflict), whereas unipolar pessimism was associated with the opposite pattern. This occurred both in general reports of social support and aversive experiences, and in the specific case of marriage, both cross-sectionally and longitudinally. These effects cannot be attributed solely to the common method variance, nor can they be explained fully through the overlapping effects of depressive symptoms. Importantly, when tested simultaneously there was evidence of both interpersonal benefits of optimism and liabilities of pessimism, in associations with general social support, general negative interpersonal experiences, and both positive and negative aspects of marital quality. However, only the benefits of optimism were apparent longitudinally.

The independent associations of optimism and pessimism with social functioning and relationship quality are consistent with the view that optimism confers resilience in a way that cannot be simply attributed to low pessimism. However, forcing statistical independence between the inversely correlated measures of optimism and pessimism creates the possibility that the resulting predictors reflect method variance rather than the constructs of substantive interest (c.f., Maydeu-Olivares & Coffman, 2006). Consistent with prior research, the correlations between the unipolar scales in the present samples ranged from r = −.66 (women, Study 3) to r = −.44 (Study 1, sample 4), with a median of r = −.52. However, it is important to note that all significant effects for the independent associations in analyses that considered optimism and pessimism simultaneously were also significant when these scales were considered separately. Also, effects for unidimensional optimism were essentially identical to results for the total LOT (i.e., bipolar optimism). Thus, the independent effects of unipolar optimism seem to reflect substantive associations rather than simple method artifacts resulting from forced separation of optimism and pessimism, and support the view that optimism is a unique resilience factor.

Currently, most studies of risk or resilience conceptualize personality factors such as optimism and features of the social environment as separate classes of influence on health and well-being. In contrast, the interpersonal perspective encourages the combined consideration of personality and key relationship contexts. The present studies illustrate the initial steps in applying the interpersonal perspective as a general theory-driven approach to the commonly observed aggregation of personality and social-environmental risk and resilience factors. Once such aggregations are described through these initial applications of interpersonal concepts and methods, the broader perspective suggests that studies of psychosocial influences on emotional adjustment and physical health should model personality and social relationship factors together, rather than only consider them separately, because their aggregation may more directly reflect the processes through which they influence health and well-being.

Further, this perspective suggests that psychosocial risk and resilience are conferred through recurring patterns of interpersonal experience involving variations in affiliation and control. Specifically, resilience may be conferred through greater exposure to affiliative experiences, lower levels of cold and quarrelsome experiences, less frequent and pronounced effortful exertion of influence or control over others, and less exposure to unwelcome control from others. In contrast, risk is conferred by the opposite pattern of experiences (Smith et al., 2010). These recurring experiential manifestations of the dynamic aggregation of personality and social-environmental risk or resilience factors may capture influences on health and well-being more directly than do the static and separate risks implied by concepts and measures reflecting personality characteristics or social relationship factors when they are considered alone. Hence, studies directly assessing such daily experiences (c.f., Kamarck et al., 2012; Moskowitz, 2010; Raikkonen et al., 1999) and examining the specific transactional processes linking personality-based risk or resilience factors with social-environmental factors may further explicate these proximal mechanisms. This integrative and dynamic approach may extend our understanding of psychosocial influences on emotional adjustment and physical health, and perhaps suggest refined approaches to intervention.

highlights.

Dispositional optimism predicts health and well-being.

Interpersonal processes may contribute to these associations.

The interpersonal perspective provides a useful integrative framework.

Optimism is associated with a warm and dominant interpersonal style.

Optimism is associated with greater social support and marital quality.