Abstract
Contemporary perspectives on relationship commitment posit that intimates decide whether or not to maintain a relationship based on their commitment to that relationship, and that they base such commitment partially on their current satisfaction with that relationship. Nevertheless, given that ending a relationship requires knowing about both the current state of the relationship and the likely future state of the relationship, we propose that people base their commitment to a relationship more on their expected future satisfaction with the relationship than on their current satisfaction with that relationship. Six studies provided evidence for these ideas. Study 1 demonstrated that expected satisfaction is shaped by not only current satisfaction but also several unique indicators of the likelihood of future satisfaction, including anticipated life events, plans to improve the relationship, and individual differences. Then, using a combination of cross-sectional, experimental, and longitudinal methods, Studies 2–6 demonstrated that (a) expected satisfaction was a stronger predictor of relationship commitment, maintenance behaviors, and/or divorce than was current satisfaction and (b) expected satisfaction mediated the association between current satisfaction and these outcomes. These findings highlight not only the need to incorporate expected satisfaction into extent perspectives on commitment, but also the importance of expectations for decision-making processes more broadly.
Keywords: Expectations, Commitment, Satisfaction, Romantic Relationships, Interdependence theory
One of the most important and challenging interpersonal decisions people face is whether to continue or dissolve a close relationship. Given that forming and maintaining strong social bonds is essential for survival, well-being, and the fulfillment of personal goals (see Baumeister & Leary, 1995; Feeney & Collins, 2014; Fitzsimons, Finkel, & vanDellen, 2015), people should benefit to the extent that they choose to maintain relationships that help them meet these needs. Nevertheless, not all relationships are equally helpful in this regard. In fact, some relationships are marked by hostility, selfishness, distrust, and/or physical violence (e.g., Amar & Alexy, 2005; Campbell, 1999; Hellmuth & McNulty, 2008; McNulty & Russell, 2010; Reis, Clark, & Holmes, 2004; Spanier & Margolis, 1983; Weissman, 1987), and poor quality relationships are harmful for both physical and mental health (Proulx, Helmes, & Buehler, 2007; Robles, Slatcher, Trombello, & McGinn, 2014). Thus, optimizing personal well-being often requires choosing whether or not to end a relationship.
Existing theoretical perspectives (e.g., Jacobson & Margolin, 1979; Karney & Bradbury, 1995; Kelley, 1983; Kenrick, Groth, Trost, & Sadalla, 1993; Markman, 1979; Rusbult, 1980a; Shackelford & Buss, 1997) suggest that intimates base this decision on the extent to which they are currently satisfied with that relationship. Nevertheless, nearly all relationships encounter problems (McGonagle, Kessler, & Schilling, 1992) that lead to dissatisfaction with those relationships (McNulty, O’Mara, & Karney, 2008; McNulty & Russell, 2010), and a robust body of work demonstrates that many intimates remain in their relationships despite fluctuating or declining relationship satisfaction (Arriaga, 2001; Karney & Bradbury, 1997; Kurdek, 1998; Lavner & Bradbury, 2010; Le, Dove, Agnew, Korn, & Mutso, 2010; McNulty & Karney, 2001; McNulty, O’Mara, & Karney, 2008; Meltzer, McNulty, Jackson & Karney, 2014; Stewart, Copeland, Chester, Malley, & Barenbaum, 1997; for review, see Karney & Bradbury, 1995). The inclination to remain in a relationship despite current dissatisfaction may at times be functional—such commitment can motivate intimates to engage in processes that may prove to benefit the relationship (Rusbult, Olsen, Davis, & Hannon, 2001). Nevertheless, not all attempts to improve a relationship will be successful, and thus deciding whether or not to dissolve a relationship requires considering what the future of the relationship will be like. For these reasons, we argue that intimates may base their commitment to a relationship on their expectations regarding the likelihood that they will be satisfied in the future rather than whether or not they are satisfied at the present; such expected satisfaction is likely based partially on current satisfaction but should also reflect additional factors relevant to the future.
The remainder of this introduction is organized into four sections. The first section briefly reviews theoretical and empirical work on relationships suggesting that intimates base their commitment to a relationship primarily on their current satisfaction with that relationship. The second section, in contrast, reviews theoretical and empirical challenges to this perspective that suggest that current satisfaction provides limited information on which to base relationship decisions and thus expected satisfaction serves as a more functional and proximal source of commitment. The third section identifies unique sources of expected satisfaction that may cause it to diverge from current satisfaction. Finally, the fourth section describes three cross-sectional studies, three experiments, and one longitudinal study that tested these ideas.
Current Conceptualizations of Commitment
Numerous theoretical perspectives address whether or not intimates will be committed to, and thus maintain, their close relationships, and most of these perspectives emphasize the role of intimates’ current satisfaction with their relationship. For example, behavioral theories (e.g., Jacobson & Margolin, 1979; Markman, 1979), the vulnerability-stress-adaptation model (Karney & Bradbury, 1995), and evolutionary perspectives (e.g., Kenrick et al., 1993; Shackelford & Buss, 1997), all suggest that intimates’ commitment to a relationship is determined by how satisfied they currently are with that relationship. These perspectives are based primarily on interdependence theory, one of the most influential theoretical perspectives in relationship science (Finkel & Simpson, 2015). According to contemporary conceptualizations of interdependence theory (e.g., Agnew, Arriaga, & Wilson, 2008; Kelley, Holmes, Kerr, Reis, Rusbult, & Van Lange, 2003), intimates should be more committed to a relationship to the extent that they—(a) have previously invested resources into that relationship, (b) believe the relationship is better than their alternatives to the relationship, and (c) are currently satisfied with the relationship. Regarding the latter, two factors determine intimates’ current satisfaction with the relationship: (a) the frequency and strength of the relationship rewards and costs they have experienced with the relationship thus far and (b) how those experiences compare to their interpersonal standards (Rusbult & Van Lange, 2003). More specifically, intimates should be more satisfied with a relationship to the extent that their prior experiences with that relationship meet or exceed their standards.
Empirical evidence is consistent with these ideas. For example, cross-sectional studies demonstrate that current relationship satisfaction is associated with greater relationship rewards and fewer relationship costs (e.g., Arriaga, 2001; Rusbult, 1980b; Rusbult, Johnson, & Morrow, 1986), and that relationship commitment is associated with greater relationship investments, fewer relationship alternatives, and greater current relationship satisfaction (e.g., Duffy & Rusbult, 1986; Gaertner & Foshee, 1999; Sanderson & Kurdek, 1993). Longitudinal studies have provided similar support (e.g., Bui, Peplau, & Hill, 1996; Impett, Beals, & Peplau, 2001; Le, Dove, Agnew, Korn, & Mutso, 2010; Rusbult, 1983; VanderDrift, Wilson, & Agnew, 2013). For example, Rusbult (1983) asked participants to report their current relationship satisfaction, investments, perceived alternatives, and commitment approximately every three weeks for seven months and found that intimates’ commitment to their relationship over the course of the seven months was negatively associated with their perceived alternatives and positively associated with their investments and current relationship satisfaction. Likewise, Bui and colleagues (1996) demonstrated that intimates who were more satisfied with, or perceived fewer alternatives to, their romantic relationship were more likely to remain in that relationship 15 years later, and that their commitment to that relationship mediated these effects. Most notably, Le and Agnew (2003) conducted a meta-analysis of 60 samples that tested the main tenets of interdependence theory and, crucially, documented that current relationship satisfaction appeared to be the strongest predictor of commitment.
Limitations of Current Conceptualizations of Commitment
Nevertheless, there is reason to believe that current satisfaction may not be the most functional or proximal source of relationship commitment. Most people encounter problems in their relationships and thus experience corresponding declines in relationship satisfaction (Arriaga, 2001; Karney & Bradbury, 1997; Kurdek, 1998; Lavner & Bradbury, 2010; Le et al., 2010; Meltzer et al., 2014; Stewart et al., 1997; for review, see Karney & Bradbury, 1995). According to current conceptualizations of commitment, such declines in relationship satisfaction should be accompanied by corresponding declines in commitment. Yet, many intimates remain committed to their relationships despite declines in current satisfaction (for review, see Karney & Bradbury, 1995).
We argue that this disconnect emerges because the primary function of current satisfaction is not to motivate commitment but to signal problems with the relationship that have already been experienced. According to Carver and Scheier’s (1998) theory of self-regulation, positive emotions signal to people that they have met their goals (e.g., are in a quality relationship) whereas negative emotions signal to people that they are not meeting their goals (e.g., are not in a quality relationship). From this perspective, current satisfaction is a summary of the positive and negative emotions already experienced in the relationship (see Rusbult & Van Lange, 2003) that may lead intimates to decide to attempt to resolve their problems or not. And given such attempts may be successful or unsuccessful, positive and negative emotions, and thus this overall summary, tend to fluctuate over time as people encounter problems and either successfully resolve them or not (Arriaga, 2001; McNulty & Karney, 2001; McNulty et al., 2008).
In contrast, it is the primary function of expected emotions to inform decisions about the future (Dennett, 1991; see Olson, Roese, & Zanna, 1996). For example, expectancy-value theories (Atkinson, 1957; Feather, 1982) have long highlighted the important role that expectations play in the decision-making process, positing that people base their decisions on the likelihood that such decisions will produce desired outcomes, and a large body of research is consistent with this idea (for reviews, see Seligman, Railton, Baumeister, & Sripada, 2013; Wigfield, Tonks, & Klauda, 2009). Similarly, research on affective forecasting (Chapman & Coups, 2006; Gilbert & Ebert, 2002; Green et al., 2013; for review, see Gilbert & Wilson, 2007) demonstrates that people consider the future emotional consequences when making decisions. Indeed, research outside the domain of relationships suggests expected evaluations and emotions are more important to decisions than are current ones (Bushman et al., 2001; DeWall et al., 2016; Freedman et al., 1967; Gross, 1998; Manucia et al., 1984; Silverman, 1967; Mellers et al., 1997; Resulaj et al., 2009; Tice et al., 2001; Vogel et al., 2005; for review, see Baumeister et al., 2007).
In line with these ideas, the fluctuating nature of relationship satisfaction highlights the fact that an informed decision regarding whether to remain in or dissolve a relationship requires considering the likelihood that it will be satisfying in the future. In the next section, we develop the argument that expected satisfaction is a stronger predictor of commitment than is current satisfaction by arguing expected satisfaction is based on not only current satisfaction, which reflects prior experience, but also three additional factors relevant to the future of the relationship: plans to improve the relationship, any anticipated life events, and individual differences.
The Sources and Function of Expected Relationship Satisfaction
Most expectations are based partly on prior experience (Olson et al., 1996), and expected satisfaction is no exception. Not only does Carver and Scheier’s (1998) theory of self-regulation posit that current emotions signal to people whether or not they have met their goals, as noted earlier, it also posits that the primary function of such affective experiences is to shape expectations for the future. Thus, not only should current satisfaction serve as a summary of the extent to which people’s prior experiences have met or not met their relationship goals, it should also partially inform people’s expectations about whether they will continue to meet or not meet their goals in the future. Consistent with these ideas, numerous studies indicate that current emotions play a powerful role in predicting people’s expectations regarding their future success (Erber, 1991; Feather, 1966; Feather & Saville, 1967; Forgas & Moylan, 1987; Johnson & Tversky, 1983; Ross, Lepper, Strack, & Steinmetz, 1977), and several studies of relationships specifically indicate that intimates’ relationship expectations are shaped by their current satisfaction (Lemay, 2016; Lemay, Lin, & Muir, 2015; McNulty & Karney, 2002, 2004; Neff & Geers, 2013). This overlap between current and expected satisfaction likely accounts for the associations between current satisfaction and commitment observed in prior research (Le & Agnew, 2003).
Nevertheless, we argue that there are several independent sources of expected satisfaction that can cause it to functionally diverge from current satisfaction and thus independently predict commitment. One such source is anticipated life changes. People experience numerous predictable events in their lives that have nontrivial implications for their relationships. The most obvious of these are those that directly involve the relationship, such as having a child, beginning to cohabitate with a partner, becoming engaged or married, or moving closer or further away from a partner. But even events that do not directly involve the relationship, such as beginning a new job, losing a job, becoming ill, or moving to a new residence can alter life experiences in ways that change the relationship (e.g., McCubbin & Patterson 1983). Although the direction and extent to which such changes shape expected satisfaction may vary from person to person, it is likely that most people at least recognize the possibility that such changes will affect their relationships. For example, research on the transition to parenthood indicates that intimates often expect that becoming a new parent will strain their romantic relationships (Lawrence, Nylen, & Cobb, 2007), and the extent to which new mothers expect their relationship quality will decline depends on how much assistance they expect their partners will provide (Ruble, Fleming, Hackel, & Stangor, 1988).
A second unique source of expected satisfaction that can cause it to diverge from current satisfaction and more strongly predict commitment is plans to improve the quality of the relationship. People often attempt to strengthen their relationships by disclosing their thoughts and feelings, behaving in a positive manner, providing emotional and instrumental support, and spending additional time with their partners (Feeney & Collins, 2014; Stafford, 2011). Similarly, people regularly attempt to resolve existing relationship problems by negotiating, compromising, sacrificing, and attempting to regulate one another’s behavior (Overall, Fletcher, Simpson, & Sibley, 2009; Rusbult, Verette, Whitney, Slovik, & Lipkus, 1991). Given such behaviors are often directly motivated by the desire to improve the relationship (e.g., Knee, Patrick, Vietor, Nanayakkara, & Neighbors, 2002), it is likely that they are followed by the expectation that their satisfaction with the relationship will improve. Research on couples who have attended therapy provides some evidence for this possibility by showing that, compared to waitlisted controls, couples involved in therapy expect their relationship satisfaction to improve (e.g., Carson, Carson, Gil, & Baucom, 2004).
A final unique source of expected satisfaction that can cause it to diverge from current satisfaction and more strongly predict commitment are individual differences. For example, one of the defining features of insecure attachment is a tendency to be less optimistic regarding the availability of close others (see Shaver & Mikuliner, 2002). Likewise, people with high self-esteem tend to be more confident that other will accept them (e.g., Murray, Holmes, MacDonald, & Ellsworth, 1998), and people with elevated levels of shyness tend to be less confident in their ability to resolve interpersonal problems (e.g., Baker & McNulty, 2010). Finally, research also indicates that individual differences in neuroticism are associated with more pessimistic interpersonal expectations (McNulty, 2008). Although such individual differences are associated with current satisfaction (e.g., Karney & Bradbury, 1995) and thus they may exert their influence on expected satisfaction through their effect on current satisfaction, they also likely have an independent effect on intimates’ expectations. For example, McNulty (2008) demonstrated that individuals high in neuroticism had more negative relationship expectations than individuals low in neuroticism, even after controlling for their current relationship satisfaction.
Overview of the Current Studies
Given these theoretical arguments, we made three predictions. First, we predicted that anticipated life events, plans to improve the relationship, and individual differences (e.g., neuroticism, self-efficacy) would predict expected satisfaction above and beyond current satisfaction, thereby leading expected relationship satisfaction to diverge from current satisfaction. Second, given this uniqueness, we predicted that expected relationship satisfaction would predict commitment and that this effect would be stronger than the corresponding effect of current satisfaction on commitment. Finally, given that intimates base their relationship expectations partly on their current levels of relationship satisfaction, we predicted that expected satisfaction would mediate the association between current satisfaction and relationship commitment that has been frequently observed in other research (Le & Agnew, 2003). 1
We conducted six studies to test these predictions. In Study 1, people in romantic relationships reported anticipated life events, plans to improve the quality of their relationship, individual differences (e.g., neuroticism, self-efficacy), current relationship satisfaction, and expectations for future relationship satisfaction. We tested whether anticipated life events, plans to improve the relationship, and individual differences explained variance in expected satisfaction that was unique from current satisfaction. The next five studies tested whether current satisfaction exerted unique and stronger effects on commitment than did current satisfaction and mediated the effects of current satisfaction that did emerge. In Study 2, people in romantic relationships reported their current relationship satisfaction, expectations for future relationship satisfaction, and current commitment to their relationship. In Study 3, we asked people to report their commitment to a hypothetical relationship after imagining that the relationship (a) was either currently satisfying or unsatisfying and (b) would be either satisfying or unsatisfying in the future. In Study 4, we assessed intimates’ commitment with their actual relationships after successfully manipulating their levels of current satisfaction and expected satisfaction with that relationship. To address concerns that any effects in these studies reflected semantic confounds related to the use of self-report measures of cognition, Studies 5 and 6 assessed the implications of current and expected satisfaction for two important behavioral indicators of commitment—relationship maintenance behaviors (Study 5) and divorce (Study 6). Specifically, in Study 5, newlywed couples reported their current and expected relationship satisfaction, participated in a task that assessed their tendency to attend to attractive others, and reported the extent to which their partners engaged in constructive relationship maintenance behaviors. In Study 6, newlywed couples reported their current relationship satisfaction, expectations for future satisfaction, and whether or not they remained married every six months for approximately four years. In an effort to maintain consistency despite the different designs and data obtained across studies, we always use a subsequent analysis to directly compare the magnitude of the association between expected satisfaction and commitment to the association between current satisfaction and commitment.
Study 1
Given that little prior research has identified sources of expectations for future relationship satisfaction, we first assessed intimates’ levels of current relationship satisfaction, expected relationship satisfaction, and several hypothesized predictors of each (e.g., anticipated life events, plans to improve the relationship, current rewards/costs), to examine whether such factors uniquely predict intimates’ current versus expected relationship satisfaction. Based on our theoretical rationale, we predicted that anticipated life events, plans to improve the relationship, and individual differences (e.g., neuroticism, relationship self-efficacy) would be associated with expected relationship satisfaction beyond their potential influence on current relationship satisfaction. Furthermore, based on interdependence theory and supporting research, we predicted that intimates’ current relationships costs and rewards would be associated with current relationship satisfaction; however, we predicted that current relationships costs and rewards would not be associated with expected relationship satisfaction beyond their potential influence on current relationship satisfaction.
Methods
Participants
Participants were 110 individuals (53 men, 57 women) who were in a relationship for at least three months and were recruited using the Mechanical Turk service on amazon.com (MTurk). This sample size was obtained because a power analysis anticipating medium effect sizes indicated that the power to detect an association between relationship satisfaction and the predicted determinants of satisfaction was greater than .95. Ten participants were excluded from analyses because they failed attention checks. The remaining 100 participants (48 men, 52 women) had a mean age of 35.33 years (SD = 11.15). Participants reported being in a relationship for an average of 98.53 (SD = 107.91) months. Forty-five (45%) participants were married, 48 (48%) reported they were in an exclusive relationship, five (5%) reported they were engaged, and two (2%) reported they were in a casual relationship. The majority of participants (n = 80; 80%) identified as White or Caucasian, 8 (8%) identified as Black or African American, 5 (5%) identified as Hispanic or Latino/a, 4 (4%) identified as Asian, and the remaining 3 (3%) identified as another ethnicity or two or more ethnicities. The majority of participants (n = 91; 91%) identified as heterosexual, five (5%) identified as gay or lesbian, three (3%) identified as bisexual, and one (1%) identified as other.
Procedure
Participants received $0.50 for completing the study online. After providing informed consent, participants completed the following measures that were presented through the university’s online participation site.
Measures
Current relationship satisfaction
Participants’ current relationship satisfaction was assessed with a version of the Quality Marriage Index (QMI; Norton, 1983) that was modified to ask participants about their relationship rather than their marriage. This measure requires individuals to report agreement with six items that assess relationship satisfaction (e.g., “My relationship with my partner makes me happy”). Five items ask participants to respond according to a 7-point scale, whereas one item asks participants to respond according to a 10-point scale. Thus, scores could range from 6 to 45, with higher scores reflecting greater relationship satisfaction. All items were summed. Internal consistency was high. (Coefficient alpha was .96.)
Expected relationship satisfaction
Participants’ expectations for their future relationship satisfaction were assessed with a modified version of the QMI (Norton, 1983) that was developed for this study. This version of the QMI instructs participants to report their “expectations for [their] relationship in the future,” using items modified to assess expected rather than current satisfaction (e.g., “I expect my relationship with my partner will make me happy”). All items were summed. Internal consistency was high. (Coefficient alpha was .97.)
Anticipated life events
To assess intimates’ anticipated life events, participants were asked to complete a modified version of the Life Experiences Survey (LES; Sarason, Johnson, & Siegel, 1978; see Neff & Broady, 2011). This version of the LES presents participants with 60 major life events across nine domains (e.g., health, legal, finances). For each event, participants were first asked the likelihood of each event occurring in the next year (0 = not at all likely to happen, 6 = very likely to happen). Because the implications of each event may differ for each participant (e.g., people may differ in whether they believe the birth of a new child will benefit or harm their romantic relationships), participants were also asked the anticipated impact of each event (−3 = this will make my relationship with my romantic partner much worse, 0 = this will not affect my relationship with my romantic partner, 3 = this will make my relationship with my romantic partner much better). Weighted event scores were created that were a product of the likelihood and impact scores, and these weighted scores were summed. Accordingly, positive scores indicated that participants anticipated more positive changes to the relationship whereas negative scores indicated that participants anticipated more negative changes to their relationships. Analyses controlled for the level of each component of this product (i.e., likelihood and impact).
Plans to improve the relationship
To assess intimates’ plans to improve their relationship, participants were asked the report the extent to which they planned to improve nine specific aspects of their relationship that are common sources of manageable conflict (e.g., sex life, amount of time spent together, the way you resolve disagreements; Storaasli & Markman, 1990) from 1 (do not plan to improve) to 7 (do plan to improve). All items were summed. Internal consistency was high. (Coefficient alpha was .94.)
Individual differences
We asked participants to complete the Experiences in Close Relationships-Short Form (Wei, Russell, Mallinckrodt, & Vogel, 2007; 12 items that range from 1 to 7; coefficient alpha was .84 for both attachment anxiety and avoidance) to assess attachment anxiety and avoidance, the Neuroticism subscale of the Big Five Personality Inventory-Short (Goldberg, 1999; 10 items that range from 1 to 5; coefficient alpha was .87), the Rosenberg self-esteem scale (1965; 10 items that range from 1 to 4; coefficient alpha was .89), and a relationship self-efficacy scale (Bradbury, 1989; seven items that range from 1 to 5; coefficient alpha was .89).
Current relationship rewards and costs
Given that the goal of the current study was to identify unique factors that predict current and expected relationship satisfaction, and given that interdependence theory posits that current relationship rewards and costs should determine current relationship satisfaction, we assessed current relationship rewards and costs with two face-valid items used by Rusbult and colleagues (Rusbult, Johnson, & Morrow, 1986). For both items, participants were first instructed to think about the current state of their relationship. To assess rewards, participants then responded to the following item: “The good traits your partner possesses and the good things about your relationship are termed ‘rewards.’ How rewarding is your relationship right now?” from 1 (not at all rewarding) to 7 (extremely rewarding). To assess costs, participants responded to the following item: “The bad traits your partner possesses and the bad things about your relationship are termed ‘costs.’ How costly is your relationship right now?” from 1 (not at all costly) to 7 (extremely costly).
Results
Descriptive statistics and correlations are presented in Table 1. Men and women did not differ in their current levels of satisfaction with their relationships, their expected levels of satisfaction, anticipated life events, plans to improve the relationship, attachment anxiety or avoidance, neuroticism, self-esteem, self-efficacy, or current relationship rewards or costs (all ps > .15).
Table 1.
Descriptive Statistics and Correlations among Variables in Study 1
| 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | |
|---|---|---|---|---|---|---|---|---|---|---|---|
| (1) Current satisfaction | -- | ||||||||||
| (2) Expected satisfaction | .86** | -- | |||||||||
| (3) Life events | .30** | .31** | -- | ||||||||
| (4) Plans | .06 | .16 | .17 | -- | |||||||
| (5) Attachment anxiety | −.49** | −.58** | −.07 | .13 | -- | ||||||
| (6) Attachment avoidance | −.63** | −.66** | −.20 | −.07 | .51** | -- | |||||
| (7) Neuroticism | −.31** | −.38** | .04 | .11 | .60** | .44** | -- | ||||
| (8) Self-esteem | .30** | .42** | −.00 | −.05 | −.54** | −.52** | −.71** | -- | |||
| (9) Self-efficacy | .64** | .64** | .02 | −.12 | −.55** | −.49** | −.41** | .40** | -- | ||
| (10) Current rewards | .81** | .73** | .30** | .23* | −.36** | −.53** | −.22* | .30** | .51** | -- | |
| (11) Current costs | .10 | .11 | .03 | .01 | −.11 | −.08 | −.08 | .04 | .01 | −.01 | -- |
|
| |||||||||||
| M | 36.52 | 37.27 | 8.00 | 32.81 | 17.96 | 15.29 | 24.05 | 31.24 | 25.88 | 5.73 | 4.35 |
| SD | 8.15 | 8.34 | 93.05 | 15.14 | 8.38 | 6.81 | 7.91 | 5.43 | 6.56 | 1.27 | 2.71 |
Note.
p < .05,
p < .01.
Despite the high reliability of both measures, current satisfaction was, as expected, not perfectly correlated with expected satisfaction. Instead, just over 25% of the variance in expected satisfaction was unique from current satisfaction. With the exception of plans and costs, all hypothesized predictors of both current and expected satisfaction were associated with both current and expected satisfaction. Nevertheless, these zero-order correlations do not control for the strong overlap between current and expected satisfaction and thus cannot address questions regarding the unique and independent sources of each. The primary analyses described next addressed these issues.
Hypothesized unique predictors of expected satisfaction
To examine whether the six factors we identified as possible unique predictors of expected satisfaction were indeed independent predictors of intimates’ expected relationship satisfaction, we conducted six regression models in which we separately regressed expected satisfaction scores onto one of the following factors: (a) weighted anticipated life event scores (controlling for likelihood and impact of the events), (b) plans to improve the relationship, (c) attachment anxiety and avoidance, (d) neuroticism, (e) self-esteem, and (f) relationship self-efficacy. To ensure that each factor was a unique predictor of expected satisfaction, each analysis controlled for current relationship satisfaction.
Results indicated that all six sources were uniquely associated with expected relationship satisfaction. Specifically, intimates’ expected satisfaction was positively associated with (a) anticipated life events, b = 0.01, SE = 0.01, t(95) = 2.10, p = .04, (b) plans to improve the relationship, b = 0.06, SE = 0.03, t(97) = 2.07, p = .04, (c) self-esteem, b = 0.26, SE = 0.08, t(97) = 3.35, p < .01, and (d) relationship self-efficacy, b = 0.18, SE = 0.08, t(97) = 2.22, p = .03, and negatively associated with (e) attachment anxiety, b = −0.16, SE = 0.06, t(96) = −2.98, p < .01, (f) attachment avoidance, b = −0.18, SE = 0.08, t(96) = −2.31, p = .02, and (g) neuroticism, b = −0.13, SE = 0.06, t(97) = −2.34, p = .02. A subsequent analysis that entered all sources simultaneously with current satisfaction demonstrated that our hypothesized factors accounted for 8% of the total variance in expected satisfaction, which was 31% of the variance in expected satisfaction that was unique from current satisfaction.
To identify whether these same factors also uniquely predicted current satisfaction, we repeated the same six regression analyses but this time regressed current satisfaction scores onto each of the factors and expected satisfaction scores. Only relationship self-efficacy was uniquely associated with current satisfaction, b = 0.18, SE = 0.08, t(97) = 2.29, p = .02; current satisfaction was not significantly associated with anticipated positive life events, b = 0.00, SE = 0.01, t(95) = 0.35, p = .73, plans to improve the relationship, b = −0.04, SE = 0.03, t(97) = −1.48, p = .14, attachment anxiety, b = 0.04, SE = 0.06, t(96) = 0.59, p = .58, attachment avoidance, b = −0.13, SE = 0.08, t(96) = −1.53, p = .13, neuroticism, b = 0.02, SE = 0.06, t(97) = 0.38, p = .71, or self−esteem, b = −0.10, SE = 0.08, t(97) = −1.23, p = .22.
Hypothesized unique predictors of current satisfaction
To examine whether the two factors we identified as possible unique predictors of current satisfaction were indeed independent predictors of intimates’ current relationship satisfaction, we conducted two regression models in which we separately regressed current satisfaction scores onto (a) rewards and (b) costs. To ensure that each factor was a unique predictor of expected satisfaction, each analysis controlled for expected relationship satisfaction.
Consistent with predictions, current satisfaction was positively associated with relationship rewards, b = 2.48, SE = 0.42, t(96) = 5.94, p < .01. Unexpectedly, but consistent with other research (Argyle & Furnham, 1983; Rusbult, 1983; Rusbult, et al., 1986; see Clark & Grote, 1998), current satisfaction was not associated with relationship costs, b = 0.10, SE = 0.14, t(96) = 0.75, p = .46.
To examine whether these same two factors uniquely predicted expected satisfaction, we repeated the same two regression analyses but this time regressed expected satisfaction scores onto each of the factors and current satisfaction scores. Consistent with predictions, neither current relationship rewards, b = 0.63, SE = 0.57, t(96) = 1.11, p = .27, nor costs, b = 0.11, SE = 0.16, t(96) = 0.71, p = .48, uniquely predicted expected satisfaction.
Discussion
Study 1 provides support for the idea that intimates can expect future levels of relationship satisfaction that differ from the satisfaction they are currently experiencing and evidence of several unique sources of such differences. Whereas current rewards uniquely predicted intimates’ current satisfaction and not their expected satisfaction, anticipated life events, plans to improve the quality of the relationship, attachment anxiety, attachment avoidance, neuroticism, and self-esteem uniquely predicted how satisfied intimates expected to be in the future and not their current satisfaction. Only relationship self-efficacy predicted both current and expected satisfaction, and relationship costs were unassociated with neither current, nor expected, satisfaction. Although this last finding is inconsistent with the predictions made by interdependence theory, previous research examining the implications of relationship costs has been mixed, with several studies revealing null associations between relationship costs and current satisfaction (Argyle & Furnham, 1983; Rusbult, 1983; Rusbult et al., 1986; see Clark & Grote, 1998). Overall, these current results provide evidence that intimates rely on different sources of information to determine their current versus expected satisfaction.
Particularly notable is the fact that three individual differences examined here, neuroticism, attachment insecurity, and self-esteem, were associated with current satisfaction when expected satisfaction was ignored but unassociated with relationship satisfaction once expected satisfaction was controlled. These findings join others (e.g., Fisher & McNulty, 2008; Little, McNulty, & Russell, 2010) in suggesting that the robust effects of these individual difference variables on relationship outcomes may be attributable to expected rather than current satisfaction. As addressed in more detail in the General Discussion section, future research may benefit from directly addressing this issue.
Nevertheless, with respect to the current questions, Study 1 did not test our two other primary hypotheses. That is, Study 1 did not examine whether intimates’ expectations for their future satisfaction would predict their commitment to the relationship more strongly than would their current satisfaction with that relationship and whether expected satisfaction would mediate any associations between current satisfaction and commitment. Armed with evidence that expected satisfaction is unique from current satisfaction and has unique sources, Study 2 addressed both of these predictions.
Study 2
Given that no prior research has examined the implications of naturally occurring expectations for future satisfaction with a relationship for commitment to that relationship, we conducted a cross-sectional study that assessed intimates’ levels of current relationship satisfaction, expected relationship satisfaction, and commitment. Based on our theoretical rationale, we predicted that intimates’ expectations for their future satisfaction would more strongly predict their commitment to the relationship than would their current satisfaction with that relationship, and that expected satisfaction would mediate any association between current satisfaction and commitment to that relationship. Further, given that interdependence perspectives posit that satisfaction with a current relationship and perceptions of alternatives to that relationship are independent predictors of commitment, and given that alternatives capture expected outcomes outside the relationship, we additionally assessed and controlled intimates’ perceptions of the quality of their alternatives to the current relationships.
Methods
Participants
Participants were 111 individuals (35 men, 75 women, 1 transgender) who were recruited using the Mechanical Turk service on amazon.com (MTurk). This sample size was obtained because it was the number of participants we obtained during one week of data collection. An a priori power analysis based on correlations obtained in previous research (Le and Agnew, 2003) indicated that it would provide power > .99 to detect an effect of current satisfaction on commitment, and a post-hoc power analysis indicated that the power to detect a difference between the effects of current and expected satisfaction on commitment was .87. Participants had a mean age of 32.74 years (SD = 10.78). All participants had been involved in a romantic relationship for at least three months (M = 68.31, SD = 82.12). Forty-eight (43%) participants were married, 44 (40%) reported they were in an exclusive relationship, 14 (13%) reported they were engaged, and five (5%) reported they were in a casual relationship. The majority of participants (n = 79; 71%) identified as White or Caucasian, 12 (11%) identified as Black or African American, 9 (8%) identified as Asian, 5 (5%) identified as Hispanic or Latino/a, 1 (1%) identified as American Indian or Alaska Native, and the remaining 5 (5%) identified as another ethnicity or two or more ethnicities. The majority of participants (n = 101; 91%) identified as heterosexual, four (4%) identified as gay or lesbian, three (3%) identified as bisexual, and three (3%) identified as other or indicated that they did not know.
Procedure
Participants received $0.20 for completing the study online. After providing informed consent, participants completed the following measures that were presented through the university’s online participation site.
Measures
Relationship commitment
Participants’ commitment to their relationship was assessed with the commitment subscale of Rusbult and colleagues’ (1998) Investment Model Scale. This measure requires individuals to report agreement with 7 items that assess relationship commitment (e.g., “I am committed to maintaining my relationship with my partner”) using a 9-point Likert response scale from 1 (do not agree at all) to 9 (agree completely). Appropriate items were reversed and all items were summed. Internal consistency was acceptable. (Coefficient alpha was .88.)
Current relationship satisfaction
Participants’ current relationship satisfaction was assessed with the modified version of the QMI that was used in Study 1. Internal consistency was high. (Coefficient alpha was .97.)
Expected relationship satisfaction
Participants’ expectations for their future relationship satisfaction were assessed with the modified expectations version of the QMI that was used in Study 1. Internal consistency was high. (Coefficient alpha was .98.)
Relationship alternatives
We assessed and controlled for relationship alternatives to examine whether the implications of intimates’ expectations about their future satisfaction for their relationship commitment are independent from how satisfying they expect alternative relationships would be. Participants’ relationship alternatives were assessed with the alternatives subscale of Rusbult and colleagues’ (1998) Investment Model Scale. This measure requires individuals to report agreement with 5 items that assess relationship alternatives (e.g., “My needs for intimacy, companionship, etc., could easily be fulfilled in an alternative relationship”) using a 9-point Likert response scale from 1 (do not agree at all) to 9 (agree completely). All items were summed. Internal consistency was acceptable. (Coefficient alpha was .92.)
Results
Descriptive statistics and correlations are presented in Table 2. Men and women did not differ in their current levels of satisfaction with their relationships, t(106) = −1.53, p = .13; however, women (M = 53.27, SD = 11.82) were more committed to their relationships than were men (M = 47.66, SD = 12.53), t(108) = −2.27, p = .03, women (M = 39.23, SD = 7.94) expected to be marginally happier with their relationships in the future than did men (M = 36.03, SD = 9.82), t(108) = −1.82, p = .07, and men (M = 23.80, SD = 10.27) reported more desirable alternatives to their current relationship than did women (M = 18.20, SD = 11.39), t(108) = 2.48, p = .02.
Table 2.
Descriptive Statistics and Correlations among Variables in Study 2
| 1 | 2 | 3 | 4 | |
|---|---|---|---|---|
| (1) Relationship commitment | -- | |||
| (2) Current satisfaction | .48** | -- | ||
| (3) Expected satisfaction | .61** | .74** | -- | |
| (4) Alternatives | −.47** | −.21** | −.24** | -- |
|
| ||||
| M | 51.59 | 37.35 | 38.23 | 20.01 |
| SD | 12.27 | 8.81 | 8.63 | 11.26 |
Note.
p < .01.
To examine whether expectations for satisfaction was a stronger predictor of relationship commitment than was current relationship satisfaction, we regressed participants’ commitment scores onto their current relationship satisfaction scores and expected satisfaction scores. In this analysis, current satisfaction was no longer significantly associated with commitment, b = 0.07, SE = 0.16, t(106) = 0.45, p = .65. Consistent with predictions, however, expectations for future satisfaction remained significantly associated with commitment, b = 0.81, SE = 0.16, t(106) = 5.02, p < .01. More importantly, a direct test (Fisher, 1921) indicated that expectations for satisfaction was a significantly stronger predictor than was current satisfaction, z = 3.09, p < .01. Finally, a subsequent analysis that regressed relationship commitment scores onto current satisfaction scores, expected satisfaction scores, and relationship alternatives scores indicated that alternatives were negatively associated with commitment, b = −0.37, SE = 0.08, t(105) = −4.61, p < .01, expected satisfaction remained positively associated with commitment, b = 0.72, SE = 0.15, t(105) = 4.83, p < .01, and current satisfaction remained not significantly associated with commitment, b = 0.04, SE = 0.15, t(105) = 0.25, p = .80.
To examine whether expected satisfaction mediated the effects of current satisfaction on commitment, we computed asymmetric confidence intervals for the mediated effect by following the procedure described by MacKinnon and colleagues (2007). This procedure tends to yield greater power than traditional bootstrapping approaches (Tofighi & MacKinnon, 2016). The first step was to demonstrate that current satisfaction is associated with the expected mediator—expectations for future satisfaction. To address this, we regressed participants’ expected satisfaction scores onto their current satisfaction scores. Consistent with the first criterion necessary for establishing mediation, current satisfaction was significantly associated with expected satisfaction, b = 0.73, SE = 0.06, t(107) = 11.40, p < .01. The second step was to demonstrate that expected satisfaction was associated with commitment, controlling for current satisfaction, which was shown in the previous section. Finally, we multiplied these two effects together to obtain an estimate of the mediated effect, b = 0.59, and computed the 95% confidence interval (0.35: 0.86), which was significant, supporting mediation. Notably, the fact that current satisfaction was not associated with commitment after controlling for expected satisfaction rules out the alternative mediational path—that current satisfaction mediates the association between expected satisfaction and commitment. Indeed, this alternative indirect effect was not significant (−0.18: 0.29).
Discussion
Study 2 provides initial support for our predictions. Specifically, intimates’ (a) current satisfaction with their relationship was positively associated with their commitment to that relationship in bivariate analyses, (b) current satisfaction was no longer associated with commitment once expectations for future satisfaction were controlled, (c) expected satisfaction was more strongly predictive of commitment than was current satisfaction, and (d) expected satisfaction appeared to mediate the association between their current satisfaction and their commitment to that relationship. Further, such effects emerged independently of intimates’ perceived relationship alternatives.
Nevertheless, Study 2 is limited in two important ways. First, Study 2 was correlational, obviating the ability to draw causal conclusions. In particular, in the context of such correlational designs, mediational analyses have limited ability to demonstrated that expected satisfaction mediates the relationship between current satisfaction and commitment (see Fiedler, Schott, Meiser, 2011). Second, given that items from both the commitment and expected satisfaction measures asked participants to think about the future, yet items from the current satisfaction scale asked participants to think about the present, it is possible that the stronger association between commitment and expected, rather than current, satisfaction obtained in Study 2 was due to this shared temporal focus. Study 3 addressed both of these issues by employing an experimental design and assessing both current and future commitment.
Study 3
Study 3 employed an experimental design in which participants were asked to report either their current or future commitment (between-person) after imagining they were in a relationship that (a) was currently satisfying or unsatisfying (between-person) and (b) would be either satisfying or unsatisfying in the future (between-person). This design is similar to the methods first used in the original test of the investment model (Rusbult, 1980b), except we manipulated relationship satisfaction rather than relationship costs, additionally manipulated expectations for future relationship satisfaction, and assessed present or future commitment rather than just present commitment. We predicted the manipulation of expectations for future satisfaction would predict commitment more strongly than the manipulation of current satisfaction, regardless of how commitment was assessed.
Methods
Participants
Participants were 400 individuals (166 men, 232 women, 1 transgender, 1 did not report) who were recruited using the Mechanical Turk service on amazon.com (MTurk). This sample size was obtained because a power analysis based on correlations obtained in Study 2 and experimental manipulations of related constructs in previous research (e.g., Rusbult, 1980b) indicated that the power to detect a difference between the effects of current and expected satisfaction on commitment was > .99. Twelve participants were excluded from analyses because they failed attention checks. The remaining 388 participants (162 men, 224 women, 1 transgender, 1 did not report) had a mean age of 33.79 years (SD = 10.78). The majority of participants were involved in a romantic relationship (n = 364; 94%) and the average relationship length was 82.43 months (SD = 107.38).2 The majority of participants (n = 291; 75%) identified as White or Caucasian, 40 (10%) identified as Black or African American, 23 (6%) identified as Asian, 19 (5%) identified as Hispanic or Latino/a, 2 (1%) identified as American Indian or Alaska Native, 1 (< 1%) identified as Native Hawaiian, or Pacific Islander, and the remaining 12 (3%) identified as another ethnicity or two or more ethnicities. The majority of participants (n = 339; 87%) identified as heterosexual, 17 (4%) identified as gay or lesbian, 26 (7%) identified as bisexual, and 6 (2%) identified as other or did not know.
Procedure
Participants received $0.10 for completing the study online. After providing informed consent, participants were randomly assigned to one of two expectations conditions, one of the two current satisfaction conditions, and one of the two commitment conditions (i.e., all manipulations were crossed). First, participants were asked to imagine that they were in a relationship that was (a) either currently satisfying or unsatisfying and (b) would be either satisfying or unsatisfying in the future. Specifically, participants read instructions that stated, “Imagine you are currently completely (un)satisfied with your romantic relationship and assume that you will be completely (un)satisfied with the relationship in the future.” Next, participants were asked to report how committed they would be to that relationship “this moment” or how committed they would be to that relationship “in the future.” Participants were then debriefed and paid for their participation.
Measures
Relationship commitment
All participants reported their commitment using a 7-point Likert response scale that ranged from 1 (very uncommitted) to 7 (very committed).
Results
Descriptive statistics and correlations are presented in Table 3. Women (M = 5.33, SD = 1.99) and men (M = 5.29, SD = 1.86) did not differ in regard to how committed they believed they would be, t(384) = −0.18, p = .86.
Table 3.
Descriptive Statistics and Correlations among Variables in Study 3
| 1 | 2 | 3 | 4 | |
|---|---|---|---|---|
| (1) Current satisfaction manipulation | -- | |||
| (2) Expected satisfaction manipulation | .04 | -- | ||
| (3) Commitment condition | .05 | −.14** | -- | |
| (4) Relationship commitment | .24** | .50** | −.11* | -- |
|
| ||||
| M | -- | -- | -- | 5.31 |
| SD | -- | -- | -- | 1.93 |
Note.
p < .05,
p < .01.
Consistent with previous research and the results from Study 2, intimates in the high current satisfaction condition (M = 5.81, SD = 1.62) were more committed to their relationships than were those in the low current satisfaction condition (M = 4.88, SD = 2.07), t(386) = 4.91, p < .01. Subsequent analyses indicated this effect was further moderated by whether they reported their current or future commitment, t(384) = 2.05, p = .04; although the current satisfaction manipulation did predict both current and future commitment, the manipulation was unexpectedly more strongly associated with future commitment, t(384) = 5.01, p < .01, than current commitment, t(384) = 2.15, p = .03.
Also consistent with the results from Study 2, intimates in the high expectations condition (M = 6.24, SD = 1.18) were more committed to their relationships than were those in the low expectations condition (M = 4.32, SD = 2.08), t(386) = 11.21, p < .01. Subsequent analyses indicated this effect was not further moderated by whether participants reported their current or future commitment, t(384) = 1.33, p = .18.
We tested the primary prediction that the expected satisfaction manipulation predicted commitment more strongly than the current satisfaction manipulation using a model that regressed participants’ commitment scores onto a dummy-code for the current satisfaction manipulation (0 = low, 1 = high) and a dummy-code for the expected satisfaction manipulation (0 = low, 1 = high). In this analysis, both expected satisfaction, b = 1.87, SE = 0.17, t(385) = 11.34, p < .01, and current satisfaction, b = 0.85, SE = 0.16, t(385) = 5.16, p < .01, significantly predicted commitment. Nevertheless, consistent with predictions, a direct test indicated that expected satisfaction was a significantly stronger predictor than was current satisfaction, z = 4.02, p < .01. This was true regardless of whether intimates reported their current commitment, z = 3.18, p < .01, or future commitment, z = 2.24, p = .03.
Discussion
Study 3 provides experimental support for our predictions. Consistent with previous research (e.g., Rusbult, 1980b), people who imagined being in a relationship that was currently satisfying reported that they would be more committed to that relationship than were people who imagined being in a relationship that was currently unsatisfying. Further, consistent with the results from Study 2, people who imagined being in a relationship that would be satisfying in the future reported that they would be more committed to that relationship than were people who imagined being in a relationship that would be unsatisfying in the future. But most importantly, consistent with predictions, expected satisfaction predicted commitment more strongly than did current satisfaction, regardless of whether it was assessed with respect to the present or future. In other words, the stronger predictive power of expectations emerged experimentally and is not an artifact of the way commitment is measured.
Study 4
Study 4 sought to provide experimental evidence that expected satisfaction with a real relationship predicts commitment to that relationship more strongly than does current satisfaction with that relationship. Using a sample of intimates in real relationships, we experimentally manipulated intimates’ expectations for their future satisfaction with their romantic relationship by having them describe qualities of their relationship that they expected to be either good or bad in the future; or we experimentally manipulated intimates’ current satisfaction with their relationship by having them describe qualities of their relationship that were currently either good or bad. Then, participants reported how committed they were to that relationship. We predicted that (a) the manipulation of expected satisfaction would predict commitment more strongly than the manipulation of current satisfaction and (b) any effects of the current satisfaction manipulation would be mediated by expected satisfaction.
Methods
Participants
Participants were 250 individuals (89 men, 160 women, 1 transgender) involved in a romantic relationship for a minimum of three months who were recruited using the Mechanical Turk (MTurk) service on amazon.com. We first recruited 200 participants, which an a priori power analysis indicated would have power > .80 to detect a difference between the effects of current and expected satisfaction on commitment, and after finding trends consistent with our predictions subsequently recruited an additional 50 participants.3 Two participants were excluded from analyses because they failed attention checks, indicating that they did not take the study seriously. The remaining 248 participants (160 women, 87 men, 1 transgender) had a mean age of 31.71 years (SD = 9.70). All participants had been involved in a romantic relationship for at least three months (M = 76.85, SD = 88.30). The majority of participants (n = 188; 76%) identified as White or Caucasian, 21 (9%) identified as Black or African American, 10 (4%) identified as Hispanic or Latino/a, 7 (3%) identified as Asian, 6 (2%) identified as American Indian or Alaska Native, and the remaining 16 (7%) identified as another ethnicity or two or more ethnicities. The majority of participants (n = 209; 84%) identified as heterosexual, nine (4%) identified as gay or lesbian, twenty-four (10%) identified as bisexual, and six (2%) identified as other or did not know.
Procedure
Participants received $0.20 for completing the study online. After providing informed consent, participants were randomly assigned to one of two expectations conditions or one of the two current satisfaction conditions using procedures that have been used to effectively manipulate current relationship satisfaction in previous research (e.g., Hofmann, Finkel, & Fitzsimons, 2015; Morry, 2005).
The current satisfaction manipulations asked participants to describe three qualities of their relationship that were currently either positive or negative. Specifically, those in the high current satisfaction condition read instructions that stated, “there are many things in our romantic relationships that work out well. Please think about the current state of your relationship… please think about and list three things that are currently good about your relationship” and those in the low current satisfaction condition read instructions that stated, “there are many things in our romantic relationships that do not work out so well. Please think about the current state of your relationship… please think about and list three things that are currently not good about your relationship.”
The expectations manipulations asked participants to describe three qualities of their relationship that they expected to be either positive or negative in the future. Specifically, those who were in the high expectations condition read instructions that stated, “there are many things in our romantic relationships that work out well. Please think about how you expect your relationship will be in the future… please think about and list three things that you expect will be good about your relationship in the future” and those who were in the low expectations condition read instructions that stated, “there are many things in our romantic relationships that do not work out so well. Please think about how you expect your relationship will be in the future… please think about and list three things that you expect will not be good about your relationship in the future.”
Finally, participants were asked to complete several additional questionnaires, which included manipulation checks and a measure of relationship commitment. These manipulation checks allowed us to examine whether any effects of the current satisfaction manipulations were mediated by expected satisfaction. Participants were then debriefed and given credit for their participation.
Measures
Relationship commitment
Given that Study 3 demonstrated that the implications of expected and current relationship outcomes emerged regardless of whether intimates considered their current or future commitment, we returned to using the well-validated commitment subscale of the Investment Model Scale (Rusbult et al., 1998). Internal consistency was acceptable. (Coefficient alpha was .93.)
Manipulation checks
To determine the effectiveness of each manipulation, we assessed participants’ current relationship satisfaction and their expectations for future satisfaction after the manipulations using the modified versions of the QMI that were used in Study 1. Internal consistency of both measures was acceptable. (Coefficient alpha was .96 for current satisfaction and .97 for expected satisfaction.)
Results
Descriptive statistics and correlations are presented in Table 4. Women were more committed to their relationships (M = 52.49, SD = 13.19) than were men (M = 47.62, SD = 13.39), t(245) = −2.76, p = .01.
Table 4.
Descriptive Statistics and Correlations among Variables in Study 4
| 1 | 2 | 3 | |
|---|---|---|---|
| (1) Current satisfaction manipulation | -- | ||
| (2) Expected satisfaction manipulation | -- | -- | |
| (3) Relationship commitment | .16† | .36** | -- |
|
| |||
| M | -- | -- | 50.83 |
| SD | -- | -- | 13.43 |
Note.
p < .10,
p < .01.
Our current satisfaction manipulation appeared to be successful; participants in the high current satisfaction condition reported that they were currently more satisfied with their relationship (M = 40.23, SD = 6.17) than did participants in the low current satisfaction condition, (M = 33.17, SD = 9.43), t(118) = 4.88, p < .01. Further, consistent with our prediction that current satisfaction affects commitment through its effects on expected satisfaction, participants in the high current satisfaction condition reported that they expected to be more satisfied with their relationship in the future (M = 40.93, SD = 5.19) than did participants in the low current satisfaction condition, (M = 35.64, SD = 10.15), t(115) = 3.54, p < .01.
Our expectations manipulation appeared to be successful as well; participants in the high expectations condition reported that they expected to be more satisfied with their relationship in the future (M = 40.73, SD = 6.41) than did participants in the low expectations condition, (M = 36.87, SD = 8.97), t(125) = 2.72, p = .01. Further, consistent with the idea that intimates do not base their current satisfaction on their expectations for future satisfaction and thus that current satisfaction does not mediate the association between expectations and commitment, participants in the high expectations condition reported that they were not currently more satisfied with their relationships (M = 39.49, SD = 7.17) than were participants in the low expectations condition, (M = 38.72, SD = 7.56), t(124) = 0.58, p = .56.
Consistent with previous research and the results from Studies 2–3, intimates in the high current satisfaction condition (M = 52.85, SD = 10.49) were marginally more committed to their relationships than were those in the low current satisfaction condition (M = 48.97, SD = 14.31), t(119) = 1.71, p = .09. Adjusting the p-value associated with this effect to account for adding data after conducting one set of analyses (see Sagarin et al., 2014) indicated that paugmented ranged from .07 to .12. Without the additional fifty participants, intimates in the high current satisfaction condition were trending toward being more committed to their relationships than were those in the low current satisfaction condition, t(92) = 1.43, p = .16.
Consistent with predictions and with the results from Studies 2–3, intimates in the high expectations condition (M = 56.38, SD = 7.85) were more committed to their relationships than were those in the low expectations condition (M = 46.23, SD = 16.44), t(125) = 4.25, p < .01, paugmented = [.01, .01]. Even without the additional fifty participants, intimates in the high expectations condition were more committed to their relationships than were those in the low expectations condition, t(102) = 3.39, p < .01.
Given that participants received either the current satisfaction or expectations manipulation, we were unable to estimate a model that included both the current satisfaction and the expectations manipulations. As such, to test whether the expectations manipulation affected commitment more than the current satisfaction manipulation did, we directly compared the effect sizes of each manipulation, which indicated that the expectations manipulation had a marginally stronger effect on intimates’ commitment than did the current satisfaction manipulation, z = 1.67, p = .09, paugmented = [.08, .12]. Before the additional fifty participants, the expectations manipulation did not have a stronger effect on intimates’ commitment than did the current satisfaction manipulation, z = 1.25, p = .21.
Given that our theoretical framework suggests current satisfaction only affects commitment through its influence on expected satisfaction, we computed asymmetric confidence intervals for that mediated effect by following the procedure described by MacKinnon, Fritz, Williams, & Lockwood (2007). This procedure required conducting two additional sets of analyses. The first step was to demonstrate that the current satisfaction manipulation predicted expectations for future satisfaction. To address this, we regressed the expected satisfaction manipulation check onto a dummy-code for the current satisfaction condition (0 = low current satisfaction, 1 = high current satisfaction). Consistent with the first criterion necessary for supporting mediation, the current satisfaction manipulation significantly predicted expected satisfaction, b = 5.27, SE = 1.49, t(115) = 3.54, p < .01, paugmented = [.01, .01]. Even without the additional fifty participants, the current satisfaction manipulation significantly predicted expected satisfaction, t(89) = 3.50, p < .01. The second step was to demonstrate that expected satisfaction was associated with commitment, controlling for the current satisfaction manipulation. To address this, we regressed participants’ commitment scores onto their expected satisfaction scores and the dummy-code for their current satisfaction condition. Consistent with the second criterion necessary for supporting mediation, expected satisfaction was significantly associated with commitment, b = 0.96, SE = 0.12, t(114) = 8.27, p < .01, paugmented = [.01, .01]. Even without the additional fifty participants, expected satisfaction was significantly associated with commitment, t(88) = 7.22, p < .01. Finally, we multiplied these two effects together to obtain an estimate of the mediated effect, b = 5.05, and computed the 95% confidence interval (2.18: 8.26), which was significant, providing support for the idea that expected satisfaction mediated the effect of current satisfaction on commitment. Notably, once expected satisfaction was controlled, the current satisfaction manipulation was no longer associated with commitment, b = −1.26, SE = 1.95, t(114) = −0.65, p = .52, suggesting that expected satisfaction completely accounted for the effects of current satisfaction on commitment. Even without the additional fifty participants, the 95% confidence interval (2.39: 9.49) indicated that the mediated effect was significant. Notably, the fact that the expected satisfaction manipulation did not affect intimates’ current satisfaction rules out the alternative mediational path—that current satisfaction mediates the association between expected satisfaction and commitment. Indeed, this alternative indirect effect was not significant (−1.07: 2.14).
Discussion
Study 4 provides further evidence for the idea that expected satisfaction is a stronger predictor of commitment than is current satisfaction. Although intimates who were led to be relatively satisfied currently with their relationship were marginally more committed to that relationship than were intimates led to be relatively less satisfied currently, intimates led to hold relatively more positive expectations for their future relationship satisfaction were significantly more committed to that relationship than were intimates led to hold relatively less positive expectations. Further, a direct test of the magnitude of the effects of each manipulation indicated that expectations were a marginally stronger predictor of commitment than was current satisfaction. Finally, mediational analyses supported the prediction that expected satisfaction would account for the effect of the current satisfaction manipulation. In sum, although Study 4 provided experimental evidence for that oft-observed association between current satisfaction and commitment, like Studies 2–3, it nevertheless provided evidence for the superiority of expected satisfaction in predicting commitment.
Nevertheless, even considered together, Studies 2–4 are limited in two important ways. First, although these studies revealed the implications of current and expected satisfaction for the proximal outcome specified by interdependence theory—relationship commitment—none of these studies examined the implications for the distal outcomes specified by interdependence theory—relationship maintenance. Second, by relying on self-reports of cognitions, it is possible that the associations between expected satisfaction and commitment observed in Studies 2–4 were due to common-method variance or common language used in the commitment and expected satisfaction measures. Study 5 addressed these issues by examining the association between self-reports of expectations and partner-reported and observed relationship maintenance processes.
Study 5
Study 5 expanded on Studies 1–4 by examining whether expected marital satisfaction, compared to current marital satisfaction, was more strongly associated with two specific relationship maintenance processes that indicate greater commitment—the extent to which spouses (a) avoided attending to tempting attractive alternative partners (see Maner, Gailliot, & Miller, 2009; Miller, 1997) and (b) attempted to constructively resolve problems in their relationships (see Rusbult, Verette, Whitney, Slovik, & Lipkus, 1991). Specifically, a sample of newlywed couples reported their current and expected marital satisfaction, participated in a visual dot-probe task that assessed their tendency to avoid attractive alternative partners, and reported the extent to which their partners engaged in constructive relationship maintenance behaviors. Based on the theoretical analysis described above and the results from the previous studies, we predicted that intimates’ expected marital satisfaction would predict these relationship maintenance processes more strongly than would their current marital satisfaction. Further, we predicted that intimates’ expected marital satisfaction would mediate the association between their current marital satisfaction and these relationship maintenance processes.
Methods
Participants
Participants in Study 5 were 63 newlywed couples participating in an ongoing broader study of marriage. Participants were recruited through invitations sent to eligible couples who had applied for marriage licenses in the county where the study took place. Couples who responded were screened in a telephone interview to ensure they met the following eligibility criteria: (a) they had been married for less than 3 months, (b) they were at least 18 years of age, and (c) they spoke English and had completed at least 10 years of education (to ensure comprehension of the questionnaires). This sample size was the maximum number of couples we were able to recruit in 8 months. A post-hoc power analysis indicated that the power to detect a difference between the effects of current and expected satisfaction on relationship maintenance processes was .60.
Husbands were 32.40 years old (SD = 8.90) on average and had received 16.34 years (SD = 3.28) of education. Sixty-three percent were White or Caucasian, twenty-five percent were Black or African American, and the remaining twelve percent were another or two or more ethnicities. Forty percent were Christian, 42% were agnostic, atheist, or reported that they were not religious, and the remaining 18% reported another religious affiliation. Seventy-three percent were employed full time and 14% were full-time students. Wives were 33.02 years old (SD = 9.78) on average and had received 17.90 years (SD = 2.69) of education. Sixty-six percent were White or Caucasian, 24% were Black or African American, and the remaining 10% were another or two or more ethnicities. Forty-four percent were Christian, 39% were agnostic, atheist, or reported that they were not religious, and the remaining 17% reported another religious affiliation. Sixty-seven percent were employed full time and 9% were full-time students.
Procedure
Couples first completed a series of questionnaires that included measures of current marital satisfaction, expected future marital satisfaction, the extent to which their spouses engaged in constructive relationship maintenance behaviors, and other questionnaires that are unrelated to the current hypotheses. All measures were completed online using Qualtrics survey software. After completing these questionnaires, participants attended a laboratory session where they completed a visual dot-probe task that assessed their automatic attention to attractive faces of their preferred gender. Couples were paid $100 for their participation.
Measures
Current marital satisfaction
Current relationship satisfaction was assessed with the QMI (Norton, 1983). Internal consistency was acceptable. (Coefficient alpha was above .94 for both husbands and wives.)
Expected future marital satisfaction
Expected future marital satisfaction was assessed with the same modified version of the QMI (Norton, 1983) used in previous studies, except this version asked participants about their marriage instead of their relationship. Internal consistency was acceptable. (Coefficient alpha was above .90 for both husbands and wives.)
Constructive problem-solving behavior
To assess spouses’ tendencies to engage in constructive problem-solving behavior, we asked each participant to report the extent to which their partners engaged in three specific conflict resolution behaviors (i.e., “How often does your partner make constructive proposals to solve the problem?” “…offer to change his or her behavior to solve the problem?” “…avoid the problem?” reverse-scored) when facing a relationship problem, using a 7-point scale from 1 (never) to 7 (frequently). We used partner-reports rather than self-reports to ensure that any obtained associations between current/expected marital satisfaction and relationship maintenance would not be due to common-method variance associated with reporting own satisfaction and behavior. Internal consistency was slightly lower than desired. (Coefficient alpha was .64 for husbands’ reports and .71 for wives’ reports.)
Automatic attention to attractive alternative partners
Attractive alternative relationship partners can threaten one’s current relationship, and thus committed intimates often avoid tempting alternative partners by diverting their attention away from such attractive others (e.g., Maner et al., 2009). To assess the extent to which participants diverted attention from such alternative others, participants completed a visual dot-probe task adapted by Maner and colleagues (e.g., Maner, Holm-Denoma, Van Orden, Gailliot, Gordon, & Joiner, 2006; Maner, Rouby, & Gonzaga, 2008) to assess how quickly they shift their attention away from attractive and average male and female faces. This computer task first instructed participants to focus their attention on a fixation cross (“X”) that appeared for 1000 ms in the center of the computer screen. Next, once the fixation cross disappeared, a picture of either an attractive male, average male, attractive female, or average female face appeared for 500 ms in one of the four quadrants of the computer screen. Next, once the face disappeared, a picture of a categorization object (circle or square) appeared in one of the four quadrants of the computer screen. Participants were instructed to press the “a” (circle) or “k” (square) key when they identified the categorization object. The speed with which participants respond reflects the attention captured by the face on the screen. Participants first completed a practice block of 20 trials and then a block of 20 target trials. An index of spouses’ attention to alternatives was formed that was an average of their reaction times to trials in which they saw an attractive face of the gender that they preferred (e.g., the average reaction time to trials with attractive female faces for heterosexual men). Higher scores indicate that participants took longer to divert their attention from attractive faces of their preferred gender.
Results
Descriptive statistics and correlations are presented in Table 5. Men and women did not differ in regards to their current levels of marital satisfaction, t(122) = −0.71, p = .68, their expectations for future marital satisfaction, t(124) = −0.00, p = .83, their reports of how constructive their partners were during problem solving discussions, t(124) = −0.79, p = .43, or how long it took to divert the attention from attractive faces of their preferred gender, t(124) = 0.38, p = .71.
Table 5.
Descriptive Statistics and Correlations among Variables in Study 5
| 1 | 2 | 3 | 4 | |
|---|---|---|---|---|
| (1) Current satisfaction | -- | -- | -- | -- |
| (2) Expected satisfaction | .55** | -- | -- | -- |
| (3) Constructive behavior | .15 | .28** | -- | -- |
| (4) Attention to alternatives | −.06 | −.19** | .03 | -- |
|
| ||||
| M | 41.56 | 42.24 | 15.25 | 647.99 |
| SD | 5.55 | 4.42 | 3.60 | 147.93 |
Note.
p < .10,
p < .05,
p < .01.
To test whether expected marital satisfaction was a stronger predictor of the relationship maintenance behaviors than was current marital satisfaction, we estimated two, two-level models in HLM 7.01, in which each maintenance behavior (i.e., constructive behavior, attention to alternatives) was regressed separately onto current marital satisfaction scores and expected marital satisfaction scores. The non-independence of couples’ data was controlled in the second level of the model that allowed for a randomly varying intercept. First, we regressed intimates’ constructive behavior scores, as reported by their partners, onto their current marital satisfaction scores and expected marital satisfaction scores. In this analysis, intimates’ current satisfaction was not significantly associated with their constructive behavior, b = −0.02, SE = 0.07, t(59) = −0.33, p = .75; however, consistent with predictions, their expectations for future satisfaction were significantly positively associated with their constructive behavior, b = 0.23, SE = 0.08, t(59) = 3.13, p < .01. Importantly, a direct test (Fisher, 1921) indicated that their expected satisfaction was a significantly stronger predictor than was their current satisfaction, z = 3.42, p < .01.
Second, we regressed intimates’ attention to alternatives scores onto their current marital satisfaction scores, expected marital satisfaction scores, and the mean latency for average male and female faces to control for their average reaction time. In this analysis, intimates’ current satisfaction was not significantly associated with the extent to which they attended to attractive faces of their preferred gender, b = 2.13, SE = 3.49, t(57) = 0.61, p = .54; however, consistent with predictions, their expectations for future satisfaction were significantly negatively associated with the extent to which they attended to attractive faces of their preferred gender, b = −7.46, SE = 3.01, t(57) = −2.48, p = .02. Further, expected satisfaction was a significantly stronger predictor than was current satisfaction, z = 3.14, p < .01.
Although bivariate correlations between current satisfaction and both relationship maintenance behaviors failed to reach significance, we next examined whether expectations for future satisfaction would mediate the association between current satisfaction and those behaviors. To do so, we computed asymmetric confidence intervals for the mediated effect by following the procedure described by MacKinnon and colleagues (2007). The first step was to demonstrate that current satisfaction is associated with the expected mediator—expected satisfaction—by regressing participants’ expected satisfaction scores onto their current satisfaction scores. Consistent with the first criterion necessary for supporting mediation, current satisfaction was significantly associated with expected satisfaction, b = 0.43, SE = 0.12, t(60) = 3.66, p < .01. The second step was to demonstrate that expected satisfaction was associated with both relationship maintenance behaviors, controlling for current satisfaction, which was shown in the previous section. Finally, we multiplied these two effects together to obtain an estimate of the mediated effect and computed the 95% confidence intervals. Both sets of confidence intervals indicated that the mediated effect was significant (for constructive behavior, b = 0.10, 95% CI = 0.03: 0.19; for attention to alternatives, b = −7.03, 95% CI = −7.06: −0.31). Notably, the fact that current satisfaction was not associated with either of the relationship maintenance behaviors after controlling for expected satisfaction rules out the alternative mediational paths—that current satisfaction mediates the association between expected satisfaction and relationship maintenance. Indeed, these alternative indirect effects were not significant (for constructive behavior, 95% CI = −0.07: 0.05; for automatic attention, 95% CI = −1.98: 4.13).
Discussion
Study 5 further demonstrates the relative importance of expected over current relationship satisfaction by demonstrating that expected marital satisfaction, compared to current marital satisfaction, was a significantly stronger predictor of two different relationship maintenance behaviors that are behavioral indicators of commitment—behaving constructively during problem-resolution discussions and diverting attention from attractive others. Finally, mediational analyses suggested that expected satisfaction mediated the association between current satisfaction and those relationship maintenance behaviors. By using partners’ reports of behavior and reaction time measures, the results from Study 5 suggest that the associations between self-reports of expectations and commitment revealed in Studies 2–4 were not simply the result of common-method variance.
Nevertheless, Studies 2–5 are limited in one important way. Specifically, although these studies revealed the implications of current and expected satisfaction for proximal outcome specified by interdependence theory—relationship commitment—none of these studies examined the implications for another distal outcome specified by interdependence theory—relationship stability. Study 6 addressed this issue.
Study 6
Study 6 expanded on Studies 1–5 by examining whether expected marital satisfaction predicted the likelihood of divorce more strongly than did current marital satisfaction. Specifically, a sample of newlywed couples reported their current and expected marital satisfaction and whether they were still married every 6 to 8 months for the first five years of their marriages. Based on the theoretical analysis described above and the results from the previous studies, we predicted that intimates’ expected satisfaction with their marriage would predict divorce more strongly than would their current satisfaction with that marriage. Further, we predicted that intimates’ expected satisfaction would mediate the association between their current satisfaction and their likelihood of divorce. Finally, given that we assessed naturally-occurring rather than manipulated current and expected satisfaction, we once again assessed and controlled for intimates’ perceptions of their alternatives to the marriage.
Methods
Participants
Participants in Study 6 were 135 heterosexual newlywed couples participating in broader longitudinal study of marriage. Participants were recruited through advertisements placed in community newspapers and bridal shops and through invitations sent to eligible couples who had applied for marriage licenses in counties near the study location. Couples who responded were screened in a telephone interview to ensure they met the following eligibility criteria: (a) they had been married for less than 6 months, (b) neither partner had been previously married, (c) they were at least 18 years of age, (d) they spoke English and had completed at least 10 years of education (to ensure comprehension of the questionnaires), and (e) did not yet have children (because a larger aim of the study was to examine the transition to parenthood). This sample size was the maximum number of couples we were able to recruit in 1 year. A post-hoc power analysis indicated that the power to detect a difference between the effects of current and expected satisfaction on relationship maintenance behaviors was .78.
At the start of the study, husbands were 25.90 years old (SD = 4.57) on average and had received 15.69 years (SD = 2.38) of education. Ninety-one percent were Caucasian and 76% were Christian. Seventy percent were employed full time and 26% were full-time students. Wives were 24.21 years old (SD = 3.59) on average and had received 15.92 years (SD = 2.29) of education. Ninety-three percent were Caucasian and 83% were Christian. Fifty-six percent were employed full time and 28% were full-time students.
Procedure
At baseline, couples were mailed a packet of questionnaires that included measures of current marital satisfaction, expectations for future marital satisfaction, perceived alternatives to the marriage, and other questionnaires that are unrelated to the current hypotheses. Couples completed those questionnaires at home and brought them to a laboratory session unrelated to the current analyses. At approximately 6- to 8-month intervals, couples were re-contacted, asked if they were still married, and mailed a packet of questionnaires that contained the same measures of current satisfaction, expected satisfaction, and alternatives, along with a postage-paid return envelope, and a letter reminding them to complete the questionnaires separate from one another. Couples were paid $80 for their baseline assessment and $50 for participating in each subsequent assessment. Analyses are based on up to seven assessments of current satisfaction, expected satisfaction, and alternatives, and eight assessments of divorce.
Measures
Current marital satisfaction
Current relationship satisfaction was assessed with the QMI (Norton, 1983). Internal consistency was acceptable. (Across all phases, coefficient alpha was above .87 for both husbands and wives.)
Expected future marital satisfaction
Expected future marital satisfaction was assessed with the same modified version of the QMI (Norton, 1983) used in Study 5. Internal consistency was acceptable. (Across all phases, coefficient alpha was above .83 for both husbands and wives.)
Perceived alternatives
Interdependence perspectives suggest it is not merely alternative relationship partners, but perceptions of alternative partners, lost investments/resources, and alternative lifestyles that predict commitment (e.g., Rusbult, 1980a; Thibaut & Kelley, 1959). Thus, we assessed and controlled for a measure likely to broadly capture all such perceived alternatives to the marriage developed by Frye, McNulty, and Karney (2008). This measure requires individuals to report agreement with 5 items (e.g., “I need to make this marriage last, whether or not I find it satisfying”) using a 7-point Likert response scale from 1 (not at all agree) to 5 (completely agree). All items were summed. Internal consistency was acceptable. (Across all phases, coefficient alpha was above .80 for both husbands and wives.)
Divorce
Participants were contacted at approximately 6- to 8-month intervals and were asked whether or not they were still married.
Results
Descriptive statistics and bivariate correlations between each type of satisfaction and whether a couple divorced across the duration of the study appear in Table 6. Growth curve models revealed that newlyweds’ current satisfaction, b = −0.49, SE = 0.08, t(1220) = −6.37, p < .01, and expected satisfaction, b = −0.34, SE = 0.07, t(1124) = −4.83, p < .01, both declined over the first five years of marriage. Husbands and wives’ did not differ in their current, t(1227) = −0.37, p = .71, or expected satisfaction, t(1131) = −0.72, p = .48. Consistent with the idea that motives operate more strongly on expected than current satisfaction (Karney & Frye, 2002), a multilevel regression analysis indicated that spouses’ average expected satisfaction was greater than their average current satisfaction, t(1110) = 5.67, p < .01. Nineteen (14%) couples divorced during the study; 3 couples had divorced by Time 2, 6 additional couples had divorced by Time 3; 3 additional couples had divorced by Time 4; 2 additional couples had divorced by Time 5; 1 additional couple had divorced by Time 6; 2 additional couple had divorced by Time 7; and 2 additional couples had divorced by Time 8.
Table 6.
Descriptive Statistics and Correlations among Variables in Study 6
| 1 | 2 | 3 | 4 | |
|---|---|---|---|---|
| (1) Divorce | -- | |||
| (2) Current satisfaction | −.11** | -- | ||
| (3) Expected satisfaction | −.11** | .70** | -- | |
| (4) Alternatives | −.05 | .06 | .06 | -- |
|
| ||||
| M | 14% | 40.50 | 41.27 | 23.94 |
| SD | -- | 5.64 | 4.70 | 7.98 |
Note.
p < .01.
We addressed the role of current and expected satisfaction in predicting divorce at the next wave of assessment in several two-level models using the HLM 7.01 computer program; we regressed a dummy-code of next-wave divorce onto husbands’ and wives’ prior-wave relationship satisfaction scores, where these effects were constrained to be equal across husbands and wives, and wave of assessment was controlled to ensure any effects that emerged were not due to fluctuations in these variables over time caused by other factors (see Bolger & Laurenceau, 2013). Because the dependent variable was binary, we specified a Bernoulli outcome distribution. The non-independence of repeated assessments was controlled in the second level of the model with a randomly varying intercept.
In the first model, we examined the association between current satisfaction and divorce at the next assessment. Consistent with current conceptualizations of interdependence theory, spouses’ current satisfaction with their relationship was negatively associated with the likelihood of divorce at the next wave of assessment, b = −0.04, SE = 0.02, t(458) = −2.39, p = .02. A subsequent analysis indicated that spouses’ current satisfaction with their relationship remained negatively associated with the likelihood of divorce at the next wave of assessment when controlling for spouses’ perceived alternatives, b = −0.04, SE = 0.02, t(432) = −2.15, p = .03.
In a second model, we examined the association between expected satisfaction and divorce at the next assessment by repeating the same analysis except this time using prior-wave expected satisfaction scores rather than current satisfaction scores. Consistent with our predictions and the results from the previous studies, results indicated that spouses’ expectations for future satisfaction with their relationship were also negatively associated with the likelihood of divorce at the next wave of assessment, b = −0.07, SE = 0.02, t(392) = −4.11, p < .01. A subsequent analysis indicated that spouses’ expectations for future satisfaction with their relationship remained negatively associated with the likelihood of divorce at the next wave of assessment when controlling for spouses’ perceived alternatives, b = −0.07, SE = 0.02, t(387) = −3.88, p < .01.
In a third analysis, we examined whether expectations for future satisfaction were a stronger predictor of divorce than was current satisfaction by repeating the same analysis described above except this time regressing a dummy-code of next-wave divorce onto prior-wave current relationship satisfaction scores and prior-wave expected future satisfaction scores simultaneously. Results indicated that spouses’ current satisfaction with their relationship was no longer associated with the likelihood of divorce at the next time of assessment once expected satisfaction was controlled, b = 0.05, SE = 0.04, t(370) = 1.25, p = .21; however, consistent with our predictions, expected satisfaction was negatively associated with the likelihood of divorce even after current satisfaction was controlled, b = −0.10, SE = 0.04, t(370) = −2.65, p = .01. Further, and crucially, a direct test comparing the size of these two effects indicated that expected satisfaction was a significantly stronger predictor than was current satisfaction, z = 2.34, p = .02. Subsequent analyses indicated that expected satisfaction remained negatively associated with divorce, b = −0.08, SE = 0.04, t(365) = −2.20, p = .03, and current satisfaction remained not associated with divorce, b = 0.04, SE = 0.04, t(365) = 0.83, p = .41, when controlling for spouses’ perceived alternatives.
Finally, we examined whether spouses’ expected future satisfaction mediated the association between their current satisfaction and the likelihood of divorce by once again computed asymmetric confidence intervals for the mediated effect by following the procedure described by MacKinnon and colleagues (2007). The first step was to demonstrate that current satisfaction was associated with the expected mediator—expectations for future satisfaction. To test this, we estimated a three-level model using the HLM 7.01 computer program in which we regressed participants’ expected satisfaction scores onto their current satisfaction scores and the time of assessment. Consistent with the first criterion necessary for establishing mediation, current satisfaction was significantly associated with expected satisfaction, b = 0.51, SE = 0.05, t(698) = 9.35, p < .01. The second step was to demonstrate that expected satisfaction predicted next-wave divorce, controlling for current satisfaction, which was demonstrated in the previous section. Finally, we multiplied these two effects together to obtain an estimate of the mediated effect, b = −0.05, and computed the 95% confidence interval (−9.29E-2: −1.34E-2) that indicated that the mediated effect was significant. Notably, the fact that current satisfaction was not associated with commitment after controlling for expected satisfaction, as reported in the previous section, rules out the alternative mediational path—that current satisfaction mediates the association between expected satisfaction and commitment. Indeed, this alternative indirect effect was not significant (−2.15E-2: 1.00E-1).
General Discussion
Given that close relationships can sometimes be a source of great fulfillment and happiness, yet at other times lead to disappointment and distress, the decision to remain in or leave a close relationship often has profound consequences for intimates’ well-being. How do they make this decision? Existing theoretical perspectives (e.g., Jacobson & Margolin, 1979; Karney & Bradbury, 1995; Kelley, 1983; Kenrick et al., 1993; Markman, 1979; Rusbult, 1980a; Shackelford & Buss, 1997), suggest that intimates primarily base this decision on the extent to which they are currently satisfied with that relationship. Nevertheless, nearly all relationships will inevitably encounter problems that cause intimates to become less satisfied with those relationships and many intimates remain committed to those relationships despite growing dissatisfaction. Given that intimates’ expectations regarding the likelihood that they will be satisfied in the future should be based on the implications of staying versus leaving the relationship, the current work examined whether such expected satisfaction is a more proximal and stronger predictor of commitment than is their current satisfaction.
Study 1 provided support for the idea that expected satisfaction is based on not only current satisfaction but also several other unique sources of information relevant to commitment decisions. Whereas current rewards uniquely predicted intimates’ current satisfaction and not their expected satisfaction, anticipated life events, plans to improve the quality of the relationship, attachment anxiety, attachment avoidance, neuroticism, and self-esteem uniquely predicted how satisfied intimates expected to be in the future and not their current satisfaction. Only relationship self-efficacy uniquely predicted both current and expected satisfaction and relationship costs were unassociated with either current or expected satisfaction. Study 2 examined intimates’ naturally occurring current satisfaction, expected satisfaction, and commitment. Although current satisfaction predicted commitment before expected satisfaction was controlled, (a) only expected satisfaction predicted commitment once both predictors were considered simultaneously, (b) expected satisfaction was a stronger predictor of commitment than was current satisfaction, and (c) expected satisfaction mediated the effect of current satisfaction on commitment but not vice versa. In Study 3, people imagined being in a relationship that was either currently highly satisfying or highly unsatisfying and would be either satisfying or unsatisfying in the future. The extent to which the relationship would be satisfying in the future determined relationship commitment more than the extent to which the relationship was currently desirable. These effects emerged regardless of whether people reported how committed to the relationship they were currently or how committed they expected to be in the future. In Study 4, intimates were more committed when they were led to believe their actual relationships were more satisfying currently or would be more satisfying in the future. Nevertheless, the manipulation of expected satisfaction was a stronger predictor of relationship commitment than was the manipulation of current satisfaction and expected satisfaction completely mediated the effects of the current satisfaction manipulation on commitment. Studies 5 and 6 extended these effects to relationship maintenance processes that result from commitment. Specifically, Study 5 demonstrated that, compared to current satisfaction, expected satisfaction was a significantly stronger predictor of two different relationship maintenance behaviors: behaving constructively during problem-resolution discussions and diverting attention from attractive others. Further, expected satisfaction mediated the association between current satisfaction and those relationship maintenance behaviors. Finally, Study 6 used a longitudinal study of newlywed couples who reported their current and expected relationship satisfaction and whether they were still married every 6 to 8 months for the first five years of their marriages. Although current satisfaction predicted divorce before expected satisfaction was controlled, (a) only expected satisfaction predicted divorce once both predictors were considered simultaneously, (b) expected satisfaction was a stronger predictor of dissolution than was current satisfaction, and (c) expected satisfaction mediated the effect of current satisfaction on dissolution.
Implications
These findings have important theoretical implications. Most notably, the current studies join a growing body of research demonstrating the importance of relationship expectations. Expectations are essential for survival and reproduction and thus some have argued that the human brain is essentially an “anticipation machine” (Dennett, 1991; see Olson et al., 1996). Given the crucial role of relationships for both survival and reproduction (Bowlby, 1969; Baumeister & Leary, 1995), it is not surprising that people form interpersonal expectations and that such expectations play an important role in determining interpersonal outcomes (see Joel, MacDonald, & Plaks, 2013). Indeed, not only did expected satisfaction account for the effects of robust association between current satisfaction and commitment in Studies 2–6, expected satisfaction was associated with several individual difference variables that previous research has demonstrated are crucial to relationship functioning—self-esteem (Murray, Holmes, & Collins, 2006), neuroticism (Karney & Bradbury, 1995), and attachment insecurity (Shaver & Mikuliner, 2002). Given that such individual difference variables presumably affect relationship satisfaction through specific processes such as behavior and perception (Karney & Bradbury, 1995; McNulty, 2008), future research may benefit from examining whether expected satisfaction accounts for the effects of these individual difference variables on specific aspects of relationship functioning as well. Consistent with this possibility, expectations are a proximal process through which a variety of interpersonal processes, such as attraction (Wang, Hahn, Fisher, DeBruine, & Jones, 2014; Hill, 2007), mate selection (Fletcher, Kerr, Li, & Valentine, 2014; Li et al., 2013), sexual behavior (McNulty & Fisher, 2008), conflict resolution (McNulty & Karney, 2002; Overall, Fletcher, Simpson, & Sibley, 2009; Srivastava et al., 2006), and supportive exchanges (Collins & Feeney, 2004; Lemay & Clark, 2008), operate on relationship functioning. Further, expectations appear to mediate the interpersonal effects of both attachment insecurity (Little et al., 2010) and neuroticism (Fisher & McNulty, 2008) and are also a key component of trust (Holmes, 2002; Holmes & Rempel, 1989; Kramer & Carnevale, 2001; Rempel, Holmes, & Zanna, 1985), which serves as the foundation of many theories of relationships (Simpson, 2007). In sum, theories of attraction and close relationships may benefit to the extent that researchers continue examining not only intimates’ current experiences, but also what they expect to experience in the future.
Second, the current results suggest a need to expand interdependence theory. For decades, interdependence theory (Blau, 1964; Levinger, 1976; Rusbult, 1980a; Thibaut & Kelley, 1959) has been the prevailing theory to explain why some intimates remain committed to, and thus maintain, their close relationships, whereas others do not (Finkel & Simpson, 2015; Fletcher & Overall, 2010; Simpson & Winterheld, 2012). Specifically, modern conceptualizations of interdependence theory suggest that intimates tend to be more committed to, and thus more likely to maintain, their relationships to the extent that they (a) have invested more resources into the relationship, (b) believe they do not have more desirable alternatives to the relationship, and (c) are currently more satisfied with the relationship. Nevertheless, given that commitment reflects the extent to which intimates desire their relationships to continue in the future, that desire should be based on how satisfying those intimates believe their relationships will be in the future, not how satisfying they believe their relationships are currently. Consistent with this idea, and consistent with perspectives emerging in other domains that highlight the relative importance of anticipated outcomes over current ones (DeWall et al., 2016; for review, see Baumeister et al., 2007), the results from the current studies demonstrate that expected satisfaction with the current relationship is a more important and proximal predictor of commitment and relationship maintenance processes than is current satisfaction with that relationship. In fact, although current satisfaction demonstrated bivariate associations with commitment/maintenance across all studies except for Study 5, it was unassociated with commitment/maintenance in Studies 2 and 6 once the important association between expected satisfaction and commitment/maintenance had been controlled. In those studies, current satisfaction predicted commitment indirectly by playing an important role in shaping intimates’ expectations. Nevertheless, given that expected rather than current satisfaction predicted commitment whenever expected and current satisfaction were different, these results also suggest that current satisfaction may have little effect on commitment when intimates anticipate changes in their relationship satisfaction (e.g., due to pregnancy, career changes). These results are consistent with findings involving the role of expectations regarding other key aspects of interdependence theory. For instance, Goodfriend and Agnew (2008) demonstrated that intimates’ planned relationship investments accounted for variance in commitment and dissolution above and beyond that accounted for by their past and current investments. Future research may benefit from more directly integrating expectations into interdependence perspectives.
Third, the current studies suggest the need to examine how people’s expectations for future satisfaction with aspects of their lives in domains other than their romantic relationships may affect commitment-related decisions in those domains. Not only has interdependence theory inspired research on romantic relationships, it has inspired research on relationships with friends (Rusbult, 1980b), family members (Myers & Bryant, 2008), coworkers (Bishop & Scott, 2000), and neighbors (Nunkoo & Ramkissoon, 2011). Not surprisingly, like research on romantic relationships, that work suggests that current satisfaction drives commitment-related decisions. Furthermore, research outside of the domain of relationships has similarly revealed the importance of satisfaction with other aspects of life for commitment-related decisions in those domains. For example, compared to people who are less satisfied, people who are more satisfied with their jobs are less likely to quit those jobs (Farrell & Rusbult, 1981), people who are more satisfied with their education tend to drop out of school less (Alexander, Entwisle, & Horsey, 1997), people who are more satisfied with their athletic teams tend to train harder (Scanlan, Russell, Beals, & Scanlan, 2003), and people who are more satisfied with consumer products are more likely to remain loyal customers (Gustafsson, Johnson, & Roos, 2005). Nevertheless, given anticipated life changes, plans to improve these domains, and individual differences may better forecast the future and thus lead expected satisfaction to diverge from current satisfaction in these domains as well, future research may benefit by examining whether expected satisfaction with these other domains are a stronger predictor of commitment than is current satisfaction.
Finally, the current results have important practical implications. Understanding why intimates do not leave unsatisfying relationships has been a clear and important goal of relationship research for decades (e.g., Becker, 1960; Johnson, 1973, 1982), and interdependence theory has addressed this issue in ground-breaking ways. Most notably, such perspectives have helped us understand and potentially assist intimates who remain in their relationships despite low satisfaction because they perceive no better alternatives or have invested considerable resources. Indeed, Rusbult and Martz (1995) demonstrated such processes in a sample of women victimized by violence. The current results continue this tradition by additionally demonstrating that some dissatisfied intimates may remain in unsatisfying relationships because they optimistically believe that the relationship will be satisfying in the future. Consistent with this idea, Baker and colleagues (Baker, Cobb, McNulty, Lambert, & Fincham, 2016) demonstrated that victimization was associated with relationship dissolution among victims who did not optimistically believe they were capable of resolving their relationship conflicts but unassociated with dissolution among victims who optimistically believed they were capable of resolving their relationship conflicts. The current studies join these in highlighting perceptions of the future as a key target for interventions aimed at either encouraging, or discouraging, relationship stability.
Study Strengths and Limitations
Several strengths of the current studies enhance our confidence in the results reported here. First, our hypotheses were supported in six independent samples, reducing the likelihood that the results were due to sampling error. Second, Studies 2 and 3 experimentally manipulated expectations for future satisfaction, enhancing our confidence in the role of relationship expectations in causing commitment. Third, Studies 5 and 6 provided evidence that relationship expectations not only determine intimates’ commitment to their relationships, but also the extent to which they maintain (Study 5), and the likelihood of remaining in (Study 6), those relationships. Further, examining the implications of expectations for these behavioral indicators of commitment helped ensure that the observed associations between self-reports of expectations and commitment were not the result of common-method variance or common temporal focus. Fourth, the results replicated across individuals in varying stages of relationships, from dating university students to newlywed couples, ensuring that the results obtained were not unique to individuals at certain stages in their relationships (see Russell, Baker, & McNulty, 2013). Finally, Studies 2 and 6 provided evidence that the implications of expected relationship satisfaction are independent from perceived alternatives to the relationship.
Nevertheless, several factors limit the conclusions that can be drawn from these studies until they can be replicated and extended. First, participants were primarily White. Although we are not aware of any theoretical reasons that the implications of current and expected relationship satisfaction for relationship commitment should vary across people of different ethnicities, generalizations to other samples should be made with caution. Second, none of these studies specified or assessed the length of time on which intimates were basing their expectations. For example, when asked their expectations for the future, some participants may have been thinking about the following days, while others may have been thinking about the following years. Future research may benefit by examining the implications of prospecting different lengths of time. Finally, although the experimental nature of Studies 3–4 provided strong support for our meditational model, like any theoretical proposition, genuine mediation cannot be proved unequivocally (see Fiedler et al., 2011).
Conclusion
Contemporary perspectives on relationship commitment posit that intimates’ commitment to a relationship is partially based on their current satisfaction with that relationship. However, given that intimates’ expectations about their future satisfaction take into account additional information about the future, such expectations should be a more functional and proximal source of commitment. Consistent with this idea, results from the current studies suggest that current satisfaction is not the key predictor of commitment it is assumed to be. Instead, these results suggest that how satisfied intimates expect to be with a relationship in the future more strongly determines their commitment and accounts for the effects of their current satisfaction.
Footnotes
After developing the current predictions and presenting preliminary findings at the 2015 annual convention of the Society for Personality and Social Psychology, the authors learned that Edward Lemay had independently developed and was testing a similar theoretical model. After a brief meeting, the two parties agreed to continue their pursuits independently.
The overall pattern of results did not change if we excluded those who were not in romantic relationships. Specifically, both expected satisfaction, b = 1.84, SE = 0.17, t(361) = 10.87, p < .01, and current satisfaction, b = 0.89, SE = 0.17, t(361) = 5.26, p < .01, significantly predicted commitment, and expected satisfaction was a significantly stronger predictor than was current satisfaction, z = 3.64, p < .01.
Simmons, Nelson, and Simonsohn (2011) raised concerns about collecting additional data after testing hypotheses, particularly with doing so repeatedly. In an effort to reduce type II errors associated with abandoning hypotheses too quickly, Sagarin, Ambler, and Lee (2014) demonstrated that it can be acceptable to continue collecting additional data after discovering marginal results if the researchers also report the results of the original sample and augmented p-values. We report original, updated, and augmented results, along with original, new, augmented p-values.
Author Note: Portions of this article were presented at the main conference of the International Association for Relationship Research (2016), 37th Annual Meeting of the Society of Southeastern Social Psychologists (2016), and the Annual Convention for the Association for Psychological Science (2015).
Contributor Information
Levi R. Baker, University of North Carolina at Greensboro
James K. McNulty, Florida State University
Laura E. VanderDrift, Syracuse University
References
- Agnew CR, Arriaga XB, Wilson JE. Committed to what? Using the Bases of Relational Commitment Model to understand continuity and changes in social relationships. In: Forgas JP, Fitness J, editors. Social relationships: Cognitive, affective and motivational processes. New York: Psychology Press; 2008. pp. 147–164. [Google Scholar]
- Alexander KL, Entwisle DR, Horsey CS. From first grade forward: Early foundations of high school dropout. Sociology of Education. 1997;70:87–107. [Google Scholar]
- Arriaga XB. The ups and downs of dating: Fluctuations in satisfaction in newly-formed romantic relationships. Journal of Personality and Social Psychology. 2001;80:754–765. doi: 10.1037//0022-3514.80.5.754. [DOI] [PubMed] [Google Scholar]
- Ajzen I. From intentions to actions: A theory of planned behavior. In: Kuhl J, Beckman J, editors. Action-control: From cognition to behavior. Heidelberg: Springer; 1985. pp. 11–39. [Google Scholar]
- Amar AF, Alexy EM. ‘Dissed’ by dating violence. Perspectives in Psychiatric Care. 2005;41:162–171. doi: 10.1111/j.1744-6163.2005.00032.x. [DOI] [PubMed] [Google Scholar]
- Argyle M, Furnham A. Sources of satisfaction and conflict in long-term relationships. Journal of Marriage and the Family. 1983;45:481–493. [Google Scholar]
- Atkinson JW. Motivational determinants of risk-taking behavior. Psychological Review. 1957;64:359–372. doi: 10.1037/h0043445. [DOI] [PubMed] [Google Scholar]
- Baker LR, Cobb RA, McNulty JK, Lambert NM, Fincham FD. Remaining in a situationally aggressive relationship: The role of relationship self-efficacy. Personal Relationships. 2016;23:591–604. [Google Scholar]
- Baker L, McNulty JK. Shyness and marriage: Does shyness shape even established relationships? Personality and Social Psychology Bulletin. 2010;36:665–676. doi: 10.1177/0146167210367489. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Baker LR, McNulty JK, Overall NC, Lambert NM, Fincham FD. How do relationship maintenance behaviors affect individual well-being?: A contextual perspective. Social Psychological and Personality Science. 2013;4:282–289. [Google Scholar]
- Baumeister RF, Leary MR. The need to belong: Desire for interpersonal attachments as a fundamental human motivation. Psychological Bulletin. 1995;117:497–529. [PubMed] [Google Scholar]
- Baumeister RF, Vohs KD, DeWall CN, Zhang L. How emotion shapes behavior: Feedback, anticipation, and reflection, rather than direct causation. Personality and Social Psychology Review. 2007;11:167–203. doi: 10.1177/1088868307301033. [DOI] [PubMed] [Google Scholar]
- Becker HS. Notes on the concept of commitment. American Journal of Sociology. 1960;66:32–40. [Google Scholar]
- Bishop JW, Scott DK. An examination of organizational and team commitment in a self-directed team environment. Journal of Applied Psychology. 2000;85(3):439–450. doi: 10.1037/0021-9010.85.3.439. [DOI] [PubMed] [Google Scholar]
- Blau PM. Exchange and power in social life. New York: John Wiley & Sons, Inc; 1964. [Google Scholar]
- Bolger N, Laurenceau J. Intensive longitudinal methods: An introduction to diary and experience sampling research. New York, NY, US: Guilford Press; 2013. [Google Scholar]
- Bowlby J. Disruption of affectional bonds and its effects on behavior. Canada’s Mental Health Supplement. 1969;59:12. [Google Scholar]
- Bradbury TN. Unpublished doctoral dissertation. University of Illinois; Urbana-Champaign: 1989. Cognition, emotion, and interaction in distressed and nondistressed marriages. [Google Scholar]
- Bui KT, Peplau LA, Hill CT. Testing the Rusbult model of relationship commitment and stability in a 15-year study of heterosexual couples. Personality and Social Psychology Bulletin. 1996;22:1244–1257. [Google Scholar]
- Bushman BJ, Baumeister RF, Phillips CM. Do people aggress to improve their mood? Catharsis beliefs, affect regulation opportunity, and aggressive responding. Journal of Personality and Social Psychology. 2001;81:17–32. [PubMed] [Google Scholar]
- Campbell WK. Narcissism and romantic attraction. Journal of Personality and Social Psychology. 1999;77:1254–1270. [Google Scholar]
- Carson JW, Carson KM, Gil KM, Baucom DH. Mindfulness-based relationship enhancement. Behavior Therapy. 2004;35:471–494. [Google Scholar]
- Carver CS, Scheier MF. On the self-regulation of behavior. New York, NY, US: Cambridge University Press; 1998. [Google Scholar]
- Chapman GB, Coups EJ. Emotions and preventive health behavior: Worry, regret, and influenza vaccination. Health Psychology. 2006;25:82–90. doi: 10.1037/0278-6133.25.1.82. [DOI] [PubMed] [Google Scholar]
- Clark MS, Grote NK. Why aren’t indices of relationship costs always negatively related to indices of relationship quality? Personality and Social Psychology Review. 1998;2:2–17. doi: 10.1207/s15327957pspr0201_1. [DOI] [PubMed] [Google Scholar]
- Collins NL, Feeney BC. Working models of attachment shape perceptions of social support: Evidence from experimental and observational studies. Journal of Personality and Social Psychology. 2004;87:363–383. doi: 10.1037/0022-3514.87.3.363. [DOI] [PubMed] [Google Scholar]
- Dennett DC. Consciousness explained. New York, NY: Little, Brown and Co; 1991. [Google Scholar]
- DeWall CN, Baumeister RF, Chester DS, Bushman BJ. How often does currently felt emotion predict social behavior and judgment? A meta-analytic test of two theories. Emotion Review. 2016;8:136–143. [Google Scholar]
- Duffy S, Rusbult CE. Satisfaction and commitment in homosexual and heterosexual relationships. Journal of Homosexuality. 1986;12:1–23. doi: 10.1300/j082v12n02_01. [DOI] [PubMed] [Google Scholar]
- Farrell D, Rusbult CE. Exchange variables as predictors of job satisfaction, job commitment, and turnover: The impact of rewards, costs, alternatives, and investments. Organizational Behavior & Human Performance. 1981;28:78–95. [Google Scholar]
- Fazio RH, Jackson JR, Dunton BC, Williams CJ. Variability in automatic activation as an unobtrusive measure of racial attitudes: A bona fide pipeline? Journal of Personality and Social Psychology. 1995;69:1013–1027. doi: 10.1037//0022-3514.69.6.1013. [DOI] [PubMed] [Google Scholar]
- Feather NT. Expectancy-value approaches: Present status and future directions. In: Feather NT, editor. Expectations and Actions: Expectancy-Value Models in Psychology. Hillsdale, NJ: Erlbaum; 1982. [Google Scholar]
- Feeney BC, Collins NL. A theoretical perspective on the importance of social connections for thriving. In: Mikulincer M, Shaver PR, Mikulincer M, Shaver PR, editors. Mechanisms of social connection: From brain to group. Washington, DC, US: American Psychological Association; 2014. pp. 291–314. [Google Scholar]
- Fiedler K, Schott M, Meiser T. What mediation analysis can (not) do. Journal of Experimental Social Psychology. 2011;47:1231–1236. [Google Scholar]
- Finkel EJ, Simpson JA. Relationship science, 2015. Current Opinion in Psychology. 2015;1:5–9. [Google Scholar]
- Fishbein M, Ajzen I. Belief, Attitude, Intention, and Behavior: An Introduction to Theory and Research. Reading, MA: Addison-Wesley; 1975. [Google Scholar]
- Fisher RA. On the probable error of a coefficient of correlation deduced from a small sample. Metron. 1921;1:3–32. [Google Scholar]
- Fisher TD, McNulty JK. Neuroticism and marital satisfaction: The mediating role played by the sexual relationship. Journal of Family Psychology. 2008;22:112–122. doi: 10.1037/0893-3200.22.1.112. [DOI] [PubMed] [Google Scholar]
- Fitzsimons GM, Finkel EJ, vanDellen MR. Transactive goal dynamics. Psychological Review. 2015;122:648–673. doi: 10.1037/a0039654. [DOI] [PubMed] [Google Scholar]
- Fletcher GO, Overall NC. Intimate relationships. In: Baumeister RF, Finkel EJ, Baumeister RF, Finkel EJ, editors. Advanced social psychology: The state of the science. New York: Oxford University Press; 2010. pp. 461–494. [Google Scholar]
- Fletcher GO, Kerr PG, Li NP, Valentine KA. Predicting romantic interest and decisions in the very early stages of mate selection: Standards, accuracy, and sex differences. Personality and Social Psychology Bulletin. 2014;40:540–550. doi: 10.1177/0146167213519481. [DOI] [PubMed] [Google Scholar]
- Freedman JL, Wallington SA, Bless E. Compliance without pressure: The effect of guilt. Journal of Personality and Social Psychology. 1967;7:117–124. [Google Scholar]
- Gaertner L, Foshee V. Commitment and the perpetration of relationship violence. Personal Relationships. 1999;6:227–239. [Google Scholar]
- Gilbert DT, Ebert JJ. Decisions and revisions: The affective forecasting of changeable outcomes. Journal of Personality and Social Psychology. 2002;82:503–514. [PubMed] [Google Scholar]
- Gilbert DT, Wilson TD. Prospection: Experiencing the future. Science. 2007;317:1351–1354. doi: 10.1126/science.1144161. [DOI] [PubMed] [Google Scholar]
- Goldberg LR. A broad-bandwidth, public-domain, personality inventory measuring the lower-level facets of several five-factor models. In: Mervielde I, Deary I, De Fruyt F, Ostendorf F, editors. Personality psychology in Europe. Vol. 7. Tilburg, The Netherlands: Tilburg University Press; 1999. pp. 7–28. [Google Scholar]
- Goodfriend W, Agnew CR. Sunken costs and desired plans: Examining different types of investments in close relationships. Personality and Social Psychology Bulletin. 2008;34:1639–1652. doi: 10.1177/0146167208323743. [DOI] [PubMed] [Google Scholar]
- Green JD, Davis JL, Luchies LB, Coy AE, Van Tongeren DR, Reid CA, Finkel EJ. Victims versus perpetrators: Affective and empathic forecasting regarding transgressions in romantic relationships. Journal of Experimental Social Psychology. 2013;49:329–333. [Google Scholar]
- Gross JJ. The emerging field of emotion regulation: An integrative review. Review of General Psychology. 1998;2:271–299. [Google Scholar]
- Gustafsson A, Johnson MD, Roos I. The effects of customer satisfaction, relationship commitment dimensions, and triggers on customer retention. Journal of Marketing. 2005;69:210–218. [Google Scholar]
- Hellmuth JC, McNulty JK. Neuroticism, marital violence, and the moderating role of stress and behavioral skills. Journal of Personality and Social Psychology. 2008;95:166–180. doi: 10.1037/0022-3514.95.1.166. [DOI] [PubMed] [Google Scholar]
- Hill SE. Overestimation bias in mate competition. Evolution and Human Behavior. 2007;28:118–123. [Google Scholar]
- Hofmann W, Finkel EJ, Fitzsimons GM. Close relationships and self-regulation: How relationship satisfaction facilitates momentary goal pursuit. Journal of Personality and Social Psychology. 2015;109:434–452. doi: 10.1037/pspi0000020. [DOI] [PubMed] [Google Scholar]
- Holmes JG. Interpersonal expectations as the building blocks of social cognition: An interdependence theory perspective. Personal Relationships. 2002;9:1–26. [Google Scholar]
- Holmes JG, Rempel JK. Trust in close relationships. In: Hendrick C, Hendrick C, editors. Close relationships. Thousand Oaks, CA, US: Sage Publications, Inc; 1989. pp. 187–220. [Google Scholar]
- Impett EA, Beals KP, Peplau LA. Testing the investment model of relationship commitment and stability in a longitudinal study of married couples. Current Psychology: A Journal for Diverse Perspectives on Diverse Psychological Issues. 2001;20:312–326. [Google Scholar]
- Jacobson NS, Margolin G. Marital therapy: Strategies based on social learning and behavior exchange principles. New York: Brunner/Mazel; 1979. [Google Scholar]
- Joel S, MacDonald G, Plaks JE. Romantic relationships conceptualized as a judgment and decision-making domain. Current Directions in Psychological Science. 2013;22:461–465. [Google Scholar]
- Johnson MP. Commitment: A conceptual structure and empirical application. The Sociological Quarterly. 1973;14:395–406. [Google Scholar]
- Johnson MP. Social and cognitive features of the dissolution of commitment to relationships. In: Duck S, editor. Personal relationships 4: Dissolving personal relationships. London: Academic Press; 1982. pp. 51–73. [Google Scholar]
- Karney BR, Bradbury TN. The longitudinal course of marital quality and stability: A review of theory, methods, and research. Psychological Bulletin. 1995;118:3–34. doi: 10.1037/0033-2909.118.1.3. [DOI] [PubMed] [Google Scholar]
- Karney BR, Bradbury TN. Neuroticism, marital interaction, and the trajectory of marital satisfaction. Journal of Personality and Social Psychology. 1997;72:1075–1092. doi: 10.1037//0022-3514.72.5.1075. [DOI] [PubMed] [Google Scholar]
- Karney BR, Frye NE. ‘But we’ve been getting better lately’: Comparing prospective and retrospective views of relationship development. Journal of Personality and Social Psychology. 2002;82:222–238. [PubMed] [Google Scholar]
- Kelley HH. Love and commitment. In: Kelley HH, Berscheid E, Christensen A, Harvey JH, Huston TL, Levinger G, McClintock E, Peplau LA, Peterson DR, editors. Close relationships. New York: W. H. Freeman; 1983. pp. 265–314. [Google Scholar]
- Kelley H, Holmes J, Kerr N, Reis H, Rusbult C, Van Lange P. An atlas of interpersonal situations. New York: Cambridge; 2003. [Google Scholar]
- Kelley HH, Thibaut JW. Interpersonal relations: A theory of interdependence. New York: Wiley; 1978. [Google Scholar]
- Kelley HH, Thibaut JW. Interpersonal relations: A theory of interdependence. New York: Wiley; 1978. [Google Scholar]
- Kenrick DT, Groth GE, Trost MR, Sadalla EK. Integrating evolutionary and social exchange perspectives on relationships: Effects of gender, self-appraisal, and involvement level on mate selection criteria. Journal of Personality and Social Psychology. 1993;64:951–969. [Google Scholar]
- Knee CR, Patrick H, Vietor NA, Nanayakkara A, Neighbors C. Self-determination as growth motivation in romantic relationships. Personality and Social Psychology Bulletin. 2002;28:609–619. [Google Scholar]
- Kurdek LA. The nature and predictors of the trajectory of change in marital quality over the first 4 years of marriage for first-married husbands and wives. Journal of Family Psychology. 1998;12:494–510. doi: 10.1037//0012-1649.35.5.1283. [DOI] [PubMed] [Google Scholar]
- LaGaipa JJ. Interpersonal attraction and social exchange. In: Duck S, editor. Theory and practice in interpersonal attraction. San Diego, CA: Academic Press; 1977. pp. 129–164. [Google Scholar]
- Lavner JA, Bradbury TN. Patterns of change in marital satisfaction over the newlywed years. Journal of Marriage and Family. 2010;72:1171–1187. doi: 10.1111/j.1741-3737.2010.00757.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Lawrence E, Nylen K, Cobb RJ. Prenatal expectations and marital satisfaction over the transition to parenthood. Journal of Family Psychology. 2007;21:155–164. doi: 10.1037/0893-3200.21.2.155. [DOI] [PubMed] [Google Scholar]
- Le B, Agnew CR. Commitment and its theorized determinants: A meta-analysis of the investment model. Personal Relationships. 2003;10:37–57. [Google Scholar]
- Le B, Dove NL, Agnew CR, Korn MS, Mutso AA. Predicting nonmarital romantic relationship dissolution: A meta-analytic synthesis. Personal Relationships. 2010;17:377–390. [Google Scholar]
- Lemay EP., Jr The forecast model of relationship commitment. Journal of Personality and Social Psychology. 2016;111:34–52. doi: 10.1037/pspi0000052. [DOI] [PubMed] [Google Scholar]
- Lemay EP, Jr, Clark MS. How the head liberates the heart: Projection of communal responsiveness guides relationship promotion. Journal of Personality and Social Psychology. 2008;94:647–671. doi: 10.1037/0022-3514.94.4.647. [DOI] [PubMed] [Google Scholar]
- Lemay EP, Jr, Lin JL, Muir HJ. Daily affective and behavioral forecasts in romantic relationships: Seeing tomorrow through the lens of today. Personality and Social Psychology Bulletin. 2015;41:1005–1019. doi: 10.1177/0146167215588756. [DOI] [PubMed] [Google Scholar]
- Levinger G. A social psychological perspective on marital dissolution. Journal of Social Issues. 1976;32:21–47. [Google Scholar]
- Li NP, Yong JC, Tov W, Sng O, Fletcher GO, Valentine KA, … Balliet D. Mate preferences do predict attraction and choices in the early stages of mate selection. Journal of Personality and Social Psychology. 2013;105:757–776. doi: 10.1037/a0033777. [DOI] [PubMed] [Google Scholar]
- Little KC, McNulty JK, Russell VM. Sex buffers intimates against the negative implications of attachment insecurity. Personality and Social Psychology Bulletin. 2010;36:484–498. doi: 10.1177/0146167209352494. [DOI] [PMC free article] [PubMed] [Google Scholar]
- MacKinnon DP, Fritz MS, Williams J, Lockwood CM. Distribution of the product confidence limits for the indirect effect: Program PRODLIN. Behavior Research Methods. 2007;39:384–389. doi: 10.3758/bf03193007. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Maner JK, Gailliot MT, Miller SL. The implicit cognition of relationship maintenance: Inattention to attractive alternatives. Journal of Experimental Social Psychology. 2009;45:174–179. [Google Scholar]
- Maner JK, Holm-Denoma JM, Van Orden KA, Gailliot MT, Gordon KH, Joiner TJ. Evidence for Attentional Bias in Women Exhibiting Bulimotypic Symptoms. International Journal of Eating Disorders. 2006;39:55–61. doi: 10.1002/eat.20222. [DOI] [PubMed] [Google Scholar]
- Maner JK, Rouby DA, Gonzaga GC. Automatic inattention to attractive alternatives: The evolved psychology of relationship maintenance. Evolution and Human Behavior. 2008;29:343–349. [Google Scholar]
- Manucia GK, Baumann DJ, Cialdini RB. Mood influences on helping: Direct effects or side effects? Journal of Personality and Social Psychology. 1984;46:357–364. [Google Scholar]
- Markman HJ. Application of a behavioral model of marriage in predicting relationship satisfaction of couples planning marriage. Journal of Consulting and Clinical Psychology. 1979;47:743–749. doi: 10.1037//0022-006x.47.4.743. [DOI] [PubMed] [Google Scholar]
- McCubbin HI, Patterson JM. The Family Stress Process: The Double ABCX Model of adjustment and adaptation. Marriage & Family Review. 1983;6:7–37. [Google Scholar]
- McGonagle KA, Kessler RC, Schilling EA. The frequency and determinants of marital disagreements in a community sample. Journal of Social and Personal Relationships. 1992;9:507–524. [Google Scholar]
- McNulty JK. Neuroticism and interpersonal negativity: The independent contributions of perceptions and behaviors. Personality and Social Psychology Bulletin. 2008;34:1439–1450. doi: 10.1177/0146167208322558. [DOI] [PubMed] [Google Scholar]
- McNulty JK, Fisher TD. Gender differences in response to sexual expectancies and changes in sexual frequency: A short-term longitudinal study of sexual satisfaction in newly married couples. Archives of Sexual Behavior. 2008;37:229–240. doi: 10.1007/s10508-007-9176-1. [DOI] [PubMed] [Google Scholar]
- McNulty JK, Karney BR. Attributions in marriage: Integrating specific and global evaluations of a relationship. Personality and Social Psychology Bulletin. 2001;27:943–955. [Google Scholar]
- McNulty JK, Karney BR. Expectancy confirmation in appraisals of marital interactions. Personality and Social Psychology Bulletin. 2002;28:764–775. [Google Scholar]
- McNulty JK, Karney BR. Positive expectations in the early years of marriage: Should couples expect the best or brace for the worst? Journal of Personality and Social Psychology. 2004;86:729–743. doi: 10.1037/0022-3514.86.5.729. [DOI] [PubMed] [Google Scholar]
- McNulty JK, O’Mara EM, Karney BR. Benevolent cognitions as a strategy of relationship maintenance: “Don’t sweat the small stuff” but it is not all small stuff. Journal of Personality and Social Psychology. 2008;94:631–646. doi: 10.1037/0022-3514.94.4.631. [DOI] [PubMed] [Google Scholar]
- McNulty JK, Russell VM. When ‘negative’ behaviors are positive: A contextual analysis of the long-term effects of problem-solving behaviors on changes in relationship satisfaction. Journal of Personality and Social Psychology. 2010;98:587–604. doi: 10.1037/a0017479. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Mellers B, Schwartz A, Ritov I. Emotion-based choice. Journal of Experimental Psychology: General. 1999;128:332–345. [Google Scholar]
- Meltzer AL, McNulty JK, Jackson G, Karney BR. Sex differences in the implications of partner physical attractiveness for the trajectory of marital satisfaction. Journal of Personality and Social Psychology. 2014;106:418–428. doi: 10.1037/a0034424. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Miller RS. Inattentive and contented: Relationship commitment and attention to alternatives. Journal of Personality and Social Psychology. 1997;73:758–766. [Google Scholar]
- Morry MM. Relationship satisfaction as a predictor of similarity ratings: A test of the attraction-similarity hypothesis. Journal of Social and Personal Relationships. 2005;22:561–584. [Google Scholar]
- Murray SL, Holmes JG, Collins NL. Optimizing assurance: The risk regulation system in relationships. Psychological Bulletin. 2006;132:641–666. doi: 10.1037/0033-2909.132.5.641. [DOI] [PubMed] [Google Scholar]
- Murray SL, Holmes JG, MacDonald G, Ellsworth PC. Through the looking glass darkly? When self-doubts turn into relationship insecurities. Journal of Personality and Social Psychology. 1998;75:1459–1480. doi: 10.1037//0022-3514.75.6.1459. [DOI] [PubMed] [Google Scholar]
- Myers SA, Bryant LE. The use of behavioral indicators of sibling commitment among emerging adults. Journal of Family Communication. 2008;8:101–125. [Google Scholar]
- Neff LA, Broady EF. Stress resilience in early marriage: Can practice make perfect? Journal of Personality and Social Psychology. 2011;101:1050–1067. doi: 10.1037/a0023809. [DOI] [PubMed] [Google Scholar]
- Neff LA, Geers AL. Optimistic expectations in early marriage: A resource or vulnerability for adaptive relationship functioning? Journal of Personality and Social Psychology. 2013;105:38–60. doi: 10.1037/a0032600. [DOI] [PubMed] [Google Scholar]
- Norton R. Measuring marital quality: A critical look at the dependent variable. Journal of Marriage and the Family. 1983;45:141–151. [Google Scholar]
- Nunkoo R, Ramkissoon H. Developing a community support model for tourism. Annals of Tourism Research. 2011;38:964–988. [Google Scholar]
- Olson JM, Roese NJ, Zanna MP. Expectancies. In: Higgins ET, Kruglan-ski AW, editors. Social psychology: Handbook of basic principles. New York: Guilford Press; 1996. pp. 211–238. [Google Scholar]
- Overall NC, Fletcher GO, Simpson JA, Sibley CG. Regulating partners in intimate relationships: The costs and benefits of different communication strategies. Journal of Personality and Social Psychology. 2009;96:620–639. doi: 10.1037/a0012961. [DOI] [PubMed] [Google Scholar]
- Proulx CM, Helms HM, Buehler C. Marital quality and personal well-being: A meta-analysis. Journal of Marriage and Family. 2007;69:576–593. [Google Scholar]
- Reis HT, Clark MS, Holmes JG. Perceived Partner Responsiveness as an Organizing Construct in the Study of Intimacy and Closeness. In: Mashek DJ, Aron AP, Mashek DJ, Aron AP, editors. Handbook of closeness and intimacy. Mahwah, NJ, US: Lawrence Erlbaum Associates Publishers; 2004. pp. 201–225. [Google Scholar]
- Rempel JK, Holmes JG, Zanna MP. Trust in close relationships. Journal of Personality and Social Psychology. 1985;49:95–112. [PubMed] [Google Scholar]
- Resulaj A, Kiani R, Wolpert DM, Shadlen MN. Changes of mind in decision-making. Nature. 2009;461:263–266. doi: 10.1038/nature08275. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Robles TF, Slatcher RB, Trombello JM, McGinn MM. Marital quality and health: A meta-analytic review. Psychological Bulletin. 2014;140:140–187. doi: 10.1037/a0031859. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Rosenberg M. Society and the adolescent self-image. Princeton, NJ: Princeton University Press; 1965. [Google Scholar]
- Ruble DN, Fleming AS, Hackel LS, Stangor C. Changes in the marital relationship during the transition to first time motherhood: Effects of violated expectations concerning division of household labor. Journal of Personality and Social Psychology. 1988;55:78–87. doi: 10.1037//0022-3514.55.1.78. [DOI] [PubMed] [Google Scholar]
- Rusbult CE. Satisfaction and commitment in friendships. Representative Research in Social Psychology. 1980a;11:96–105. [Google Scholar]
- Rusbult CE. Commitment and satisfaction in romantic associations: A test of the investment model. Journal of Experimental Social Psychology. 1980b;16:172–186. [Google Scholar]
- Rusbult CE. A longitudinal test of the investment model: The development (and deterioration) of satisfaction and commitment in heterosexual involvements. Journal of Personality and Social Psychology. 1983;45:101–117. [Google Scholar]
- Rusbult CE, Johnson DJ, Morrow GD. Predicting satisfaction and commitment in adult romantic involvements: An assessment of the generalizability of the investment model. Social Psychology Quarterly. 1986;49:81–89. [Google Scholar]
- Rusbult CE, Martz JM. Remaining in an abusive relationship: An investment model analysis of nonvoluntary dependence. Personality and Social Psychology Bulletin. 1995;21:558–571. [Google Scholar]
- Rusbult CE, Martz JM, Agnew CR. The Investment Model Scale: Measuring commitment level, satisfaction level, quality of alternatives, and investment size. Personal Relationships. 1998;5:357–391. [Google Scholar]
- Rusbult CE, Olsen N, Davis JL, Hannon P. Commitment and relationship maintenance mechanisms. In: Harvey JH, Wenzel A, editors. Close romantic relationships: Maintenance and enhancement. Mahwah, NJ: Erlbaum; 2001. pp. 87–113. [Google Scholar]
- Rusbult CE, Van Lange PAM. Interdependence, interaction, and relationships. Annual Review of Psychology. 2003;54:351–375. doi: 10.1146/annurev.psych.54.101601.145059. [DOI] [PubMed] [Google Scholar]
- Rusbult CE, Verette J, Whitney GA, Slovik LF, Lipkus I. Accommodation processes in close relationships: Theory and preliminary empirical evidence. Journal of Personality and Social Psychology. 1991;60:53–78. [Google Scholar]
- Russell VM, Baker LR, McNulty JK. Attachment insecurity and infidelity in marriage: Do studies of dating relationships really inform us about marriage? Journal of Family Psychology. 2013;27:242–251. doi: 10.1037/a0032118. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Sagarin BJ, Ambler JK, Lee EM. An ethical approach to peeking at data. Perspectives on Psychological Science. 2014;9:293–304. doi: 10.1177/1745691614528214. [DOI] [PubMed] [Google Scholar]
- Sanderson B, Kurdek LA. Race and gender as moderator variables in predicting relationship satisfaction and relationship commitment in a sample of dating heterosexual couples. Family Relations: An Interdisciplinary Journal of Applied Family Studies. 1993;42:263–267. [Google Scholar]
- Sarason IG, Johnson JH, Siegel JM. Assessing the impact of life changes: Development of the Life Experiences Survey. Journal of Consulting and Clinical Psychology. 1978;46:932–946. doi: 10.1037//0022-006x.46.5.932. [DOI] [PubMed] [Google Scholar]
- Scanlan TK, Russell DG, Beals KP, Scanlan LA. Project on elite athlete commitment (PEAK): II. A direct test and expansion of the sport commitment model with elite amateur sportsmen. Journal of Sport & Exercise Psychology. 2003;25:377–401. [Google Scholar]
- Seligman MP, Railton P, Baumeister RF, Sripada C. Navigating into the future or driven by the past. Perspectives on Psychological Science. 2013;8:119–141. doi: 10.1177/1745691612474317. [DOI] [PubMed] [Google Scholar]
- Shackelford TK, Buss DM. Marital satisfaction in evolutionary psychological perspective. In: Sternberg RJ, Hojjat M, Sternberg RJ, Hojjat M, editors. Satisfaction in close relationships. New York, NY, US: Guilford Press; 1997. pp. 7–25. [Google Scholar]
- Shaver PR, Mikulincer M. Attachment-related psychodynamics. Attachment & Human Development. 2002;4:133–161. doi: 10.1080/14616730210154171. [DOI] [PubMed] [Google Scholar]
- Silverman IW. Incidence of guilt reactions in children. Journal of Personality and Social Psychology. 1967;7:338–340. doi: 10.1037/h0025073. [DOI] [PubMed] [Google Scholar]
- Simmons JP, Nelson LD, Simonsohn U. False-positive psychology: Undisclosed flexibility in data collection and analysis allows presenting anything as significant. Psychological Science. 2011;22:1359–1366. doi: 10.1177/0956797611417632. [DOI] [PubMed] [Google Scholar]
- Simpson JA. Psychological foundations of trust. Current Directions in Psychological Science. 2007;16:264–268. [Google Scholar]
- Simpson JA, Winterheld HA. Person-by-situation perspectives on close relationships. In: Deaux K, Snyder M, Deaux K, Snyder M, editors. The Oxford handbook of personality and social psychology. New York, NY, US: Oxford University Press; 2012. pp. 493–516. [Google Scholar]
- Spanier GB, Margolis RL. Marital separation and extramarital sexual behavior. Journal of Sex Research. 1983;19:23–48. [Google Scholar]
- Srivastava S, McGonigal KM, Richards JM, Butler EA, Gross JJ. Optimism in close relationships: How seeing things in a positive light makes them so. Journal of Personality and Social Psychology. 2006;91:143–153. doi: 10.1037/0022-3514.91.1.143. [DOI] [PubMed] [Google Scholar]
- Stafford L. Measuring relationship maintenance behaviors: Critique and development of the revised relationship maintenance behavior scale. Journal of Social and Personal Relationships. 2011;28:278–303. [Google Scholar]
- Stewart AJ, Copeland AP, Chester NL, Malley JE, Barenbaum NB. Separating together: How divorce transforms families. New York, NY, US: Guilford Press; 1997. [Google Scholar]
- Storaasli RD, Markman HJ. Relationship problems in the early stages of marriage: A longitudinal investigation. Journal of Family Psychology. 1990;4:80–98. [Google Scholar]
- Thibaut JW, Kelley HH. The social psychology of groups. Oxford, England: John Wiley; 1959. [Google Scholar]
- Tice DM, Bratslavsky E, Baumeister RF. Emotional distress regulation takes precedence over impulse control: If you feel bad, do it! Journal of Personality and Social Psychology. 2001;80:53–67. [PubMed] [Google Scholar]
- Tofighi D, MacKinnon DP. Monte Carlo confidence intervals for complex functions of indirect effects. Structural Equation Modeling: A Multidisciplinary Journal. 2016;23:194–205. [Google Scholar]
- Dunbar Evolution of Social Behaviour Patterns in Primates and Man. Proceedings of the British Academy. 88:119–143. [Google Scholar]
- VanderDrift LE, Wilson JE, Agnew CR. On the benefits of valuing being friends for non-marital romantic partners. Journal of Social and Personal Relationships. 2013;30:115–131. [Google Scholar]
- Vogel DL, Wester SR, Wei M, Boysen GA. The Role of Outcome Expectations and Attitudes on Decisions to Seek Professional Help. Journal of Counseling Psychology. 2005;52:459–470. [Google Scholar]
- Wang H, Hahn AC, Fisher CI, DeBruine LM, Jones BC. Women’s hormone levels modulate the motivational salience of facial attractiveness and sexual dimorphism. Psychoneuroendocrinology. 2014;50(2):46–251. doi: 10.1016/j.psyneuen.2014.08.022. [DOI] [PubMed] [Google Scholar]
- Wei M, Russell DW, Mallinckrodt B, Vogel DL. The Experiences in Close Relationship Scale (ECR)-short form: Reliability, validity, and factor structure. Journal of Personality Assessment. 2007;88:187–204. doi: 10.1080/00223890701268041. [DOI] [PubMed] [Google Scholar]
- Weissman MM. Advances in psychiatric epidemiology: rates and risks for major depression. American Journal of Public Health. 1987;77:445–451. doi: 10.2105/ajph.77.4.445. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Wigfield A, Tonks S, Klauda SL. Expectancy-value theory. In: Wenzel KR, Wigfield A, Wenzel KR, Wigfield A, editors. Handbook of motivation at school. New York, NY, US: Routledge/Taylor & Francis Group; 2009. pp. 55–75. [Google Scholar]