Abstract
People differ in their tendencies to labor over decisions and to make choices that maximize their outcomes—a difference known as maximization. Here, we used two independent, 3-year longitudinal studies of newlywed couples to demonstrate that this individual difference in decision making has important implications for romantic relationships. Consistent with the idea that maximizers are more likely to compare their current romantic partners to potential alternative partners’ readily observable qualities, such as their physical attractiveness and status, results demonstrated that intimates’ maximization moderated the implications of these sex-differentiated variables for marital satisfaction. Specifically, maximizing men who had attractive (vs. unattractive) wives were more satisfied at the start of their marriages. Likewise, maximizing women who had high (vs. low) status husbands experienced less steep declines in satisfaction over time. These findings demonstrate that maximization has important implications for long-term romantic relationships by accentuating the effects of readily observable partner qualities on relationship outcomes.
People often form pair bonds that sustain over an extended period of time (Hazan & Diamond, 2000). Not only do such pair bonds benefit reproduction (Symons, 1979), they play a crucial role in people’s overall physical and mental health (Holt-Lunstad, Smith, & Layton, 2010; Proulx, Helms, & Buehler, 2007; Robles, Slatcher, Trombello, & McGinn, 2014). Indeed, merely being married reduces mortality risk (House, Landis, & Umberson, 1988). In fact, the effect of poor relationship quality on mortality is as strong as the effects of better-known risk factors, such as smoking and alcohol use, and even stronger than other important factors, such as sedentariness and obesity (Holt-Lunstad et al., 2010). Although numerous factors can affect relationship quality (Karney & Bradbury, 1995), individual characteristics of the partners themselves play a significant role (McNulty, 2013). Given the benefits associated with maintaining satisfying long-term relationships, it is important that intimates choose suitable long-term relationship partners.
Nevertheless, there are fundamental differences in the way that people make such choices. One notable way people differ in decision making is the extent to which they labor over and attempt to maximize their outcomes. Drawing on the theory of bounded rationality (Simon, 1955), Schwartz and colleagues (2002) provided evidence that some people aim to make the best choices and, even after making their choices, continue to exhaustively consider all possible alternatives (i.e., maximizers), whereas other people aim to make choices that meet their standards without necessarily being the best (i.e., satisficers). For example, maximizers are more likely to flip through all available radio stations to find the one playing the “best” song and may even continue to check other stations, whereas satisficers are more likely to settle on the first station that is playing a “good enough” song.
In the context of romantic relationships, then, maximizers should attempt to evaluate all possible potential partners and aim to choose the “best” partner while continuing to evaluate alternative partners. Satisficers, in contrast, should seek out partners that meet their standards, or are considered “good enough” partners, without continuing to evaluate other options. Indeed, when evaluating online dating profiles, maximizers examine more profiles than do satisficers (Yang & Chiou, 2010; see also Schwartz, 2004). Maximization, however, should predict more than merely laboring over a choice. Given that people in relationships face alternative partners on a daily basis (Kelley & Thibaut, 1978; Lydon & Karremans, 2015; McNulty, Meltzer, Makhanova, & Maner, 2018; Rusbult, 1980), the tendency to continually consider options may have important implications for how people evaluate their relationships—and this may be especially true in the context of long-term relationships, where people consider significantly more options over an extended period of time. Indeed, a large body of research demonstrates that maximizing tendencies are negatively associated with people’s satisfaction in a variety of decision-making domains (Besharat, Ladik, & Carrillat, 2014; Bruine de Bruin, Dombrovski, Parker, & Szanto, 2016; Dahling & Thompson, 2012; Newman, Schug, Yuki, Yamada, & Nezlek, 2018).
The goal of the present research was to examine whether people’s tendencies to maximize versus satisfice are associated with their relationship outcomes, and whether such an association depends on the qualities or traits of their partners. In pursuit of this goal, the remainder of this introduction is divided into three sections. The section “Maximization in the Context of Long-Term Romantic Relationships” reviews rational choice theory and highlights the potential negative implications of maximization for people’s long-term romantic relationships. The section “The Moderating Role of Important Partner Qualities” raises the possibility that the association between people’s maximizing tendencies and their long-term relationship outcomes may be moderated by their partners’ readily observable traits—specifically, their physical attractiveness and status. Given the sex-differentiated implications of partner physical attractiveness and partner status for long-term relationship outcomes, this section additionally raises the possibility that any such effects may be further moderated by sex. The section “Overview of the Current Study” summarizes the current research, which uses data from two independent, 3-year longitudinal studies of newlywed couples to test the prediction that people’s maximizing tendencies moderate the sex-differentiated association between partner physical attractiveness and intimates’ marital satisfaction trajectories as well as the sex-differentiated association between partner status and intimates’ marital satisfaction trajectories.
Maximization in the Context of Long-Term Romantic Relationships
Drawing from critiques of rational choice theory (see Simon, 1955), Schwartz and colleagues (2002) have argued that some individuals (i.e., maximizers) seek to make optimal choices—that is, they consider all potential options to make choices that exceed their standards. In contrast, other people (i.e., satisficers) make choices that are deemed good enough—that is, they make choices that meet their standards. 1 Moreover, and critical with respect to the current question, their desire to achieve the best possible outcomes leads maximizers to continue considering all possible alternative options even after making a final decision whereas satisficers are less likely to do so (Schwartz et al., 2002). This postdecision tendency to consider alternatives is associated with increased interest in, and sensitivity to, social comparison feedback such that maximizers (vs. satisficers) are more likely to compare their choices with other people’s choices and derive satisfaction based on those comparisons (Iyengar, Wells, & Schwartz, 2006; Schwartz et al., 2002; Sparks, Ehrlinger, & Eibach, 2012). Consequently, maximizers experience more regret and report lower satisfaction with their choices (Besharat et al., 2014; Bruine de Bruin et al., 2016; Newman et al., 2018; Schwartz et al., 2002).
Notably, maximizers are particularly likely to experience negative outcomes in decision-making domains, where they are unable to conduct an exhaustive search of all possibilities before making their final choice (Newman et al., 2018; Schwartz et al., 2002). One such domain may be the mating market, where maximizers are unable to consider all possible long-term partners and thus subsequently question whether they have chosen the “best” long-term partner. Notably, relationships may be a novel and important domain in which to examine such decision-making tendencies because such choices are longer lasting relative to most other decision-making domains. Moreover, after choosing a long-term partner, intimates continue to face a barrage of potential alternative partners on a daily basis—many of whom appear desirable (particularly in the age of social media and photo editing; see Li, van Vugt, & Colarelli, 2018), which could have potential negative implications for their long-term relationship outcomes (Kelley & Thibaut, 1978; McNulty et al., 2018; Rusbult, 1980). To the extent that maximizers (vs. satisficers) attend more to such potential alternative partners and more frequently compare their current partners to those alternatives, maximizers may be at greater risk of experiencing poorer relationship outcomes. Empirical evidence is consistent with this notion. Mikkelson and Pauley (2013) demonstrated that, in a sample of partnered undergraduate students, maximizers attended more to relationship alternatives and reported lower relationship satisfaction than did satisficers.
The Moderating Role of Important Partner Qualities
It is possible, however, that some maximizers may be buffered from experiencing such negative relationship outcomes and may even experience more positive relationship outcomes. Drawing on the basic tenets of social comparison theory (Festinger, 1954), when people compare their outcomes to others with lesser outcomes, they experience greater satisfaction; in contrast, when people compare their outcomes to others with better outcomes, they experience reduced satisfaction. Drawing on the basic tenets of interdependence theory (Kelley & Thibaut, 1978), people are most satisfied with their relationships to the extent that their relationships meet or exceed their standards (i.e., to the extent that their relationship benefits outweigh their relationship costs). Taken together, these theories suggest that maximizers whose long-term partners compare less favorably to other people’s partners or potential alternatives may be most susceptible to experiencing lower relationship satisfaction because such maximizers learn, through the process of frequent upward social comparisons, that they did not choose the “best” partner; thus, their relationships fail to meet their standards. In contrast, maximizers whose long-term partners compare relatively more favorably to other people’s partners or potential alternatives may actually experience higher relationship satisfaction because such maximizers learn, through the process of frequent downward social comparisons, that they indeed chose the “best” partner; thus, their relationships meet their standards.
Given that information regarding potential alternative partners is generally limited, most comparisons are likely based on readily observable qualities. According to the Ideal Standards Model (Fletcher, Simpson, Thomas, & Giles, 1999), people evaluate their partners against three key traits: partner warmth, partner attractiveness, and partner status. Nevertheless, only two of these key traits are readily observable in potential alternatives and thus readily comparable to people’s current partners—partner attractiveness and partner status. Indeed, physical attractiveness (e.g., facial symmetry) is one of the first and strongest predictors of initial attraction (Walster, Aronson, Abrahams, & Rottman, 1966), and people demonstrate high agreement regarding who is attractive (Langlois et al., 2000). Likewise, status is easily observed in terms of readily observable features such as people’s dress, homes, cars, and jobs (Fletcher et al., 1999)—status cues that are strongly representative of people’s income (Fieder et al., 2005; Li, Bailey, Kenrick, & Linsenmeier, 2002). By observing and comparing their partners’ attractiveness and status to potential alternatives’ attractiveness and status, maximizers may surmise the extent to which they have obtained the “best” partner. Satisficers, in contrast, attend less to alternatives and engage in fewer comparisons, which may attenuate the extent to which their partners’ attractiveness and status affect their satisfaction. Accordingly, we might expect intimates’ maximizing tendencies to moderate the effects of partner attractiveness and partner status on relationship outcomes; maximizers (vs. satisficers) may be less satisfied with their partners, to the extent that their partners are relatively unattractive or have relatively low status (i.e., income), but more satisfied with their partners to the extent that their partners are relatively attractive or have relatively high status (i.e., income).
Considering Sex Differences
If maximization tendencies do indeed interact with such readily observable partner qualities to predict relationship outcomes, it is important to consider sex differences for such effects. A large body of research suggests that men and women differentially value both partner physical attractiveness and partner status (Buss, 1989; Buss & Barnes, 1986; Kenrick, Sadalla, Groth, & Trost, 1990; Li et al., 2002; Li et al., 2013; Meltzer, McNulty, Jackson, & Karney, 2014b). Whereas men value long-term partner physical attractiveness more than do women (Buss & Schmitt, 1993; Feingold, 1990, 1992; Kenrick et al., 1990; Li et al., 2002; Li et al., 2013), women value long-term partner status more than do men (Buss, 1989; Buss & Barnes, 1986; Kenrick et al., 1990; Li et al., 2002; Li et al., 2013). Accordingly, we might expect the effects of partner attractiveness and status on long-term relationship outcomes to be further moderated by sex such that maximizing men (vs. maximizing women) are more satisfied to the extent that their partners are relatively attractive, and maximizing women (vs. maximizing men) are more satisfied to the extent that their partners have relatively high status.
Interestingly, consistent with the idea that sex-differentiated effects of partner qualities for relationship evaluations may be moderated by intimates’ maximizing tendencies, recent work suggests that such sex-differentiated effects are relatively small and sometimes fail to emerge. In one speed-dating study, for example, although men stated that long-term partner physical attractiveness was more important to them than did women, and women stated that long-term partner status was more important to them than did men, partner attractiveness and status did not differentially predict men’s and women’s partner choices (Eastwick & Finkel, 2008; also see Eastwick, Luchies, Finkel, & Hunt, 2014). It may be that such sex-differentiated effects are particularly likely to emerge in certain contexts (e.g., Li & Meltzer, 2015; Meltzer, McNulty, Jackson, & Karney, 2014a) such as among maximizers. Satisficers, in contrast, may be less susceptible to any implications of a partner’s failure to meet sex-differentiated standards—likely due, at least in part, to the fact that such individuals attend less to alternative partners and less frequently compare their partners with others who fare favorably on such standards (see Mikkelson & Pauley, 2013). Consistent with this possibility, Eastwick and colleagues (2014) acknowledged that “a moderator may emerge to explain [such inconsistent sex-differentiated effects]” (p. 433).
Overview of the Current Study
The current study aimed to test whether the associations between two readily observable partner traits—physical attractiveness and status—and intimates’ long-term relationship satisfaction would depend on intimates’ sex and maximizing tendencies. Given that newly married couples recently made the ultimate relationship decision (i.e., choosing a lifelong partner), newlywed couples may be the most ideal sample in which to examine these associations. Indeed, couples often experience dramatic changes during the early years of marriage (Lavner & Bradbury, 2010), and these changes strongly impact subsequent marital success (Huston, Caughlin, Houts, Smith, & George, 2001). Thus, we pooled the data from two independent, 3-year longitudinal studies of newlywed couples 2 to test our predictions. At baseline in both studies, husbands and wives completed measures of maximizing tendencies and marital satisfaction, and we obtained objective measures of both partners’ physical attractiveness and status. Then, every 4 to 6 months for the subsequent 3 years, husbands and wives again completed a measure of marital satisfaction. Given both studies’ parallel designs, we combined and simultaneously analyzed all data to maximize power (and, as will be seen in the “Results” section, our key effects were not moderated by study). We predicted that partner physical attractiveness would differentially predict maximizing (but not satisficing) men’s and women’s marital satisfaction. Likewise, we predicted that partner status would differentially predict maximizing (but not satisficing) men’s and women’s marital satisfaction. Notably, these associations could potentially emerge on intimates’ initial satisfaction and/or changes in their satisfaction over time.
Method
Participants
Participants in Study 1 were 113 husbands and 112 wives (comprising 113 heterosexual newlywed couples) participating in a broader longitudinal study in Northern Texas (one wife did not complete the maximization measure and thus could not be included in the current analyses). Participants in Study 2 were 119 husbands and wives (comprising 120 newlywed couples) participating in a broader longitudinal study in Northern Florida (one couple self-identified as a same-sex couple and, given our sex-differentiated predictions, could not be included in the current analyses). In each study, recruitment was initially planned for 12 months but was extended for one additional month (but no longer due to time and monetary limitations) to increase sample size. A post hoc power analysis that accounted for the repeated observations (multiple ratings within each participant; Snijders & Bosker, 2012) suggested that the 3,955 observations—or, effectively 856 observations—were, not surprisingly, relatively nonindependent (intraclass correlation coefficient [ICC] = .48); nevertheless, the study had nearly ample power to detect the smallest, significant effect of interest (power = .75).
Couples in both studies were recruited via invitations sent to couples in the surrounding area who had recently applied for marriage licenses; participants in Study 2 were additionally recruited via fliers and Facebook advertising. As part of the broader goals of the studies, eligibility required that all participants (a) had been married for less than 4 months in Study 1 and less than 3 months in Study 2, (b) were at least 18 years of age, and (c) spoke English (to ensure comprehension of the questionnaires). Study 1 additionally required that both couple members were in their first marriage.
At baseline, husbands and wives in Study 1 were 27.97 years (SD = 5.61 years) and 26.96 years (SD = 4.86 years) of age, respectively, and had completed 15.20 years (SD = 2.77 years) and 15.83 years (SD = 2.86 years) of education, respectively. Seventy-one percent of husbands and 53% of wives were employed full time and 13% of husbands and wives were full-time students. The sample was relatively diverse compared with other samples of newlywed couples (Karney, Kreitz, & Sweeney, 2004); 48% of husbands and wives self-identified as Caucasian, 28% of husbands and 26% of wives self-identified as Black or African American, 16% of husbands and wives self-identified as Hispanic or Latino/a, 3% of husbands and 5% of wives self-identified as Asian, and 5% of husbands and wives self-identified as another race/ethnicity. Couples had been together for an average of 39.47 months (SD = 33.28 months) prior to marriage and 24% of the couples had children.
At baseline, husbands and wives in Study 2 were 32.09 years (SD = 9.88 years) and 30.15 years (SD = 8.09 years) of age, respectively, and had completed 15.86 years (SD = 2.32 years) and 16.19 years (SD = 2.35 years) of education, respectively. Seventy-one percent of husbands and 61% of wives were employed full time and 20% of husbands and 19% of wives were full-time students. Seventy-seven percent of husbands and 78% of wives self-identified as Caucasian, 13% of husbands and 12% of wives self-identified as Black or African American, 3% of husbands and 4% of wives self-identified as Hispanic or Latino/a, 1% of husbands and wives self-identified as Asian, 2% of husbands and 5% of wives identified as having two or more ethnicities, and 4% of husbands self-identified as another race/ethnicity. Couples had been together for an average of 43.34 months (SD = 31.41 months) prior to marriage and 27% of the couples had children.
Procedure
After enrolling in each study, participants completed a battery of questionnaires via Qualtrics.com or through the mail prior to a laboratory session. These questionnaires included a consent form approved by each study’s local human subjects review board; measures of maximizing tendencies, income, and global marital satisfaction; additional measures beyond the scope of the current analyses; and a letter instructing all spouses to complete their questionnaires independently of one another. At their laboratory session, we took each spouse’s photograph, which provided objective information regarding intimates’ physical attractiveness (as described in the next section). All couples received US$100 for completing these questionnaires and the corresponding session.
Across the subsequent 3 years, we recontacted couples at approximately 6-month (for a total of seven assessments in Study 1) or 4-month (for a total of 10 assessments in Study 2) intervals, and again mailed each spouse a global marital satisfaction questionnaire, along with a letter of instruction reminding spouses to complete their questionnaires independently. Couples received a check (Study 1 = US$30, Study 2 = US$25) after completing each follow-up assessment.
Measures
Physical attractiveness
For each study, a group of trained research assistants (Study 1, N = 5; Study 2, N = 4) used the photographs to rate each intimate’s facial attractiveness on a scale ranging from 1 to 10, where higher ratings indicate more physically attractive faces. We centered each intimate’s face in the photo and instructed coders to rate only the facial attractiveness of each person. To ensure that each intimate’s attractiveness ratings were made independent of his or her partner’s attractiveness, coders rated all husbands first followed by all wives. As other research has suggested (see Meltzer et al., 2014a), objective ratings of physical attractiveness help minimize the influence of factors confounded with own or partner perceptions of attractiveness. Consistent with findings that people within and across cultures show very high levels of agreement regarding who is attractive (Langlois et al., 2000), our coders demonstrated adequate levels of agreement (Study 1: ICC = .82 for husbands, ICC = .92 for wives; Study 2: ICC = .85 for husbands, ICC = .86 for wives).
Status
Prior research has demonstrated that income is a readily observable indicator of objective status (Fieder et al., 2005; Li et al., 2002). Thus, at baseline, we assessed intimates’ yearly income. Specifically, participants indicated how much money they earned during the previous year before accounting for taxes or other deductions (reported in the thousands). One husband in Study 1 and one wife in Study 2 reported extremely high incomes (each were more than 8 standard deviations above the sample mean) and thus we truncated their incomes to match the next highest income reported (US$215K). Across both studies, 11 (4.7%) husbands and 22 (9.5%) wives failed to provide their yearly incomes and thus could not be included in analyses that utilized income as a predictor (though we replaced each missing value with the sample average in analyses that utilized income as a covariate to maximize our sample size).
Maximizing tendencies
At baseline, we assessed intimates’ maximizing tendencies using the Maximization Scale (Schwartz et al., 2002), which assesses individual differences in the tendency to satisfice versus maximize across numerous different domains and thus captures intimates’ general maximizing tendencies. Intimates indicated the extent to which they agreed with 13 statements using a 7-point scale (1 = “Strongly disagree,” 7 = “Strongly agree). Higher scores reflect a greater tendency to maximize; lower scores reflect a greater tendency to satisfice. In the current study, internal consistency of this measure was modest (Study 1: α = .70; Study 2: α = .69) but similar to what others have demonstrated (Schwartz et al., 2002). Given known issues surrounding Cronbach’s alpha (McNeish, 2018), however, this somewhat low reliability is likely an underestimate of the measure’s true reliability and, as will be seen in the “Results” section, did not undermine our ability to detect effects.
Marital satisfaction
At baseline and all follow-up assessments, we assessed global marital satisfaction using the Quality Marriage Index (Norton, 1983). Intimates indicated the extent to which they agreed or disagreed with six general statements about their marriage (e.g., “My relationship with my partner makes me happy”). Five items ask intimates to respond according to a 7-point scale (1 = “Strongly disagree,” 7 = “Strongly agree”), whereas one item asks intimates to respond according to a 10-point scale (1 = “Very unhappy,” 10 = “Perfectly happy”), yielding scores from 6 to 45. Higher scores reflect greater satisfaction with the marriage. Internal consistency was high across all assessments in both studies (all αs ≥ .90).
Covariates
To ensure that partner physical attractiveness, partner status, and intimates’ maximizing tendencies did not appear to be associated with intimates’ marital satisfaction only because they are associated with related factors, we assessed and controlled several covariates. Specifically, given that physical attractiveness and status are associated with age (see Meltzer et al., 2014b) and social skills/extraversion (Langlois et al., 2000; Meier, Robinson, Carter, & Hinsz, 2010), we assessed partner age and extraversion at baseline and controlled for each (Study 1 utilized the 10-item version of the International Personality Item Pool’s [Goldberg, 1999] Extraversion subscale, α = .88; whereas Study 2 utilized the 60-item version, α = .93). Across both studies, two husbands did not provide their age and one wife did not complete the extraversion measure. In addition, given that the association between partner physical attractiveness and marital satisfaction depends on intimates’ own physical attractiveness (McNulty, Neff, & Karney, 2008; Meltzer, McNulty, Novak, Butler, & Karney, 2011), we aimed to minimize the likelihood that any association between partner physical attractiveness, intimates’ sex, maximizing tendencies, and marital satisfaction was not due to intimates’ own attractiveness. Thus, we controlled for objective ratings of intimates’ own facial attractiveness in analyses that utilized partner attractiveness as a predictor. Likewise, to minimize the likelihood that any association between partner income, intimates’ sex, maximizing tendencies, and marital satisfaction was not due to intimates’ own income, we controlled for intimates’ own income in analyses that utilized partner income as a predictor. Finally, many intimates in both studies self-identified as full-time students, which would negatively affect the validity of yearly income as a measure of partner status; thus, we assessed and additionally controlled intimates’ student status (coded such that nonstudents = −1 and full-time students = 1) in analyses that utilized partner income as a predictor.
Results
Descriptive Statistics and Preliminary Analyses
Descriptive statistics for and zero-order correlations among all predictors and covariates are presented in Table 1. A few results are worth highlighting. First, consistent with other samples of long-term couples (e.g., Kenrick & Keefe, 1992; Meltzer et al., 2014b), husbands (vs. wives) on average were older, t(228) = 5.75, p < .001, and reported a higher income, t(202) = 3.16, p = .002. Moreover, husbands (vs. wives) reported higher maximizing tendencies, t(230) = 1.99, p = .048. Second, husbands’ and wives’ physical attractiveness fell near the midpoint of the scale and, although they did not differ from one another, t(230) = −0.31, p = .758, they were positively associated with one another, supporting our decision to control for intimates’ own attractiveness. Third, partner age was negatively associated with partner physical attractiveness among both husbands and wives, again, supporting our decision to control for partner age. Finally, wives’ income was negatively associated with husbands’ maximizing tendencies, supporting our decision to control for partner income.
Descriptive Statistics for and Zero-Order Correlations Among All Predictors and Covariates.
Note. Correlations for husbands are presented below the diagonal and correlations for wives are presented above the diagonal. Correlations between spouses are presented in bold along the diagonal. Different subscripts in the same column denote sex-differentiated means (p < .05).
p < .10. *p < .05. **p < .01. ***p < .001.
Describing the Trajectory of Marital Satisfaction
We used the mixed model function in SPSS 23 to account for the nested nature of our data. Specifically, we used two-level cross models, where intimates were nested within dyads and intimates and assessments were crossed to account for the fact that both couple members completed all assessments at approximately the same time (see Kenny, Kashy, & Cook, 2006). Before testing our key predictions, we first examined the trajectory of intimates’ marital satisfaction over time. Specifically, we regressed marital satisfaction onto husbands’ and wives’ Intercept and Time estimates, in which Time represents year of assessment and was coded from 0 to 3 (so that the intercept represented initial marital satisfaction), and the intercepts and slopes were allowed to vary across intimates. We additionally examined whether any effects differed across husbands and wives, and controlled for study (Study 1 = −1, Study 2 = 1) to account for idiosyncratic differences across studies.
Consistent with other studies of newlywed couples (Meltzer et al., 2014b; Meltzer et al., 2011), results demonstrated that, on average, intimates reported relatively high levels of initial marital satisfaction (π = 41.14, SE = 0.29) that did not differ across husbands and wives, β = 0.08, 95% confidence interval (CI95%) = [−0.21, 0.37], t(224.48) = 0.53, p = .595. 3 Moreover, intimates’ marital satisfaction, on average, decreased linearly over the first 3 years of marriage, and wives (π = −2.06, SE = 0.27) experienced steeper declines than did husbands (π = −1.50, SE = 0.22), β = −0.23, CI95% = [−0.47, −0.09], t(127.29) = −2.87, p = .005, effect-size r = .25. Notably, there was substantial between-subjects variability in all random effects (all ps < .001), suggesting that some husbands and wives began their marriages with higher or lower levels of marital satisfaction than others and that some husbands and wives experienced more or less change in their satisfaction than others. The primary analyses examined whether partner attractiveness and partner status accounted for this variability, and whether such effects depend on intimates’ maximizing tendencies and sex.
Do Maximizing Tendencies and Partner Physical Attractiveness Differentially Predict Husbands’ and Wives’ Marital Satisfaction?
Our second set of analyses examined whether intimates’ maximizing tendencies and sex moderated the association between intimates’ partners’ physical attractiveness and their marital satisfaction trajectories. Meltzer and colleagues (2014a) have argued that sex-differentiated effects of partner physical attractiveness are most likely to emerge among couples with wives who are of a reproductively viable age; thus, we excluded any couples with wives older than 35 years of age (n = 33). We then reestimated the growth-curve model described above but additionally included partner attractiveness (standardized across intimates and studies), intimates’ sex (coded such that husbands = −1 and wives = 1), maximizing tendencies (standardized across intimates and studies), all two-way interactions, and the crucial three-way interaction (Partner Attractiveness × Sex × Maximization) as predictors, as well as the interactions between all predictors and time (to test whether the three-way interaction predicted changes in intimates’ satisfaction). Moreover, consistent with other work examining sex-differentiated implications of partner attractiveness (see Meltzer et al., 2014b), we additionally controlled for intimates’ own attractiveness as well as their partners’ age, income, and extraversion 4 (each standardized across intimates and studies). Results demonstrated that the Partner Attractiveness × Sex × Maximization effect did not interact with Time (p = .848); thus, we removed all interactions involving Time (though we continued to control Time and the Time × Sex interaction, given that intimates experienced sex-differentiated declines in satisfaction).
The results of this analysis are presented in Table 2. As can be seen, among intimates with average maximizing tendencies, the Partner Attractiveness × Sex interaction trended toward significance (p = .149). Nevertheless, this two-way interaction was qualified by the three-way interaction, suggesting that intimates’ maximizing tendencies marginally moderated the sex-differentiated association between partner attractiveness and intimates’ initial marital satisfaction. 5 We deconstructed this marginal three-way interaction by estimating the simple Sex × Partner Attractiveness interactions for satisficers (1 SD below the sample mean) and maximizers (1 SD above the sample mean). Among satisficers, the Partner Attractiveness × Sex interaction was not significant, β = 0.08, CI95% = [−0.45, 0.62], t(295.52) = 0.31, p = .758, indicating that partner attractiveness was not differentially associated with satisficing husbands’ and wives’ initial marital satisfaction. In fact, partner attractiveness was unassociated with satisficing intimates’ initial marital satisfaction, β = −0.13, CI95% = [−0.73, 0.47], t(400.81) = −0.42, p = .676. Among maximizers, in contrast, the Partner Attractiveness × Sex interaction emerged as significant, β = −0.62, CI95% = [−1.19, −0.05], t(315.24) = −2.15, p = .032, effect-size r = .12, indicating that partner attractiveness was differentially associated with maximizing husbands’ and wives’ initial marital satisfaction. This three-way interaction is depicted in Figure 1—as can be seen, maximizing husbands (vs. wives) were more satisfied at the start of their marriages to the extent that they had relatively attractive partners.
Effects of Intimates’ Maximizing Tendencies, Sex, and Partner Attractiveness on Intimates’ Initial Marital Satisfaction.
Note. 90% CIs are provided. For sake of clarity, key effect is bolded. CI = confidence interval.
p < .10. *p < .05. **p < .01. ***p < .001.

Interactive effects of intimates’ partner attractiveness, sex, and maximizing tendencies on initial marital satisfaction.
We conducted four exploratory supplemental analyses to examine the robustness of this three-way interaction. First, we tested whether it was further moderated by study; it was not, γ = −0.02, CI95% = [−0.48, 0.45], t(276.98) = −0.08, p = .940. Second, we examined whether it held when we no longer controlled intimates’ own attractiveness or their partners’ age, income, and extraversion; it did, β = −0.36, CI90% = [−0.707, −0.004], t(364.09) = −1.67, p = .096, effect size r = .09. Third, we explored whether it continued to predict intimates’ marital satisfaction at the end of the study (three years into marriage; to test this, we recentered Time such that 0 represents 3 years, and controlled for the association between partner physical attractiveness, sex, maximizing tendencies, and the null associations with changes over time); it did not, β = −0.45, CI95% = [−1.38, 0.47], t(205.42) = −0.96, p = .337, suggesting that such effects diminish over time. Finally, given that age can serve as a proxy for fertility (e.g., Dunson, Colombo, & Baird, 2002), we additionally explored whether a similar pattern emerged using partner age instead of partner physical attractiveness; a similar pattern indeed emerged although partner age differentially affected maximizing husbands’ (vs. wives) changes in satisfaction over time rather than their initial marital satisfaction (see the supplemental material for the results of this exploratory analysis).
Do Maximizing Tendencies and Partner Status Differentially Predict Husbands’ and Wives’ Marital Satisfaction?
Our final set of analyses examined whether intimates’ maximizing tendencies and sex moderated the association between intimates’ partners’ status and their marital satisfaction trajectories. Specifically, we reestimated the original growth-curve model but additionally included partner income (standardized across intimates and studies), intimates’ sex (coded such that husbands = −1 and wives = 1), maximizing tendencies (standardized across intimates and studies), all two-way interactions, and the crucial three-way interaction (Partner Income × Sex × Maximization) as predictors, as well as the interactions between all predictors and time (to test whether the three-way interaction predicted changes in intimates’ satisfaction). Similar to our previous set of analyses, we additionally controlled for intimates’ own income, their partners’ age, physical attractiveness, and extraversion (each standardized across intimates and studies 4 ) as well as their partners’ student status.
The results of this analysis are presented in Table 3. As can be seen, unlike partner physical attractiveness, the Partner Income × Sex × Maximization interaction was not significantly associated with intimates’ initial satisfaction, though it was marginally associated with intimates’ changes in satisfaction. That is, the sex-differentiated association between partner income and intimates’ changes in marital satisfaction during the first 3 years of marriage depended on their maximizing tendencies 5 . We deconstructed this three-way interaction by estimating the simple Partner Income × Sex interactions for satisficers (1 SD below the sample mean) and maximizers (1 SD above the sample mean). Among satisficers, the Partner Income × Sex interaction was not significant, β = −0.13, CI95% = [−0.45, 0.19], t(159.01) = −0.79, p = .430, indicating that partner income was not differentially associated with changes in satisficing husbands’ and wives’ marital satisfaction. In fact, partner income was unassociated with changes in satisficing intimates’ marital satisfaction, β = −0.31, CI95% = [−0.71, 0.09], t(206.60) = −1.51, p = .132. Among maximizers, in contrast, the Partner Income × Sex interaction emerged as marginally significant, β = 0.41, CI90% = [0.02, 0.81], t(200.73) = 1.73, p = .085, effect-size r = .12, indicating that partner income was differentially associated with changes in maximizing husbands’ and wives’ marital satisfaction. This three-way interaction is depicted in Figure 2—as can be seen, maximizing wives (vs. husbands) experienced less steep declines in satisfaction across the first 3 years of marriage to the extent that their partners earned relatively high incomes.
Effects of Intimates’ Maximizing Tendencies, Sex, and Partner Status on Intimates’ Marital Satisfaction Trajectories.
Note. 90% CIs are provided. For sake of clarity, key effects are bolded. CI = confidence interval.
p < .10. *p < .05. ***p < .001.

Interactive effects of intimates’ partner income, sex, and maximizing tendencies on changes in marital satisfaction across the first 3 years of marriage.
We conducted three exploratory supplemental analyses to examine the robustness of this three-way interaction. First, we tested whether it was further moderated by study; it was not, γ = −0.04, CI95% = [−0.31, 0.22], t(192.81) = −0.28, p = .784. Second, we examined whether it held when we no longer controlled intimates’ own income or their partners’ age, attractiveness, and extraversion; it did, β = 0.28, CI90% = [0.01, 0.55], t(201.78) = −1.74, p = .084, effect-size r = .12. Finally, we explored whether it predicted intimates’ marital satisfaction at the end of the study (three years into marriage); it did not, β = 0.57, CI95% = [−0.27, 1.42], t(214.44) = 1.912, p = .265, suggesting that it takes more than 3 years for partner income to differentially affect maximizing husbands’ and wives’ marital satisfaction at a given point in time.
Discussion
Rationale and Summary of Results
After choosing a long-term partner, intimates face a barrage of potential alternative partners. Given that maximizers (vs. satisficers) more frequently attend to such potential alternatives and compare the readily observable, desirable traits of those alternatives (e.g., physical attractiveness, status) to their current partners’ traits, they may be at greater risk of experiencing poorer relationship outcomes, to the extent that their partners compare less favorably to such alternatives. Nevertheless, if their partners compare more favorably to such alternatives, maximizers (vs. satisficers) may experience better relationship outcomes. Given the sex-differentiated preferences for partner physical attractiveness and partner status (e.g., Buss, 1989; Buss & Barnes, 1986; Kenrick et al., 1990; Li et al., 2002; Meltzer et al., 2014b; but also see Eastwick & Finkel, 2008), we predicted that maximizing men (vs. women) would be more satisfied to the extent that their partners are relatively attractive, and that maximizing women (vs. men) would be more satisfied to the extent that their partners have relatively high status. We pooled the data from two independent, 3-year longitudinal studies of newlywed couples to test these predictions. Results were consistent with predictions. Maximizing men (vs. women) were more satisfied at the start of their marriages to the extent that they had physically attractive partners; in contrast, satisficing men and women were no more or less satisfied at the start of their marriages to the extent that they had physically attractive partners. Likewise, maximizing women (vs. men) experienced less steep declines in marital satisfaction across the first three years of marriage to the extent that their partners had high incomes; in contrast, satisficing men and women experienced no more or less steep declines in marital satisfaction to the extent that their partners had high incomes. It is worth noting, however, that our higher order associations emerged as marginally (rather than traditionally) significant; results should thus be interpreted with caution until they can be replicated.
An astute reader may have noticed that maximizing women with attractive (vs. unattractive) partners were relatively less satisfied at the start of their marriages, and maximizing men with high (vs. low) status partners experienced relatively steeper declines in satisfaction over time. Although we did not predict such simple effects a priori, these findings are consistent with other research. Other scholars have demonstrated, for example, that women with attractive (vs. unattractive) partners are more concerned about sexual infidelity (White, 1980), which appears to be a valid concern—attractive, partnered men not only desire more frequent extra-pair relations (White, 1980), they engage in more frequent sexual infidelity (Gangestad & Thornhill, 1997; cf. Rhodes, Morley, & Simmons, 2013). Likewise, other scholars have demonstrated that increases in partnered women’s income leads to declines in their partners’ overall well-being over time (Rogers & DeBoer, 2001), and is associated with an increased likelihood of marital dissolution (Ono, 1998; Teachman, 2010). As the current work demonstrates, however, individual differences such as women’s maximizing tendencies can moderate these associations. Indeed, partner attractiveness was negatively associated with maximizing wives’ initial marital satisfaction but unassociated with satisficing wives’ initial marital satisfaction, and partner income was positively associated with maximizing husbands’ declines in marital satisfaction but unassociated with satisficing husbands’ declines in marital satisfaction.
Implications and Future Directions
These results have several important theoretical implications. Perhaps most notably, the current research demonstrates the critical implications of an important individual difference in decision-making styles for long-term romantic relationships—maximization. The choice of a long-term partner is a relatively enduring one that has notable implications for people’s overall physical and mental health (Holt-Lunstad et al., 2010; Proulx et al., 2007; Robles et al., 2014). For these reasons, decision making in this domain is critical; thus, it is likely that people attend to certain qualities in potential partners that will maximize their outcomes. Nevertheless, there are individual differences in such maximizing tendencies, and the current research provides novel evidence that such differences have implications for long-term relationship outcomes. Moreover, the current findings suggested that maximizers are not always successful in maximizing their outcomes. Indeed, maximizing men with unattractive partners were relatively dissatisfied with their marriages, and maximizing women with low-status partners experienced relatively steeper declines in marital satisfaction over time. These negative outcomes are likely due to maximizers’ tendencies to frequently attend to relationship alternatives (Mikkelson & Pauley, 2013). Although we were unable to directly test this mechanism in the current research, future research may benefit from doing so. Future research may also benefit from continuing to examine additional ways that maximization influences romantic relationships. For example, given men’s relative preference for sexual novelty (Little, DeBruine, & Jones, 2014), maximizing men (vs. women) may be at greater risk of lower sexual satisfaction in their long-term relationships, though this effect may be buffered by their partners’ willingness to introduce sexual novelty. Likewise, maximization may play an important role in individuals’ decision to marry such that maximizers may engage in longer courtships or be more hesitant to agree to marriage.
The current research also has implications for our understanding of maximization more generally. In contrast to most previous research demonstrating that maximizers (vs. satisficers) experience more negative outcomes following decision-making processes (Besharat et al., 2014; Bruine de Bruin et al., 2016; Newman et al., 2018), the current studies are among the first (at least to our knowledge) to demonstrate that some maximizers can actually experience more positive outcomes—especially in decision-making domains where an exhaustive search of all possibilities is impossible (Newman et al., 2018; Schwartz et al., 2002). Indeed, maximizing men were more satisfied to the extent that they had attractive partners, and maximizing women were more satisfied to the extent that they had high-status partners. Future research may benefit from further exploring other domains in which maximizers experience similar positive outcomes.
Readers familiar with the maximization literature may have noticed that, in the current research, maximization was, on average, unassociated with intimates’ initial satisfaction or changes in marital satisfaction (see the simple effects of maximization in Tables 2 and 3). Decisions regarding who to marry are critically important, and may be one of the most important decisions that people make in their lives. Given the enduring nature of such a decision, it is possible that newly married couples are less susceptible to such negative outcomes (at least at the beginning) because they may more heavily weigh the potential costs and benefits of their decision, and because they hold relatively positive illusions (Murray, Holmes, & Griffin, 1996). Indeed, we are unaware of other research that has examined the implications of maximization for such important life decisions. Although Mikkelson and Pauley (2013) demonstrated that maximizers (vs. satisficers) were less satisfied with their relationships, they utilized samples of undergraduate women, and the implications of choosing dating partners in young adulthood are notably less consequential than the implications of choosing marriage partners in adulthood. Of course, in the current research, the null association between intimates’ maximizing tendencies and their relationship outcomes were moderated by the quality of their partners.
The current findings also help reconcile inconsistencies in support for evolutionary-based theories such as sexual strategies theory (Buss & Schmitt, 1993) and parental investment theory (Trivers, 1972). According to such theories, partner attractiveness should more positively impact men’s (vs. women’s) long-term relationship outcomes, and partner status should more positively impact women’s (vs. men’s) long-term relationship outcomes. Although such effects have emerged in numerous studies (e.g., Li et al., 2013; Meltzer et al., 2014b), they have failed to emerge in other studies (Eastwick & Finkel, 2008; Eastwick et al., 2014). It is worth noting that the sex-differentiated effect of partner attractiveness trended toward significance in the current research, and the sex-differentiated effect of partner status emerged in the predicted direction. Nevertheless, both effects emerged more strongly for maximizers than for satisficers. It may thus be that unknown sample differences in maximization have accentuated sex differences in prior studies that demonstrated the predicted effect (e.g., Li et al., 2013; Meltzer et al., 2014b) and attenuated those differences in research that failed to demonstrate the predicted effect (e.g., Eastwick & Finkel, 2008, though see Meltzer et al., 2014a). Considering maximizing tendencies and other important individual differences in future research may provide more consistent results.
Finally, future research may benefit from considering the function of individual differences in maximizing tendencies—that is, why some people expend the time and energy necessary to labor over their decision making, such as choosing a suitable long-term partner, whereas other people do not. It may be that such individual differences reflect differences in individual needs, such as those stemming from different life histories. According to life history theory, the harshness and unpredictability of people’s childhood environments can affect their psychological and behavioral functioning in adulthood (e.g., Belsky, Steinberg, & Draper, 1991; Simpson, Griskevicius, Kuo, Sung, & Collins, 2012). People who are exposed to unpredictable early environments tend to be more opportunistic, be more impulsive, and are more likely to seek immediate gratification; thus, they may also be more likely to adopt satisficing tendencies when choosing a relationship partner. That is, they may be more likely to choose a “good enough” partner. People who are exposed to stable early environments, in contrast, tend to be long-term planners who delay gratification for later and potentially larger payoffs; thus, they may be more likely to adopt maximizing tendencies when choosing a relationship partner. That is, they may consider all possible partners in hopes of choosing the “best” partner. In other words, people’s maximizing tendencies may be a result of their early environmental exposure. Future research may benefit from examining this possibility, as well as whether early environmental experiences similarly moderate the effects of partner physical attractiveness and partner status on long-term relationship outcomes.
Strengths and Limitations
Several strengths of this research enhance our confidence in the findings reported here. First, the studies drew from, and did not vary across, two independent studies of marriage, which allowed for increased power. Second, in contrast to using newly formed or hypothetical relationships, the current study utilized samples of participants who were all young, married couples for whom the measured outcomes were real and consequential. Finally, analyses in the current study controlled numerous potential confounds (i.e., own attractiveness, own income, partner age, partner extraversion, and partner student status), helping to decrease the possibility that the results were spurious or suppressed due to associations with those variables. Nevertheless, supplemental analyses also demonstrated that the key effects continued to emerge in uncontrolled models.
Despite these strengths, several factors limit interpretations of the current findings until they can be replicated and extended. First, whereas the relative homogeneity of our two samples enhances our confidence in the pattern of associations that emerged, this lack of variability limits our ability to generalize these findings to other samples of couples (e.g., short-term couples, older married couples, nonheterosexual couples). Maximizing tendencies, for example, may similarly affect relatively shorter term relationships (e.g., dating relationships). Likewise, it is possible that maximizing tendencies may not moderate the association between partner attractiveness and relationship satisfaction among older couples. Given that the current predictions were derived from evolutionary perspectives and the notion that partner attractiveness is important to relationship outcomes due to its association with fertility and successful reproduction, partner attractiveness may no longer differentially affect older men’s and women’s long-term relationships (for a related discussion, see Meltzer et al., 2014a). Future research may benefit from examining this possibility and the extent to which the current findings generalize. Second, although we had a priori, theoretically driven predictions for two different traits that should matter for intimates’ relationship satisfaction in contextually different ways, the predicted effects emerged as marginally (rather than traditionally) significant. Nevertheless, there are notable challenges associated with conducting longitudinal, dyadic research (e.g., stringent inclusion criteria, resource intensiveness; see Finkel, Eastwick, & Reis, 2015) as well as known statistical difficulties of detecting moderator effects (McClelland & Judd, 1993); thus, we believe that the results reported here warrant notable consideration. Of course, future research would benefit from replicating the current results using a larger sample that is specifically designed to test these associations. Finally, the data presented here are correlational and thus are unable to support strong causal conclusions. Although we were able to control some variables that could have been responsible for the associations observed here, other potential third variables remained uncontrolled.
Conclusion
The current research demonstrated that intimates’ maximization has important implications for relationships by accentuating the effects of readily observable partner qualities. Specifically, sex-differentiated implications of partner attractiveness and partner status—hallmarks of evolutionary psychological research—emerged among maximizers but not among satisficers. The current work is among the first studies to demonstrate that some maximizers can actually experience more positive outcomes following a major life decision. Given that choosing a long-term partner is relatively permanent, at least relative to other decision-making domains, it is important that future research continue to explore the implications of maximization for long-term relationships.
Supplemental Material
Supplemental_Material_R3_AB – Supplemental material for Maximizing Tendencies in Marriage: Accentuating the Implications of Readily Observable Partner Characteristics for Intimates’ Satisfaction
Supplemental material, Supplemental_Material_R3_AB for Maximizing Tendencies in Marriage: Accentuating the Implications of Readily Observable Partner Characteristics for Intimates’ Satisfaction by Juliana E. French and Andrea L. Meltzer in Personality and Social Psychology Bulletin
Footnotes
Acknowledgements
We would like to thank James K. McNulty for granting us access to his data.
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: Preparation for this article was supported by start-up funds awarded to Andrea L. Meltzer from Southern Methodist University (Study 1) and by a National Science Foundation Grant BCS-1251520 awarded to James K. McNulty (Study 2).
Notes
Supplemental Material
Supplemental material is available online with this article.
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
