Introduction

Human beings are inherently social creatures who benefit from relationships with other people (Orth-Gomer 2009). Connecting with others by sharing the intimate details of our lives has important implications for our well-being (Cutrona 1996). When people are able to share their problems, they enjoy better mental and physical health (Kulik and Mahler 1989; Sarason and Sarason 2009). Some have suggested that the most promising interventions to improve well-being are those that focus on strengthening individuals’ social supports rather than on reducing their exposure to stressors (Krieger 2001). Although receiving support is clearly important, support from certain persons seems more beneficial than support from others. For married individuals in the United States the spouse is the most important support provider (Dakof and Taylor 1990), and research on U.S. samples has found many mental, physical, and relational benefits for individuals in a supportive intimate relationship, from lower reactivity to stressful events (Grewen et al. 2003) to better marital quality (Acitelli and Antonucci 1994). Though the assumption has been that these benefits are driven by support receipt, research suggests there are also benefits of providing support (Brown et al. 2003). Given that spouses are likely taking on both the roles of support provider and recipient, it is plausible some of the benefits assigned to being a support recipient may be masking a more reciprocal process. Spouses thus may benefit not only from receiving sensitive support from their partner, but from returning the favor as well.

In light of the benefits of support, it is imperative to understand how the benefits of support receipt and provision may operate differently for husbands and wives, given that males and females in the U.S. today are socialized to play different roles when it comes to supportive interactions. While females are typically encouraged to seek support from others and accept it, males are typically socialized to be less reliant upon others and to take a more assertive or expert role in problem situations (Hook et al. 2003; Reevy and Maslach 2001). Such gender differences in socialization processes may shape typical supportive interactions between spouses. Wives may find it easier to seek out support and derive benefits from support (Fiori and Denckla 2012), whereas men may find it more beneficial to provide support. A more complete understanding of how gender relates to supportive roles in a marriage, and how marital outcomes are associated with those roles, will thus not only address key questions about the role of gender in support, but the results of this study may be of particular significance to married couples in the U.S. and around the world. Given that supportive behaviors have been linked to decreased mortality for older adults in the United States (Brown, et al. 2003), understanding more about supportive interactions in marriage is a timely and important issue in light of the dramatically rising number of older adults both in the U.S. (Bengston 2001) and worldwide (Mathers et al. 2001).

Given the importance of supportive interactions in marriage (Grewen et al. 2003), it is critical to understand whether support provision or receipt (or both) is linked to marital outcomes for males and females. The present study sought to answer the question of how the acts of providing or receiving sensitive support were related to positive marital outcomes (i.e., support satisfaction, greater love, and less conflict) for each spouse, using a series of path analyses, specifically Actor Partner Interdependence Models, to account for the linked nature of married couples (APIM; Cook and Kenny 2005; Kenny et al. 2006). Though many studies have examined gender differences in support, a majority have highlighted specific types of support provided by males and females (Acitelli and Antonucci 1994; Cutrona 1996; Verhofstadt et al. 2007), rather than examining links between gender and the ability to sensitively match one’s support efforts to the needs and desires of one’s partner, a critical piece of Cutrona’s (1990) Theory of Optimal Matching. Using this theory, the present study investigated whether providing or receiving sensitive support was more beneficial for men and women by examining paths between husband and wife sensitive support provision (independent variables) and each spouse’s reported support satisfaction, love and conflict (dependent variables). Paired t-tests were also conducted to test for gender differences among all study variables. It should be noted that similar to the current study, which assesses couples from the Midwestern portion of the U.S., the majority of the studies referenced in this paper are based in the U.S. We have explicitly noted those studies in which data were collected outside the U.S.

Theoretical Perspectives on Optimal Support Provision

Although support is critical to well-being, not all spouses are equally adept at providing it. To understand what makes a spouse proficient at support provision, we draw upon the Theory of Optimal Matching (Cutrona 1990; Cutrona and Russell 1990), which emphasizes the importance of the match between support attempts and the recipients’ needs. Optimal matching is best demonstrated when the support provider is sensitive to the needs of the person eliciting support (Cutrona and Suhr 1992), operationalized as the ability to recognize an emotional cue and react or respond in a helpful or appropriate manner (Fruzzetti and Worrall 2010). Building upon this theory, Rini and Dunkel Schetter (2010) highlighted the importance of the support provider’s sensitivity by pointing out that providers often face a difficult dilemma—the ever changing needs of the support recipient. They cite this as a primary reason for why the support provider should focus on sensitively responding to the current needs of the recipient, rather than seeking to continually provide a specific type of support. Although different situations may elicit needs for different types of support, the need for sensitive support provision remains constant. Yet, males and females often differ both in their respective supportive needs (Reevy and Maslach 2001), as well as their ability to respond sensitively to current needs of the partner. Samter (2002) noted that females are more likely than males to offer support both spontaneously and when requested, and place greater value on the management of others’ emotions.

To best capture the sensitivity of the support provided by both males and females during a supportive interaction, Verhofstadt et al. (2007), drawing upon a study of 32 married couples in Belgium, advocated for the combined use of insider (i.e., self report) and outsider (i.e., observational measure) perspectives, as outside observers bring an unbiased perspective and, thus, tend to be better at identifying supportive behaviors than those within the romantic relationship. Further, Rini and Dunkel Schetter (2010) posited it is possible to assess optimal matching by examining recipients’ post hoc appraisals of whether the support received met their needs. As support satisfaction has been linked to numerous health benefits (Heffner et al. 2004), it is clear why also capturing individual’s perception of support is critical. Utilizing both outsiders and insiders better captures relationship dynamics, such as support, rather than relying solely on one reporter (Melby et al. 1995).

The Benefits of Receiving and Providing Support

The importance of capturing and understanding sensitive support provision is best demonstrated by the numerous benefits associated with having a sensitive spousal support provider, particularly for the spouses’ relationship. For example, Cutrona (1996) proposed that support is linked to both positive and negative aspects of marital quality, increasing the strength of the former and lessening the frequency of the latter. Thus, sensitive spousal support is thought to both foster love and minimize conflict because supportive gestures promote trust and appreciation and strengthen the commitment between partners—a commitment that may help the couple survive and prevent future conflicts. Cutrona’s theory indicates that the goodwill engendered by consistent sensitive and supportive acts may work to both prevent the occurrence of conflict, while simultaneously increasing the love in the marriage.

The above-mentioned benefits of support have each been associated with the reception of sensitive support. Rini and Dunkel Schetter (2010) maintained their focus on the needs of the support recipient is consistent with the view that a person’s psychological experience of a social interaction governs the effect of said interaction (Reis et al. 2004). Consistent with this perspective, when the support recipient perceives that he or she has received helpful support, we would expect positive outcomes (increased love, decreased conflict). However, not all of the benefits of support are reserved for the recipient. Supportive interactions in marriage are dyadic, with one spouse playing the role of provider 1 day and recipient the next. Further, receiving support is highly correlated with the provision of support, perhaps due to norms of reciprocity in the exchange of benefits (Knoll et al. 2007). Therefore, focusing only on support receipt and not provision may produce erroneous results (Knoll et al. 2007). For example, significant findings typically associated with reception of sensitive support may actually be capturing to some degree the effects of support provision. Consequently, simultaneous assessment of both support provision and receipt is essential to understand the independent and combined effects of receiving and providing sensitive support on couples.

Why would providing support to another produce benefits for the support provider? Several studies have sought to explain why helping another would have favorable effects for the helper (Brown et al. 2005; Brown et al. 2003; Kleiboer et al. 2006). Supporting a spouse increases the provider’s perception of being needed and valuable, which should be associated with increases in well-being and self-esteem (Knoll et al. 2007), even when the support provision is purely altruistic (Bateson 1998). As to the physical health benefits of support provision, Brown and colleagues (2003) found reduced mortality for older adults who reported providing support to others. Surprisingly, receiving support had no effect on mortality once support provision was taken into account. The authors concluded that the oft-cited beneficial effects of support may mostly be associated with providing rather than receiving. As Gleason et al. (2003) found that giving support to partners was linked with less negative mood in the evening, it is likely providing support would be linked to less conflict in the marriage. Further, Williamson and Clark (1989) found providing support was not only related to more positive moods for the provider, but also to more positive self-evaluation. In line with Cutrona’s (1996) work, these positive moods would likely translate into more positive marital interactions, which would engender greater love between spouses. Taken together, these findings illustrate the importance of considering not only the benefits of receiving helpful support from a spouse, but also the benefits of being a sensitive support provider.

The Role of Gender in Support Provision

Despite the fact that marital support is important for both spouses, research has shown that men and women differ in their actual supportive interactions. Although the bulk of the literature examining gender differences in support has focused on differences in the types of support men and women typically provide and prefer to receive (Acitelli and Antonucci 1994; Cutrona 1996; Verhofstadt et al. 2007), many have begun to question whether these differences may be an artifact of how support has been traditionally studied. For example, using observational and diary data, Neff and Karney (2005) found spouses did not differ on the amount of support provided. Further, Verhofstadt et al. (2007) argued that gender differences may vary depending on how support is measured. They found gender differences when using only self-report measures, such that wives provided more and better support, but observational assessments did not reveal such differences. In support of this finding, the observational research that has been done with married couples has found few gender differences in spousal support behaviors (Neff and Karney 2005; Verhofstadt et al. 2005). The inconsistency in findings across studies measuring support via self-reports versus observational methods underscores the importance of utilizing both methods.

Due to the inconsistency in these findings, we refer back to Rini and Dunkel Schetter (2010) point that the type of support provided is not the crucial element determining its successful delivery, rather we should focus on its sensitivity. Gender differences in support provision exist and cannot be overlooked. Yet, the role of the spouse (provider vs. recipient), not the type of support provided, may be the critical gender difference. In a study of recently married men and women from both the U.S. and mainland China and Taiwan, Xu and Burleson (2001) found that across both samples, women reported desiring significantly greater levels of support from their spouses than did men. Moreover, both U.S. and Chinese wives received less emotional and esteem support than did husbands. Therefore, despite desiring more support than men, it appears that many women actually received less support than did men, highlighting the potential significance of men’s support when provided adequately. Scholars have suggested these gender differences are a result of traditional gender roles that propose that women seek nurturing and understanding in relationships, whereas men seek to take control (Mickelson et al. 1995; Tannen 1990). Therefore, it may be that wives benefit when they receive sensitive support as this serves as an indicator that they are in a relationship in which they are understood. Husbands, conversely, may benefit from providing support as it serves as an indicator of their own agency in the relationship. The better support they provide, the better they feel about themselves and in turn their marriages because effective support may be linked to a greater sense that they are helping provide a solution. Thus, the role of provider versus recipient may be a more meaningful gender difference to focus on in spousal support interactions.

Current Study

Although spousal support has been examined for years, few studies have simultaneously examined the links between sensitive support provision and receipt in couples and their effects on individual and marital outcomes. Given the fact that studies have shown both providing (Brown et al. 2003) and receiving support (Strazdis and Broom 2007) are linked to positive outcomes, examining the combination of the conjoint effects of both on marital outcomes for men and women is a logical next step. Supportive interactions in a marriage are dyadic and the benefits of support are likely allocated to both recipient and provider (Knoll et al. 2007). Thus, simultaneous assessment of both acts is necessary to understand the independent and unique effects of receiving and providing sensitive support. To capture this dyadic nature of support in marriage, the current study utilized the actor-partner interdependence models (APIMs; Cook and Kenny 2005). This approach allowed us to answer two questions pertinent to understanding how both sensitive support provision (actor effect) and receipt (partner effect) are related to husbands and wives feeling more supported and enjoying more love and less conflict.

First, we were interested in examining how sensitive support provision and receipt was associated with each spouse’s self-reported support satisfaction and how this relationship varied by participant gender. We sought to examine whether the provision (actor effect), receipt (partner effect), or the combination of the two (conjoint effect) was significantly associated with husbands and wives feeling satisfied with the support they received. An examination of the conjoint effect provided a unique dyadic understanding of the critical role of simultaneously providing and receiving sensitive support. In order to determine how our independent variables (wife and husband sensitive support provision) were associated with our dependent variables (wife and husband support satisfaction), we made use of a path analysis using APIM. This allowed us to simultaneously assess the impact of both actor and partner effects on multiple outcomes (Kenny et al. 2006). Second, we sought to determine how sensitive support provision and receipt were associated with marital love and conflict for each spouse. We examined whether actor/partner effects or the conjoint effect of both providing and receiving sensitive support was linked with greater love and less conflict for both husbands and wives. Again we conducted a path analysis using APIM to examine the relationship between our independent variables (wife and husband sensitive support provision) and our dependent variables (wife love, husband love, wife conflict, husband conflict). To answer both research questions, we tested the following two hypotheses using APIM as outlined by Kenny et al. (2006) and McIsaac et al. (2008).

  • Hypothesis 1: Based on findings that support receipt is most critical for women (Fiori and Denckla 2012), we predict that husbands’ sensitive support provision (the partner effect for wives) will be associated with greater support satisfaction for wives. Conversely, for males, given the dyadic nature of support interactions and the continual shifting between the roles of provider and recipient (Knoll et al. 2007) we predict that both wives’ and husbands’ sensitive support provision (the conjoint effect of both providing and receiving sensitive support) will be related to greater support satisfaction.

  • Hypothesis 2: We predict that for wives, husbands’ sensitive support provision (or the partner effect for wives) will be significantly associated with greater love and less conflict. We further hypothesize that for husbands, both wives’ and husbands’ sensitive support provision (the conjoint effect of both providing and receiving sensitive support) will be associated with both greater marital love and less marital conflict. Therefore, the conjoint effect of actor and partner behaviors will predict variance in both marital love and marital conflict for men.

Method

Participants

A total of 58 married couples were drawn from a larger study that examined the effects of the marital relationship on children’s development and sibling relationships; only procedures and measures relevant to this study are discussed here. Families were recruited from birth records, newspaper advertisements, church bulletins, daycares, and preschools. To be eligible for the study, couples had to meet the following criteria: a) identify as happily married, b) both spouses agreed to participate, c) have a 2-year-old child, and d) have another child in preschool or elementary school. Recruitment materials indicated that couples could call in to the lab if they considered themselves eligible based on these criteria. After an initial phone screening, all couples who were deemed eligible based on these criteria were scheduled for their first visit.

As seen in Table 1, couples were married for an average of 8.7 years (SD = 3.4). Most spouses were White (93 % of husbands, 97 % of wives) and middle class (most couples had a combined annual income between $70,000 and $90,000). Husbands were approximately 37 years old (SD = 4.6 years) and had at least some college education. Wives were approximately 35 years old (SD = 4.5 years) and all had completed some college. Paired t-tests revealed that husbands and wives only differed significantly by age, t(56) = −3.32, p < .01, such that husbands were 1.5 years older than their wives. This slight age difference is common both in the U.S. (Sweeney 2002) and abroad (Casterline et al. 1986). Of the 58 families that participated, one couple terminated early; therefore the final sample size was 57 couples.

Table 1 Participant demographics characteristics (N = 57)
Full size table

Procedure

Families participated in two laboratory visits, one focusing on the couple and the other on the family. During the first visit, the basis of the current study, couples participated in several marital communication tasks (a baseline picture viewing task, a problem solving task, a compassionate love task) and finished with a support task, the focus of the current study. At the end of the visit, the couple received a questionnaire that assessed both marital and individual functioning. Each couple was given $50 and each child received a small toy as compensation.

Measures

Spousal Support Sensitivity

To examine spousal support sensitivity, we used a 20-min procedure similar to one developed by Cutrona et al. (1997). This task was divided into two 10-min sessions in which each partner had a turn playing the speaker and listener roles. The order of who spoke first was counterbalanced across visits. As the speaker, spouses were told to describe any issue they had been struggling with or that had been causing them stress (e.g., issues with work, friends, health). To ensure the task elicited support and not conflict, the speaker was asked to not choose a topic: (1) that the couple had argued about in the past, and (2) that the speaker blamed the spouse for. In the listener role, the partner was asked to respond normally as if having the conversation at home. Using a coding system adapted from Frosch et al. (1998), two coders evaluated the listener’s support sensitivity on a 7-point scale that ranged from 1 (not characteristic) to 7 (extremely characteristic), with each spouse coded by different coders. A ‘1’ reflected minimal sensitivity, such that the spouse displayed little regard or consideration for their partner, and may have ignored the spouse’s comments or shown a delayed or negative response; a ‘4’ reflected moderate sensitivity, such that the individual showed moderate levels of responsiveness and sensitive support, with comments and needs being responded to fairly often (sometimes neutrally and sometimes sensitively),; and a ‘7’ reflected very high sensitivity, including empathic responsiveness and sensitive behaviors, and attentiveness to the desires, needs, and actions of the spouse. Intercoder reliability was good for both wives (r = .89, p < .001) and husbands (r = .85, p < .001).

Support Satisfaction

Immediately following the support task, both spouses completed a 21-item support satisfaction scale (1 = strongly disagree to 4 = strongly agree; see Appendix A). This scale assessed each partner’s satisfaction with the support they received during the support task (e.g., “My spouse was sensitive to my feelings”; α = .89 for husbands; α = .84 for wives).

Marital Love and Conflict

Marital love and conflict were measured using subscales from Braiker and Kelley (1979) Intimate Relations Questionnaire. To assess marital love, the 10-item love subscale assessed spouses’ degree of belonging, closeness, and interdependence (e.g., “To what extent did you have a sense of belonging with your spouse?”; α = .82 for husbands; α = .83 for wives) on a 9-point scale (1 = not at all to 9 = very much). To assess marital conflict, the 5-item conflict subscale assessed spouses’ overt behavioral conflict and communication of negative affect (e.g., “How often did you and your spouse argue with one another?”; α = .79 for husbands; α = .75 for wives) on a 9-point scale (1 = not at all to 9 = very much).

Results

Descriptive Statistics

Intercorrelations and descriptive statistics are presented in Table 2. As expected among this happily married sample, means for both husbands’ (7.45, SD = .93) and wives’ love (7.74, SD = .85) were relatively high, and means for both husbands’ (3.69, SD = 1.27) and wives’ conflict (3.89, SD = 1.27) were relatively low. Paired t-tests were conducted to test for gender differences among all study variables. Gender differences were found for sensitive support provision, t(56) = 2.30, p < .05, such that wives provided more sensitive support than did husbands, and for love, t(56) = 2.21, p < .05, such that wives reported more love than their husbands did.

Table 2 Descriptive statistics and intercorrelations for study variables for husbands and wives (N = 57)
Full size table

Sensitive Support Provision/Receipt and Spousal Support Satisfaction

To address our first hypothesis, we examined the associations between sensitive support provision and self-reported support satisfaction by gender using APIM (Cook and Kenny 2005; Kenny et al. 2006; McIsaac et al. 2008). The data were analyzed using MPlus 6.12 (Muthen and Muthen 2009). We tested our first hypothesis by comparing the relative fit of a fully saturated model (i.e., a perfect fitting model with zero degrees of freedom; Cook and Kenny 2005) to one in which the actor and partner effects were constrained to equal zero. In our fully saturated model, we allowed the predictor variables (spouses’ sensitive support provision) to correlate with one another, controlling for the effect of one predictor on the outcome as the effect of the other was estimated, and vice versa. Outcome residuals were also allowed to correlate, allowing for the possibility that the relationship between outcomes may be due to dyadic factors other than those in the model. Figure 1 shows the fully saturated model. APIM requires the outcome variables are non-independent (significant within dyad correlation), and the predictor variables are mixed (had both within- and between dyad variability), as indicated by a significant intra-class correlation (ICC) that is between, but not equal to, -1 or 1 (Kenny et al. 2006). These assumptions were met here, as the ICC calculated for husband and wife support sensitivity was .44, p < .05 (see Table 2).

Fig. 1
figure 1

The fully saturated model. Wife and husband support sensitivity as predictors of self reported satisfaction. Unstandardized coefficients presented with standardized coefficients in parentheses (N = 57)

Full size image

To test for actor/partner or conjoint effects for each spouse, the fully saturated model was compared with a model in which actor and partner effects were constrained to be equal. To address the links between sensitive support and support satisfaction, a fully saturated model was compared to two constrained models—one for each spouses’ support satisfaction. A significant chi-square value in the new constrained model indicated the actor and partner pathways were significantly different from each other, suggesting the outcome variable was uniquely predicted by the actions of one relationship member. Non-significant chi-square values indicated the actor and partner effects were equivalent, meaning the outcome was conjointly affected, and the pooled regression coefficient (i.e., regression coefficient averaged across actor and partner effects in the constrained model) was used to capture this effect (Cook and Kenny 2005).

Hypothesis 1 stated that for females, husbands’ sensitive support provision would be linked to support satisfaction, and for males, the conjoint effect of both providing and receiving sensitive support (both wives’ and husbands’ sensitive support provision) would be related to greater support satisfaction. For wives’ support satisfaction, the chi-square value of the constrained model was significant, suggesting that actor and partner effects were significantly different from one another in predicting this outcome (see Table 3). Therefore, as hypothesized, husband sensitivity, but not wife sensitivity, was uniquely and positively related to wife support satisfaction. For husbands’ support satisfaction, we found the chi-square value of the constrained model was significant, indicating the actor and partner effects were significantly different from one another. Husband sensitivity, but not wife sensitivity, was uniquely and positively associated with his support satisfaction. Thus, Hypothesis 1 was not supported for males, as it was husbands’ own sensitive support provision and not the conjoint effect of providing and receiving sensitive support that was associated with support satisfaction.

Table 3 Actor, partner, and conjoint effects of sensitive support provision on support satisfaction and marital quality (N = 57)
Full size table

Sensitive Support Provision and Receipt and Marital Love and Conflict

To test Hypothesis 2, which stated that for females, husbands’ sensitive support provision would be significantly associated with greater love and less conflict for women, and that for males, the conjoint effect of both providing and receiving sensitive support (both husbands’ and wives’ sensitive support provision) would be significantly related to greater marital love and less marital conflict, we again fit APIMs to determine whether actor/partner or conjoint effects were present. All APIM assumptions were met for this research question. Figure 2 shows the fully saturated model. A similar procedure to that conducted for Hypothesis 1 was followed for Hypothesis 2, which considered four outcome variables (wife love, husband love, wife conflict, husband conflict). The fully saturated model was compared to four constrained models, with each of these four models constraining actor and partner pathways to a different outcome. Each constrained model was only allowed to constrain actor and partner pathways to a single outcome variable.

Fig. 2
figure 2

The fully saturated model. Wife and husband support sensitivity as predictors of self-reported domains of marital quality (Braiker and Kelley 1979). Unstandardized coefficients presented with standardized coefficients in parentheses (N = 57)

Full size image

For wives’ love, the chi-square value of the constrained model was significant, suggesting actor and partner effects were significantly different from one another in predicting this outcome (see Table 3). Results supported Hypothesis 2 for females as husband sensitivity uniquely and positively predicted wife love, such that greater husband sensitivity was associated with wives reporting more love. Next, for wife conflict, the chi-square value of the constrained model was not significant, indicating wife conflict was conjointly influenced by sensitivity of the actor and the partner (see Table 3), suggesting only partial support for our hypothesis. Results supported our prediction that the partner effect of greater husbands’ sensitive support provision would be predictive of greater love for wives, but we found no such association for wives’ conflict. Rather, results indicated that it was the conjoint effect of both providing and receiving sensitive support that predicted less conflict for wives.

For husband love, the chi-square value of the constrained model was significant, suggesting that actor and partner effects were significantly different from one another in predicting this outcome (see Table 3). Specifically, husband sensitivity was uniquely and positively linked to husband love, whereby greater sensitive support provision was associated with husbands reporting more love. Finally, for husband conflict, the chi-square value of the constrained model was significant, indicating a significant difference in actor and partner effects on husband conflict (see Table 3). The fully saturated model revealed husband sensitivity was uniquely and negatively related to husband conflict, such that less husband sensitivity was associated with greater self-reported conflict for husbands. Failing to support our Hypothesis 2 prediction of conjoint effects for husbands, we instead found that only husbands’ sensitive support provision was linked to husbands reporting more love and less conflict.

Discussion

Studies have shown that support from others is critical to mental and physical health (Sarason and Sarason 2009), and that the spouse is the most important support provider for married individuals (Dakof and Taylor 1990). Although spousal support is known to be important, few researchers have simultaneously assessed the acts of both providing and receiving support in couples, or considered their joint effect on marital outcomes. Moreover, little is known concerning how gender is related to the relationship between support provision/receipt and positive marital outcomes. To gain a more complete understanding of optimal support in marriage and its effects by gender, the current study utilized both observational and self-reports of support and APIM to simultaneously examine both actor and partner effects of support sensitivity on support satisfaction, love, and conflict in a sample of married U.S. couples.

Results revealed intriguing gender differences. Husband sensitivity predicted both husbands’ and wives support satisfaction. For wives, husbands’ sensitivity also predicted their love and the conjoint effects of husband and wife sensitivity predicted their conflict. These findings are consistent with Acitelli and Antonucci’s (1994) work on the benefits of receiving support from an intimate partner. For husbands, on the other hand, their own sensitivity while providing support uniquely predicted their love and conflict, which is in line with previous work on how support provision affects individuals’ moods (Gleason et al. 2003; Williamson and Clark 1989). Thus, not only do these findings contribute to and extend the scant literature implicating the importance of support provision for well-being (e.g., Brown et al. 2005; Knoll et al. 2007), they underscore how critical it is to both examine spousal support from a dyadic perspective and to consider the role that gender plays in the links between support and marital outcomes.

Husbands as Support Providers: The Critical Piece

Due to the fact that wives have been found to desire significantly greater levels of support than husbands (Xu and Burleson 2001), we anticipated and found that the reception of sensitive support would be most critical for wives’ marital outcomes, as wives felt more satisfied with their support and happier with their marriages when their husbands provided sensitive support. Given the dyadic nature of support interactions, the continual shifting between the roles of provider and recipient, and the fact that receiving and providing support are often highly correlated (Knoll et al. 2007), we expected the conjoint effect of both providing and receiving sensitive support would be critical for husbands’ outcomes. However, the findings revealed the independent actor effects for husbands and partner effects for wives mostly trumped the importance of the conjoint effects. These findings bring with them unanswered questions. Why was husband support provision so vital for both spouses’ marital outcomes? Why was sensitive support from wives not as important as it appeared to be from husbands?

One explanation for why husbands and wives would benefit differentially from support provision versus receipt draws upon spouses’ differential motivation within their intimate relationships. Women are more likely to seek nurturing and understanding in relationships and men are more likely to seek to take control (Mickelson et al. 1995; Tannen 1990). Therefore, wives may benefit when their husbands provide them sensitive support as it may serve as an indicator that they have a partner who is capable of collaborating with them to solve problems. Husbands, however, may benefit from providing support as it validates their role as “expert” and simultaneously serves as evidence of their agency in the relationship (Knoll et al. 2007). Thus, through support provision, men may obtain an outlet through which to make a meaningful but socially acceptable contribution to their wives and thus their marriages. In comparison, men who provide minimal and/or insensitive support may not benefit as much because they feel ineffective in their role as problem-solver. These feelings of uselessness might create a more negative mood (Gleason et al. 2003), which in turn could contribute to greater conflict with their spouse.

As to why spouses would not benefit as much when the roles are reversed (wives provided and husbands received), perhaps because wives are taught and reinforced by society to nurture their partners (Joseph and Bhatti 2005; Tannen 1990), doing so does not provide unique benefits. We noted that wives had a higher mean score on observed sensitivity as well as a smaller range in observed sensitivity scores, which suggests wives were generally unlikely to be insensitive. Wives’ restricted variability in sensitive support provision suggests that providing support is both expected and reliably demonstrated in women and may be unlikely to contribute to variation in marital outcomes for either spouse. Samter (2002) conclusions about gender differences in cognitive complexity may help to explain these findings. According to Samter, because men are socialized in a way that values their ability to lead and entertain others, whereas women are taught to express and manage emotions in themselves and others, men may have more difficulty than women discerning others’ emotions, motivations, and perspectives. Thus, for those men who are able to adequately provide sensitive support, a greater effect is likely felt at both the individual and couple level. Thus, the differential benefits for husbands and wives of providing versus receiving support may be explained by men’s and women’s socialization.

A related explanation for why husbands and wives may have benefitted more from providing and receiving support, respectively, is that it may reflect their different preferences for intimate communication (e.g., supportive interactions). Hook et al. (2003) stated that men prefer intimate interactions that involve “doing” rather than “being” (p. 465), as men are expected to behave assertively and assume positions of expertise, preserving their independence in both actions and words. Consequently, men may prefer playing the role of provider and benefit more from this role due to societal expectations. Samter (2002) added that both men and women regard requests for emotional help to be inconsistent with the traditional male gender role, and men’s support seeking behaviors are more likely to elicit responses from both genders that minimize the problem or change the topic. Conversely, Hook and colleagues argued that women generally prefer intimate interactions in which they share their emotions. Thus, husbands who provide more sensitive support likely offer wives a more conducive forum in which to express themselves, and thus greater husband support sensitivity is related to wives’ marital functioning. Future studies may want to consider whether differential gender preferences with regards to intimate communication may offer further insight into the links between support and marriage.

Strengths and Limitations

Our confidence in the findings of this study is strengthened by its methodological strengths and design. First, our multi-method approach was critical. Including observations of support allowed us to examine both support provision and receipt in addition to spouses’ reports of their support satisfaction, love, and conflict. Had we not examined both provision and receipt, we would have failed to capture the importance of husbands’ sensitive support provision for both spouses’ support satisfaction and marital outcomes. Additionally, our utilization of APIM allowed us to appropriately examine our dyadic data and simultaneously consider both actor and partner effects, as well as consider the conjoint effects of both provision and receipt.

However, certain limitations suggest these results should be interpreted with caution. First, the cross sectional design of the study precludes us from commenting on the direction of effects for our findings. Although our direction of analysis was theoretically driven by Rini and Dunkel Schetter (2010) discussion of sensitivity as an independent variable, we recognize that spouses’ marital love and conflict may also be predictive of their sensitive support provision. For example, husbands who report greater love for their spouses may be more inclined to provide sensitive support to their wives. Next, the smaller sample was comprised mostly of well-educated, White, middle-class couples in the United States. Future research should examine how support is related to the marital relationship in more diverse samples. For example, researchers have found African American families tend to be more matriarchal (Jarrett 1994), perhaps suggesting that husbands’ support provision may not be as critical for them as wives may have more influence in the relationship. Readers should be also aware that findings from this U.S. sample may not generalize to married couples worldwide. To note, because participants were required to self-identify as happily married in order to participate in the current study, the potential for bias in self-reported support receipt exists. It is reasonable to assume that respondents may have felt pressure to report that they were, indeed, happily married. Therefore, the self-selection of participants in our sample results in the need for caution when interpreting results. Future work should additionally consider examining the timing of support provision as some have claimed it is critical for understanding spousal support (Neff and Karney 2005). It may be that husbands do benefit from wives’ sensitive support provision, when provided at critical times. Finally, though we attempted to minimize any lingering effects from the conflict task (including a compassionate love task and break before the support task), we may have captured some effects of prior conflict on support. As most support situations do occur during stressful times (Cutrona 1996), we believe our findings are still valid.

In spite of these limitations, the results of the current study shed important light upon the way spouses support one another and outcomes related to that support. We found husbands’ ability to provide sensitive support was critical not only to their wives’ marital outcomes, but to their own as well. In light of these findings, researchers should attempt to uncover both the factors related to husbands’ ability to provide such support and why husbands appear to derive greater benefit than their wives from sensitive support provision. In conclusion, these findings contribute to the limited literature implicating the importance of support provision in marriage, and suggest that simultaneous consideration of husbands’ and wives’ support receipt and provision is critical when explicating support satisfaction, as gender appears to act as a powerful filter through which support is experienced.