Can Expressing Positivity Elicit Support for Negative Events? A Process Model and Review

Model of Positive Expressivity’s Effects on Support for Negative Experiences: Overview

We propose a theoretical model (see Figure 1) in which seekers’ expressions of positivity when seeking support for negative events exert causal effects on providers’ support behavior. We define support as the amount of support (i.e., emotional support, instrumental/practical support, helping behaviors, responsiveness, or caregiving) that providers extend or intend to extend. Although receiving higher levels of support does not always produce favorable outcomes (e.g., Bolger & Amarel, 2007; Bolger et al., 2000; Gleason et al., 2003; Howland & Simpson, 2010), accumulating longitudinal and experimental evidence suggests that receiving support is beneficial when people want, need, or try to elicit support from a partner (Bar-Kalifa & Rafaeli, 2013, 2015; Crockett et al., 2017; Deelstra et al., 2003; Feinstein, 1988, as cited in Dunkel-Schetter et al., 1992; Girme et al., 2013; Merluzzi et al., 2016; Siewert et al., 2011). For example, findings from daily diary studies indicate that receiving less support than one wants or needs is far more costly to an individual’s well-being than is receiving too much support (Bar-Kalifa & Rafaeli, 2013, 2015; Siewert et al., 2011). Seekers who receive more (vs. less) responsive support (according to coders) when discussing their personal problems experience more favorable changes immediately after the support interaction (e.g., increased positive affect, felt acceptance, and engagement in positive reframing; and decreased negative affect) and over time (e.g., increased engagement in constructive coping and self-efficacy), in part because of their accurate detection of the partner’s responsive support (Lemay & Neal, 2014). Thus, the decision to examine support quantity as an outcome serves our aim of evaluating the proposed model’s principal tenet—that expressing positivity when seeking support for negative events or experiences shapes support from potential providers.

OPEN IN VIEWER

Figure 1.

Theoretical process model linking a support seeker’s positive expressivity when seeking support for negative personal events to a provider’s support.

However, operationalizing support provision in terms of support quantity may restrict conclusions about the sensitivity or helpfulness of the support that positive expressivity might elicit. The quality of support received is likely to be consequential because support offers benefits primarily when it is responsive to the recipient’s needs (Maisel & Gable, 2009). As such, after considering evidence relevant to support quantity, we also consider evidence that bears on support quality.

Our model includes three pathways through which expressing positivity should increase support (i.e., support-eliciting pathways): by strengthening providers’ prorelational motives, increasing their positive mood states, and enhancing their expected support attempt effectiveness. It also includes one pathway through which expressing positivity may decrease support (i.e., a support-suppressing pathway): by decreasing providers’ appraisals of the seeker’s need for support. Each of the three support-eliciting pathways (Figure 1, Paths A–B, C–D, and E–F) operates by addressing one of the challenges providers face when providing support for negative events. First, strengthening the providers’ prorelational motives (i.e., their desire to bond with seekers) should push providers to weigh the costs that support provision has on to their self-regulatory resources less heavily than (a) their desire to be a supportive partner or (b) the costs of not being a supportive partner (e.g., impeding intimacy; Gable et al., 2012; Path A–B). Second, increasing providers’ positive mood states helps combat their unpleasant experiences of listening to negative disclosures (e.g., negative mood states; distress/discomfort; Path C–D). Third, increasing providers’ expectations of the effectiveness of their support attempts helps alleviate providers’ uncertainty that their support attempts will be successful (Path E–F). In contrast, the support-suppressing pathway operates by dampening the seeker’s distress signaled through negative expressivity, such that expressions of positivity undermine providers’ appraisals of the seeker’s need for support (Path G–H). In the following sections, we elaborate on and describe evidence for each path in the model.

Strengthening Provider’s Prorelational Motives

The first proposed support-eliciting pathway operates by strengthening the provider’s prorelational motives (Path A), which in turn increases support (Path B).

Promoting Positive Mood States in Providers

The second support-eliciting pathway involves a process whereby seekers’ expressions of positivity increase providers’ positive mood (Figure 1, Path C), which in turn enhances providers’ supportive behavior (Figure 1, Path D).

Increasing Provider’s Expected Support Attempt Effectiveness

The third support-eliciting pathway in our model predicts that seekers’ positive expressivity increases providers’ expectations of their support attempt’s effectiveness (Figure 1, Path E), which in turn increase support provision (Figure 1, Path F).

Undermining Provider Need Appraisals

The final indirect pathway in our model is a support-suppressing pathway. This pathway holds that positive expressivity may undermine providers’ appraisals of the seeker’s distress or need for support (Figure 1, Path G), which in turn decreases provided support (Path H).

Summary and Evaluation of Evidence for Indirect Paths

Overall, the proposed model is well supported by existing research, with varying degrees of evidence for different indirect pathways. A wealth of correlational and experimental work supports Path A (positive expressivity to prorelational motives) and Path B (prorelational motives to support provision). Importantly, two studies also provide correlational evidence that prorelational motives mediate the direct path from positive expressivity to support provision (provider-reported and coder-rated) among romantic partners in daily life and in the lab (Gordon et al., 2012). Thus, the evidence for the prorelational motives pathway appears to be relatively strong, but experimental work that tests the mediational pathways is still needed to confirm causality.

In terms of the paths linking positive expressivity to support provision through positive mood states, some work provides evidence for Path C (positive expressivity to positive mood states) between strangers and between close relationship partners. Strong empirical evidence exists for Path D (positive mood states to support provision), and one experiment found evidence of mediation when the dependent variable was providers’ support intentions (Telle & Pfister, 2012, Study 2). However, given that no evidence of mediation was found when examining positive expressivity’s effects on supportive behavior—and that positive mood states are known to increase favorable self-evaluations (Forgas et al., 1984)—it seems possible that these findings reflect that positive expressivity ultimately translates to providers’ more benevolent self-perceptions. Until future work rules out this possibility, the evidence for this indirect path is tentative.

Evidence for Path E (positive expressivity to expected support effectiveness) is predominantly correlational, but some experimental data provide suggestive evidence for this path. Given that most of the experimental evidence involves dependent variables related to observers’ judgments of the seeker’s competency, future work that investigates the effects of positive expressivity on observers’ perceptions of their own competence in providing effective support is necessary to build strong evidence for this link in the model. More research is also needed to bolster confidence in the proposed path from expected support effectiveness to support provision (Path F), given that most of the evidence for this path stems from theory, a few correlational studies, and indirect manipulations of expected support effectiveness (e.g., manipulations of the seeker’s ostensible coping efforts). Furthermore, we did not uncover any studies that tested the full indirect path from positive expressivity to support provision via expected support attempt effectiveness.

The evidence for Path G (positive expressivity to need appraisals) is limited to correlational findings and to research suggesting that positive expressivity may work against negative expressivity’s signal for help. Experimental work that directly tests the effect of seekers’ expressions of positivity when seeking support for negative events on providers’ need appraisals is needed to corroborate Path G. In contrast, strong evidence—comprising both correlational and experimental findings—exists for Path H (need appraisals to support provision). Little work has tested the indirect path from positive expressivity to support provision via need appraisals. Experimental work is needed to substantiate this indirect pathway.

In sum, considerable evidence gathered from the literatures on support, emotion, helping, and close relationships supports the proposed model. In our view, the evidence is strongest for Paths A, B, C, D, and G, and more limited for Paths E, F, and H. Having described work that bears on the indirect pathways in our model, next, we review evidence that bears on the direct path—that is, on the hypothesized link between expressing positivity and receiving support for negative experiences (Figure 1, Path I).

Evidence of a Support-Eliciting Effect

Evidence of a Support-Suppressing Effect

Although there is substantial evidence that supports our proposition that expressing positivity can elicit support, a handful of studies examining relevant variables have found evidence consistent with a support-suppressing effect of positivity. Some of this evidence comes from investigations we described when reviewing evidence of a support-eliciting effect (i.e., investigations that provide evidence of both support-eliciting and support-suppressing effects of positivity). A few studies we have not yet reviewed provide additional evidence of a support-suppressing effect.

Studies Showing No Relation Between Positivity and Support

Although limited in number, a few null effects also exist in the published literature. Some null effects emerged in studies we have already described because they also demonstrated some support-eliciting or support-suppressing effects. Others emerged in studies we have not discussed up to this point.

Studies That Bear on Support Quality

In the sections above, we have focused on evidence linking seekers’ positive expressivity to the amount of support that seekers receive, or that others provide (or intend to provide) them. However, researchers have pointed out the importance of considering support quality (e.g., Feeney & Collins, 2015). Responsiveness—the degree to which a response demonstrates caring, understanding, and validation (Reis et al., 2004)—might be the gold standard construct to capture quality because of the aforementioned link between responsive support seeker outcomes (Maisel & Gable, 2009). Unfortunately, none of the studies that we reviewed examined whether expressions of positivity elicit support that seekers perceive as responsive. Nonetheless, in an attempt to glean some insight into the effects of positive expressivity on support quality, we examined three subsets of studies from the larger set of evidence reviewed here that speak to positive expressivity’s links with the quality of support received.

Because high-quality, responsive support is often defined as support that meets the seeker’s needs (e.g., Cutrona & Russell, 1990; Cutrona & Suhr, 1992), we first considered studies that assessed support that met specific and clear needs the recipient had in a support-seeking/giving situation (e.g., providing requested feedback on a cover letter; picking up dropped papers). All the studies that we located that met this criterion provided at least some evidence consistent with positive expressivity enhancing support quality. Among the correlational evidence, Ratcliff et al. (2012) found that participants who saw immigrants in a protest march (Studies 1 and 2) or lesbian, gay, bisexual, and transgender (LGBT) members in an LGBT community (Study 3) expressing more (vs. less) authentic pride reported stronger intentions to support immigrants (e.g., by working to reduce the inequalities they face) or the LGBT community (e.g., by putting a progay sticker on their cars). Perceiving the expression of hubristic pride, in contrast, was negatively associated with support intentions (Ratcliff et al., 2012, Study 1). Genevsky and Knutson’s (2015) work on charitable giving showed when people displayed more (vs. less) happiness in their photos that accompanied online requests for financial help, they received money more quickly and were more likely to receive the amount of money they had requested, but positive text within the request predicted neither loan success nor loan rate.

The experimental research from the subset of studies that bear on support quality showed that (a) an ostensible job applicant’s expression of gratitude for participants’ previous feedback on their cover letter increased the likelihood of participants’ providing additional feedback on the applicant’s cover letter or the amount of time participants worked on the cover letter in a second round of feedback (Grant & Gino, 2010, Studies 1 and 4); (b) a fundraising director’s expression of gratitude for fundraisers’ work increased the number of calls fundraisers subsequently made to potential donors (Grant & Gino, 2010, Study 3); (c) a confederate’s warm (vs. neutral) tone of voice when interacting with strangers increased the likelihood that the strangers would pick up the confederate’s dropped papers (Goldman & Fordyce, 1983); and (d) participants who imagined a hypothetical scenario in which a help seeker (stranger or acquaintance) displayed happiness (vs. neutral affect) were more likely to report that they would attempt to address the seeker’s need (e.g., by driving the seeker home, if she or he needed a ride home; by lending the seeker a cell phone, if she or he needed to make a call; Telle & Pfister, 2012, Study 2), but happiness (vs. sadness) decreased the amount of money participants (providers) allocated to a hypothetical person in an imagined negative event scenario (Telle & Pfister, 2012, Study 1). Overall, these studies suggest that expressions of positivity may elicit supportive action that meets needs that seekers or recipients appear to have, based on the support context. However, it was not always clear in these studies whether recipients actually wanted such help. For example, someone who drops papers likely needs to retrieve those papers but may prefer to do so on their own.

In a second effort to identify evidence that was suggestive of support quality, we considered studies that assessed support recipients’ perceptions of the support they received—namely, felt support or perceived helpfulness. All studies we located that met this criterion were correlational studies, which revealed positive associations between positivity and recipient’s reports of support quality and/or showed that positive expressions positively predicted support, which in turn was associated with the recipient’s/seeker’s perception of helpful support. Liu et al. (2017) found that when seekers included both positive and negative expressions in their posts to an online support group, they were more likely to receive emotionally supportive responses from other group members, which in turn predicted an increased likelihood that the seeker subsequently thanked the provider by name. Assuming expressed gratitude reflected actual grateful feelings, this finding suggests seekers perceived the emotional support they received to be a helpful benefit (Algoe, 2012; Algoe et al., 2008). Miczo and Burgoon (2008) found that seekers’ positive expressivity (e.g., pleasant affective tone of voice; interpersonal warmth toward the provider) predicted both their support-providing romantic partners’ expressed admiration of the seeker’s coping ability—which in turn was positively associated with the seeker’s feelings that the provider understood them—and seekers’ perceptions of the provider’s helpfulness. Two other correlational studies found that seekers who engaged in more (vs. less) “positive support-seeking” behaviors (as rated by coders) received more support from their partners (according to coders) and also felt more supported/perceived their partner as more helpful (Don & Hammond, 2017; Overall et al., 2010).

Our final consideration of evidence bearing on support quality was guided by Feeney and Collins’s (2015) assertation that high-quality support should confer benefits to seekers. We examined a subset of studies that reported associations between positive expressivity and support, and also provided insight into whether the provided support was associated with some positive outcome for the seeker. All studies that we located that met this criterion demonstrated positive associations between positive expressivity and a construct (e.g., a certain type of support or provider’s psychological experience) that in turn predicted desirable seeker outcomes. Results from these studies revealed that (a) seekers’ displays of appreciation for providers (as reported by providers) were positively associated with providers’ appreciative feelings, which in turn positively predicted providers’ responsive behavior (as rated by coders), and when providers were more (vs. less) responsive, seekers felt more appreciated (Gordon et al., 2012, Study 3); (b) seekers’ “positive support-seeking” was positively associated with providers’ informational support provision, which was, in turn, positively associated with seekers’ positive affect postdiscussion (when controlling for support-seeking behaviors and preinteraction positive affect; Don & Hammond, 2017); and (c) seekers’ positive expressivity was positively associated with their support-providing friend’s expressions of esteem for them, which was, in turn, positively associated with the seekers’ satisfaction with their discussion (Agne & White, 2004).

Thus, the evidence that exists concerning support quality suggests that expressing positivity when seeking support for negative events elicits high-quality support.

Summary and Evaluation of Evidence Regarding the Direct Path (Figure 1, Path I)

The strong majority of studies reviewed above provides converging evidence that positive expressivity can elicit more and (in the investigations that examined quality) better support in negative contexts. Yet, some studies revealed that positive expressivity suppressed support or showed no relation between positive expressivity and support. Before addressing the implications of these findings, we first evaluate the strength of the evidence for support-eliciting effects of positivity, then evidence for support-suppressing effects, and finally we consider studies showing null effects.

In evaluating the strength of the evidence for the direct support-eliciting effect in our model, it is notable that relevant empirical studies employed a diverse array of designs and methods. These included observational lab studies, daily diary studies, experiments, and the use of machine learning techniques. Furthermore, the work that we have reviewed looked across a variety of relationship types, spanning from strangers to friends and romantic partners, and involved a wide variety of negative experiences for which people sought support (e.g., financial need, chronic pain, cancer). Support-eliciting evidence emerged in studies including close partners who were talking about ongoing stressors and more distant others/strangers who had more mundane help needs. Several types of verbal and nonverbal expressions of positivity are represented in the evidence (e.g., happiness, affection, gratitude), and—looking across the correlational work in particular—these types of positive expressivity were assessed by different sources (i.e., seekers’, providers’, and coders’ ratings). In addition, this evidence includes various indices of support, such as provider- and coder-rated reports of support and behavioral measures of support or helping. The diverse methods, measures, and contexts employed in this set of studies give us confidence that a direct support-eliciting effect of positivity can—and often does—emerge (Figure 1, Path I). Many more studies examined positivity’s effects on support quantity than on support quality. However, the studies that did permit us to investigate positivity’s effects on support quality by and large suggested that positivity increased support quality.

Overall, we find the range of studies that provided evidence relevant to our predicted direct effect impressive, especially given that the potential support-eliciting function (or impact more generally) of positive expressivity in negative contexts has not been focal in the literature and that studies from which we reviewed evidence often were conducted to test other hypotheses. Nevertheless, there are some important limitations of the evidence that exist to date regarding the direct support-eliciting effect in our model. Although the evidence presented here does span a variety of relationship types, the bulk of the evidence involves nonclose relationship contexts. This is especially true of the experimental evidence bearing on support quality. This is unfortunate, given that people typically seek and receive support from close relationship partners (American Psychological Association, 2018; Rimé, 2009). Indeed, close relationship theorists (e.g., Barbee, 1990; Clark, 2002) have speculated that processes related to negative event support and helping may unfold differently between close relationship partners and strangers. Relatedly, most experimental evidence bearing on support quality involves participants providing help to individuals who faced a negative event and/or had an unmet need but were not likely highly distressed. This experimental evidence nonetheless complements the correlational evidence linking the expression of positivity when disclosing about more upsetting negative events to high-quality support. Moreover, the experimental evidence relies heavily on static and/or hypothetical representations of emotional expressivity. Yet, outside of the lab, social interactions typically involve dynamic exchanges between partners (van Kleef, 2009). Thus, it is unclear how well the evidence that comes from studies involving nonclose relationship partners, hypothetical scenario paradigms, or static representations of positive expressivity would generalize to support transactions in the real world. This gap in the literature will be important for future research to fill.

Although fewer studies provided evidence of support-suppressing (vs. both support-eliciting and null) effects, the studies that showed support-suppressing effects shared some notable features. In particular, they employed paradigms in which people disclosed about a real, meaningful, and highly emotional negative life event: bereavement. Some studies showing a support-eliciting effect for positivity also involved people talking about a real, ongoing, and personally meaningful problems or stressors with a close relationship partner (Agne & White, 2004; Don & Hammond, 2017; Miczo & Burgoon, 2008; Overall et al., 2010). However, these studies often—though not always—involved stressors less severe than bereavement (e.g., “everyday problems or troubles”; Agne & White, 2004, p. 5). ⁴ It seems plausible that positivity’s effects may vary across stressor features such as severity, although additional work is needed to test this possibility. We return to the importance of stressor severity in our discussion of moderators.

Only a few studies showed no relation between positivity and support. This could be in part because null effects may not always be published (although as mentioned earlier, most of the studies we reviewed were conducted to test other hypotheses, so their publication likely hinged more on the findings related to the authors’ main hypotheses). Furthermore, we are cautious in interpreting the null effects that were reported; they may be the product of low power or methodological features. It is noteworthy, though, that most of the studies in which null effects were observed employed similar paradigms involving interactions between strangers and used self-reported support intentions as the only index of support. In a few cases, null effects emerged in one analysis or using one measure, but significant support-eliciting or support-suppressing effects emerged in other analyses or using other measures within the same study. The variability of the findings across (and sometimes within) investigations points to an important avenue for future research: to untangle critical moderators that determine when and for whom positivity has support-eliciting effects, support-suppressing effects, or no effects on support.

The work reviewed here provides an exciting jumping-off point for future research that bolsters and extends our understanding of positive expressivity’s potential value in support-seeking contexts. We see two avenues as especially critical for further investigation. First, work must address the effects of different kinds of positivity on support provision. Second, research must examine moderators of positivity’s effects on support. We now consider each of these topics in turn.

Families of Positivity: Direct Effects on Support

Of the positivity families we proposed, the empirical work included in our review most commonly focused on (in no particular order) partner-oriented positivity (gratitude, appreciation, positive regard), self-oriented positivity (pride, positive evaluations of one’s own coping efforts, and self-compassion), or stressor-oriented positivity (optimism, humor, and positive attitudes about the negative event). Investigations less commonly captured incidental positivity.

Nearly all the studies included in our review of the direct path that involved partner-oriented positivity revealed that partner-oriented positivity predicted more support (for an exception, see Gordon et al., 2012, Study 3, wherein partner-oriented positivity did not predict support directly, but did so indirectly and in a way consistent with a support-eliciting effect). For example, in Grant and Gino’s (2010) investigation of the effects of gratitude, seekers who expressed (vs. did not express) gratitude elicited more tangible support.

Similarly, except for one study showing hubristic pride was negatively related to support (Ratcliff et al., 2012), all the studies that involved self-oriented positivity revealed that self-oriented positivity predicted more support. In one such study, seekers who expressed positive evaluations of their handling of a personal problem received more help from their romantic partner (provider), according to coders (Miczo & Burgoon, 2008).

In contrast, evidence for the link between stressor-oriented positivity and support was inconsistent: Whereas humor and positive attitudes about the negative events consistently predicted more support intentions (e.g., Whitsett et al., 2010), optimism was either negatively related (e.g., Vollmann et al., 2007) or unrelated (e.g., Vollmann & Renner, 2010) to support.

Fewer studies examined incidental positivity (pleasant and warm tones of voice, general happiness), but in the handful of studies that did examine this family, incidental positivity uniformly predicted increased support. Telle and Pfister (2012, Study 2), for example, showed a hypothetical seeker’s happy (vs. neutral) facial expressions increased providers’ intentions to provide tangible support.

It is also worth noting that several studies treated positivity as a unitary construct or did not specify the type of positivity expressed. For example, in studies conducted by Capps and Bonanno (2000) and Wang et al. (2015), positivity was operationalized as positive emotion or positive thought. In Keltner and Bonanno’s (1997) work, expressions of positivity were not contextualized, such that the referent of positivity was unclear: Silent videos displayed laughter as people described their experiences with bereavement, but the absence of verbal content obscures what people were laughing about. Interestingly, there was about as much research focused on positivity with an ambiguous referent as research focused on partner-, self-, or stressor-oriented positivity. Furthermore, among the studies that examined positivity with an ambiguous referent, evidence for the link between positivity and support was mixed: The unidentifiable/unclassifiable positivity in these studies sometimes predicted increased support (Liu et al., 2017; Monin et al., 2009; Wang et al., 2015), sometimes predicted decreased support (Capps & Bonanno, 2000; Keltner & Bonanno, 1997; Telle & Pfister, 2012, Study 1), and sometimes was not related to support (Capps & Bonanno, 2000). These observations are notable for two reasons. First, they point to the need for a taxonomy of positive expressions (for a similar call for an agreed-upon taxonomy of positive emotion more generally, see Keltner, 2019). Second, they may suggest that the referent of positivity is a meaningful way of grouping specific types of positivity into families that have differential social consequences when expressed within distress-related contexts.

Seeker Features

Positivity should be particularly likely to elicit support when expressed by seekers who chronically express negativity (Forest et al., 2014). In addition to negatively affecting providers’ moods (e.g., Neumann & Strack, 2000), listening to disclosures from someone who chronically expresses negativity undermines provider efficacy beliefs and prorelational sentiments: Partners of people who chronically (vs. seldom) express negativity report believing that their support efforts are less likely to be effective and feeling resentful that their partner takes more than they give in the relationship (Forest et al., 2014).

In addition, positivity should be particularly useful at eliciting support for seekers with low (vs. high) self-esteem. One reason is that people with low self-esteem (LSEs) express more negativity than their high self-esteem counterparts (HSEs; see Wood & Forest, 2016), so self-esteem may function similarly to the chronic levels of negative expressivity just described. Moreover, LSEs tend to feel comfortable with (Wood et al., 2009) and do little to repair their negative moods (Heimpel et al., 2002; Wood et al., 2009). This may lead providers to doubt their ability to help LSE seekers. Indeed, providers often feel bad about their interactions and themselves after attempting to support LSEs (Marigold et al., 2014). LSE seekers might especially benefit from expressing positivity that increases providers’ expected support effectiveness because doing so should address the two related reasons their partners report withholding support: they feel like lack the skills necessary to provide support and they think their partner is too difficult to help (Feeney & Collins, 2003). A similar pattern may operate for people experiencing depression, who also express high levels of negativity (Rottenberg & Vaughan, 2008) and whose partners also report that their partner is too difficult to help (Feeney & Collins, 2003).

Provider Features

We expect that positivity is especially likely to elicit support from providers who are higher (vs. lower) in attachment avoidance—people who are uncomfortable with closeness and who strive to maintain autonomy and emotional independence from their partners (Hazan & Shaver, 1994). Avoidantly attached people tend to provide less responsive support than their secure counterparts (Feeney & Collins, 2001; Feeney et al., 2013; Simpson et al., 1992). Low levels of prorelational sentiments seem to underlie this effect: When asked about why they do not provide support, one reason that highly avoidant people cite is their lack of concern for their partner (Feeney & Collins, 2003; Feeney et al., 2013). However, highly avoidant individuals become increasingly willing to engage in intimacy-promoting behaviors with their partners after positive, intimacy-related relationship experiences (Stanton et al., 2017). Seekers’ positive expressivity—especially partner-focused positivity (e.g., gratitude; Y. Park et al., 2019) and incidental positivity, we suspect—may function as a positive relational experience that could make avoidant providers feel comfortable and motivated to provide support. It is less obvious to us that provider’s attachment anxiety would moderate positivity’s effects because people high in attachment anxiety tend to provide high levels of support to their partners (even if it is sometimes intrusive; Feeney & Collins, 2001).

Because providers may feel more competent when they have had experience with a particular stressor (Dunkel-Schetter & Skokan, 1990; Suitor et al., 1995), we also expect that positive expressivity will be especially likely to elicit support from providers who have not had experience with a particular stressor. Positive expressions that heighten provider expectations of effectiveness (e.g., self-oriented positivity, incidental positivity) should be particularly valuable for eliciting support from such providers. Such expressions should also be particularly valuable for eliciting support from providers who doubt their abilities as support providers for other reasons—for example, perhaps because they perceive the seeker as inconsolable (Karimiha et al., 2015) or lack confidence in their abilities as a romantic partner more generally (Weiser & Weigel, 2016), or have a history of failed attempts at support provision from other relationships. Providers with LSE may be one such group who doubt their support effectiveness (possibly with good reason): Others tend not to seek support from partners with LSE because they perceive LSE providers as inefficacious (Cavallo & Hirniak, 2019).

Relationship Features

Positivity may be especially useful as a support elicitor for people in newer (vs. more established) relationships. Newer partners tend to have less confidence in their knowledge of their partners (Swann & Gill, 1997), which may lead them to doubt the effectiveness of their support attempts. Furthermore, enacting prorelational behavior requires more self-control resources among partners in newer (vs. older) relationships (Kammrath et al., 2015). Thus, when faced with the option of devoting their resources to providing support or to furthering other goal pursuits, having enhanced prorelational motives should increase the likelihood that providers prioritize supporting their partner (Mills et al., 2004). We therefore expect that expressing types of positivity that strengthen providers’ expectations of effectiveness and/or their prorelational motives will boost support particularly strongly within newer relationships.

Positivity may also hold special support-eliciting value among people in relationships that are relatively lower (vs. higher) in closeness and satisfaction. People in less close and satisfying relationships tend to provide worse support to one another than do closer, more satisfied partners (e.g., Dunkel-Schetter & Skokan, 1990). Almost by definition, such effects seem likely to emerge because of differences in prorelational sentiments. Expectations of effectiveness may also play a role: Couples who are less close may feel less certain about their ability to understand their partners’ needs and meet them. By expressing positivity that bolsters providers’ prorelational sentiments and efficacy beliefs, seekers should increase the likelihood of eliciting support in the context of less close and satisfying relationships.

Finally, positivity may be less valuable—and in fact, may undermine support—in relationships in which need appraisals or sensitivity to a partner’s needs is low. Low need appraisals may often come about in seeker–provider relationships characterized by low responsibility for meeting each other’s needs (e.g., exchange relationships, weaker vs. stronger communal relationships). People in exchange (vs. communal) relationships (Clark et al., 1989) and people in lower (vs. higher) quality relationships (Chen et al., 2015) tend to be less attentive to one another’s needs. If need appraisals are high, expressing positivity may lower them somewhat, such that providers underestimate but still do recognize the seeker’s need for support, but if need appraisals are low, expressing positivity could lower them so much that it seems as though no support is needed.

Stressor features

The severity of the stressor for which the seeker is soliciting support strikes us as particularly important to consider. Severe stressors such as losing a loved one or experiencing a serious illness (Holmes & Rahe, 1967) should make obvious that a seeker needs support, and we expect that these need appraisals will persist even if the seeker expresses positivity. Severe stressors may also worsen providers’ moods (reducing their ability to help with the potentially unpleasant task of providing support; Cunningham et al., 1990; Manucia et al., 1984) and leave providers doubting their ability to provide effective support (e.g., Kennedy-Moore & Watson, 2001). Positive expressivity should be helpful in boosting providers’ mood, and stressor-focused and incidental positivity, in particular, should boost provider efficacy beliefs. Therefore, we expect that positivity will have its largest support-eliciting effects for stressors that are more versus less severe.

One exception to this may be when the type or intensity of positivity expressed seems inappropriate for a stressor of high severity. In such cases, positivity may undermine support. This may be why Keltner and Bonanno (1997) found that observers were less willing to comfort people who laughed genuinely while talking about their bereavement experiences than people who did not laugh. Given that genuine laughter signals amusement (Keltner & Buswell, 1997), perhaps this finding emerged because observers felt uneasy about people laughing as they spoke about their bereavement experiences, and consequently were less willing to offer them comfort.

The chronicity of the stressor is also likely to matter. When stressors are chronic and seekers continually need support, providers’ supportiveness can deteriorate over time (e.g., Kaniasty & Norris, 1993; Norris & Kaniasty, 1996). Indeed, people experiencing chronic illness or major crises do not always get the support they need (e.g., Bolger et al., 1996; Dunkel-Schetter & Bennett, 1990; Moyer & Salovey, 1999). We suspect that provider efficacy beliefs and prorelational sentiments decrease when stressors are chronic. Thus, positive expressivity should be especially valuable in eliciting support for chronic (vs. acute) stressors.

In addition, the degree to which the stressor is controllable may affect positivity’s utility. Providers tend to feel angrier and less inclined to offer assistance when they perceive seekers’ problems as within their own (i.e., seekers’) control versus outside of seekers’ control (e.g., Dooley, 1995; Weiner, 1980a, 1980b). When seekers express positivity in their support-seeking for controllable stressors, they should boost providers’ mood and prorelational sentiments, thereby increasing providers’ support.

Support Communication Features

Positivity’s effects on support are likely to hinge on features of the support-seeking communication in which seekers engage. In our view, one particularly important variable is the authenticity of seekers’ positive expressions.

There is good reason to expect that inauthentic positive expressions—or expressions that providers perceive to be inauthentic, regardless of the accuracy of such perceptions—could undermine support. People tend to perceive inauthentic expressions as inappropriate (van Kleef, 2016), and they judge individuals expressing fake (vs. genuine and vs. neutral) expressions to be less likable, attractive, and trustworthy (Krumhuber et al., 2007). Expressing positivity in ways that appear insincere or inauthentic may engender unfavorable social evaluations (Algoe & Zhaoyang, 2015; Keltner & Bonanno, 1997), thereby reducing providers’ prorelational motivation. Greenaway and Kalokerinos (2017) aptly articulated, “Individuals should be mindful to express positive emotion authentically if their aim is to reap social rewards” (p. 152). Therefore, if seekers express positivity that they do not genuinely feel—or if providers perceive those expressions to be disingenuous—seekers may not reap the potential benefits of positive expressivity. Indeed, experimental data suggest displaying fake (vs. genuine) happiness may decrease the amount of support people elicit—at least from some providers: When fundraisers (trained actors) expressed fake (vs. genuine) happiness while soliciting support on behalf of a charity organization, people low on dialectical thinking donated less money and were less willing to volunteer their time (Hideg & van Kleef, 2017).

It seems likely that support-suppressing effects of inauthentic positivity may be especially likely to emerge when seekers fake (or appear to be faking) partner-oriented positivity; doubting the authenticity of affection or praise from one’s partner may lead people to feel hurt and relationally insecure (Lemay & Clark, 2008a, 2008b; Lemay & Dudley, 2009), and raise suspicions about the partner’s motives for expressing such positivity (Ackerman et al., 2011; Kaplar & Gordon, 2004). Thus, perceiving a seeker’s partner-oriented positivity as inauthentic may weaken providers’ prorelational motives and consequently decrease their supportive behavior.

Although our focus in this article has been on positivity, the degree of negativity that seekers express in their support-seeking communications is likely to be extremely important. As we mentioned earlier, support-seeking attempts will often involve expression of negative thoughts or feelings, which signal one’s need for help and can elicit support (Fischer & Manstead, 2008; S. M. Graham et al., 2008; Karnaze & Levine, 2018; Nesse & Ellsworth, 2009). Negative expressions nonetheless can create important challenges that providers must overcome when attempting to support upset individuals. If seekers can help providers overcome the challenges that negativity brings, seekers should garner more support. However, we would not advise that seekers attempt to deal with these challenges by inhibiting their negative expressivity. Inhibiting negativity should weaken providers’ perception that the seeker needs support, and therefore restrict seekers’ likelihood of eliciting support. In addition, emotion suppression predicts unfavorable outcomes for the discloser, their partner, and their relationship (Bastian et al., 2012; Butler et al., 2003). Instead, we propose that including positive expressions in their communications might allow seekers to express the negativity they are feeling and still have a high chance of receiving support.

Scholars have theorized that the interplay between positivity and negativity is important for personal and relational well-being (e.g., Filipowicz et al., 2011; Gottman et al., 1998; Gottman & Levenson, 2000). Surprisingly, though, few studies have examined how negative expressivity and positivity might combine to shape support provision.

We expect positive expressivity’s support-eliciting effects to be most pronounced when seekers also express high levels of negativity. At high levels of negativity, the addition of positive expressivity not only seems unlikely to undermine providers’ already heightened need appraisals but also becomes especially valuable in helping providers overcome the escalated challenges of supporting highly distressed seekers. Highly negative disclosures may lead providers to doubt their effectiveness (Gosnell & Gable, 2017; Kennedy-Moore & Watson, 2001), induce negative moods in providers (Neumann & Strack, 2000), and decrease providers’ prorelational sentiments toward the seeker under some conditions (e.g., Forest & Wood, 2012; but see also S. M. Graham et al., 2008).

To be clear, although negative expressivity poses challenges to its recipients, we do not recommend that seekers refrain from expressing negativity to providers. Not only is expressing negativity often an effective way to elicit support (e.g., Hobfoll & Lerman, 1989; Kaniasty & Norris, 1995), but emotion suppression also predicts negative outcomes for the discloser, their partner, or their relationship (Butler et al., 2003). Although suppressing or inhibiting negative expressivity seems unwise, we suspect that combining negativity with positive expressions might effectively elicit support. In future work, researchers should examine the effects of positive expressivity at different levels of negative expressivity. Future work should also consider how expressions of both negativity and positivity shape distress-related support, examining—for example—the optimal combination of negativity and positivity to elicit support and whether the net valence of the support-seeking message affects support. In addition, researchers should test for the possibility of curvilinear effects of positive expressivity when negativity is held constant. Both very high and very low levels of positive expressivity may stifle providers’ support: High levels could lead providers to underestimate the seeker’s need for support, and very low levels may not sufficiently help providers overcome the challenges of supporting individuals in distress.

Cultural Context

The cultural context in which support transactions are occurring has important implications for the ways in which positivity affects support. In particular, cultural values and norms regarding the experience and expression of emotion are likely to govern how positivity is received and support is provided in response. For example, high arousal positive affect (e.g., excitement, elevation) is valued more strongly in Northern American contexts (e.g., among European Americans) than in many East Asian contexts (e.g., among Taiwanese, Koreans; Tsai et al., 2006). Thus, high arousal positive affect may have support-eliciting effects in Northern American contexts but support-suppressing effects in some East Asian contexts. Indeed, some work has shown that online microloan requesters displaying more (vs. less) excitement in their display photos receive loans from proportionately more North American lenders and less from Taiwanese lenders (B. Park et al., 2019). Moreover, positivity’s effects on support may vary across cultural contexts, depending on how normative the expressed type of positivity is. Culturally normative types of positivity—such as socially disengaging positivity (e.g., pride) in the United States (but not in Eastern Asia; Kitayama et al., 2006; Markus & Kitayama, 1991), socially engaging positive emotion (e.g., respect, friendly feelings) in Eastern Asia, and self-compassion in Thailand (but not in Taiwan; Neff et al., 2008)—may have the most potential to elicit support. Within close relationships, support-eliciting effects of gratitude may be weaker among Chinese (vs. European North American) people, in part because thanking close others signals relationship distance and heightens negative feelings (e.g., offense; Zhang et al., 2018).

There also may be cultural differences in the extent to which providers face challenges when seekers come to them for support, which would have implications for positivity’s effects on support. Whereas the desire to maximize positive affect and minimize negative affect is particularly strong in Western cultural contexts (e.g., the United States), the value of positivity (and devaluation of negativity) is less strong within East Asian contexts (e.g., China, Japan; Miyamoto et al., 2017; Sims et al., 2015). Therefore, a boost in positive mood states due to the seeker’s expressions of positivity might be less useful to providers with collectivistic (vs. individualistic) cultural heritages. Countries with cultures that have strong communal norms (e.g., India) often feel morally responsible for helping individuals in need—regardless of their personal feelings about the individual, relationship to the individual, or costs they would incur by extending help (Miller et al., 2017; Miller & Bersoff, 1998). We suspect the utility of expressing positivity due to its ability to strengthen prorelational motives would be constrained in cultures with strong communal norms, and therefore be less likely to affect support.

In developing our model, we based our theorizing on the idea that when people want help coping with personal stressors, they often try to obtain it by disclosing about the stressful event or explicitly requesting advice or comfort (i.e., they seek explicit support). As such, the proposed model of positive expressivity’s effects on support may be specific to cultural contexts in which such explicit support-seeking is normative—specifically, Western cultural contexts. In some cultural contexts, explicit support-seeking is rare. For example, Asian/Latin Americans (vs. European/European Americans) are less likely to seek explicit support, and they accurately expect soliciting this type of support from family/friends will be less helpful (H. S. Kim et al., 2006; Mojaverian & Kim, 2013; Taylor et al., 2007 for a review, see H. S. Kim et al., 2008). Relational concerns explain these cultural differences: Asian/Asian Americans because of the negative relational implications doing so might have (e.g., jeopardizing social harmony, burdening the provider; H. S. Kim et al., 2006; Taylor et al., 2004). Instead, Asians and Asian Americans tend to seek and benefit from receiving implicit support (i.e., coping assistance that does not depend on disclosing the stressor or resulting negative feelings, such as comfort that comes from spending time with close others; Taylor et al., 2007). Future work should focus on positive expressivity within distress-related implicit support-seeking contexts, with particular consideration to the types of positivity that seekers spontaneously express (positivity types within the stressor-oriented positivity family seem unlikely within implicit support-seeking contexts) and their effects on support.

Although the proposed model does not accommodate all the complex relations involving culture, social support transactions, and emotional expressivity, we see the perspective on which our model rests—that seekers can maximize their support outcomes by enacting behaviors that help providers overcome challenges of behaving supportively—as a cross-culturally applicable framework for advancing scholarship in this area. By considering (a) barriers that may prevent people from serving as a source of support and (b) support-seeking behaviors that could reduce such barriers, researchers can generate novel and culturally specific predictions about which seeker behaviors have the potential to elicit support.

Develop and Validate Families of Positive Expressivity

Devising conceptually meaningful and empirically supported positivity type groupings is an essential next step in gaining a fuller understanding of how and when different types of seeker-expressed positivity affect support. Doing so will enable more systematic investigations of the effects of expressed positivity within support-seeking contexts. Such investigations are critical in refining our understanding of how and why expressing different types of positivity affects providers’ psychological experience of support interactions and their resulting (un)supportive behavior.

We proposed four theoretically derived families of positivity (i.e., partner-oriented, self-oriented, stressor-oriented, and incidental positivity) that we think merit further investigation. This raises several questions that future research should examine: Do positive expressions indeed group into the four families we propose? If so, which types of positivity do people naturally express when seeking support for their personal stressors? Which family-specific positive expressions elicit or suppress support, and via which mechanisms? More generally, disentangling type-specific effects of positivity will be a high priority not only in scholarship on support elicitation but also in the broader literature on positive emotion (Keltner, 2019; Sauter, 2010).

Investigate the Downstream Consequences of the Support That Expressed Positivity Elicits

Whereas the current article focuses on whether, why, and under what conditions expressing positivity might affect seekers’ receipt of support, the effects of upset seekers’ expressions of positivity likely extend beyond providers’ supportive behavior in that particular support interaction. For example, expressing positivity in one interaction may elicit support in future interactions (Marigold et al., 2014, Study 6). Furthermore, given the well-documented stress-buffering effects of social support (see Cohen & Pressman, 2004; Thoits, 2011), and the evidence linking received or enacted support with desirable seeker/recipient outcomes in contexts of adversity in which support is sought (e.g., Bar-Kalifa & Rafaeli, 2013; McLaren & High, 2019), we expect that enhancing the quantity and quality of the support that seekers receive will confer personal and relational benefits. However, very few of the studies we reviewed investigated the downstream consequences of the support that seekers’ positivity elicited or suppressed (for exceptions, see the studies described on pp. 44–45, in the third category of studies that bear on support quality), and none tested stress-buffering processes. Future work should examine the effects of positivity-elicited (or suppressed) support on a broad array of seeker outcomes including seekers’ perceived responsiveness of the support, indices of coping, physical and mental health, relationship quality with the provider, and other components of thriving described in Feeney and Collins’s (2015) conceptual framework for thriving through relationships (e.g., indices of eudaimonic well-being such as personal growth and having a sense of purpose in life). This work should test not only for main effects of support on these outcomes, but also for stress-buffering effects of the support.

Specify the Relations Among Mechanistic Variables

The conceptual model put forth here focuses on four parallel pathways through which expressions of positivity might affect support provision. However, it seems possible that the relations among these pathways may be more complex than those described in the proposed model. For example, positivity families that increase providers’ positive mood states may trigger a process through which positive mood states increase providers’ expectations about the effectiveness of their support (Kavanagh & Bower, 1985). Such sequential effects should ultimately increase providers’ support. Researchers should investigate both parallel and serial mediated pathways in studies testing positivity’s indirect effects on support.

Uncover Other Support-Eliciting Seeker Behaviors

The proposed model focuses on the expression of positivity as one type of behavior seekers can enact to elicit support. Moving forward, an exciting direction for support research will be to build on the idea that seekers play an active role in support transactions (Barbee & Cunningham, 1995; Feeney & Collins, 2015; Feeney et al., 2017) by examining what other behaviors seekers can enact to elicit support. Behaviors that affect the mechanisms in our theoretical model should shape providers’ supportive behavior. For example, conveying care for the provider’s well-being (e.g., asking about their day; Cortes & Wood, 2019) or communicating that one sees the self and partner as interconnected (e.g., using “we” language; Fitzsimons & Kay, 2004) may increase the provider’s prorelational sentiments. Behaviors that suggest the task of providing support is manageable (e.g., focusing on one or two ways in which one has been negatively affected by an upsetting event, rather than describing a multitude of issues) may enhance the provider’s expected support effectiveness. Thus, we suspect that enacting these types of behaviors when seeking support for negative experiences is also likely to help seekers garner support.

Understanding the behaviors seekers can enact to increase their receipt of support might offer insight into points of intervention for individuals who struggle to marshal adequate support. Such interventions should not be attempted until research has carefully uncovered relevant behaviors, their mechanisms, moderators, and boundary conditions. Then, perhaps practitioners could teach undersupported individuals to enact such controllable behaviors when seeking support to maximize their support outcomes.