Abstract
Growing evidence indicates that whether critical and hostile behavior harms relationships depends on how partners respond. The current studies test a key behavioral indicator of partners’ responsiveness by examining whether partners’ withdrawal when actors exhibit negative-direct behavior predicts within-person and longitudinal declines in perceived partner responsiveness and relationship satisfaction. Test of Actors’ negative-direct × Partners’ withdrawal interactions indicated that partners’ withdrawal in the context of actors’ negative-direct behavior when targeted for change during conflict discussions (Study 1, N = 162 dyads) and during daily interactions (Study 2, N = 151 dyads) predicted lower perceived partner responsiveness and relationship satisfaction. This Actor × Partner effect did not emerge when actors were pushing for change during conflict (Study 1) and was more consistent predicting perceived partner responsiveness. These results illustrate the importance of Actor × Partner effects and indicate that actors’ own destructive behavior provides an important context to diagnose partners’ responsiveness.
The challenges of interdependence create ample opportunity for critical, hostile, and hurtful behavior to emerge in relationships. The impact of these negative-direct behaviors varies across contexts (McNulty, 2016). Negative-direct behavior damages relationships when it reflects disproportionate responses or resistance to change problems, but improves relationships when it addresses and motivates change in serious problems (Overall & McNulty, 2017). Research establishing these contextual effects has examined the effects of one person’s negative-direct behavior on relationship problems and satisfaction. Yet, important relationship outcomes are likely shaped not only by individuals’ (actor effects) or their partners’ (partner effects) behavior, but also by how individuals’ and partners’ behavior interact (Actor × Partner effects; Kelley & Thibaut, 1978; Rusbult & Van Lange, 2008).
Actor × Partner interactions are rarely formally tested, despite extant models indicating that interactive behavioral effects are important. The influential work examining demand/withdraw patterns suggests that poor relationship outcomes occur when actors’ demands for change are met with partners’ withdrawal (Christensen & Heavey, 1990). Integrating the demand/withdraw pattern with the contextual effects of negative-direct behavior, we propose that partners’ withdrawal in the context of actors’ negative-direct behavior will be particularly diagnostic that partners are unresponsive. In the following sections, we outline the theoretical and empirical basis of this proposed Actor × Partner process, clarify how this process advances extant research, and present two dyadic studies testing whether Actors’ negative-direct behavior × Partners’ withdrawal during couples’ conflict discussions (Study 1) and daily life (Study 2) predict immediate within-person and longitudinal declines in actors’ perceived partner responsiveness and relationship satisfaction.
Why Actor Negative-Direct Behavior × Partner Withdrawal Should Predict Perceived Partner Responsiveness and Relationship Satisfaction
The longitudinal effects of criticism and hostility have been principally examined by assessing negative-direct behavior within couples’ conflict interactions. Many studies indicate that greater negative-direct behavior predicts declines in actors’ relationship satisfaction (see Karney & Bradbury, 1995). However, some studies suggest that negative-direct behavior can sustain actors’ satisfaction (e.g., Cohen & Bradbury, 1997; Gottman & Krokoff, 1989; Karney & Bradbury, 1997; McNulty & Russell, 2010), whereas others find no association between negative-direct behavior and later satisfaction (Lavner et al., 2016).
Consistent with contextual models of communication (Karney & Bradbury, 1995; McNulty, 2016), these inconsistencies arise because the effects of actors’ negative-direct behavior depend on how partners respond. Criticism and hostility directed toward problematic partner behavior can erode actors’ relationship satisfaction when partners do not attempt to correct their behavior, whereas negative-direct behavior can improve actors’ satisfaction when it motivates partners to change problematic behavior, thereby improving relationship quality (e.g., Baker & McNulty, 2015; McNulty & Russell, 2010; Overall et al., 2009). Accordingly, prior findings indicate that actors’ negative-direct behavior will be harmful when partners are not motivated, able, or secure enough, to respond in ways that redress actors’ dissatisfaction (see McNulty, 2016; Overall & McNulty, 2017).
Although partners’ responsiveness is central to these contextual patterns, prior work examining negative-direct behavior has not assessed the behavioral responses of partners within conflict interactions that signal partners are unresponsive, or examined whether perceptions of partners’ responsiveness is an important outcome of these dynamics. We propose that partners’ withdrawal is a central indicator of low partner responsiveness. Withdrawal involves disengaging from couples’ interactions and psychologically or physically distancing from the partner (Christensen & Heavey, 1990; Gottman & Levenson, 1992). Withdrawal has been characterized as the most damaging conflict behavior probably because partner withdrawal invalidates the actor’s experiences and conveys an unwillingness to engage and thus be responsive (Gottman, 1998). Accordingly, we propose that a principal outcome of partners’ withdrawal when actors are exhibiting negative-direct behavior is actors’ lower perceived partner responsiveness.
Perceived partner responsiveness captures how much people feel their partners are caring and understanding (Reis et al., 2004), and is a powerful determinant of relationship satisfaction (Canevello & Crocker, 2010; Lemay & Clark, 2008). Partners are perceived to be more responsive when they accommodate during conflict (Wieselquist et al., 1999), sacrifice when encountering conflicting goals (Visserman et al., 2018), provide support when needed (Maisel & Gable, 2009), and respond positively to self-disclosure (Gable et al., 2006). Perceived partner responsiveness is generated from these types of “diagnostic contexts” because, in these situations, partners need to put aside their own goals, needs, and hurt feelings to respond in supportive or relationship-promoting ways (Reis et al., 2004; Rusbult & Van Lange, 2008). Thus, partners’ behavior in these situations provides diagnostic information about how responsive partners are and will likely be in the future. Similarly, actors’ negative-direct behavior offers a salient situational context for actors to diagnose partners’ responsiveness because partners need to overcome inevitable hurt, anger, and destructive impulses to remain caring and understanding in the face of criticism and hostility (Rusbult et al., 1991). Partners’ withdrawal when actors are behaving negatively, however, should communicate partners’ unwillingness to care for and understand actors, and thus undermine actors’ perceived partner responsiveness and relationship satisfaction.
Testing Actor Negative-Direct Behavior × Partner Withdrawal Effects across Contexts: Conflict Discussions and Daily Life
The proposed costs arising from Actor negative-direct behavior × Partner withdrawal align with the poor outcomes associated with the most studied dyadic communication pattern during conflict: demand/withdraw. Demand/withdraw involves the actor criticizing and pressuring for change, whereas the targeted partner disengages and withdraws (Christensen & Heavey, 1990; Heavey et al., 1993). Demand/withdraw during conflict is theorized to hinder problem solving and reduce intimacy (Baucom et al., 2011; McGinn et al., 2009; Papp et al., 2009; Weger, 2005). Accordingly, demand/withdrawal is associated with lower satisfaction across many cross-sectional (see Burrell et al., 2014; Schrodt et al., 2014) and some longitudinal (e.g., Christensen & Shenk, 1991; Heavey et al., 1993) studies.
Our conceptualization and investigation of Actor negative-direct × Partner withdrawal behavior expand prior examinations of demand/withdraw in several ways. First, demand/withdraw specifically refers to actors exhibiting demanding behavior (often including criticism and hostility) when they are agents pushing for change in their partner, and their targeted partner withdraws. In the context of actors as agents of change and partners as targets of change, partners’ withdrawal reflects an unwillingness to understand the actors’ relationship concerns, which predicts lower actors’ satisfaction (Weger, 2005). This aligns with our framework that partners’ withdrawal in the context of actors’ negative-direct behavior offers actors diagnostic information about partners’ responsiveness although this has not been previously tested. In Study 1, we test whether Actors’ negative-direct behavior as agents of change × Partners’ withdrawal as targets of change predict declines in actors’ perceived partner responsiveness and relationship satisfaction across time.
Second, we propose that partners’ withdrawal when actors behave negatively will be diagnostic of partners’ responsiveness across a range of contexts, not just when actors are agents pushing for change. With regard to conflict discussions, actors’ negative-direct behavior when actors are targets of their partners’ desired change has been shown to be particularly destructive, given negative-direct behavior by targets of change conveys defensiveness, resistance to improve relationships, and a lack of commitment (Overall, 2018; Overall et al., 2009). In the context of actors’ negative-direct behavior as targets of change, partners’ withdrawal will also convey that partners are unwilling to be caring and understanding when actors are behaving particularly destructively. Moreover, given these contexts involve partners as agents wanting change, partners’ withdrawal may be especially indicative of partners’ lack of caring by signaling that partners are giving up on improving the relationship. Thus, we test and expect that Actors’ negative-direct behavior as targets of change × Partners’ withdrawal as agents of change will also predict actors’ lower perceived partner responsiveness and satisfaction across time.
Third, partners’ withdrawal in the context of actors’ negative-direct behavior in more routine daily interactions should also be diagnostic of partners’ responsiveness. In daily life, actors can behave in critical and hurtful ways not only during conflict (Overall et al., 2014), but also when they are stressed (e.g., Timmons et al., 2017), need support (e.g., Overall et al., 2016), or feel lower acceptance or self-esteem (e.g., Murray et al., 2003). Just as in conflict discussions, actors’ daily hurtful, critical, and unpleasant behavior should provide an important context in which partners’ willingness to be caring and understanding is put to the test. Thus, partners’ withdrawal in the context of actors’ daily negative-direct behavior should also diagnose partners’ responsiveness and shape actors’ relationship satisfaction.
Providing suggestive evidence for this pattern in daily life, two studies have shown that demand/withdraw during daily interactions is concurrently associated with more negative affect and lower marital satisfaction (Caughlin & Huston, 2002; Papp et al., 2009). Other studies have also shown that daily negative-direct behavior and withdrawal, such as ignoring or spending less time with a partner, is perceived to have a harmful impact on the relationship (Drigotas et al., 1995; Overall et al., 2010), which is consistent with research examining global self-reports of these behaviors (e.g., Berry & Willingham, 1997; Rusbult et al., 1991). However, no prior research has examined whether partners’ withdrawal in the context of actors’ daily negative-direct behavior predicts lower perceived partner responsiveness, or whether this Actor × Partner behavioral pattern has longitudinal implications. Study 2 provides these novel tests by examining whether Actors’ negative-direct behavior × Partners’ withdrawal during daily interactions is associated with (a) immediate within-person, and (b) longitudinal declines in actors’ perceived partner responsiveness and relationship satisfaction.
Fourth, we extend the approach typically used to assess demand/withdraw patterns and these types of behaviors in daily life by modeling Actor × Partner interaction effects. Prior research assessing demand/withdraw has relied on summed dyadic scores (e.g., Caughlin, 2002; Heavey et al., 1993, 1995) or global dyadic-level ratings (e.g., Christensen & Sullaway, 1984; Donato et al., 2014; Weger, 2005; for exceptions, see Kurdek, 1995 and Ross et al., 2019). Other self-report assessments of withdrawal involve rating how partners have responded when actors have been critical or hurtful (Rusbult et al., 1991), which implies a dyadic interaction pattern. However, failure to isolate Actor × Partner effects means that the relative main versus interactive effects of actor and partner behavior are unclear; any effects could primarily arise from actors’ negative-direct or partners’ withdrawal, rather than the interactive combination of these behaviors. Consistent with prior contextual (interactive) examinations revealing that the effects of negative-direct conflict behavior depend on how partners’ respond (McNulty, 2016), we test the main and interactive effects of actors’ negative-direct behavior and partners’ withdrawal on immediate within-person and longitudinal declines in actors’ perceived partner responsiveness and relationship satisfaction.
Current Research
The prior research reviewed above indicates that (a) partners’ responses to actors’ negative-direct behavior determine actors’ relationship outcomes, (b) demand/withdraw patterns can damage relationships, and (c) partners’ responses in diagnostic contexts shape actors’ perceptions of partners’ responsiveness. We integrate and expand these separate lines of research to propose that partners’ withdrawal in the context of actors’ negative-direct behavior should be particularly diagnostic of partners’ responsiveness. We present two dyadic longitudinal studies testing whether the combination of Actors’ negative-direct behavior × Partners’ withdrawal during couples’ conflict (Study 1) and daily (Study 2) interactions predict immediate within-person (Study 2) and longitudinal (Studies 1 and 2) declines in actors’ perceived partner responsiveness and relationship satisfaction. This investigation expands prior theory and research by (a) considering whether prior contextual effects of negative-direct behavior emerge from partners’ behavior within couples’ interactions, (b) broadening examinations of demand/withdraw by testing a more general Actor × Partner pattern across conflict and daily contexts, and (c) extending the typical focus on relationship satisfaction to include a key outcome—perceived partner responsiveness—that is pivotal to how partners’ behavior in diagnostic contexts shapes relationship outcomes.
Study 1
Study 1 involved video-recording couples discussing two areas of conflict that specified conflict roles: one dyad member as the agent pushing for change and the other as the target being asked to change. Observational coders rated how much each dyad member exhibited negative-direct and withdrawal behavior. Participants reported their perceptions of partners’ responsiveness and relationship satisfaction four times across the following year. We tested the effects of Actors’ negative-direct behavior × Partners’ withdrawal in the context specified by demand-withdraw (actors’ negative-direct behavior as agents of change and partners’ withdrawal as targets) and in the context where negative-direct behavior is particularly damaging (actors’ negative-direct behavior as targets of change and partners’ withdrawal as agents).
Method
Participants
Of the 180 recruited couples, 161 couples provided longitudinal data, resulting in longitudinal analyses based on 322 conflict discussions. This sample is larger than most prior studies using similar methods, but the aims of the study when funded did not focus on Actor × Partner effects, and thus a priori power analyses were not conducted. See online supplemental material (OSM) for power considerations and prior use of this sample. Participants’ age ranged from 18 to 45 years (M = 23.21, SD = 4.27). Couples were mostly married (43%) or cohabiting (32%), with the remainder in serious dating relationships. Mean relationship length was 3.01 years (SD = 2.23). Couples were paid NZ$70 for initial participation and NZ$30 for each longitudinal assessment.
Procedure
During an initial laboratory session, participants completed questionnaire measures, and identified and ranked in importance their three most serious relationship conflicts that involved desiring change in their partner’s thoughts, feelings, or behavior. This procedure was designed to differentiate conflict roles involving one partner (the agent) wanting change in the other (the target). Partners rarely (4.5%) identified the same issues in which case those topics were discarded. Couples had two 7-min discussions about the top-ranked problem identified by the (a) male partner (man agent, woman target), and (b) female partner (woman agent, man target; discussion order counterbalanced). See OSM for more procedural and descriptive details.
Measures
Participants completed scales assessing perceived partner responsiveness and relationship satisfaction at the initial session and then through post every 3 months for the following year. 1 All items were averaged to construct scores (αs > .80; see Table 1).
Descriptive Statistics of All Measures.
Note. All measures represent averages across items on 1 to 7 Likert-type scales. Observed communication represents averages of independent coders’ ratings for each 30-s segment of the conflict discussions, and then these 14 segment scores averaged across the conflict discussion. Daily measures represent averages of daily assessments across the 21-day diary period. Thus, the highest scores represent consistently high levels of each type of behavior across the discussion (Study 1) or across days (Study 2). In Study 1, zero-order correlations between perceived responsiveness and satisfaction ranged from .57 to .70 across the five time points (average r = .64). In Study 2, the correlation between perceived responsiveness and satisfaction was .50 at Time 1 and .52 at Time 2, and ranged from .56 to .74 within each daily assessment (average r = .65).
Perceived partner responsiveness
Participants rated six items developed from Fletcher and colleagues (1999) to assess partner’s warmth and trustworthiness, which are central indicators of responsiveness (e.g., “understanding,” “supportive,” “kind”; 1 = not at all like my partner, 7 = very much like my partner).
Relationship satisfaction
Participants rated four items developed by Rusbult et al. (1998) to assess satisfaction (e.g., “I feel satisfied with our relationship”; 1 = strongly disagree, 7 = strongly agree).
Observational coding
Two to four coders independently rated how much each participant exhibited negative-direct and withdrawal behavior for each 30-s segment of the discussion (1 = low, 7 = high). Negative-direct behavior included derogating, criticizing, or blaming the partner; expressing anger and hostile affect; invalidating the partner; and/or being domineering. Withdrawal included avoiding discussing the problem; ignoring, refusing, or dismissing the partner’s issues or concerns; disengaging and withdrawing from the discussion. Agents and targets were coded in separate viewings (order counterbalanced). Coders’ ratings were averaged for each 30-s segment (intraclass correlation coefficients [ICCs] > .85), and then averaged across the discussion (αs > .80) to create negative-direct behavior and withdrawal scores. See OSM for more details.
Results
Growth curve analyses were conducted using the MIXED procedure in SPSS 21 (see Kenny et al., 2006). The first stage in growth curve analyses involves specifying the trajectory of actors’ perceived partner responsiveness (or relationship satisfaction) across time by modeling the multiple ratings across the year as a function of an intercept and a slope representing time. Time was coded 0 at the initial session through to 4 for the final 12-month follow-up; thus, the intercept represents initial levels of perceived partner responsiveness (or satisfaction) and the slope of time represents whether perceived partner responsiveness (or satisfaction) increased, decreased, or remained stable across the year. The intercept and slope of time were modeled as random and allowed to covary within and across couple members.
The second stage involves assessing whether independent variables—in this case actors’ negative-direct behavior (mean-centered), partners’ withdrawal (mean-centered), and the interaction between actors’ negative-direct behavior and partners’ withdrawal—predict the trajectory of actors’ perceived partner responsiveness (or relationship satisfaction). Given the number of model parameters, we conducted separate analyses for each discussion when individuals were agents versus targets of change (see OSM for syntax). Additional analyses modeling the two discussions simultaneously showed analogous results (see OSM). 2
Table 2 displays the results. The model examining actors’ negative-direct behavior as agents of change and partners’ withdrawal as targets of change is shown in the top of Table 2. Partners’ withdrawal was associated with lower initial actors’ perceived partner responsiveness but did not predict change in actors’ perceived responsiveness or satisfaction across time. The interactions of Actors’ negative-direct behavior as agents × Partners’ withdrawal as targets were not significant.
The Effects of Actor Negative-Direct Behavior and Partner Withdrawal During Couples’ Conflict Discussions on Change in Actors’ Perceived Partner Responsiveness and Relationship Satisfaction Over the Following Year (Study 1).
Note. The longitudinal effects presented in bold are shown in Figure 1. Effect sizes (r) were computed using Rosenthal and Rosnow’s (2007) formula: r = √[t 2/(t 2 + df)]. Degrees of freedom ranged from 108.79 to 315.58. CI = confidence interval.
† p < .06. *p < .05. **p < .01. ***p < .001.
The model examining actors’ negative-direct behavior as targets of change and partners’ withdrawal as agents of change is shown in the bottom of Table 2. Partners’ withdrawal was associated with lower initial actors’ perceived partner responsiveness, and there was more evidence for Actors’ negative-direct × Partners’ withdrawal effects on perceived partner responsiveness (p = .056) and satisfaction (p = .011) across time. Shown in Figure 1, only when actors exhibited greater negative-direct behavior as targets (+1 SD) and partners exhibited greater withdrawal as agents (+1 SD) did actors experience declines in perceived partner responsiveness (Panel A; B = –.10, t = –3.47, p = .001, r = .13) and satisfaction (Panel B; B = –.12, t = –3.20, p = .002, r = .12). In contrast, there were no declines in perceived responsiveness or satisfaction when actors’ negative-direct behavior was high but partners’ withdrawal was low (B = –.02, t = –0.77, p = .444, r = .03; B = .02, t = 0.69, p = .489, r = .03), or when actors’ negative-direct behavior was low and partners’ withdrawal was high (B = .00, t = 0.13, p = .895, r = .00; B = .00, t = 0.07, p = .943, r = .00) or low (B = –.00, t = –1.57, p = .117, r = .06; B = –.06, t = –1.78, p = .076, r = .07, for perceived responsiveness and relationship satisfaction, respectively).
The effects of actors’ negative-direct behavior (as targets of change) and partners’ withdrawal (as agents of change) during couples’ conflict discussions on actors’ perceived partner responsiveness (Panel A) and relationship satisfaction (Panel B) across time (Study 1).
Given existing evidence that perceived responsiveness arising from couple dynamics shapes relationship satisfaction, and our proposition that perceived responsiveness is a central outcome of partners’ withdrawal when actors behave destructively, supplementary analyses examined whether perceived responsiveness might link the Actor × Partner effects to the more global outcome of relationship satisfaction. Although the interaction effect predicting relationship satisfaction appeared larger than perceived responsiveness, the interaction effect needs to be considered in the context of the main effects. Partners’ withdrawal was significantly associated with lower initial levels of perceived partner responsiveness, but not relationship satisfaction, and these initial differences continued (and expanded) across time (see Figure 1). Mediated moderation analyses using the RMediation Package (Tofighi & MacKinnon, 2011) provided supplementary support that high actors’ negative-direct behavior as targets and high partners’ withdrawal as agents may predict reductions in actors’ relationship satisfaction at least partly through reductions in perceived responsiveness (indirect effect = –0.05, 95% confidence interval [CI] = [–0.083, –0.023]). Moreover, there were no indirect effects for the other three dyadic combinations (high actor targets’ negative-direct behavior/low partner agents’ withdrawal = –0.01, 95% CI = [–0.039, 0.017]; low actor targets’ negative-direct behavior/high partner agents’ withdrawal = 0.00, 95% CI is [–0.029, 0.033]; low actor targets’ negative-direct behavior/low partner agents’ withdrawal = –0.02, 95% CI = [–0.050, 0.006]). We consider the consistency of the effects for perceived partner responsiveness and relationship satisfaction across the studies in the discussion.
Alternative Actor × Partner effects
Additional analyses detailed in the OSM ruled out the possibility that the results arose from other dyadic patterns. First, tests of Actors’ withdrawal × Partners’ negative-direct behavior revealed no significant longitudinal interaction effects. Second, tests of Actors’ negative-direct × Partners’ negative-direct behavior (negative reciprocity) revealed that this dyadic combination predicted similar declines in actors’ perceived responsiveness (but not satisfaction). Nonetheless, simultaneously modeling negative reciprocity and Actors’ negative-direct behavior × Partners’ withdrawal revealed that the longitudinal effects of Actors’ negative-direct behavior as targets × Partners’ withdrawal as agents shown in Figure 1 remained. Finally, tests of Actors’ withdrawal × Partners’ withdrawal (mutual withdrawal) revealed no significant longitudinal effects. Thus, the effects in Figure 1 were not due to alternative dyadic patterns.
Study 2
Study 2 extended the examination of Actors’ negative-direct behavior × Partners’ withdrawal beyond conflict discussions by testing whether this dyadic pattern during daily interactions predicted lower actors’ perceived partner responsiveness and relationship satisfaction. To maximize power, we combined two samples that followed the same procedures. Each day for 3 weeks, both couple members rated the level of their negative-direct and withdrawal behavior, perceived partner responsiveness, and relationship satisfaction. Participants completed questionnaire assessments of perceived partner responsiveness and relationship satisfaction prior to completing the daily sampling procedure and then 9 months later. We tested whether Actors’ negative-direct behavior × Partners’ withdrawal during daily interactions predicted both (a) immediate within-person, and (b) longitudinal declines in actors’ perceived responsiveness and satisfaction.
Method
Participants
Study 2 used integrative data analysis to pool two samples that involved very similar methods and measures (Curran & Hussong, 2009). Sample 1 involved 78 heterosexual couples involved in serious relationships (43.6% married/cohabitating) that averaged 2.57 years in length (SD = 1.96), and whose average age was 22.44 years (SD = 4.81). Sample 2 involved 73 heterosexual couples involved in serious relationships (47% married/cohabitating) that averaged 3.20 years in length (SD = 3.56) and whose average age was 23.61 years (SD = 6.87). Participants completed a 3-week daily sampling procedure and were asked to complete follow-up measures 9 months later. Due to couples failing to respond, opting out of participation, or dissolving, the longitudinal data consisted of 50 couples from Sample 1 and 57 couples from Sample 2. Thus, daily analyses were conducted on a sample of 151 couples and longitudinal analyses were conducted on a sample of 107 couples. We focused on integrative data analyses, given that this approach maximizes statistical power and produces more reliable, stable estimation of effects (Hussong et al., 2013). The OSM presents analyses testing sample differences. Two of the within-person daily effects were stronger in Sample 1, but the effects were significant in both samples (see OSM). No sample differences emerged for longitudinal outcomes (see OSM). See OSM for power considerations and prior use of these samples.
Materials and procedure
During an initial session, couples completed scales assessing perceived partner responsiveness and relationship satisfaction and were given instructions for completing the daily sampling procedure. Nine months later, participants completed online assessments of perceived partner responsiveness and relationship satisfaction. All scale items were averaged to construct overall scores (α > .77; see Table 1).
Perceived partner responsiveness
In both samples, participants rated the same items as in Study 1 (e.g., “understanding,” “supportive,” “kind”) according to how closely their partner matched expectations (1 = does not match at all, 7 = very completely matches my ideal). These samples included this measure because research has shown that evaluations of these key indicators of partners’ responsiveness involve assessing whether partners meet expectations (Fletcher et al., 1999; Overall et al., 2006; see Fletcher et al., 2020).
Relationship satisfaction
In Sample 1, participants completed the five-item satisfaction scale used in Study 1 (Rusbult et al., 1998). In Sample 2, participants completed a seven-item scale assessing relationship quality (Fletcher et al., 2000), including satisfaction, intimacy, trust, and love (e.g., “How satisfied are you with your relationship?” 1 = not at all, 7 = extremely). The two measures did not include identical items, but are similar in format and length and are widely used to assess relationship evaluations. Moreover, the effects did not vary across samples (see OSM), suggesting that no differences arose due to the different measures. Thus, we applied integrative data analysis to examine the effects on relationship satisfaction across time (Hussong et al., 2013; Overall, 2020).
Daily assessments
Participants completed a web-based record for 21 consecutive days. On average, participants completed 19.3 (Sample 1) and 19.1 (Sample 2) records. All items were averaged to construct daily scores.
Daily behavior
Participants rated the degree to which they enacted negative-direct behavior (“I acted in a way that could be hurtful to my partner” “I was critical or unpleasant toward my partner” r = .88, R C = .76, R 1F = .69) and withdrawal (“I wanted to be left alone and/or spend less time with my partner” “I withdrew from my partner and did my own thing” r = .79, R C = .65, R 1F = .67; 1 = not at all, 7 = very much). These items have been used in prior studies to assess these behaviors (e.g., Overall, 2020; Overall & Sibley, 2010).
Perceived partner responsiveness
Participants rated two items assessing partners’ acceptance, caring, understanding, and validation, which constitute core components of perceived responsiveness (Reis et al., 2004). Items were very similar across Sample 1 (“My partner felt acceptance and understanding toward me” “My partner felt value and respect for me” r = .95, R C = .80, R 1F = .80) and Sample 2 (“My partner accepted and loved me” “My partner valued and respected me” r = .85, R C = .63, R 1F = .63; 1 = not at all, 7 = very much).
Relationship satisfaction was assessed with one item in Sample 1 (“Today, I was satisfied with our relationship”) and Sample 2 (“How satisfied were you with your relationship today?”; 1 = not at all, 7 = very much).
Results
Daily analyses
Following procedures outlined by Kenny et al. (2006) to analyze repeated measures dyadic data, we modeled the degree to which (a) actors’ negative-direct behavior, (b) partners’ withdrawal, and the (c) interaction between actors’ negative-direct behavior and partners’ withdrawal predicted actors’ perceived partner responsiveness (or relationship satisfaction) on the same day, controlling for actors’ perceived partner responsiveness (or satisfaction) the prior day to ensure that any effects that emerged did not stem from lingering effects of the previous day (Bolger & Laurenceau, 2013). Predictor variables were person-centered. To ensure the effects assessed daily variations in actors’ and partners’ behavior, we also modeled the between-person effects of actors’ negative-direct behavior and partners’ withdrawal (Bolger & Laurenceau, 2013; see OSM for syntax).
Focusing on the within-person effects in Table 3, significant main effects of both actors’ negative-direct behavior and partners’ withdrawal emerged, but these were qualified by significant Actor × Partner interactions (shown in bold). Shown by the dashed lines in Figure 2, days of high actors’ negative-direct behavior and high partners’ withdrawal were associated with lower actors’ perceived responsiveness (left side of Figure 2; b = –.36, t = –26.52, p < .001, r = .36) and relationship satisfaction (right side of Figure 2; b = –.41, t = –25.73, p < .001, r = .35) compared with days of low actors’ negative-direct behavior and high partners’ withdrawal. Shown by the solid lines, high actors’ negative-direct behavior and low partners’ withdrawal was also associated with lower actors’ perceived responsiveness and satisfaction (b = –.22, t = –11.93, p < .001, r = .17 and b = –.26, t = 12.26, p < .001, r = .18), compared with low actors’ negative-direct behavior and low partners’ withdrawal, but these dips were not as strong as days involving high actors’ negative-direct behavior and high partners’ withdrawal. Supplementary tests of indirect effects (see Study 1) supported that the negative within-person effects of high actor negative-direct behavior and high partner withdrawal on actors’ relationship satisfaction through perceived partner responsiveness was stronger (indirect effect = –0.19, 95% CI = [–0.206, –0.170]) than high actor negative-direct behavior and low partner withdrawal (indirect effect = –0.11, 95% CI = [–0.132, –0.094]).
The Effects of Daily Actor Negative-Direct Behavior and Partner Withdrawal on Actors’ Daily Perceived Partner Responsiveness and Relationship Satisfaction (Study 2).
Note. The significant interaction effects presented in bold are presented in Figure 2. Effect sizes (r) were computed using Rosenthal and Rosnow’s (2007) formula: r = √[t 2/(t 2 + df)]. Degrees of freedom ranged from 127.85 to 4823.24. CI = confidence interval.
***p < .001.
The effects of actors’ negative-direct behavior and partners’ withdrawal on actors’ daily perceived partner responsiveness (left panel) and daily relationship satisfaction (right panel; Study 2).
The within-person analyses supported that the lowest levels of perceived partner responsiveness and relationship satisfaction arose on days when high levels of partners’ withdrawal occurred in the context of high levels of actors’ negative-direct behavior. At the between-person level, however, the Actors’ negative-direct × Partners’ withdrawal interaction was not significant (see Table 3). Thus, average levels of this dyadic combination across the 3-week period did not predict daily perceived partner responsiveness and relationship satisfaction once the immediate within-person declines arising from dynamics that day were accounted for. This may indicate that between-couple differences have a smaller impact on situational or in-the-moment judgments of perceived responsiveness (i.e., people’s assessments on a particular day). These daily behaviors may nonetheless culminate to shape broader assessments of partners’ general responsiveness across time, which we tested next.
Longitudinal analyses
We first examined whether Actors’ negative-direct behavior × Partners’ withdrawal averaged across the 21 daily samples predicted longitudinal changes in actors’ perceived partner responsiveness. Applying multilevel modeling for distinguishable dyads (Kenny et al., 2006), we estimated the effects of actors’ negative-direct behavior, partners’ withdrawal, and the Actors’ negative-direct behavior × Partners’ withdrawal interaction on actors’ perceived responsiveness 9 months later, controlling for actors’ perceived responsiveness assessed at the initial session (see OSM for syntax). We centered perceived responsiveness at 9 months on initial levels so that the predicted values inform whether perceived responsiveness increased, reduced, or remained the same across time. Analogous analyses tested the effects on actors’ relationship satisfaction.
Shown in Table 4, Actors’ negative-direct behavior × Partners’ withdrawal predicted actors’ perceived responsiveness across time. Shown in Figure 3, actors’ high versus low negative-direct behavior was associated with lower actors’ perceived responsiveness when partners’ withdrawal was high (dashed line b = –.41, t = –4.41, p < .001, r = .29), but not when partners’ withdrawal was low (solid line b = –.04, t = –.46, p = .649, r = .03). Thus, as in the daily analyses, it was the particular combination of partners’ withdrawal in the context of actors’ negative-direct behavior that predicted the lowest levels of perceived partner responsiveness. There were no significant main or interaction effects predicting relationship satisfaction across time. Despite the nonsignificant interaction on relationship satisfaction, supplementary tests of indirect effects supported that high actors’ negative-direct behavior and high partners’ withdrawal was associated with reductions in actors’ relationship satisfaction through reductions in perceived responsiveness (indirect effect = –0.12, 95% CI = [–0.215, –0.052]), but this was not the case for high actors’ negative-direct behavior and low partners’ withdrawal (indirect effect = –0.01, 95% CI = [–0.067, 0.040]).
The Effects of Average Levels of Daily Actor Negative-Direct Behavior and Partner Withdrawal on Actors’ Perceived Partner Responsiveness and Relationship Satisfaction 9 Months Later.
Note. The interaction effects presented in bold are presented in Figure 3. Effect sizes (r) were computed using Rosenthal and Rosnow’s (2007) formula: r = √[t 2/(t 2 + df)]. Degrees of freedom ranged from 105.25 to 207.86. CI = confidence interval.
*p < .05. **p < .01. ***p < .001.
The effects of actors’ negative-direct behavior and partners’ withdrawal across daily life on actors’ perceived partner responsiveness 9 months later (Study 2).
Alternative Actor × Partner effects
Additional daily analyses revealed that the dyadic patterns of (a) Actors’ withdrawal × Partners’ negative-direct behavior, (b) Actors’ negative-direct × Partners’ negative-direct behavior, and (c) Actors’ withdrawal × Partners’ withdrawal had similar effects to Actors’ negative-direct behavior × Partners’ withdrawal. These results indicate that the combination of any type of actor and partner “destructive” behavior could undermine actors’ daily relationship outcomes. To assess the unique effects of each pattern, we ran additional analyses, simultaneously modeling Actors’ negative-direct behavior × Partners’ withdrawal with each of the alternative dyadic patterns. The effects of Actors’ negative-direct behavior × Partners’ withdrawal remained significant for both perceived partner responsiveness and relationship satisfaction, whereas the alternative Actor × Partner interaction effects on relationship satisfaction were no longer significant. Furthermore, no longitudinal effects of Actors’ withdrawal × Partners’ negative-direct behavior, negative reciprocity, or mutual withdrawal emerged. Thus, the effects in Tables 3 and 4, and Figures 2 and 3, were not due to alternative dyadic patterns (see OSM for details).
General Discussion
The current studies provide support for a key premise derived from interdependence theory (Kelley & Thibaut, 1978; Rusbult & Van Lange, 2008): Actors’ own behavior likely represents an important context for understanding the meaning and implications of partners’ behavior. The results across studies provide initial evidence that greater actors’ negative-direct behavior, in combination with greater partners’ withdrawal when targeted for change during conflict (Study 1) and during daily interactions (Study 2), was associated with lower perceived partner responsiveness and relationship satisfaction. This Actor × Partner pattern is consistent with influential models of communication behavior and relationship processes that (a) highlight that the outcomes of negative-direct behavior depend on how partners respond (McNulty, 2016), (b) illustrate the importance of dyadic behavioral patterns during conflict (Christensen & Heavey, 1990), and (c) emphasize the pivotal role of perceived partner responsiveness (Reis & Clark, 2013; Reis et al., 2004). The results expand these distinct lines of work by showing that dyadic patterns involving Actors’ negative-direct behavior × Partners’ withdrawal (a) contribute to when negative-direct behavior will be harmful for relationships, (b) are important in both conflict and daily contexts, and (c) shape evaluations of partners’ responsiveness. In the following, we consider the consistency of effects across studies and discuss how the current studies advance understanding and generate directions for future research.
Actor × Partner Interactions Inform Contextual Models of Relationship Behavior
The current studies support our proposition that actors’ behavior reflects a key contextual feature that shapes the meaning and impact of partners’ behavior. It is increasingly clear that ignoring contextual factors can produce null and inconsistent effects, thereby impeding understanding of how behavior affects relationships (see McNulty, 2016). The extant evidence demonstrates that actors’ negative-direct behavior damages relationships when it is disproportionate, reflects resistance to problems, or directed toward partners who are incapable of being responsive (e.g., Baker & McNulty, 2015; McNulty & Russell, 2010; Overall, 2018; Overall et al., 2009). In contrast, actors’ negative-direct behavior can sustain satisfaction when targeted at serious problems which partners are open and motivated to change (Overall & McNulty, 2017). Although prior research demonstrating these contextual effects has focused on situational factors (e.g., problem severity, conflict role) or partners’ dispositions (e.g., attachment insecurity, depressive symptoms), a central element connecting these factors involves whether partners are responsive to the needs of the situation. The current studies more directly test this central element by examining immediate behavioral dynamics that should be diagnostic of partners’ responsiveness—partners’ withdrawal.
Providing evidence that partners’ withdrawal is a key indicator of partners’ relative responsiveness, greater partners’ withdrawal was concurrently associated with lower perceived partner responsiveness across studies. However, illustrating the importance of examining how partners’ behaviors combine, neither actors’ negative-direct behavior nor partners’ withdrawal independently predicted longitudinal changes in relationship evaluations, despite both types of behavior—particularly partner withdrawal—commonly identified as important risk factors for relationship decline (Gottman, 1998; Karney & Bradbury, 1995). Instead, partners’ withdrawal only had detrimental effects across time when actors were exhibiting negative-direct behavior, and thus when partners’ responsiveness was put to the test. These Actor × Partner effects indicate that many contextual effects focused on actors’ negative-direct behavior shown in prior research may involve partners’ behavioral responses within the situations in which actors’ negative-direct behavior arises.
Considering actors’ behavior as a context to understand the meaning and impact of partners’ behavior is foundational to interdependence-based models (Kelley & Thibaut, 1978; Rusbult et al., 1991) and the perceived partner responsiveness framework (Reis & Clark, 2013; Reis et al., 2004). Yet, most prior research adopting these perspectives considers the situation as providing the opportunity to diagnose partners’ responsiveness, such as whether partners behave constructively during conflict (e.g., Wieselquist et al., 1999), provide support when needed (e.g., Maisel & Gable, 2009), or reciprocate and capitalize on personal disclosures (e.g., Gable et al., 2006; Laurenceau et al., 1998). Yet, within all of these situations, actors’ behavior also provides context to determine whether partners’ behavior is responsive, including actors’ own conflict behavior (shown in the current studies), the type of support actors’ seek/need (Girme et al., 2013), or the level and type of actors’ disclosure (Gable et al., 2006; Laurenceau et al., 1998). Even models that identify actors’ behavior as central to understanding partners’ accommodative behavior have not measured the Actor × Partner effects that directly examine the implied dyadic process (Rusbult et al., 1991). Our application of Actor × Partner effects provides preliminary evidence that actors’ negative-direct behavior provides the opportunity for partners to reveal their relative responsiveness.
Beyond Demand/Withdraw: Actors’ Negative-Direct Behavior × Partners’ Withdrawal Across Conflict and Daily Contexts
An influential body of work assessing demand/withdraw during conflict highlights the importance of the dyadic patterns we examined. However, prior work testing demand/withdraw has (a) tended to focus on a specific context involving actors demanding change and targeted partners withdrawing during conflict, (b) rarely modeled Actor × Partner effects to test whether the effects of demand/withdraw arise from the interaction between actors’ and partners’ behavior, and (c) primarily examined relationship satisfaction as an outcome. By testing the effects of Actors’ negative-direct behavior × Partners’ withdrawal across conflict and daily contexts, and examining perceived responsiveness as a key additional outcome, the current studies identify a broader behavioral dynamic that may be particularly informative in evaluating relationships when partners’ responsiveness is put to the test.
Study 1 tested the longitudinal effects of Actors’ negative-direct behavior × Partners’ withdrawal observed during conflict discussions in two conflict contexts: actors as agents wanting change in targeted partners (the context specified by demand/withdraw patterns) and actors as targets responding to partners’ desires for change (the context in which negative-direct behavior has been shown to be particularly harmful). The dyadic interaction involving high levels of actors’ negative-direct behavior as targets of change and partners’ withdrawal as agents was associated with longitudinal declines in perceived partner responsiveness and relationship satisfaction. Actors’ negative-direct behavior as agents and partners’ withdrawal as targets did not have these longitudinal effects, despite this context aligning with the conceptualization of demand/withdraw during conflict.
The inconsistent effects across conflict contexts could be due to limited power to detect Actor × Partner interactions. However, based on the existing literature, we also think there are important theoretical reasons for the distinctions across conflict contexts. Indeed, inconsistent effects in testing demand/withdraw patterns using less analytically complex tests are not unusual. Despite demand/withdraw being robustly associated with lower concurrent relationship satisfaction (see Burrell et al., 2014; Schrodt et al., 2014), some studies have found that demand/withdraw predicts declines in satisfaction (Gottman & Levenson, 2000; Heavey et al., 1993, 1995; Kurdek, 1995), whereas others have found null or even reverse longitudinal effects (e.g., Caughlin, 2002; Donato et al., 2014; Heavey et al., 1993, 1995) These prior inconsistent longitudinal effects of demand/withdraw, and the null effects of this combination in Study 1 (Actors’ negative-direct behavior as agents of change × Partners’ withdrawal as targets), could arise because negative-direct behavior when agents are trying to change serious problems can express investment and motivate problem improvement (McNulty & Russell, 2010; Overall, 2018; Overall et al., 2009; also see Heavey et al., 1993). These potential benefits of negative-direct behavior, when enacted by agents of change, might have counterbalanced the detrimental effects of targeted partners’ withdrawal, leading to null effects.
Compared with negative-direct behavior by agents, negative-direct behavior by targets of change have been shown to be more damaging because targets’ negative-direct behavior conveys a lack of commitment and resistance to improve problems (Overall, 2018; Overall et al., 2009, 2011). This prior research showing that targets’ compared with agents’ negative-direct behavior is more predictive of poorer relationship outcomes indicates that negative-direct behavior when targeted for change may create a salient context for actors to diagnose partners’ responsiveness. In this conflict context, partners likely have to resist stronger destructive urges to be responsive (Rusbult & Van Lange, 2008), and thus partners’ withdrawal as agents of change may clearly signal that partners are unwilling to be caring and understanding, especially when partners’ withdrawal also indicates resignation of problems that they themselves wanted changed. Hence, the poor outcomes associated with the particular combination of actors’ negative-direct behavior as targets and partners’ withdrawal as agents suggest that partners’ withdrawal offers important diagnostic information about partners’ responsiveness when actors behave particularly destructively.
The results across different conflict roles, and our explanation of the meaning of the behavior within those different roles, align with an established line of work providing evidence for the different meaning and implications of agent versus target behavior (see McNulty, 2016; Overall, 2018). Moreover, although Actors’ negative-direct behavior × Partners’ withdrawal predicted declines in relationship evaluations in only one of the two conflict contexts tested in Study 1, the results of Study 2 illustrated that Actors’ negative-direct behavior × Partners’ withdrawal within daily interactions predicted within-person reductions in daily perceived partner responsiveness and satisfaction, as well as declines in perceived partner responsiveness across time. The immediate daily effects are consistent with the negative evaluations of partners’ withdrawal in prior daily interaction studies (Drigotas et al., 1995; Overall et al., 2010), but provide the first demonstration of Actor × Partner behavioral patterns during daily life. By showing that Actors’ negative-direct behavior × Partners’ withdrawal during daily interactions predict immediate and longitudinal relationship outcomes, the results offer novel evidence for a broader understanding of dyadic patterns beyond conflict contexts. In sum, three of four sets of Actor × Partner tests indicate that partners’ withdrawal when actors are behaving in destructive ways during conflict and daily life predicts more negative partner evaluations that undermine relationship satisfaction.
Finally, additional analyses illustrated that the effects of Actors’ negative-direct behavior × Partners’ withdrawal were independent of other possible Actor × Partner combinations, including Actors’ withdrawal × Partners’ negative-direct behavior, negative reciprocity (Actors’ × Partners’ negative-direct behavior), and mutual withdrawal (Actors’ × Partners’ withdrawal). There was some evidence that negative reciprocity during conflict and daily life also predicted declines in perceived partner responsiveness. However, consistent with prior self-report studies showing that the effects of demand/withdraw on relationship satisfaction are independent of negative reciprocity (Caughlin & Huston, 2002; Kurdek, 1995), Actors’ negative-direct behavior × Partners’ withdrawal had independent and more robust effects. These additional analyses provide further evidence that partner’s withdrawal may be particularly diagnostic of partners’ responsiveness when actors behave destructively.
The Role of Perceived Partner Responsiveness
Extending the typical focus on relationship satisfaction as an outcome, we tested perceived partner responsiveness as a key additional outcome because (a) prior contextual effects of negative-direct behavior are linked to partners’ motivation and capacity to be responsive, (b) perceived partner responsiveness is a central framework explaining how behavioral dynamics shape relationships (Reis & Clark, 2013; Reis et al., 2004), and (c) failing to examine theoretically relevant outcomes contribute to inconsistent longitudinal effects, given more global satisfaction outcomes arise from diverse processes (see Overall & McNulty, 2017). Consistent with the theoretical relevance of partners’ withdrawal for diagnosing partners’ responsiveness, the effects were more consistent for perceived partner responsiveness than relationship satisfaction across studies, including more consistent immediate effects (four vs. two of four tests) and more consistent Actors’ negative-direct behavior × Partners’ withdrawal interaction effects (three vs. two of four tests). Supplementary indirect effect tests supported that perceived partner responsiveness is an important outcome that may shape relationship satisfaction, but applying indirect tests to correlational data do not provide strong evidence that perceived responsiveness is a causal or explanatory mechanism. Rather, our interpretation of the potential role of perceived partner responsiveness is based on extant theory and research supporting that perceived partner responsiveness is a key explanatory factor of the way in which many behavioral and affective dynamics shape relationship satisfaction (e.g., Canevello & Crocker, 2010; Lemay & Clark, 2008; Reis & Clark, 2013; Reis et al., 2004). The current results indicate that connecting the perceived partner responsiveness framework with models of communication can enhance understanding of the way both partners’ behaviors may combine to influence relationships.
Caveats and Directions for Future Research
The current results highlight the theoretical importance of testing Actor × Partner interactions, but these tests require large samples to obtain adequate statistical power and require assessing behaviors within couples’ actual interactions—time-consuming, expensive methods that constrain sample size. Limited power may have contributed to the inconsistent effects across contexts and constrained our ability to test additional moderating factors (also see Note 2). Indeed, the results in the current studies may vary according to additional contextual factors. Ross et al. (2019) found that demand/withdraw led to lower relationship quality for couples with high socioeconomic status, but sustained relationship quality among couples experiencing socioeconomic disadvantage. Ross et al. (2019) reasoned that partners’ withdrawal is costly when it reflects an unwillingness to address concerns despite having the social and economic capital to do so, but may detract attention from social and economic problems that couples cannot control or solve. Thus, when major intractable problems or vulnerabilities produce particularly aggressive and harmful conflict, partners’ withdrawal may convey restraint from reciprocal hostility and protection of the partner or relationship rather than a lack of responsiveness. Future investigations will need to adopt strategies, such as the data integration approach in Study 2, to generate the sample sizes needed to reliably detect Actor × Partner effects and examine how additional contextual features alter the meaning and thus the longitudinal impact of these dyadic patterns.
The strengths of examining naturally occurring behavior during couples’ actual interactions are accompanied by the limitations of correlational data, which prevent causal conclusions regarding explanatory pathways (as discussed above) as well as the direction of behaviors across partners. Prior research indicates that behaviors are likely to reciprocally cycle across dyad members: actors’ demand predicts greater subsequent partners’ withdrawal and vice versa (Baucom et al., 2010, 2011, 2015). Thus, partners’ withdrawal could occur in response to actors’ negative-direct behavior and lead to declines in perceived responsiveness as we theorized, but actors’ negative-direct behavior could also reflect greater dissatisfaction, resistance, and anger in response to partners’ withdrawal. This latter possibility suggests a different interpretation of the results. Rather than partners’ withdrawal when actors behave destructively providing diagnostic information about partners’ responsiveness, perhaps partners’ withdrawal is only linked to perceived partner responsiveness when actors are angered by and are critical of their partners’ behavior as indexed by actors’ greater negative-direct behavior. Yet, if this latter dynamic was responsible for the effects, then Actors’ negative-direct behavior × Partners’ negative-direct behavior (negative reciprocity) should be just as predictive of declines in actors’ perceived responsiveness, given that actors’ negative-direct behavior would similarly indicate that actors are angered by and are critical of their partners’ behavior. However, Actors’ negative-direct × Partners’ withdrawal produced more consistent and robust effects than Actors’ negative-direct × Partners’ negative-direct behavior, which suggests it is unlikely that the Actors’ negative-direct × Partners’ withdrawal effects only occurred because actors were responding more negatively to their partners’ behavior. Future research could integrate the strengths of longitudinal and experimental designs by assessing whether interventions designed to reduce partners’ withdrawal in response to actors’ negative-direct behavior help sustain perceived partner responsiveness and relationship satisfaction across time.
Conclusion
Extant theory and research indicate that (a) whether critical and hostile behavior harms relationships depends on how partners respond, (b) dyadic patterns should shape relationship outcomes, and (c) behavioral dynamics are likely to inform perceptions of partners’ responsiveness. The current studies provide evidence that testing Actor × Partner interactions may advance understanding of when and why important behaviors undermine relationships. Dyadic and longitudinal analyses observing behavior across different conflict roles (Study 1) and during daily life (Study 2) indicated that partners’ withdrawal in the context of actors’ negative-direct behavior when targeted for change during conflict discussions (Study 1) and during daily interactions (Study 2) predicted lower perceived partner responsiveness and relationship satisfaction. These results indicate that partners’ withdrawal when individuals behave in destructive ways signals that partners are not responsive when individuals are at their worst and undermines relationship satisfaction.
Supplemental Material
Actor_ND_X_Partner_Withdrawal_Supplementary_Materials_RandR_v1 - Partners’ Withdrawal When Actors Behave Destructively: Implications for Perceptions of Partners’ Responsiveness and Relationship Satisfaction
Actor_ND_X_Partner_Withdrawal_Supplementary_Materials_RandR_v1 for Partners’ Withdrawal When Actors Behave Destructively: Implications for Perceptions of Partners’ Responsiveness and Relationship Satisfaction by Eri Sasaki and Nickola Overall in Personality and Social Psychology Bulletin
Supplemental Material
Supplemental Material, Sasaki_Online_Appendix - Partners’ Withdrawal When Actors Behave Destructively: Implications for Perceptions of Partners’ Responsiveness and Relationship Satisfaction
Supplemental Material, Sasaki_Online_Appendix for Partners’ Withdrawal When Actors Behave Destructively: Implications for Perceptions of Partners’ Responsiveness and Relationship Satisfaction by Eri Sasaki and Nickola Overall in Personality and Social Psychology Bulletin
Footnotes
Authors’ Note
Acknowledgments
We thank Matthew Hammond, Garth Fletcher, Melissa Grouden, Helena Struthers, Rosabel Tan, Kelsey Deane, Yuthika Girme, Desmond Packwood, Briar Douglas, Phoebe Molloy, Shuai Han, Lucy Travaglia, David Pirie, and Jan Trayes for their contribution to data collection and behavioral observation coding.
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research was supported by a Royal Society of New Zealand Marsden Fund grant (UOA0811) and University of Auckland Science Faculty Research Development Fund (FRDF) grant (3626244) awarded to Nickola C. Overall, and a Victoria University Wellington grant (200759) awarded to Garth J. O. Fletcher.
Supplemental Material
Supplemental material for this article is available online.
Notes
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
