Humour Use Between Spouses and Positive and Negative Interpersonal Behaviours During Conflict

The present research investigated the relation between the use of positive, negative and instrumental humour in the context of romantic relationships and relational well-being as assessed by positive and negative patterns of conflict resolution behaviour. A sample of 116 heterosexual married couples completed scales of relational humour use as well as conflict resolution behaviour. Behaviour of couples while attempting to resolve a relationship based conflict was also coded by independent raters. Actor-Partner Interdependence Model (APIM) analyses showed patterns of actor and partner effects for each type of humour use. Specifically, positive humour use of both partners predicted more positive conflict resolution, whereas negative humour use of both partners predicted less positive conflict resolution. Additionally, instrumental humour use of both partners seemed to predict greater apathy during conflict resolution. Implications for considering couple humor use, assessed for both partners of the dyad, for understanding relational well-being are discussed.

relationship that has developed over time may be systematically associated with the quality of these dyadic interactions (e.g., the different ways partners may attempt to resolve interpersonal conflict).In other words, are partners that typically use different types of humour when communicating to each other, for example, more or less likely to effectively resolve relationship conflict?In the present research we assessed different types of humour intimates typically use with each other in their relationships, and predicted that these patterns of couple humour use should be associated with quality of conflict resolution (both self-reported and observed) for both partners.

Humour and Relationship Well-Being
Given that humour has been linked with the formation and maintenance of interpersonal relationships in the past (Kuiper & McHale, 2009;Lefcourt, 2001;Li, Griskevicius, Durante, Jonason, Pasisz, & Aumer, 2009;Shiota, Campos, Keltner, & Hertenstein, 2004;Ziv & Gadish, 1989), and that patterns of conflict resolution in romantic relationships are indicators of the current and future quality of the relationship (e.g., Holmes & Murray, 1996), it follows that humour use may play a role in conflict resolution as well.Indeed, humour use has been identified as a means of communicating underlying affection in the face of conflict; it may also be used to defuse tension during conflict situations, or as a means of letting go of the conflict at hand (Kane, Suls, & Tedeschi, 1977;Long & Graesser, 1988).Partners may also use humour to soften the blow of mild criticism, or to approach a sensitive topic that appears threatening to both partners, by teasing each other (Keltner, Capps, Kring, Young, & Heerey, 2001).
Certain types of humour, however, may be conducive to conflict resolution while others may be detrimental.Indeed, dyadic analyses have shown that individuals whose partners use more affiliative (positive) humour while discussing a relationship based conflict reported greater resolution of differences and felt more satisfied in their relationship (Campbell et al., 2008).Past research has also shown that the use of affiliative humour in discussions with one's partner is connected with decreases in negative mood (Howland & Simpson, 2014), and is generally more favorably received (Winterheld et al., 2013).Through such mechanisms, humour may function to stabilize relationships during times of conflict.
However, humour can also be used in a negative fashion; for instance, more aggressive forms of teasing -such as ridicule, intimidation, or manipulation -may be used in an attempt to discredit the partner's arguments or feelings (Kowalski, Howerton, & McKenzie, 2001;Long & Graesser, 1988).Past research has noted that individuals who are exposed to aggressive humour may feel more rejected by their interaction partner, as well as sadder in mood and less willing to continue the interaction (Kuiper, Kazarian, Sine, & Bassil, 2010;Winterheld et al., 2013).This could be especially detrimental in discussions of conflict, as partners who disengage from the interaction prematurely may leave conflicts unresolved.Aggressive humour has also been linked with increased counter-arguing (Bippus, Dunbar, & Liu, 2012), indicating that aggressive humour may be especially inflammatory in conflict scenarios.Finally, humour could be used to dismiss the partner's concerns, or to avoid constructive problemsolving behaviours in the relationship.In dyadic studies, individuals whose partners use aggressive (negative) humour have reported less effective resolution of differences, and also felt less satisfied in their relationship (Campbell et al., 2008); and, similarly, individuals whose partners used more aggressive humour felt increased negative mood after a discussion with them (Howland & Simpson, 2014).

Conflict Resolution Behaviour as a Marker of Relationship Well-Being
A great deal of research has shown that one important indicator of marital quality over time is the manner in which spouses communicate and interact with each other, particularly while discussing areas of disagreement (Gottman, 1994).This is not surprising given that conflict interactions are considered 'diagnostic' situations for relationships (Kelley et al., 2003), where partners' verbal and non-verbal behaviour during these interactions should provide clear signals regarding how partner's perceive each other and the relationship.It is not the existence of conflict in marriage per se that is detrimental to marital satisfaction or stability, but how spouses manage conflict when it occurs (Holmes & Murray, 1996).Conflicts can be settled positively through discussion, but in some cases can result in the escalation of arguing without resolution, or with each spouse ignoring the area of conflict in an attempt to prevent negative marital interactions (e.g., Gottman, 1994).The quality of the marriage suffers when conflicts remain unresolved.The quality of couples' conflict resolution is therefore a good indicator of the well-being of the relationship.
The Relational Humour Inventory (De Koning & Weiss, 2002), which was used in the current study, identifies and measures three main types of humour: positive, negative, and instrumental.Positive humour involves the positive aspects of humour, such as the attractiveness of humour in one's partner, closeness to one's partner, and humour appreciation.For instance, the items "My use of humor has brought me closer to my partner" and "I think that one of the attractive things about my partner is his/her sense of humour" both have high factor loadings (> .8) on the positive humour factor.Negative humour, conversely, involves humour that is used in a manipulative or aggressive manner, with factor loadings of > .75 from items such as "Sometimes I make my partner the butt of a joke" and "My partner uses humour to put me down".Finally, instrumental humour involves the use of humour to avoid tension or conflict.Items such as "Whenever I can, I prefer to use humour to avoid conflict between us" and "My partner can persuade me to do something by making me laugh" load highly (> .7)onto the instrumental humour factor.
Given these descriptions, we expected that positive humour use would likely foster more successful conflict resolution, with partners focusing on closeness and appreciation of each other.Conversely, both negative and instrumental humour use seemed likely to create difficulties in resolving conflicts; particularly, individuals whose partners routinely use these humour styles may feel victimized or manipulated.

The Current Study
The current study sought to test hypotheses regarding the typical use of different types of humour between romantic partners and how these partners resolve relationship based conflict (both self-reported and observed).Whereas prior research has established that the use of positive (i.e., affiliative) or negative (i.e., aggressive) forms of humour during actual couple interactions are associated with positive or negative outcomes respectively (e.g., feeling more, or less, close to one's partner), this body of research has not systematically focused on the actual quality of conflict resolution behaviours displayed by each partner.In the present research we therefore assessed the different types of humour used by partners with each other using the Relational Humour Inventory (De Koning & Weiss, 2002), a well validated scale targeting positive, negative, and instrumental types of humour use in relationships.We then assessed positive and negative forms of conflict resolution behaviours between partners, using both self and observational reports, as indicators of relational well-being.

Method Participants
One hundred and sixteen heterosexual married couples were recruited from the community of London, Ontario to participate in a Married Couples Survey by placing advertisements in various local community newspapers.

Couple Humor Use and Conflict Resolution
The city of London is located in Southwestern Ontario, and is an urban community with population of approximately 348,000.Individuals each received $50.00 as an honorarium for their participation.Couples reported dating an average of 35.19 months (SD = 27.79)prior to marriage, and the average length of marriage was 120.23 months (SD = 127.12),meaning that the average amount of time couples had been together was 153.16 months (SD = 124.58).The average age of participants was 38.56 years for men (SD = 11.22) and 36.7 years for women (SD = 10.71).The majority of couples were Caucasian, with an average household income per year of $55,000 -$65,000 (CDN).Sixty percent of the married couples reported they had children (M = 1.31 children per couple), ranging in number from 1 to 5.
i This data set was used in study 2 of Lackenbauer and Campbell (2012), as well as Campbell, Overall, Rubin, and Lackenbauer (2013), but those investigations did not test the hypotheses that are addressed in the current research.

Procedure
Both members of each married couple were invited to attend a laboratory session in which they separately and privately completed a booklet of questionnaires that included measures of couple humour use, conflict resolution behaviours, and other relationship perceptions.After completing the questionnaires, couples made an appointment to return to the lab in 1-2 weeks to take part in a video-recorded discussion task.Eleven couples opted not to return for this follow-up session.
ii Upon returning to the lab, participants were taken to a room containing a table and two chairs, and outfitted with a video camera housed in a small tinted dome mounted on the ceiling.A researcher provided each married couple with instructions on the discussion task.Specifically, couples were instructed to select a problem area that frequently caused friction in their relationship that they would attempt to resolve in the discussion.Participants were given 5 minutes to privately identify and agree upon a problem area in their relationship.The researcher then told couples that they may or may not be able to resolve the conflict, but to try their best to achieve this goal.Couples were given 12 minutes for their discussion.Following the discussion, spouses were asked to sign a video release form allowing the video-recorded discussion to be used for research purposes.
One couple did not sign the release form, leaving the behaviours of 104 couples available for analysis.

Materials
Relational Humour Inventory -Typical patterns of humour use between partners was assessed with the Relational Humour Inventory (De Koning & Weiss, 2002), a scale that specifically measures dyadic humor use.Three subscales assess different types of own humour use, including positive humour (e.g., "Joking with my partner makes me feel closer to him/her"; α = .82for men, α = .76for women), negative humour (e.g., "Sometimes I make my partner the butt of a joke"; α = .84for men, α = .90for women) and instrumental humour (e.g., "Whenever I can, I prefer to use humour to avoid conflict between us"; α = .80for men, α = .82for women).Items for the entire scale were anchored from 1 (total disagree -does not describe us at all) to 5 (total agree -describes us extremely well), and scores were averaged for each subscale.Higher mean scores indicated greater reported use of each type of humour, respectively.
Kansas Marital Conflict Scale -The Kansas Marital Conflict Scale (Eggeman, Moxley, & Schumm, 1985), based on Gottman's (1979) model of stages of marital interaction and conflict, was used to assess self-reported conflict behaviours using three subscales representing the phases of marital conflict that distinguish distressed Campbell & Moroz 535 reasonably quickly?"Higher average scores indicate that partners report they are able to listen to and understand each other's perspectives (α = .86for men, α = .90for women).The manner in which each person's point of view is discussed during a conflict is measured by the 15-item Arguing subscale.Sample items of the Arguing scale are: "Are you both able to express how the other feels about the issue?" and "Are you able to identify clearly the specific things about which you do agree?"Facial expressions and tone of voice are also assessed in the Arguing scale (e.g., anger, hostility, frustration).Higher scores on the Arguing scale indicate that one is able to express his or her point of view to his or her spouse, as well as identify what the disagreement is about (α = .89for men, α = .89for women).The 11-item Negotiation subscale assesses how couples come to a mutually satisfactory compromise.Participants responded to items on the negotiation scale such as: "Are you both willing to give and take in order to settle the disagreement?"and "Are you able to completely resolve it with some sort of compromise that is OK with both of you?" Higher mean scores on this subscale indicated participants reported better compromise and negotiation with their partners during a conflict (α = .87for men, α = .89for women).

Coding of Interpersonal Behaviour During Conflict Resolution
Five trained raters viewed each video-recorded discussion independently and rated the degree to which each partner engaged in eight interpersonal behaviours, four positive and four negative, using 7-point scales (anchored 1 = not at all, 7 = very much).Interrater reliability was acceptable for each item for both husbands (mean α = .

Results
For descriptive purposes, Table 1 presents the correlations between all of the study variables.In general, positive humour use was significantly positively correlated with ratings of conflict resolution behaviours at all three stages of an argument (Agenda, Arguing, and Negotiation).Negative humour use was significantly negatively correlated with conflict resolution behaviours at all three stages as well.Significant correlations were noted between partners on all scales except for instrumental humour use, indicating that couples tend to use similar types of humour and display similar behaviours when interacting with each other.
In light of this significant correlation, and in accordance with current dyadic research practices, the data analytic approach we adopted for testing our hypotheses was guided by the Actor-Partner Interdependence Model (APIM; Kashy & Kenny, 2000).According to the APIM, when individuals are involved in an interdependent relationship, their outcomes can be associated with both their own (i.e., an actor effect), and partner's (i.e., a partner effect), characteristics and inputs.The inclusion of partner effects allows us to test for the mutual influence that can exist Couple Humor Use and Conflict Resolution between persons in a relationship, and also allows us to statistically control for partner effects when estimating actor effects and vice versa.We tested all of the models reported below using multilevel modeling (MLM; also known as hierarchical linear modeling; Kenny, Kashy, & Bolger, 1998;Raudenbush & Bryk, 2002) and following the suggestions of Campbell and Kashy (2002; see also Kashy, Campbell, & Harris, 2006;Kenny, Kashy, & Cook, 2006) regarding the use of MLM with dyadic data.In the dyadic case, MLM treats the data from each partner as nested scores within a group that has an N of 2. Gender was effect coded (-1 for men, 1 for women), and all continuous predictor variables were centered on the grand mean.
To test hypotheses, we ran models with each of the self-reported and observed conflict resolution behaviours as outcome variables (a total of five models), and with the actor and partner effects of each type of self-reported relational humour style as predictors.Interactions of gender with the actor and partner effects were tested initially but removed from the models given the lack of significant interactions.
The results from these analyses are presented in Table 2. Unique actor and partner effects for each type of couple humour use emerged across the analyses.Specifically, individuals reporting a greater use of positive humour with their spouse also reported more effective problem solving behaviour, and were observed to display more positive interpersonal behaviours while discussing a relationship based conflict with their spouse.Actor effects of negative couple humour use also emerged in each analysis, suggesting that the greater use of negative humour in the relationship was associated with less effective conflict resolution behaviour (both self and observer reported).
A less consistent pattern of actor effects emerged for instrumental humour.When individuals reported using humour as a way to avoid problems in their relationship, they reported less effective conflict resolution with their spouse during the middle and end of arguments, and were observed to behave less negatively, but not more positively, toward their spouse.
Interestingly, a number of unique partner effects also emerged.Individuals reported more effective conflict resolution behaviour at the beginning and (marginally) end of arguments with their spouse, as well as were observed to behave more positively and (marginally) less negative toward their spouses, when their spouses reported using more positive humour in the relationship.Similarly, individuals reported less effective conflict resolution behaviour at all stages of an argument (marginal significance for the beginning of conflicts), when their spouses reported greater use of negative humour in the relationship.Lastly, no partner effects for instrumental humour emerged.

Discussion
This study contributes to the growing literature of humor use and relational well-being using dyadic analyses to examine conflict resolution behaviour among relationship partners.Our results show unique actor and partner effects of the use of different humour styles on both reported and observed problem-solving behaviours.Positive humour was linked with self-reports of more effective problem solving behaviour, as well as more positive interpersonal behaviours; the partners of those who used more positive humour also reported more successful problemsolving at the beginning and end of arguments, and were similarly observed to behave more positively toward their spouse during discussion of conflict.Conversely, those who used more negative humour reported less effective problem-solving behaviour, and were observed to be more negative (and less positive) toward their spouse.Again, distinct partner effects emerged, showing that those whose partners used more negative humour reported less effective conflict resolution at all stages of an argument.
The pattern of effects for instrumental humor use was not as straightforward.For instance, an actor effect of instrumental humour emerged wherein those who used more instrumental humour reported less effective problemsolving during the middle and end of arguments, and were observed to behave less negatively (but not more positively) toward their partner.This pattern of results may imply a relatively high degree of apathy for individuals that use more instrumental humor in their relationships.That is, they may care less about resolving the conflict and more about withdrawing from the conversation.These effects suggest a more nuanced relation between this particular type of humor use and relational well-being compared to those for positive and negative humor use.
The focus on dyadic behavior in this research represents an additional important contribution.Previous studies of humour style have focused primarily on self-rating variables such as mood, feelings of rejection, and emotional distress (Campbell et al., 2008;Howland & Simpson, 2014;Kuiper et al., 2010).Such information is certainly valuable, but the current study contributes uniquely to our knowledge of humour styles by examining how it can Couple Humor Use and Conflict Resolution predict behaviours.This focus on behaviour is also served well by the inclusion of both self-reported and observerrated conflict resolution behaviour measurement; without the observed behaviours, any link between humour use and problem-solving behaviour could potentially be attributed to levels of self-serving bias displayed by participants.
The current study also takes into account the reciprocal nature of interactions between partners, using dyadic analyses to control for similarities between partners and differentiate clearly between actor and partner effects.
Broadly speaking, this research has several important implications.First, it speaks to the importance of mutual influence in dyadic interactions; the partner effects observed in this study indicate that individuals' traits may not only, or primarily, predict their behaviour in relationship settings.Rather, individuals' traits can have a significant influence on their partner's behaviour, and their partner's traits can have an influence on their own behaviour as well.Secondly, this research suggests that the influence of long-standing everyday behaviours (i.e., humour use) can be important to conflict resolution behaviours; that is, humour use is important not only as it is used within the context of the argument, but as it is used in partners' everyday lives.Finally, given that the management of conflict in intimate relationships may have an effect on marital satisfaction or stability (Holmes & Murray, 1996), the current research indicates that the use of different types of humour may have a significant impact on relationship wellbeing, through the mechanism of conflict resolution.
The present research does have a few limitations.The correlational nature of the data does not allow for cause and effect conclusions to be drawn.Longitudinal designs that assess the stability of relationship well-being, as well as experimental designs that expose participants to different types of couple humor use, may be useful to help establish causal connections.Additionally, the self-reports of couple humour use are beneficial in that they reflect a pattern of humour use that has been established over time in the relationship, but it would be valuable to determine the link between self-reported and observed humour use of these different styles of humour in future research.

Conclusions
The results of this research are consistent with prior research on humor use and relational well-being, but make a few novel contributions to this literature.For instance, the research focused more systematically on conflict resolution behaviors between intimates, using both self and observer reports, than past research.Additionally, the dyadic analyses demonstrated the unique associations between one partner's humor use and the other partner's conflict resolution behaviors, suggesting the importance of the interdependence that exists between partners when considering the role of humor in relationship maintenance behaviors.Overall, how partners use different types of humor in their relationships seems to be meaningfully associated with patterns of conflict resolution behavior for both partners.Notes i) No significant differences emerged between couples with children and couples without children on the three subscales of the Kansas Marital Conflict Scale, the Relational Humour Inventory, or observed positive and negative humour use.ii) No significant differences emerged between couples who returned for the follow-up session and couples who did not on the three subscales of the Kansas Marital Conflict Scale or the Relational Humour Inventory.

Funding
We gratefully acknowledge the support of a grant from the Social Sciences and Humanities Research Council of Canada awarded to Lorne Campbell to assist in the completion of this research.

Table 1
Note.Correlations above the diagonal are for women whereas correlations below the diagonal are for men.Correlations between partners appear along the diagonal.