Both humor and friendship facilitate socio-emotional development in middle childhood, but scholars have not studied the relation between the two constructs. Specifically, researchers have not explored the relation between affiliative, aggressive and/or self-defeating spontaneous conversational humor and positive and negative friendship quality in middle childhood. Scholars theorize that affiliative spontaneous conversational humor contributes to positive friendship quality while aggressive and self-defeating spontaneous conversational humor contribute to negative friendship quality (Klein & Kuiper, 2006). Moreover, friends' social behaviors are often interdependent and "work together" to influence each child's perception of friendship (Bukowski, Motzoi, & Meyer, 2009); therefore, it is also expected that dyadic spontaneous conversational humor is associated with friendship quality in middle childhood and that dyad-level variables (i.e., behavioral similarity of dyad, duration of friendship) are related to the dyadic nature of children's spontaneous conversational humor production.

Participants were 250 fifth graders (125 dyads; M age = 10.33 years, SD = .54) from the Washington DC metropolitan area. Affiliative, aggressive, and self-defeating spontaneous conversational humor data were coded from videotaped discussion between mutually nominated, same-sex best friends. Positive and negative friendship quality data were collected via survey report. An Actor Partner Interdependence Model (APIM) was used to explore relations between actor and partner spontaneous conversational humor and actor-rated friendship quality in middle childhood (Kenny, Kashy, & Cook, 2006). Finally, multiple regression analyses were used to test the direct effects of dyad-level variables on dyadic spontaneous conversational humor.

APIM analyses revealed significant positive relations between actor affiliative spontaneous conversational humor and actor-rated absence of conflict and betrayal; actor self-defeating spontaneous conversational humor and both actor-rated intimate disclosure and total positive friendship quality; and partner self-defeating spontaneous conversational humor and actor-rated companionship. Actor affiliative and actor aggressive spontaneous conversational humor production were negatively and positively associated with actor-rated negative interactions, respectively. Five dyadic actor-partner interactions were significantly related to positive and negative friendship quality. Multiple regression analyses indicated that friendship duration was negatively associated with dyadic aggressive spontaneous conversational humor, meaning that the longer best friends reported knowing each other, the less they used interrelated aggressive spontaneous conversational humor.