Category: ADHD - Child
A growing body of studies indicates that children with ADHD experience significant and treatment-resistant friendship problems (McQuade & Hoza, 2015). Although children’s perspectives about their own friendships are often considered the gold standard when assessing friendship quality (Bagwell & Schmidt, 2011), observational measures may circumvent children’s response biases (e.g., Owens, Goldfine, Evangelista, Hoza, & Kaiser, 2007) and relatively limited verbal skills, and may capture aspects of the friendship that are less accessible through self-reports (Girard & Cohn, 2016; Schneider, 2016). The objective of the current study was to verify the psychometric properties (i.e., inter-rater reliability, construct validity, criterion validity) of a newly developed observational measure of friendship quality for children with ADHD: the Friendship Interactions Coding Scheme (FICS).
Participants were 66 target children with confirmed ADHD (45 boys, Mage = 8.7 years) and their invited friends, recruited from different clinics in Eastern Canada. Target children with ADHD and their friends were invited to the lab. They separately completed unlimited friendship nominations to confirm the reciprocal nature of their friendship and the Friendship Quality Questionnaire-short form (22 items e.g., This friend makes me feel good about my ideas). They also completed two observational tasks (i.e., toy-sharing [TST] and car-race [CRT] tasks) designed to mirror the real-world interactions of friends and to elicit a variety of social behaviors. Naïve observers later coded the filmed interactions with the FICS (i.e., timed-event sequential continuous coding: prosocial behavior, controlling behavior, aggressive behavior, rule breaking positive affect, negative affect; global ratings: closeness, cooperation).
Results indicate that inter-rater reliability was adequate (i.e., all κs ≥ 0.64 for categorical variables; all ICCs ≥ 0.81 for continuous variables). The FICS subscales were moderately intercorrelated. In the TST, observed cooperation was only positively correlated (r = .31, p < 0.05) with the friends' reports of positive friendship quality. In contrast, in the CRT, observed closeness (r =.33, p < 0.01) and aggressive behavior (r = -.29, p < 0.05) related to target children’s reports of positive friendship quality. Closeness was also significantly correlated with target children’s reports of conflict (r = -.27, p < 0.05). Prosocial behavior (r = -.35, p < 0.01) was negatively correlated with the invited friends’ reports of conflict, and positively correlated with invited friends’ reports of positive friendship quality (r = .30, p < 0.05).
Overall these results indicate that the FICS subscales seem to capture related yet independent observed aspects of friendship quality. Because the correlations between self-report and observational measures of friendship quality were weak to moderate, these two approaches may tap different dimensions of friendship quality. This highlights the importance of using a multimethod assessment when measuring friendship quality as a treatment outcome for children with ADHD.