ADHD, SOCIAL SKILLS, ED AND DEPRESSION: MODERATED MEDIATION 1

Online Supplementary Material for Self-Rated Social Skills

There may be limitations to testing our research questions with parent- and with self-report data[1]. First, because our predictor (i.e., ADHD) is based on parent-report, we may have found support for our models with parent-rated social skills as the outcome variable due to shared method variance between the predictor and the outcome. However, because our predictor is based on parent-report, it is unlikely that models with self-rated social skills would be significant due to shared method variance between the predictor and the outcome. Yet, youth with ADHD often over- and under-report the deficits they experience in diverse areas of performance (Owens et al., 2011).To determine the extent to which shared method variance accounts for our results obtained with parent-rated social skills, we repeated the analyses we conducted with parent-rated social skills as the outcome variablewith self-rated social skills as the outcome variable. We present results obtained with self-rated social skills next. Of note, our findings with self-rated social skills as the outcome largely replicate the findings we obtained with parent-rated social skills as the outcome. As such, there is reason to believe that the support we obtained for our models is not due to shared method variance.

Hypothesis 1: ED Will Mediate the Negative Association between ADHD and Self-Rated Social Skills

To test whether ADHD is negatively associated with social skills through ED, we tested one, multiple-mediational model with ERICA Self-Awareness, ERICA Emotional Control and DERS Impulse as parallel mediators. Bootstrapping results of the omnibus effect supported our model (R2= .349, F(5, 171)= 17.267, p < .001)and depression and indicated that, jointly, ADHD, ERICA Self-Awareness, ERICA Emotional Control, and DERS Impulse account for 35% of the variance in social skills. This effect was due to mediation by ERICA Self-Awareness (point estimate = .130; SE = .069; 95% CIs [.015, .288]) and Emotional Control (point estimate = -.079; SE = .041; 95% CIs [-.193, .019]). No mediation by DERS Impulse (point estimate = -.004; SE = .064; 95% CIs [-.119, .133]) was observed. These findings suggest thatERICA Self-Awareness and Emotional Control are, but DERS Impulse is not a mediatorof the negative association between ADHD and social skills.

A follow-up test of simple mediation with ERICA Self-Awareness as the mediator further supported our model (point estimate = .131; SE = .075; 95% CIs [-.001, .296]). A follow-up test of simple mediation with ERICA Emotional Control as the mediator was nonsignificant (point estimate = -.003; SE = .048; 95% CIs [-.093, .100]).

Exploratory Analysis: Does Depression Mediate the Negative Association between ADHD and Social Skills?

As previously noted, although prior evidence indicates that ED predates the development of depression and mediates the association between ADHD and depression, it is unclear whether depression is best conceptualized as a moderator of the direct and indirect (through ED) association between ADHD and social skills. Indeed, ED and depression were correlated in our sample (see Table 2). To examine whether, in line with our proposed hypothesis, depression is best conceptualized as a moderator and not a mediator, we conducted an exploratory analysis to examine whether depression mediates the negative association between ADHD and parent-rated social skills. Bootstrapping results did not support this model (point estimate = .051; SE = .062; 95% CIs [-.073, .175]), indicating that depression does not mediate the negative association between ADHD and social skills.

Thus, per our mediation results, we conclude that some aspects of ED mediate the negative association between ADHD social skills but depression does not. We thus chose to retain our proposed hypothesis, per which it may be that ADHD is associated with social skills through ED and that this process operates differently at distinct levels of depression.

Hypothesis 2: Depression Will Moderate the NegativeDirect and Indirect Association between ADHD andSelf-Rated Social Skills

To test this research question, we examined moderated mediation models with ADHD as the predictor, social skills as the outcome, ED mediating the negative association between ADHD and social skills, and depression moderating the mediational model. To this end, we evaluated whether depression moderates (1) the path between ADHD and social skills (i.e., direct path) (Figure 1) and (2) the path between ADHD and ED (i.e., indirect path) (Figure 2).

Hypothesis 2a: Depression will moderate the negative direct association between ADHD andself-rated social skills.

A test of the omnibus effect supported our hypothesized model with ERICA Self-Awareness, ERICA Emotional Control, and DERS Impulse as parallel mediators (R2= .365, F(7, 171)= 12.786, p < .001) and indicated that, jointly, ADHD, ERICA Self-Awareness, ERICA Emotional Control, DERS Impulse, and depression account for 37% of the variance in social skills. This effect was driven by depression moderating the negative direct effect of ADHD on parent-rated social skills with ERICA Self-Awareness (point estimate = .107; SE = .060; 95% CIs [.022, .270]) and Emotional Control (point estimate = -.086; SE = .046; 95% CIs [-.196, -.016])as mediators of this association. Depression did not moderate the negative direct effect of ADHD on social skills with DERS Impulse (point estimate = .021; SE = .059; 95% CIs [-.095, .156]) as a mediator of this association.

A follow-up test of moderated mediation with ERICA Self-Awareness as a single mediator was nonsignificant (ps ranging from .299 to .541). A follow-up test of moderated mediation with ERICA Emotional Control as a single mediator was nonsignificant (ps ranging from .279 to .857).

Hypothesis 2b: Depression will moderate the negativeindirect association between ADHD andself-rated social skills through ED. A test of the omnibus effect supported our hypothesized model with ERICA Self-Awareness, ERICA Emotional Control, and DERS Impulse as parallel mediators (R2= .353, F(5, 171)= 17.228, p < .001) and indicated that, jointly, ADHD, ERICA Self-Awareness, ERICA Emotional Control, DERS Awareness, and depression account for 35% of the variance in social skills. Depression moderated the negative indirect effect of ADHD on parent-rated social skills through ERICA Self-Awareness (b = 2.007, p < .001), Emotional Control(b = -.581, p = .007), and DERS Impulse (b = -1.302, p < .001)

A follow-up test of moderated mediation with ERICA Self Awareness as a single mediator indicated that, among youth with ADHD, ERICA Self-Awareness was a significant mediator of the negative effect of ADHD on social skills at non-clinical levels of depression but was not a significant mediator of the effect of ADHD on social skills at sub-clinical depression or at clinical levels of depression (see Table 5).

Follow-up test of moderated mediation with ERICA Emotional Control and DERS Impulse as single mediators was nonsignificant (95% CIs ranging from [-.107, .101] to [-.132, .125] and from [-.121, .119] to [-.257, .339], respectively).

ADHD, SOCIAL SKILLS, ED AND DEPRESSION: MODERATED MEDIATION 1

References

Klein, D. N., Dougherty, L. R., & Olino, T. M. (2005). Toward guidelines for evidence-based assessment of depression in children and adolescents.Journal of Clinical Child and Adolescent Psychology, 34(3), 412-432.

Silverman, W. K., & Ollendick, T. H. (2005).Evidence-based assessment of anxiety and its disorders in children and adolescents.Journal of Clinical Child and Adolescent Psychology, 34(3), 380-411.

[1]With regard to the mediator (i.e., ED) and moderator (i.e., depression), measuring internalizing problems via self-report is the recommended practice by experts in the field (Klein, Dougherty, & Olino, 2005; Silverman & Ollendick, 2005).