A preliminary investigation of reaction time variability in relation to social functioning in children evaluated for ADHD

Abstract

Reaction time variability (RTV) is a ubiquitous phenomenon in Attention-Deficit/Hyperactivity Disorder (ADHD). Few studies have examined RTV in relation to functional outcomes such as social impairment in children with ADHD. In this exploratory study, we investigated whether RTV is associated with social functioning in children at risk for ADHD. Specifically, we explored the association between RTV (tau derived from correct go trials of a Stop Signal task) and social functioning in 198 children ages 7–12 years referred for an ADHD evaluation. Social functioning measures included child and/or parent ratings of social competence, aggression, social problems, and impairment in relationships. In regression analyses that also included Oppositional Defiant Disorder symptoms and sex, higher RTV was significantly associated with lower ratings of social competence, and higher proactive/reactive aggression ratings on the child self-report measures. RTV was not significantly associated with parent report of social functioning or relationship impairment. This study provides preliminary evidence that RTV may be associated with social functioning in children at risk for ADHD. We propose that lapses of attention affecting cognitive control may also negatively impact social information processing thereby affecting social functioning. Replication is warranted and longitudinal studies are needed to investigate whether RTV predicts social dysfunction in ADHD.

Intra-individual variability is a consistent feature of Attention-Deficit/Hyperactivity Disorder (ADHD) which is typically indexed via reaction time dispersion during computerized tasks, i.e., reaction time variability (RTV) (Kofler et al., 2013). RTV may be determined by multiple processes, such as stimulus encoding, speed of information processing (itself varying with arousal, effort, motivation, and other state factors), speed-accuracy trade-offs, post-error slowing, motor preparation, and response execution (Ratcliff, 2002). It has been suggested that RTV reflects central nervous system integrity and is associated with cognitive functions such as top-down attention control, executive function, and monitoring (MacDonald, Li, & Backman, 2009). Meta-analytic evidence shows that RTV is a ubiquitous phenomenon in ADHD, despite considerable differences in methodology, diagnostic methods, inhibitory demands, energetic factors, volitional/motivational influences, task characteristics, and variability metric (Kofler et al., 2013), and thus may serve as a key index of psychopathology and impaired functioning.

Despite consistent data implicating RTV in ADHD, very little is known about how RTV correlates with functioning and impairment in ADHD. The majority of studies examined the association of RTV with ADHD symptoms and/or performance on neuropsychological tasks, with an emphasis on investigating RTV as a causal mechanism in ADHD. Given that ADHD is associated with numerous negative sequelae (e.g., academic, social), it stands to reason that RTV should be associated with and predict functioning in individuals diagnosed with ADHD. The limited literature to date supports such associations. Numerous studies show an association between RTV and ADHD inattention symptoms (Kofler et al., 2013), including behavioral inattention as indexed by objectively-coded off-task behavior (Antonini, Narad, Langberg, & Epstein, 2013). In the academic domain, RTV has been associated with poor reading in children with ADHD (Jacobson, Ryan, Denckla, Mostofsky, & Mahone, 2013; Ryan et al., 2017; Tamm et al., 2014). RTV has also been associated with poorer emotion regulation in preschoolers with ADHD (Sjowall, Backman, & Thorell, 2015), but not adults (Surman et al., 2015). Further, RTV predicted ratings of emotional lability, although this association was not maintained after controlling for ADHD symptoms (Banaschewski et al., 2012). A longitudinal study showed that higher RTV predicted lower clinician ratings on the Global Assessment Scale of the Dutch version of the Kiddie Schedule for Affective Disorders and Schizophrenia for School-Age Children six years later in children with ADHD (van Lieshout et al., 2017). Thus, RTV appears to be associated with some domains of impairment in children with ADHD.

One domain of functioning that has not specifically been examined in the context of RTV is social functioning, which is often impaired in children with ADHD (Hoza, 2007; Landau & Moore, 1991; Pelham & Bender, 1982). Children with ADHD may show disruptive, aggressive, and hyperactive behaviors leading to peer rejection (Nijmeijer et al., 2008). Children with these types of behaviors are often unpopular and lack reciprocal friendships (Hoza, 2007; Landau & Moore, 1991). Those with the inattentive presentation show passive, slow behaviors and may be prone to anxiety, shyness, and withdrawal that lead to compromised social skills (Nijmeijer et al., 2008). Given that social impairment significantly predicts adverse outcomes (Greene et al., 1999; Mrug et al., 2012), while adequate social functioning and peer relationships promote optimal development (Newcomb & Bagwell, 1995), it is critical to understand the underpinnings of social dysfunction in ADHD (De Boo & Prins, 2007).

It has been argued that inconsistent allocation of attention and effort plays a major role in the deficits shown by children with ADHD (Douglas, 1999). A prevailing model of RTV is that difficulties involving the allocation of effort lead to attentional lapses during the course of information processing, and therefore RTV reflects deficits in cognitive control (Leth-Steenson, Elbaz, & Douglas, 2000; Ode, Robinson, & Hanson, 2011; Vaurio, Simmonds, & Mostofsky, 2009). Cognitive control refers to processes that flexibly and adaptively allocate mental resources to permit the dynamic selection of thoughts and actions in response to context-specific goals and intentions, or, in other words, information processing and response generation (Fan, 2014). Social information processing involves encoding of cues, interpretation of cues, clarification of goals, response access or construction, response decision, and behavioral enactment (Crick & Dodge, 1994). Deficient processing at any stage may lead to disruption of social functioning in ADHD (Sibley, Evans, & Serpell, 2010; Uekermann et al., 2010). Indeed, there is significant evidence of deficits in the abilities to process information and modulate social responses according to contextual factors in children with ADHD (McQuade & Hoza, 2008; Milich & Dodge, 1984; Ros & Graziano, 2017; Rutter, 1987). Thus, we argue RTV may be associated with social dysfunction in ADHD in that it reflects cognitive factors related to the ability to process information and modulate social responses according to contextual factors. Such variability in cognitive control impacting social information processing over time may result in impaired social functioning.

While no studies have examined RTV and social functioning in ADHD, RTV correlates with social communication deficits in children with autism spectrum disorders (ASD) (Hoyland, Naerland, Engstrom, Lydersen, & Andreassen, 2017; Pinto, Rijsdijk, Ronald, Asherson, & Kuntsi, 2016). For example, RTV on an emotional continuous performance task was significantly associated with poorer social function in children with ASD (Hoyland et al., 2017). Given that children with ADHD often show deficits in social communication, ADHD and ASD may share a common genetic liability (Musser et al., 2014; Ronald, Simonoff, Kuntsi, Asherson, & Plomin, 2008), and RTV is also characteristic of ASD (Karalunas, Geurts, Konrad, Bender, & Nigg, 2014), it is plausible that RTV is similarly associated with social functioning in children with ADHD.

In the current study we explore the association between RTV and social functioning in children referred to an outpatient clinic for an ADHD evaluation. Social functioning is a multi-faceted construct and in the current study we included child and parent ratings of social acceptance and parent ratings of social problems and impairment in parent-, sibling-, and peer-relationships. We also included child self-report of aggression since aggression specifically has been linked to social information processing deficits (Dodge & Coie, 1987) and is consistently linked with peer problems (Hoza, 2007). Although this was an exploratory study, we hypothesized that higher RTV would be associated with parent and child report of poorer social functioning and greater child self-report of aggression.

Methods

Participants

Families with children ages 7–13 were recruited at an outpatient clinic specializing in pediatric ADHD. Participants who contributed data for the current analyses were 198 children. Demographic characteristics are presented in Table 1. Of the 198 children, 176 met diagnostic criteria for ADHD according to parent responses on the Kiddie Schedule for Affective Disorders and Schizophrenia for School-Age Children (KSADS) (Kaufman et al., 1997). In terms of comorbidities, 37 met criteria for ODD, 4 for conduct disorder, 2 for major depressive disorder, and 26 participants for at least one anxiety disorder. The majority of children (76.3%) were not on psychotropic medications.

Table 1.

Demographic Characteristics of Sample (n=198)

Age	M=8.71, SD=1.60
Gender	65% male
IQ	M=99.99, SD=14.06
Ethnicity/Race
Non-Hispanic White	81%
African American	15%
Hispanic	3%
Asian	1%
Native American	<1%
Income
<$30,000	20.1%
$30,001–$50,000	14.2%
$50,001–$70,000	12.1%
>$70,000	45%
declined to answer	8.6%

Procedures

The study was approved by the hospital Institutional Review Board and informed consent and assent were obtained. Parents were administered the disruptive behavior, mood, and anxiety disorder modules of the KSADS (Kaufman et al., 1997) and completed ratings. Concurrently, the child was administered the Kaufman Brief Intelligence Test-2 (Kaufman & Kaufman, 2004), Stop-Signal Test, and completed ratings.

Measures

ADHD/ODD.

Parents rate how frequently each ADHD and ODD symptom occurs on a 4-point scale (0=never, 3=very often) on the Vanderbilt ADHD Diagnostic Parent Rating Scale (VADPRS) (Wolraich et al., 2003). Internal consistency and reliability is excellent (Wolraich et al., 2003). Average scores were computed for the eighteen ADHD (α = .92), and eight ODD (α = .90) items.

Reaction Time Variability (RTV).

Reaction time data from correct go trials on the Stop-Signal Test (Logan & Cowan, 1984) was utilized to compute the ex-Gaussian parameters, sigma and tau as indices of RTV. During the Stop-Signal Test, a fixation cross was presented in the center of a computer screen for 500msec followed by a 500msec presentation of a target stimulus (airplane facing left or right). Participants press a button corresponding to the direction the target stimulus is facing. However, an auditory “stop signal” is presented on 25% of trials requiring participants to inhibit their response (stop trials). The delay between presentation of the target stimulus and the tone began at 250msec and varied according to the participant’s performance; successful inhibition resulted in increases of 50msec and unsuccessful inhibition resulted in decreases of 50msec so that the rate of inhibition was controlled to approximate 50%. Following 60 practice trials, participants completed 178 trials of 3000msec each.

Mu, sigma, and tau for the approximately 133 correct go trials (excluding practice) were computed using RTSYS, a DOS based statistical program used for the analysis of reaction time data (Heathcote, 1996). The RTSYS program uses a theoretical distribution comprised of the combination of a Gaussian distribution in the initial or left side of the distribution curve, and an exponential distribution on the latter or right side of the distribution curve as a model of reaction time distribution. This theoretical distribution, the ex-Gaussian, represents the sum of the independent Gaussian (normal) distribution and exponential random variables that comprise the exponential distribution and resulting positive skew of the curve. The ex-Gaussian distribution has three parameters: mu, the mean of the normal component, sigma, the standard deviation of the normal component, and tau, a single value that describes both the mean and the standard deviation of the exponential component. The tau component characterizes the skewness of the distribution. Higher tau values are produced by reaction time distributions with a larger rightward skew or tail (Leth-Steenson et al., 2000). Sigma represents the variability of responses both above and below the mean, that is normal distributions (Balota & Yap, 2011). Thus, tau and sigma were included as measures of RTV.

Social Functioning.

The Self-Perception Profile for Children (SPPC) (Harter, 1985) is a validated measure (Muris, Meesters, & Fijen, 2003) of perceived child competence. Parents completed an 18-item measure assessing perceptions of their child’s competence across social (α = .89), behavioral (α = .92), and academic (α = .83) domains and children completed the 24-item version assessing self-perceptions of competence in social acceptance (α = .64), behavioral conduct (α = .71), academic (α = .73), and global self-worth (α = .71) domains. We utilized the social competence (parent and child ratings of knowing how to make friends, having the skills to get others to like you, understanding what it takes to become popular, etc.) scores. Lower ratings indicate poorer competence.

Parents completed the Child Behavior Checklist (CBCL) (Achenbach, 1991), a general questionnaire consisting of 118 items scored on a 3-point scale ranging from not true to often true of the child. It is a widely-used standardized measure for evaluating maladaptive behavioral/emotional problems in children. The CBCL has alternative norms for males and females at various ages, and the T-scores generated by the scoring program use the appropriate sex norms. For the current study, the Social Problems T-score was utilized which includes items such as not being liked, being teased, not getting along with others, etc. Higher T-scores indicate more problems. It should be noted that the CBCL was only obtained on a subsample of participants (n=117).

Impairment in social relationships was captured by the VADPRS impairment ratings on a scale ranging from 1 (excellent) to 5 (problematic). Individual items assessing the child’s relationship with parent, siblings, and peers were utilized. Higher ratings indicate greater impairment.

The Reactive/Proactive Aggression Questionnaire (RPAQ) (Raine et al., 2006) is a 23-item measure assessing proactive (α = .85) and reactive (α = .82) aggression. Children rate themselves on the items which are presented on a “never, sometimes, or often” scale. Higher sum scores on each subscale reflect more aggression.

Statistical analysis

Associations between all variables of interest were assessed with Pearson (continuous variables) or Spearman (dichotomous variables) bivariate correlations (two-tailed). Social functioning variables that significantly correlated with tau or sigma (p<.05) were retained for multiple regression analyses to evaluate whether RTV was associated with social functioning. Sex and ODD were included in the models given their association with social functioning (e.g., Archer, 2004; Greene et al., 2002); we also ran an additional regression including ADHD symptoms in addition to sex and IQ. The p-values from the regression analyses were corrected for multiple comparisons using False Discovery Rate procedures (Benjamini & Hochberg, 1995).

Missing data was present for 4% of the sample on the SPPC-Child, 2.5% on the SPPC-Parent, 40.9% on the CBCL, and 4% on the VADPRS. Tests of patterns of missingness suggested data were missing completely at random (Little’s MCAR test: χ²=39.38, df=38, p=.41). Comparison of participants with complete and incomplete data revealed that the groups did not differ significantly on age, IQ, gender, race, income, ADHD symptoms, or ODD symptoms (all ps>.10).

Results

Bivariate correlations

Tau was significantly negatively correlated with child self-report of social acceptance on the SPPC and significantly positively correlated with child self-report of both proactive and reactive aggression (Table 2). The significant negative correlation between tau and sex indicates higher tau values among females than males. No significant correlations between sigma and the social functioning or aggression variables emerged (Table 2).

Table 2.

Inter-correlations between RTV and social functioning variables

	1	2	3	4	5	6	7	8	9	10	11	12	13	14
1. SST Tau	--
2. SST Sigma	.39**	--
3. SST Mu	.27**	.80**	--
4. SPPC Social Acc-C	−.17*	−.06	.03	--
5. SPPC Social Acc-P	−.07	.05	.06	.26**	--
6. CBCL Social Problems	.17	.17	.20*	−.28**	−.59**	--
7. VADPRS Parent Rel. Impair	.11	.05	.03	−.16*	−.25**	.29**	--
8. VADPRS Sibling Rel. Impair	.13	−.08	−.05	−.07	−.24**	.29**	.70**	--
9. VADPRS Peer Rel. Impair	.08	−.05	−.09	−.18*	−.63**	.50**	.52**	.53**	--
10. RPAQ Proactive Aggression	.21**	.09	−.04	−.12	−.14*	.13	.17*	.21**	.24**	--
11. RPAQ Reactive Aggression	.19**	.06	−.02	−.15*	−.15*	.06	.19**	.19**	.22**	.63**	--
12. Sex	−.19*	.03	−.05	−.08	−.08	.01	−.08	−.01	.04	.13T	.07	--
13. VADPRS ADHD symptoms	.11	−.01	−.05	−.07	−.26**	−.25**	.38**	.34**	.36**	.11	.12	.02	--
14. VADPRS ODD symptoms	.09	.01	−.01	−.13	−.34**	.34**	.55**	.50**	.43**	.16*	.19**	−.03	.56**	--
Mean SD	198.8 107.0	111.8 78.6	447.3 199.3	2.8 0.7	2.8 0.8	60.6 8.0	2.5 1.1	2.8 1.1	2.8 1.0	3.2 4.1	7.9 4.5	n/a	1.8 0.6	1.1 0.7
Min Max	35.7 569.2	7.5 546.6	183.7 164.1	1 4	1 4	50 85	1 5	1 5	1 5	0 25	0 22	n/a	.22 2.94	.00 3.00
n	198	198	198	190	193	117	192	190	193	196	196	198	193	193

Note. SST=Stop Signal Task; SPPC = Self-Perception Profile for Children; C=child; P = parent; CBCL = Child Behavior Checklist; VADPRS = Vanderbilt ADHD Diagnostic Parent Rating Scale Impair=Impairment; RPAQ = Reactive/Proactive Aggression Questionnaire ODD=oppositional defiant disorder; Rel. = relationship. Sex was coded as 0 (female) or 1 (male).

^*p<.05.

^**p<.01.

Regression analyses

After correction for multiple comparisons, RTV as indexed by tau, was significantly associated with lower child-rated social acceptance [B(SE) = −.001(.000); β=−.19, corrected p=.038], greater proactive aggression [B(SE) = .009 (.003); β=.24, corrected p=.009] and greater reactive aggression [B(SE) = .008 (.003); β=.19, corrected p=.019] (Table 3). As would be expected from the literature, ODD symptoms were associated with higher rates of proactive [B(SE) = .860(.399); β=.15, corrected p=.048] and reactive aggression [B(SE) = 1.127 (.440); β=.18, corrected p=.020], and boys had higher rates of proactive aggression [B(SE) = 1.647 (.612); β=.19, corrected p=.020] (Table 3). After including ADHD symptoms in the models, the pattern of results for RTV remained the same. Tau remained significantly associated with lower child-rated social acceptance [B(SE) = −.001(.000); β=−.19, corrected p=.030], greater proactive aggression [B(SE) = .009 (.003); β=.24, corrected p=.012] and greater reactive aggression [B(SE) = .008 (.003); β=.19, corrected p=.030] (Table 3a).

Table 3.

Multiple Regression Analyses

	Social Acceptance	Proactive Aggression	Reactive Aggression
	R2=.06	R2=.10	R2=.08
	Standardized Beta	Standardized Beta	Standardized Beta
Reaction time variability (tau)	−.19*	.24**	.19*
Oppositional defiant symptoms	−.12	.15*	.18*
Sex	−.12	.19*	.13

^*corrected p<.05.

^**corrected p<.01.

Table 3a.

Multiple Regression Analyses (including ADHD symptom severity)

	Social Acceptance	Proactive Aggression	Reactive Aggression
	R2=.06	R2=.12	R2=.08
	Standardized Beta	Standardized Beta	Standardized Beta
Reaction time variability (tau)	−.19*	.24*	.19*
ADHD symptoms	.03	−.00	−.01
Oppositional defiant symptoms	−.13	.15	.18
Sex	−.12	.19*	.13

^*corrected p<.05.

Discussion

This preliminary study explored whether RTV is associated with social functioning in children with ADHD. Although replication is warranted, particularly with a longitudinal design and objective social functioning measures, our results demonstrate that higher RTV, as indexed by tau, is associated with worse social functioning, and in particular child self-report of social acceptance, proactive aggression, and reactive aggression. We hypothesize that this association between RTV and social dysfunction is due to the possibility that RTV reflects cognitive factors related to the ability to process information and modulate social responses according to contextual factors. Specifically, tau (i.e., frequent, excessively long reaction times) likely reflects occasions when children with ADHD demonstrate fluctuations or lapses in attention (Douglas, 1999; Leth-Steenson et al., 2000). Given the cross-sectional nature of our study, we cannot directly investigate whether RTV predicts social functioning, but our work adds to a longstanding interest in the role of response latencies as indices of information processing (Posner, 1978]) and the role of attention in moderating social cognition (Posner, 2012).

Social information processing is highly automatic and decisions are made reflexively (Crick & Dodge, 1994), and therefore likely occur outside the child’s awareness. We speculate that RTV may be a useful indirect measure of implicit social cognition, and that RTV may be associated with the first two stages of social information processing – encoding of cues and interpretation of cues (Crick & Dodge, 1994). Encoding of cues involves processing relevant aspects of the stimulus through sensory input, selective attention to social cues, and encoding of cue information in short term memory. Interpretation of the social cues that were encoded leads to the formation of a mental representation or an understanding of the social event, which may be strongly influenced by previous experience and bias. Thus, we argue that a child’s perception of their competence in knowing how to make friends and having the skills to get others to like them likely directly influences encoding and interpretation of incoming social information. Similarly, a child’s self-report of aggressive responses likely reflects their intrinsic motivation for aggressive acts and how they interpret social cues. In fact, studies do show that social information functioning profiles of children with ADHD are marked by difficulties in encoding social cues (biased attention toward inappropriate cues), and interpreting social cues (biased mental representations of events) (Zentall, Cassady, & Javorsky, 2001), and that reactive aggression is linked to the encoding and interpretation stages of social information processing (Crick & Dodge, 1994; Dodge & Coie, 1987) and predicted by cognitive control processes (self-regulation of goal-directed behavior) (Giancola, Moss, Martin, Kirisci, & Tarter, 1996). The finding of proactive aggression being associated with RTV was somewhat surprising. Proactive aggression occurs in later stages of the social information processing model, is more purposeful, and driven by reward (i.e., aggressive acts to obtain a desired outcome) (Crick & Dodge, 1994). However, the RPAQ captures intrinsic motivation for aggressive acts and not the aggressive acts themselves (e.g., bullying) which may also impact the encoding and interpretation stages of social information processing. Additional research on the association of RTV and aggressive behaviors (as rated by other observers) may help disentangle these issues.

We did not include measures of emotional processing in our study which is increasingly recognized as playing a role in social functioning and social information processing specifically (Lemerise & Arsenio, 2000). Emotional processing is a key element of social cognition given its strong links to social-emotional functioning, interpersonal reciprocity, and social motivation (Schultz, 2005). Individuals with ADHD show anomalies in emotional processing including a blunted response to encoding negative stimuli associated with emotional lability (Du et al., 2006; Williams et al., 2008), failure to orient to emotional stimuli, anomalies in reward valuation, and deficient allocation of attention to emotional stimuli (Shaw, Stringaris, Nigg, & Leibenluft, 2014). Thus, an alternate hypothesis is that higher RTV may be associated with aspects of social cognition such that processes leading to inconsistent reaction times could reflect general regulatory processes that impact emotion regulation (Skirrow, McLoughlin, Kuntsi, & Asherson, 2009). In fact, RTV has been shown to be correlated with emotional regulation (Sjowall et al., 2015) and emotional lability (Banaschewski et al., 2012) in ADHD, although see study by Elmaghrabi et al. (2018) for an exception, and with negative emotional states in a non-clinical population (Ode et al., 2011). Recent models of social information processing argue for an integral role for emotional processing but the literature examining the interplay between social information processing and emotionality/regulation is limited (Lemerise & Arsenio, 2000). It will be difficult to disentangle the potential mechanistic explanation for the association between RTV and social functioning given that emotion and cognition are inextricably intertwined (Zelazo & Cunningham, 2007). Future studies will need to include measures of emotional processing and cognition, and employ complex statistical modeling to elucidate how RTV impacts social function.

It is interesting that child self-report of social functioning was associated with RTV. It has been argued that children are in the unique position to report on behaviors across different situations, and thus their perspective should be captured in any assessment (La Greca, 1990). Further, parent-child agreement is typically lower for emotional and social ratings (Eiser & Morse, 2001) and children are often better reporters of internal states that are not readily observable for parents (Moretti, Fine, Haley, & Marriage, 1985; Rey, Schrader, & Morris-Yates, 1992). Understanding the intrinsic motivation for aggression may best be understood by asking the child about their perceptions. However, others have argued that children with ADHD have a positive illusory bias (PIB) and may not be the best reporters of competence in any domain (Owens, Goldfine, Evangelista, Hoza, & Kaiser, 2007). While recent evidence calls into question how ubiquitous this phenomenon is in ADHD, (e.g., Jiang & Johnston, 2017), future work will need to examine the potential association between RTV and self-report of social competence in light of potential PIB.

Although this paper has focused on children at risk for ADHD and the potential role that RTV may play in their social functioning, it should be noted that RTV is characteristic of multiple psychiatric disorders (Kofler et al., 2013; Tamm et al., 2012). In fact, RTV has been proposed as a transdiagnostic phenotype associated with shared risk for several disorders or with symptom domains (e.g., social problems) that cut across several disorder categories (Kofler et al., 2013). Social dysfunction is characteristic of people with a range of psychiatric disorders (e.g., ASD, learning disorders, ODD, anxiety) (American Psychiatric Association, 2013), and, as noted earlier, RTV is associated with social-communication deficits in children with ASD (Hoyland et al., 2017; Pinto et al., 2016). Additional research is warranted to investigate the potential association between RTV and social functioning in a range of clinical populations to ascertain the relevance of RTV as a transdiagnostic mechanism.

Our findings from this exploratory study should be interpreted in light of study limitations. Because our data were derived from a sample of convenience rather than a research study designed to look at the relation between RTV and social functioning, our measures did not capture other relevant social functioning constructs (e.g., social withdrawal, peer rejection, exclusion, prosocial/asocial behaviors, bullying/victimization, PIB). We were also limited to parent and child ratings and were unable to collect peer or teacher ratings. Peer sociometric nominations may be especially informative since child behaviors contributing to social competence are often situation specific and, thus, less likely to be observed by adults (e.g., acts of aggression). Teacher ratings are also critical as parents may have less knowledge of their child’s social behaviors since they do not typically see their children within the context of large peer groups. Additionally, it was surprising that RTV was not significantly correlated with ADHD symptoms in the current sample which raises concern about the validity of tau in this sample. It should be noted that the direction of the association (p=.11) is similar to that reported in the literature (Kofler et al., 2013), although of lesser magnitude (most studies report correlations of .2 to .3 between RTV and ADHD symptoms). This may reflect that our sample was characterized by children referred for an ADHD evaluation and not necessarily diagnosed with ADHD. In fact, the mean score on VADPRS ADHD symptoms in the current sample was relatively low (1.8±.06; it would be expected that children with ADHD would have a score ≥2.0 on the VADPRS reflecting an average rating of ‘pretty often’ or ‘very often’ on most ADHD symptoms), and lower RTV has been reported in individuals who are described as “remitters” (with fewer ADHD symptoms) than persisters (with more ADHD symptoms) (Michelini et al., 2016).

Conclusion

Despite these limitations, our study found RTV to be independently related to child self-report of poorer competence in the domain of social acceptance as well as higher ratings of proactive and reactive aggression in children at risk for ADHD. Our data add to the limited literature indicating that RTV is associated with meaningful domains of impairment in ADHD, and could be a critical construct to measure. More work is needed to elucidate the mechanistic explanations for how RTV operates to impact social function (e.g., through its impact on cognition/social information processing and/or emotional processing) as well as the value of RTV as a transdiagnostic mechanism.