Using inspection time and ex-Gaussian parameters of reaction time to predict executive functions in children with ADHD
Introduction
Broadly characterized by difficulty with sustained attention, hyperactivity and impulsivity, Attention Deficit Hyperactivity Disorder (ADHD) is a serious risk factor for numerous negative outcomes such as frequent psychiatric comorbidity and poor academic performance (Biederman, 2005; Biederman et al., 2004; DuPaul, McGoey, Eckert, & VanBrakle, 2001; Fanti & Henrich, 2010). Though the etiology of the disorder largely remains unknown, weaker performance on tasks of executive functions (EF) and slower/more variable motor reaction time (RT) are commonly associated with the disorder (Adamo et al., 2014; Buzy, Medoff, & Schweitzer, 2009; Hervey et al., 2006; Jacobson et al., 2011; Karalunas & Huang-Pollock, 2013; Kofler et al., 2013). Current evidence suggests that these indices of cognition are not independent. RT improves with age, and this improvement is associated with improvements in working memory (WM), a key executive function (Bayliss, Jarrold, Baddeley, Gunn, & Leigh, 2005; Demetriou et al., 2014; Fry & Hale, 2000; Kail, Lervag, & Hulme, 2016; Tourva, Spanoudis, & Demetriou, 2016). There is also evidence that individual differences in RT predict individual differences in working memory capacity (Cowan et al., 2003; Karalunas & Huang-Pollock, 2013; Weigard & Huang-Pollock, 2016).
It is not clear why this should be. However, information processing during standard speeded choice response tasks is generally understood to entail several broad sub-processes or components including perceptual encoding, decision-making, and fine-motor output (Luce, 1986; Salthouse, 1996). And some (Luce, 1986; Myerson, Hale, Zheng, Jenkins, & Widaman, 2003; Rotello & Zeng, 2008; Smith & Ratcliff, 2004) have argued that these components may be inferred from the shape of the RT distribution that is produced during speeded task performance. By asking which portion(s) of the RT distribution uniquely predict(s) EF, it might be possible to identify which sub-process is most predictive of EF performance.
RT distributions are best represented by an ex-Gaussian distribution, which is formed from the integration of normal and exponential distributions, resulting in a distribution that does not have a lower tail, but does have a long upper tail (Dawson, 1988; Ratcliff & Murdock, 1976). Practically speaking, the ex-Gaussian shape appears because RTs have a lower time-limit of zero milliseconds, but no upper time-limit (Coyle, 2003). The parameters mu and sigma refer to the mean and standard deviation of the normal portion of the distribution; tau characterizes the mean of the exponentially shaped portion of the distribution (Lacouture & Cousineau, 2008).
Competing psychological constructs have been proposed for each of these components (see Matzke & Wagenmakers, 2009 for review). However, many have argued that informational encoding and motor preparation/execution are relatively automatic functions, and that the relative health or efficiency of those processes are best indexed by indices of central tendency (i.e., mu and sigma). In contrast, attentional or intentional processes are believed to be captured by the exponential tail of the distribution (i.e., tau) (Abney, McBride, & Petrella, 2013; Balota & Spieler, 1999; Gordon & Carson, 1990; Hockley, 1984; Luce, 1986; Madden et al., 1999; Moret-Tatay et al., 2016; Rotello & Zeng, 2008).
This general consensus has arisen in part because the slowest RTs are the most strongly correlated with higher order processes including intellectual ability and executive function, a phenomenon known as the “worst performance rule” (Larson & Alderton, 1990). Tau (but not mu or sigma) has been found to predict performance on WM and other EF tasks in both children and adults (Borella, de Ribaupierre, Cornoldi, & Chicherio, 2013; Karalunas & Huang-Pollock, 2013; Schmiedek, Oberauer, Wilhelm, Suss, & Wittmann, 2007; Unsworth, Redick, Lakey, & Young, 2010). And, even in simple perceptual decision tasks, compared to their same-aged peers, children with ADHD have larger tau, but similar mu and sigma (Borella et al., 2013; Epstein et al., 2011; Karalunas & Huang-Pollock, 2013; Leth-Steensen, Elbaz, & Douglas, 2000).
Our own more specific interpretation of tau is that it represents the speed at which information accumulates during the decisional sub-component that comprises a RT. Such an interpretation is heavily influenced by the diffusion model; a mathematical model of choice reaction time performance for which the parameters have an extensive and strong history of empirical validation (Ratcliff, 2002, Ratcliff, 2014; Voss, Rothermund, & Voss, 2004). Drift rate is a diffusion model parameter that specifically indexes the rate of information accumulation during a decisional process, and tau is substantively and negatively correlated with drift (Karalunas & Huang-Pollock, 2013), though the relationship is not 1:1 (Matzke & Wagenmakers, 2009).
What is clear is that the trials that form the exponential tail of the RT distribution (tau) represent some important function, which might help explain individual differences in higher order cognitive processes. That being said, roughly 30–50% of children with ADHD also experience fine motor coordination problems (Fliers et al., 2008; Kadesjö & Gillberg, 1998), which may very well influence their performance on both the speeded RT tasks and the EF tasks that commonly rely upon RT measurements. It remains a challenge to fully disambiguate encoding from motor preparation time (both are captured within mu/sigma when examining a standard motor-reaction time task, and even the diffusion model captures encoding and motor processing in a single parameter, Ter). It could be that the influence of poor motor planning is offset by adequate perceptual processing/encoding speed. Regardless, the degree to which each of these processes accounts for unique variance in the relationship between ADHD and EF deficits is not yet well understood.
It is, however, possible to isolate encoding from motor preparation by use of an “inspection time” task. During a visual inspection time task, participants are briefly shown two visual stimuli before a masking image is overlaid. Participants are then asked to make a simple forced-choice decision (e.g. which of the two lines was longer?). An adaptive staircase method (Findlay, 1978; Taylor & Creelman, 1967) is used to vary the amount of time stimuli are presented for each trial. The final output variable, inspection time (IT), refers to the shortest amount of time that a stimulus needs to be presented for an individual to achieve a pre-determined level of accuracy (e.g. 80%) (Irwin, 1984). In this way, IT represents the amount of time an individual needs to view a stimulus to perform at a specified level of accuracy, and so provides an index of perceptual encoding speed that is independent of motor response preparation (Deary & Stough, 1996).
Several studies have found negative associations between visual and auditory IT tasks and general intelligence (g) (Grudnik & Kranzler, 2001; Kranzler & Jensen, 1989) as well as with standard neuropsychological measures of executive function (Edmonds et al., 2008; Hutton, Wilding, & Hudson, 1997). Only a few studies have examined the association between IT and symptoms of ADHD, with inconsistent results. One study examining children with either Developmental Coordination Disorder (DCD) or ADHD found that children with ADHD did not differ from typically developing peers (Piek, Dyek, Francis, & Conwell, 2007), though another found a correlation between slower IT and a higher number of ADHD symptoms in their control group (Shank, Kaufman, Leffard, & Warschausky, 2010). Thus, there is some evidence that perceptual encoding speed may be slower among children with ADHD, and that slower perceptual encoding speed predicts executive functioning and other higher-order processes. However, what remains to be seen is the degree to which motor, encoding, and decision speed may separately or jointly explain EF deficits in children with ADHD.
In the current study, we utilize a visual inspection time task, and combine this with an ex-Gaussian decomposition of the RT data of the choice response times from the same task, in children with and without ADHD.
We hypothesize that compared to their same aged peers, children with ADHD will demonstrate slower perceptual encoding speed (indexed by IT), slower motor speed (indexed by mu/sigma), and longer decision-making times (indexed by tau). However, when all variables are in the model, we predict that neither motor nor encoding speed will meaningfully predict EF, and that the RT-EF association will be accounted for primarily by the speed of the decisional process. And finally, we predict that of those parameters, only tau will mediate the ADHD—EF relationship.
Section snippets
Participants
A total of 266 boys and girls between the ages of 8 and 12 years old, with and without ADHD, were community recruited from Centre, York, and Harrisburg counties of Pennsylvania to participate in a larger study on attention and learning at The Pennsylvania State University. Children were excluded based on parent report of neurological or sensorimotor disorders, pervasive developmental disorder that would preclude full participation, as well as use of non-stimulant psychoactive medications (e.g.,
Results
Table 1 provides descriptive statistics. Validating the diagnostic groups, children with ADHD had more inattentive, F (1,264) = 1727.05, p < .001, η 2 = 0.867, and hyperactive symptoms, F (1,264) = 266.30, p < .001, η 2 = 0.502, than typically developing controls. There were no group differences in age, F (1,264) = 1.63, p = .20, η 2 = 0.006, or FSIQ, F (1,264) = 0.029, p = .87, η 2 < 0.001. Children with ADHD had a smaller WM capacity (Symmetry Span: F (1,188) = 12.184, p = .001, η2 = 0.061;
Discussion
Slower mean reaction time (Barrouillet, Bernardin, & Camos, 2004; Bayliss et al., 2005; Kail, 1992) and greater intraindividual variability in reaction time (i.e. SDRT and tau) (Kofler et al., 2013; Mella, Fagot, Lecerf, & de Ribaupierre, 2015; Schmiedek et al., 2007), have long been found to be negatively associated with working memory capacity and other executive functions. However, reliance upon mean/SD of reaction time as a proxy for cognitive processing speed collapses across the several
Conclusion
Slower and more variable reaction time is frequently identified as a prominent cognitive signature in Attention Deficit Hyperactivity Disorder (ADHD). However, the nearly exclusive use of fine motor reaction time to index processing speed limits the ability to identify the source of these deficits. The use of an inspection time task in conjunction with an ex-Gaussian decomposition of the RT data allowed for a more accurate and comprehensive description of speeded performance in children with
Declarations of interest
None.
Funding
This work was supported in part by National Institute of Mental Health Grant R01 MH084947 to Cynthia Huang-Pollock. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institute of Mental Health or the National Institutes of Health.
References (81)
Attention-deficit/hyperactivity disorder: A selective overview
Biological Psychiatry
(2005)A review of the worst performance rule: Evidence, theory, and alternative hypotheses
Intelligence
(2003)- et al.
Relations between speed, working memory, and intelligence from preschool to adulthood: Structural equation modeling of 14 studies
Intelligence
(2014) - et al.
Inspection time and cognitive abilities in twins aged 7 to 17 years: Age-related changes, heritability and genetic covariance
Intelligence
(2008) - et al.
Inspection time: A flash mask to reduce apparent movement effects
Personality and Individual Differences
(1993) - et al.
Generalization of the worst performance rule across the lifespan
Intelligence
(2014) - et al.
The basis for choice reaction-time slowing in Alzheimers disease
Brain and Cognition
(1990) - et al.
Meta-analysis of the relationship between intelligence and inspection time
Intelligence
(2001) - et al.
The role of attention in the relationship between inspection time and IQ in children
Intelligence
(1997) Inspection time and its relation to intelligence
Intelligence
(1984)
Reaction time variability in ADHD: A meta-analytic review of 319 studies
Clinical Psychology Review
Inspection time and intelligence - a META-analysis
Intelligence
Reaction time variability and intelligence: A “worst performance” analysis of individual differences
Intelligence
Perceptual speed does not cause intelligence, and intelligence does not cause perceptual speed
Biological Psychology
On the heritability of inspection time and its covariance with IQ: A twin study
Intelligence
A diffusion model explanation of the worst performance rule for reaction time and IQ
Intelligence
NIMH diagnostic interview schedule for children version IV (NIMH DISC-IV)
Journal of the American Academy of Child & Adolescent Psychiatry
Psychology and neurobiology of simple decisions
Trends in Neurosciences
Cognitive correlates of developing intelligence: The contribution of working memory, processing speed and attention
Intelligence
Lapses in sustained attention and their relation to executive control and fluid abilities: An individual differences investigation
Intelligence
Interactive effects in transfer-appropriate processing for event-based prospective memory: The roles of effort, ongoing task, and PM cue properties
Memory & Cognition
Response time intra-subject variability: Commonalities between children with autism spectrum disorders and children with ADHD
European Child & Adolescent Psychiatry
Word frequency, repetition, and lexicality effects in word recognition tasks: Beyond measures of central tendency
Journal of Experimental Psychology. General
Time constraints and resource sharing in adults' working memory spans
Journal of Experimental Psychology: General
Mapping the developmental constraints on working memory span performance
Developmental Psychology
Developmental differences in intra-individual variability in children with ADHD and ASD
Journal of Child Psychology and Psychiatry
Impact of executive function deficits and attention-deficit/hyperactivity disorder (ADHD) on academic outcomes in children
Journal of Consulting and Clinical Psychology
Beyond interference control impairment in ADHD: Evidence from increased intraindividual variability in the color-stroop test
Child Neuropsychology
Intra-individual variability among children with ADHD on a working memory task: An ex-Gaussian approach
Child Neuropsychology
Conners' rating scales—3 technical manual
Working memory span tasks: A methodological review and user's guide
Psychonomic Bulletin & Review
Children's working-memory processes: A response-timing analysis
Journal of Experimental Psychology: General
Fitting the ex-Gaussian equation to reaction time distributions
Behavior Research Methods, Instruments, & Computers
Parent ratings of attention-deficit/hyperactivity disorder symptoms: Factor structure and normative data
Journal of Psychopathology and Behavioral Assessment
Preschool children with attention-deficit/hyperactivity disorder: Impairments in behavioral, social, and school functioning
Journal of the American Academy of Child and Adolescent Psychiatry
Evidence for higher reaction time variability for children with ADHD on a range of cognitive tasks including reward and event rate manipulations
Neuropsychology
Trajectories of pure and co-occurring internalizing and externalizing problems from age 2 to age 12: Findings from the National Institute of Child Health and Human Development study of early child care
Developmental Psychology
Motor coordination problems in children and adolescents with ADHD rated by parents and teachers: Effects of age and gender
Journal of Neural Transmisssion
Cited by (15)
Alterations in attentional processing in youth with misophonia: A phenotypical cross-comparison with anxiety patients
2024, Journal of Affective DisordersUnraveling the cognitive correlates of heart rate variability with the drift diffusion model
2022, International Journal of PsychophysiologyCitation Excerpt :The current tau-vmHRV associations only slightly decreased in size when adjusting for covariates, which were not significantly different in size than the analogous zero-order correlations. Also as predicted, individuals with high resting vmHRV had faster drift rates, the DDM parameter reflecting goal-directed cognitive processes (Schmiedek et al., 2007; Galloway-Long and Huang-Pollock, 2018). Importantly, this effect was limited to incongruent trials (i.e., when cognitive inhibition was needed), thus supporting the link between drift rate and cognitive inhibition as opposed to other cognitive control subcomponents such as shifting and working memory.
Interplay between state anxiety, heart rate variability, and cognition: An ex-Gaussian analysis of response times
2021, International Journal of PsychophysiologyExamining reaction time variability on the stop-signal task in the ABCD study
2023, Journal of the International Neuropsychological Society