Which Measures From a Sustained Attention Task Best Predict ADHD Group Membership?

Abstract

Difficulty with sustaining attention to a task is a hallmark of ADHD. It would be useful to know which measures of sustained attention best predict a diagnosis of ADHD. Participants were 129 children with a diagnosis of ADHD and 129 matched controls who completed the fixed Sustained Attention to Response Task (SART). The number of commission and omission errors, standard deviation of response time (SDRT), tau, fast and slow frequency variability, d-prime, and mu were able to successfully classify children with and without ADHD. The mean response time, criterion, and sigma were not able to classify participants. The best classifiers were d-prime (0.75 Area Under the Receiver Operated Characteristic), tau (.74), SDRT (0.74), omission errors (0.72), commission errors (0.71), and SFAUS (0.70). This list of the best classifier measures derived from the SART may prove useful for the planning of future studies.

Keywords

ADHD response time SART sustained attention

ADHD is a common neurodevelopmental disorder, affecting around 5.3% of school-age children (Polanczyk et al., 2007) and defined by the three symptom clusters of inattention, hyperactivity, and impulsivity (American Psychiatric Association, 2013). According to the DSM 5 diagnostic criteria, the inattention symptom cluster is characterized predominantly by poorly functioning sustained attention. Sustained attention is the ability to concentrate on a task over a period without becoming distracted. Sustained attention and distractibility feature in six of the nine DSM 5 inattentive symptom items, with the remaining three items describing organizational skills and forgetfulness. Sustained attention may therefore be considered a behavioral hallmark of the attention deficit in ADHD. Behavioral assessments are essential in clinical practice and form the basis of diagnostic evaluations for ADHD. Nevertheless, an objective cognitive test that measures sustained attention quantitatively and independently would be useful for furthering our understanding of the disorder, which can then inform cognitive, therapy response, and aetiological research.

The Sustained Attention to Response Task (SART) has been used in many studies to objectively measure sustained attention. In the fixed version of the SART, digit stimuli are presented serially and rapidly on a computer screen. It has been crafted so that the participant responds with a key press to all digits except the specified No-Go digit. The participant is required to withhold their response to the No-Go digit, which appears every nine digits. Thus, the SART provides a large amount of response time data produced under non-arousing conditions (Robertson et al., 1997). This is particularly helpful as increased response time variability is a well-documented characteristic associated with ADHD (Bellgrove et al., 2005; Kofler et al., 2013). The response time and error data from the SART yields many different measures of sustained attention performance. When planning studies, it would be useful for researchers to know which of the SART outcome measures of sustained attention are most predictive of a diagnosis of ADHD for both theoretical and practical reasons. Theoretically, knowledge of which measures best predict ADHD group membership allows for a better characterization of the sustained attention deficit in ADHD, which can help inform research. Practically, knowing which outcome measures are most predictive allows for a reduction in the number of outcome measures in statistical models and leeway for the addition of other measures of interest. This paper analyzed the SART data from a well described sample of children with the aim to identify the measures that best classified the children into those with and without an ADHD diagnosis.

One of the earliest studies investigating sustained attention in a group of hyperactive children used the Continuous Performance Task (CPT) and found that these children missed more targets (omission errors) than a control group and responded more often to non-targets (commission errors) at the slower versus faster presentation rate of stimuli (Sykes et al., 1971). A 1996 meta-analysis of 26 CPT studies found that a diagnosis of ADHD was associated with increased errors of commission (effect size Cohen’s d 0.73) and omission (0.67), and less target sensitivity (d-prime from signal detection theory), but response bias (criterion) remained similar between the two groups (Losier et al., 1996). A 2005 meta-analysis of 83 studies measuring executive functioning in ADHD found that in 77% of studies children with ADHD made more omission errors on CPTs than typically developing children (effect size Cohen’s d 0.64), and in 82% of studies children with ADHD were slower to inhibit responses on the Stop-Signal Reaction Time (0.61) (Willcutt et al., 2005). In 2012 Huang-Pollock and colleagues meta-analyzed 47 studies and found that children with ADHD made more errors of commission (effect size Cohen’s d 0.98) and omission (1.34) and were more variable (SDRT 0.93) in responding than matched-controls (Huang-Pollock et al., 2012). They also found a moderate effect size for a slower reaction time in the ADHD group (0.61). The results of these meta-analyses indicate that the errors of commission and omission are particularly instructive of difficulties with sustained attention in children with a diagnosis of ADHD. It is noted, however, that response time variability was not analyzed in several of these meta-analyses.

Error measures from CPTs may not be the best measures of performance because CPTs use clearly differentiated stimuli such as X and O, rather than just noticeable difference stimuli such as shades of gray. This may result in a reduction in reliability because differentiated stimuli are easier to detect and lead to highly accurate performance (ceiling effects) (Huang-Pollock et al., 2012). Few previous meta-analyses had reported response time as an index of performance (Huang-Pollock et al., 2012), but response time measures are extremely useful as they provide a continuum of performance. Response time, and particularly response time variability performance, may provide a more nuanced measure of performance than errors. There are a range of response time variability measures available to researchers that draw upon different analytical models (Luce, 1986), including the ex-Gaussian model and the Fast Fourier Transform (FFT) (Castellanos et al., 2005; Geurts et al., 2008; Johnson et al., 2007). The ex-Gaussian model decomposes response time distributions into the average (mu) and variability (sigma) of a Gaussian distribution component and the degree of rightward skew captured in an exponential distribution component (tau) (Leth-Steensen et al., 2000). Tau is thought to reflect occasional lapses in attention (Leth-Steensen et al., 2000). The FFT quantifies the magnitude of periodic patterns at different frequencies in a series of response times (Castellanos et al., 2005). By separately measuring spectral activity in low- and high-frequency ranges, slow and fast changes in speed of response can be measured, providing a view on sustained attention performance resolved according to timescale (Johnson et al., 2007). Two more recent meta-analyses focused on response time variability. Kofler and colleagues reported that children with ADHD were more variable in responding (including ex-Gaussian, FFT, and SDRT measures) (Hedges’ g 0.76), but were similar to typically developing children on speed of response (−0.19) (mu and mean response time, MRT), after accounting for task demands, sampling and measurement errors, publication bias, and ADHD subtype (Kofler et al., 2013). For the specific ex-Gaussian measures, Kofler et al., reported a large effect size for tau (Hedges’ g 0.99) and a moderate effect size for sigma (Hedges’ g 0.39). For the FFT measures, Kofler et al. reported medium effect sizes for the entire frequency band (Hedges’ g 0.63) and specific frequencies including Slow-4 (0.027–0.074 Hz, or 14–37 seconds) (Hedges’ g 0.59), and Slow-3 (0.074–0.2 Hz, 5–14 seconds) (Hedges’ g 0.35). Karalunas and colleagues reported that children with ADHD performed the cognitive tasks with greater response time variability at both low (i.e., <0.10 Hz, 10 seconds) (Hedge’s g 0.39) and high (0.36) frequency bands, using the FFT methodology (Karalunas et al., 2013). More recently, a review of 34 meta-analyses investigating neurocognitive performance differentiating individuals with ADHD from matched controls noted that response time variability, summarising data from several measures of variability, had the largest effect size, (0.66), when weighted by the number of studies aggregated (Pievsky & McGrath, 2018). This previous research indicates that the response time variability measures have moderate to large effect sizes for detecting ADHD versus control group differences. These meta-analyses are only able to calculate effect sizes based on measures provided frequently in the literature. An array of RTV measures is available, each of which measure subtly different forms of variation from the average speed of response and that may reflect different cognitive mechanisms. What is needed is an investigation of the different response time variability measures from the same task, from the same set of participants. An examination of error data and signal detection alongside response time variability would greatly increase our understanding of the best measures to detect group differences and to differentiate individuals with and without ADHD.

Recently, Brunkhorst-Kanaan et al. (2020) computed the Area Under the Receiver Operated Characteristic (AUROC) to test whether ex-Gaussian parameters on a CPT-like task, the Quantified Behavioral Test (QbTest), could classify adults presenting to an ADHD outpatient clinic who received a diagnosis of ADHD (n = 94) from those who underwent an evaluation and did not receive an ADHD diagnosis (n = 20) (Brunkhorst-Kanaan et al., 2020). Within this group of disconfirmed ADHD participants, eight had no diagnosis, nine suffered from depression, two from a substance use disorder, one bipolar disorder, and three from other non-named disorders. In the AUROC measure, 0.5 represents change discrimination performance, and the further the AUROC deviates from 0.5, the stronger the classification of the two groups, representing a measure of effect size. The results indicated small AUROC for all three parameters (mu: .54, sigma: .60, and tau: .57) and the authors suggested that the ex-Gaussian analysis of response time did not differentiate ADHD from other patient groups. No other forms of response time measure were analyzed. The study demonstrates that the AUROC analysis is a useful tool for determining whether a particular response time measure is effective in differentiating two groups.

Some authors have argued that for a sustained attention deficit to be concluded, time-on-task effects should be demonstrated, for example (Huang-Pollock et al., 2006; Parasuraman, 1979). Under this definition, if ADHD is associated with a sustained attention deficit then performance would decline more steeply over time than for the control group. This has been documented in some studies (Heinrich et al., 2001; Huang-Pollock et al., 2006; Johnson et al., 2007; Merrill et al., 2021), but not others (Rapport et al., 2009; van der Meere et al., 1991). No clear evidence was found for a time-on-task effect in a 1993 narrative review (Corkum & Siegel, 1993). Time-on-task effects were not evaluated in the 1996 or 2005 meta-analyses (Losier et al., 1996; Willcutt et al., 2005). In the 2012 meta-analysis, time on task group difference comparisons had moderate (omission errors, effect size Cohen’s d 0.54) to small effect sizes (commission errors 0.24, RT 0.27, SDRT 0.22) (Huang-Pollock et al., 2012). It is noted that sustained attention deficits may be measurable from early in the task. Indeed, under the DSM 5 symptom description, even starting an onerous task may be difficult for people with ADHD. In this situation, if ADHD is associated with sustained attention difficulties, a deficit in performance may be shown from the very start of the task and therefore a decrement in performance over time may not be apparent. It is also noted that participants can perform vigilance and sustained attention tasks with no decrement over time under certain circumstances, including low compared with high event rates and lower memory loading (Parasuraman, 1979), higher participant motivation (Corkum & Siegel, 1993), with greater interest in the task and with moderate arousal levels (Hancock, 2017). These findings indicate that decline in performance over time may not always be a characteristic feature of sustained attention, and there may not always be a steeper decline in performance in those with deficits in sustained attention compared to those without such deficits. Nevertheless, time on task effects should be examined when investigating sustained attention deficits in ADHD (Christakou et al., 2013; Norman et al., 2017).

We narrowed our focus here on a very simple task of sustained attention. Our aim was to examine which variables were useful in differentiating children with and without ADHD. These variables could then potentially be used for diagnostic purposes and to more clearly establish the characteristics of sustained attention control associated with ADHD. We investigated if on average, children with a diagnosis of ADHD showed difficulties with sustained attention and which measures of sustained attention classified those children with ADHD against a sample of age- and sex-matched typically developing children using an AUROC analysis. Hypothesis one was that on all measures, bar the mean response time and mu, the ADHD group would perform significantly more poorly on the SART and therefore classification could be achieved. Hypothesis two was that the errors of commission and omission, and tau would classify individuals based on their group membership.

Method

Participants

This research combines published (Bellgrove et al., 2005; Johnson et al., 2007; Johnson et al., 2008; Lewis et al., 2017) and unpublished data of 171 children with a diagnosis of ADHD and 332 typically developing children, totaling 503 participants (see Table 1). Participants who met at least one of the following criteria were excluded: missing data for age (n = 6), no IQ estimate available (n = 11), a Conners’ ADHD Index lower than 60 for those clinically diagnosed with ADHD (n = 2), higher than 60 for the typically developing participants (n = 16), or no Conners’ data available (n = 75), more than 60 omission errors made on the SART (n = 8) or commission errors made to all of the No-Go targets during the first or second halves of the SART (n = 3), both of which indicate that the participant did not understand the task instructions. After exclusions the sample included 151 children with ADHD and 231 typically developing children. The intelligence quotient was estimated using one of the WISC-III (Wechsler, 1992), the WISC-IV (Wechsler, 2004), or the WASI (Wechsler, 2011). With substantially more typically developing children in the dataset an age- and sex-matched sub-sample was created, in which a participant with ADHD was selected, and a typically developing child was randomly matched according to age and sex. This resulted in 129 children in each group.

Table 1.

Participant Information.

Group	ADHD	Typically developing controls	All
Number	129	129	258
Sex (male/female)	110/19	110/19	220/38
Age, mean (SD) range (years)	10.3 (2.8) 5.0–15.8	10.0 (2.6) 6.0–17.0	10.2 (2.7) 5.0–17.0
IQ, mean (SD) range	94.5 (14.8) 70–130	104.9 (13.7) 75–135	99.7 (15.2) 70–135
Combined-type (count, % of group)	113 (88%)
Hyperactive/impulsive (count, % of group)	6 (5%)
Primarily inattentive (count, % of group)	10 (8%)
Conners DSM IV inattentive subscale (mean, SD, range)	76.9 (6.9)	46.4 (5.5)

The children with ADHD were recruited in the Republic of Ireland and Northern Ireland and were referred to the study through consultant psychiatrists or support groups. Diagnosis was confirmed by psychiatrists using the parent form of the Child and Adolescent Psychiatric Assessment (CAPA) (Angold et al., 1995) or the Parental Account of Childhood Symptoms (PACS) (Taylor et al., 1986) using DSM-IV diagnosis (American Psychiatric Association, 1994). Parents and caregivers completed the Conners’ Parent Rating Scale—Revised: Long Version (Conners, 1997) or the Conners’ 3 (Conners, 2014). Any stimulant medication was withdrawn for at least 24 hours prior to testing. The typically developing children were recruited through schools in Dublin, Belfast, and Melbourne. Consent was obtained from parents and caregivers of all children, under the approval of local ethics committees in accordance with the Declaration of Helsinki.

Apparatus and Procedure

The Fixed Sequence Sustained to Attention to Response Task (SART) (Manly et al., 2003). A series of digits (1–9) were presented to the participant on a laptop computer screen using E-Prime in a repeating, ascending order from 1 to 9. The one to nine cycle was repeated 25 times, totaling 225 trials. Each trial consisted of a single digit presented for 313 ms, followed by a backwards mask for 125 ms to eliminate the digit appearing as a visual aftereffect, a response cue presented for 63 ms, a second mask presented for 375 ms, and lastly a fixation cross presented for 563 ms. The total inter-stimulus interval was 1,439 ms. The task took approximately 5.5 minutes. Participants were asked to respond, with a button press, to every digit except “3,” the No-Go stimulus. They were asked to respond when the response cue appeared on screen 125 ms after the digit disappeared, to limit impulsive responses and to reduce speed-accuracy trade-offs.

Response Parameters

Errors

For each participant, counts of omission—missed responses to the Go trials, and commission errors—responses to the No-Go trials were calculated. D-prime and criterion were computed using Signal Detection Theory. Using the method from Johnson et al. (2008), the digit “3,” the No-Go digit, was treated as the target in this analysis (Johnson et al., 2008). A hit was a correct inhibition to a No-Go target, a miss was a commission error, a false alarm was an omission error, and a correct rejection was a correct Go response. The measure d-prime indicates how well the participant can discriminate between targets and non-targets. A z transform of the false alarm rate was subtracted from the z transform of the hit rate to compute d-prime; d-prime = z(H) − z(F) (Stanislaw & Todorov, 1999). The measure criterion indicates the level of bias toward assuming a stimulus is a target. To compute criterion, −0.5 was multiplied by the sum of the z transforms of the hit and false alarm rates, (−(z(H) + z(F))/2) (Macmillan & Creelman, 1991). An adjustment for participants with no errors was made by assuming they made 0.5 errors (Kelly et al., 2009).

Mean response time and standard deviation of response time

Trials with commission errors, responses shorter than 100 ms, and No-Go trials were removed. Following this, the mean of RT (MRT) and standard deviation of RT (SDRT) were computed.

Ex-Gaussian

SART response time data have been more recently analyzed using the Ex-Gaussian method to produce a better model fit for response times (Johnson et al., 2015). The Ex-Gaussian method produces estimates of the average response time (mu), the variation in response time in the data fitting the normal distribution (sigma), and the extremely slow response times that better fit the exponential distribution (tau) (Geurts et al., 2008; Karalunas et al., 2014; Kofler et al., 2013; Leth-Steensen et al., 2000). Three parameters, mu, sigma, tau, were computed by applying the ex-Gaussian model based on the approach by Lacouture and Cousineau (2008).

FFT

SART response time data have also been analyzed for slow- and fast-repeating patterns, using a fast Fourier Transform (FFT) (Castellanos et al., 2005), indicating changes in response time that occur over both a slow timescale (over the 5.5 minutes) and a fast moment-to-moment timescale within a one to nine cycle, occurring anytime within the 5.5 minute task (Johnson et al., 2007). Missing data, due to omission errors, very fast responses (<100 ms), and trials in which commission errors were made, were replaced using linear interpolation. Welch’s averaged, modified periodogram method was used to perform the FFT. Trials were divided into seven segments of 75 trials of the SART, with a 50-trial overlap. Each segment was detrended, Hamming-windowed, and zero padded to 450 data points. The FFT was then applied to each segment, and subsequently the segments were averaged. Any segment of 75 data points where there were five or more consecutive interpolated trials was excluded in the FFT. If more than three segments were removed for the participant, FFT was not applied and treated as missing values.

The frequency range was divided into fast and slow frequency bands based on the peak at 0.0772 Hz (Johnson et al., 2007). This peak frequency 0.0772 Hz was chosen as it is the reciprocal of one cycle of the one to nine digit presentation (a SART cycle). The Fast Frequency Area Under the Spectra (FFAUS) encompassed all sources of variability faster than once per SART cycle, the area under the curve to the right of the peak at 0.0722 Hz. The Slow Frequency Area Under the Spectra (SFAUS) encompassed all sources of variability slower than once per SART cycle: the area under the curve to the left of the peak at 0.0772 Hz.

Half by Half Analysis

Absolute change scores

To measure the time on task effect, the first and second half of the trials were compared. The first SART cycle was removed, and the remaining trials were divided into the first half (10th to 117th trials) and the second half (118th to 225th trials). MRT, SDRT, number of omission and commission errors, d-prime, criterion, mu, sigma, and tau were computed for the first and second half. The first and second half of the FFAUS and SFAUS were computed from the first three segments and the last three segments out of the seven total segments. To compute absolute change scores from the first to the second half, each value in the first half was subtracted from the second half (second–first). A positive value indicates that there was an increase in the parameter value from the first to the second half of the SART.

Statistical Analysis

Statistical analysis was performed in R. The dependent variable was Group (ADHD vs. Typically Developing Control), a binary variable. The independent variables were the count of omission and commission errors, d-prime, criterion, MRT, SDRT, mu, sigma, tau, FFAUS, and SFAUS. Logistic regression was fitted to classify participants as ADHD and Control with one independent variable per analysis. Based on the logistic regression, each participant was predicted as belonging either to the ADHD or typically developing control groups. The performance of each model was tested by the Area Under the Curve (AUROC) from the Receiver Operating Characteristic (ROC) analysis, indicating how much each independent variable was helpful in classifying each participant to a group. An AUROC at 0.5 indicated that the performance of the model was equal to randomly classifying participants to the groups. A value greater than 0.5 indicated that the model classified the grouping more accurately, such that the independent variable contained information to discriminate the two groups. The AUROC is considered a form of effect size—of how well each parameter differentiates the two groups (Rice & Harris, 2005). AUROC 0.639 is equivalent to Cohen’s d = 0.5 (medium effect size), AUROC 0.714 is equivalent to d = 0.8 (large effect size), and AUROC 0.745 is equivalent to d = 0.93. For each independent variable, AUROC and its 95% Confidence Interval (CI) were computed using “pROC” package in R (Robin et al., 2011, 2021). The AUROC analysis was performed on the full SART and absolute change score data. I assume the correlation matrix will be ok and does not need to be proofed?

Results

Performance on the SART as described by the number of omission and commission errors, d-prime, SDRT, mu, tau, FFAUS, and SFAUS, was used to successfully classify the participants (see Figure 1 and Table 2). By examining the mean of each parameter for each group, omission, and commission errors, SDRT, tau, FFAUS, and SFAUS were higher in the children with ADHD compared with the controls. In contrast, d-prime and mu were lower in the ADHD compared with the control group. Among these parameters, omission and commission errors, d-prime, SDRT, tau, and SFAUS reached AUROC higher than 0.7 indicating better performance in classifying the two groups. The models with MRT, criterion, and sigma were operating at random in classifying the participants in the current dataset.

Table 2.

Area Under Receiver Operated Characteristic for Each SART Measure.

SART measure	Group	Mean and standard deviation	Area under the ROC and 95% confidence intervals
Mean response time (ms)	ADHDControl	M = 497, SD = 109M = 513, SD = 113	AUROC = 0.537, CI = [0.467, 0.608]
Standard deviation of response time (ms)	ADHDControl	M = 218, SD = 63M = 165, SD = 56	AUROC = 0.740, CI = [0.679, 0.801]*
Omission errors (count)	ADHDControl	M = 15.3, SD = 13.3M = 6.6, SD = 8.2	AUROC = 0.722, CI = [0.660, 0.784]*
Commission errors (count)	ADHDControl	M = 8.9, SD = 5.4M = 5.2, SD = 3.7	AUROC = 0.711, CI = [0.649, 0.773]*
D-prime	ADHDControl	M = 2.06, SD = 1.09M = 3.01, SD = 0.86	AUROC = 0.749, CI = [0.689, 0.809]*
Criterion	ADHDControl	M = 0.60, SD = 0.30M = 0.60, SD = 0.32	AUROC = 0.513, CI = [0.443, 0.584]
Fast frequency area under the spectra	ADHDControl	M = 482, SD = 264M = 339, SD = 223	AUROC = 0.680, CI = [0.615, 0.745]*
Slow frequency area under the spectra	ADHDControl	M = 2029, SD = 1,374M = 1267, SD = 1,085	AUROC = 0.701, CI = [0.637, 0.765]*
Mu	ADHDControl	M = 331, SD = 159M = 423, SD = 153	AUROC = 0.675, CI = [0.610, 0.741]*
Sigma	ADHDControl	M = 109, SD = 78M = 115, SD = 59	AUROC = 0.567, CI = [0.496, 0.638]
Tau	ADHDControl	M = 166, SD = 95M = 90, SD = 73	AUROC = 0.737, CI = [0.676, 0.798]*

Note: * denotes the SART predictors that significantly differentiated the participants by Group, because the AUROC confidence intervals did not contain the value of 0.50.

Absolute change scores were summarized in Table 3. There were no differences between the groups on how their behavioral performance changed from the first to the second halves of the SART in any of the response parameters (see Figure 2).

Table 3.

Absolute Change Score of Each SART Measure.

SART measure	Group	Mean and Standard Deviation	Area Under the ROC and 95% confidence intervals
Mean response time (ms)	ADHDControl	M = 14, SD = 77M = −2.1, SD = 67	AUROC = 0.566, CI = [0.496, 0.636]
Standard deviation of response time (ms)	ADHDControl	M = 23, SD = 51M = 12, SD = 38	AUROC = 0.562, CI = [0.491, 0.632]
Omission errors (count)	ADHDControl	M = 1.2, SD = 6.4M = 1.2, SD = 4.1	AUROC = 0.516, CI = [0.443, 0.589]
Commission errors (count)	ADHDControl	M = 0.61, SD = 2.3M = 0.50, SD = 1.8	AUROC = 0.514, CI = [0.443, 0.584]
D-prime	ADHDControl	M = −0.25, SD = 0.78M = −0.27, SD = 0.68	AUROC = 0.498, CI = [0.426, 0.569]
Criterion	ADHDControl	M = 0.04, SD = 0.35M = 0.01, SD = 0.29	AUROC = 0.536, CI = [0.465, 0.606]
Fast frequency area under the spectra	ADHDControl	M = 60, SD = 276M = 39, SD = 196	AUROC = 0.495, CI = [0.424, 0.566]
Slow frequency area under the spectra	ADHDControl	M = 280, SD = 1,810M = 257, SD = 1,054	AUROC = 0.528, CI = [0.456, 0.600]
Mu	ADHDControl	M = −10, SD = 123M = −19, SD = 106	AUROC = 0.527, CI = [0.456, 0.597]
Sigma	ADHDControl	M = 9, SD = 66M = 4, SD = 43	AUROC = 0.538, CI = [0.467, 0.609]
Tau	ADHDControl	M = 23, SD = 91M = 17, SD = 74	AUROC = 0.514, CI = [0.442, 0.585]

Discussion

Almost all the performance measures of the SART helped to classify children with and without ADHD, except for the MRT, criterion, and sigma. Children with ADHD made more errors of omission and commission, and subsequently d-prime, indicating they had more difficulty correctly identifying and responding to Go versus No-Go trials. The response time variability measures of SDRT, tau, FFAUS, and SFAUS were all useful in classifying the groups. Hypothesis one was therefore almost fully supported. The half by half analysis revealed that performance changes from the first to second half of the SART did not differ between the two groups on any of the measures, suggesting that a group difference in time-on-task effect was not apparent in this whole-group based analysis. These findings support the behavioral identification of difficulties children with ADHD have in starting, continuing with, and completing tasks that require maintenance of attention to a repetitive, non-stimulating task. The best classifier measures for group membership were d-prime (0.75), tau (.74), SDRT (0.74), omission errors (0.72), commission errors (0.71), and SFAUS (0.70), supporting hypothesis two. The SART is a short, easily administered tool that provides an objective measure of sustained attention performance in children and that can classify those with and without ADHD with a large effect size.

Children with ADHD had more difficulty in performing the task accurately compared with children without ADHD. An omission error is thought to occur with a lapse in attention while a commission error occurs when the primed response needs to be inhibited (O’Connell et al., 2009). Here children with ADHD performed the SART with more errors of both types, suggesting difficulties in maintaining attention and response inhibition. These findings are consistent with the previous meta-analytic findings focusing on error performance on sustained attention tasks (Huang-Pollock et al., 2012; Losier et al., 1996; Willcutt et al., 2005). Both error types, and d′, were able to discriminate children with and without ADHD. It is important to note that the bias to respond to a stimulus as being a target or not, the criterion measure, did not classify individuals into the two groups very well. Both errors of omission and commission are important measures to include in sustained attention studies.

The children with ADHD performed the SART with increased response time variability in all but one measure (Ex-Gaussian sigma). The SDRT can be thought of as an umbrella measure, with the other measures of sigma, tau, FFAUS, and SFAUS capturing different aspects of response time variability. This study showed that children with ADHD performed the task with larger SDRT compared with the control group, which is a consistent finding (Karalunas et al., 2013; Kofler et al., 2013). Additionally, both FFAUS and SFAUS from the FFT frequency analysis were useful in classifying the groups, with children with ADHD exhibiting larger moment-to-moment (FFAUS) and slow changes over time (SFAUS). Increased response time variability in ADHD was not specific to a particular frequency band, supporting the results of the previous meta-analysis (Karalunas et al., 2013). The similar effect sizes of the SFAUS (AUROC = 0.70) and FFAUS (AUROC = 0.68) imply that the slow-timescale variability and the moment-to-moment variability have a comparable discriminative value. It is still unclear how response time variability in the different frequency bands relates to cognitive processes. One argument is that FFAUS reflects sustained attention processes and SFAUS reflects arousal (Johnson et al., 2007). Assuming this is the case, it can be implied from the current results that deficits in both sustained attention and arousal contribute to increased response time variability in ADHD.

The ex-Gaussian analysis revealed that accounting for the skewness of the RT distribution provides another method to classify children with ADHD. Among the behavioral measures of variability in responding, only sigma was unhelpful in the classification process. The parameter tau reflects the skewness of the RT distribution, with more occurrences of infrequent extremely slow responses increasing the skew. Within the Ex-Gaussian set of analyses, increased response time variability in ADHD was explained by this more frequent occurrence of extremely slow responses rather than by variation around the mean response time (sigma). Previous studies have also reported greater tau but not greater sigma in groups of children with ADHD compared with matched control groups (Borella et al., 2013; Epstein et al., 2011; Feige et al., 2013; Leth-Steensen et al., 2000; Seymour et al., 2016; Thomson et al., 2020; Wolfers et al., 2015). Leth-Steensen et al. (2000) argued that lapses in attention lead to extremely slow responses and increased tau. These findings suggest that individuals with ADHD experience more lapses in attention during the task compared with individuals without ADHD. In the study by Brunkhorst-Kanaan and colleagues, the AUROCs for all three parameters were below 0.6 (Brunkhorst-Kanaan et al., 2020), whereas in the present study both mu and tau significantly classified participants in the two groups suggesting that the Ex-Gaussian analysis of SART data is a useful method. Some methodological differences may help explain these differences in results between the current study and the Brunkhorst-Kanaan study, including the task used and the participants chosen. The QbTest is a one-back task, in which the participant indicates whether each stimulus in a series matches the one presented just prior. This task invokes working memory whereas the SART is a simple Go/No-Go task designed to induce boredom. The Brunkhorst-Kanaan study involved adult participants referred for a formal diagnosis of ADHD, many of whom had co-morbidities. Sustained attention difficulties and increased RTV are not symptoms specific to ADHD but are noted in many psychiatric conditions. This may explain why Brunkhorst-Kanaan et al., did not find that the measures of response time variability discriminated their two groups, both of which comprised adults with confirmed psychiatric conditions. In contrast, the findings from the current study suggest that most response time variability measures successfully classified children with and without ADHD.

While it has been consistently reported that increased response time variability is associated with a diagnosis of ADHD, previous findings have been mixed regarding the speed of response. The current results confirmed that children with ADHD responded more quickly than children without ADHD when measured by mu, but not by MRT. One can view mu as a better parameter than the mean, as mu has had the skewed very slow responses removed, whereas these abnormally long responses remain in the mean calculation. Once these very slow responses were categorized into the tau calculation, the average speed of the ADHD group was faster than that of the control group. The same finding was reported by Lee and colleagues using the CPT (Lee et al., 2015). Other research has also reported contrasting findings between MRT and mu, where individuals with ADHD exhibited slower MRT and faster mu (Gu et al., 2013; Hervey et al., 2006; Hwang-Gu et al., 2019). Yet other studies have reported no group differences in mu, using a range of tasks including the N-back, flanker, oddball, and both simple and complex Go No-Go tasks (Feige et al., 2013; Seymour et al., 2016; Vainieri et al., 2009, 2022). The SART is very simple and highly predictable, leading to an increase in boredom. Higher impulsivity, one of the core behavioral symptoms of ADHD, might have led children with ADHD to respond sometimes more quickly. The degree to which the duration of the SART, which lasts for ~5 minutes, may be related to a differentiation in the mean RT and mu remains to be investigated. The current result might not be generalisable to other tasks, but in this short-lasting, fixed version of SART, faster responses were shown to be one of the characteristics of ADHD.

The half by half analysis revealed that there was no group difference in performance change across the two halves of the task. The meta-analysis by Huang-Pollock et al. (2012) showed that performance over time on CPTs only differed between individuals with and without ADHD on omission errors but not in the MRT, SDRT, or commission errors (Huang-Pollock et al., 2012). Comparing the effect sizes in that meta-analysis, the whole-task measures showed greater effect sizes relative to the half-by-half analysis. This implies that the time on task effect was not aligned with a diagnosis of ADHD, supporting the data from the current study. It is noted, however, that the SART is only 5 minutes long and may not be long enough to detect decrements in performance over time. Nevertheless, in this analysis performance difficulties were equally apparent in the first and second halves of the task for the children with the ADHD diagnosis.

In terms of limitations, most cognitive tasks have been designed to present stimuli with an unpredictable order. The fixed version of the SART is wholly predictable, engendering boredom, but making the results more difficult to generalize. Previous research has shown that there are slight performance differences in response to the fixed and random versions of the SART, where individuals tended to produce a greater number of commission errors in the random compared with fixed version of the SART (Machida et al., 2019). The neural signatures, measured using EEG, suggest that the fixed version is primarily testing the sustained attention system while the random version is pressing the response inhibition system (O’Connell et al., 2009). Future studies could incorporate the random version of the SART to examine how predictability may lead to different results, particularly with commission errors, clarifying which dimensions of the tasks affect behavioral parameters and classification potential.

Conclusion

Both the accuracy and response time variability measures were useful in classifying the children. The results imply that children with ADHD tended to respond more quickly and variably, produce more extremely slow responses, and make more errors. These behavioral findings indicate that children with ADHD experience more attentional lapses and difficulties with response inhibition, starting from the start and lasting through to the end of the task. The simple measures of the count of omission and commission errors alongside the SDRT provide substantial classification information without the need for more sophisticated modeling of RT data. This list of the best classifier measures may prove useful for the planning of future studies.

Supplemental Material

sj-docx-1-jad-10.1177_10870547221081266 – Supplemental material for Which Measures From a Sustained Attention Task Best Predict ADHD Group Membership?

Supplemental material, sj-docx-1-jad-10.1177_10870547221081266 for Which Measures From a Sustained Attention Task Best Predict ADHD Group Membership? by Keitaro Machida, Edwina Barry, Aisling Mulligan, Michael Gill, Ian H. Robertson, Frances C. Lewis, Benita Green, Simon P Kelly, Mark A. Bellgrove and Katherine A. Johnson in Journal of Attention Disorders

Footnotes

Acknowledgements

We thank all the children, parents, care-givers, and teachers for participating in this research.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The work herein was supported by grants from the Irish health Research Board, Science Foundation Ireland, Irish Higher Education Authority’s Programme for Research in Third Level Institutions, and the Australian National Health and Medical Research Council.

ORCID iDs

Keitaro Machida

Aisling Mulligan

Katherine A. Johnson

Data Availability Statement

This data set will be made available on Open Science Framework.

Supplemental Material

Supplemental material for this article is available online.

Author Biographies

Keitaro Machida is a graduate of the Melbourne School of Psychological Sciences at the University of Melbourne and currently works in Tokyo, Japan.

Edwina Barry is a consultant child psychiatrist and Clinical Senior Lecturer in Psychiatry in Trinity College Dublin, Ireland.

Aisling Mulligan is a consultant child psychiatrist and Associate Professor in Child and Adolescent Psychiatry in University College Dublin, Ireland.

Michael Gill is a consultant psychiatrist and Professor of Psychiatry in Trinity College Dublin, Ireland.

Ian H. Robertson is a clinical psychologist and Emeritus Professor of Psychology in Trinity College Dublin, Ireland.

Frances C. Lewis is a graduate of the Melbourne School of Psychological Sciences and currently works in behavioural therapy in Melbourne, Australia.

Benita Green is a clinical psychologist and Associate Lecturer in the Melbourne School of Psychological Sciences, the University of Melbourne, Australia.

Simon P. Kelly is a cognitive neuroscientist and Associate Professor in the School of Electrical and Electronic Engineering in University College Dublin, Ireland.

Mark A. Bellgrove is a cognitive neuroscientist and Professor of Psychology at Monash University, Melbourne, Australia.

Katherine A. Johnson is a developmental cognitive scientist and Associate Professor in the School of Psychological Sciences at the University of Melbourne, Australia.

References

American Psychiatric Association. (1994). Diagnostic and statistical manual of mental disorders (4th ed.). American Psychiatric Association.

American Psychiatric Association. (2013). Diagnostic and Statistical Manual of Mental Disorders (5th ed.). American Psychiatric Association.

Angold

Prendergast

Cox

Harrington

Simonoff

Rutter

(1995). The Child and Adolescent Psychiatric Assessment (CAPA). Psychological Medicine, 25, 739–753. https://doi.org/10.1017/S003329170003498X

Bellgrove

M. A.

Hawi

Kirley

Gill

Robertson

I. H.

(2005). Dissecting the attention deficit hyperactivity disorder (ADHD) phenotype: Sustained attention, response variability and spatial attentional asymmetries in relation to dopamine transporter (DAT1) genotype. Neuropsychologia, 43(13), 1847–1857. https://doi.org/10.1016/j.neuropsychologia.2005.03.011

Borella

de Ribaupierre

Cornoldi

Chicherio

(2013). Beyond interference control impairment in ADHD: Evidence from increased intraindividual variability in the color-stroop test. Child Neuropsychology, 19(5), 495–515. https://doi.org/10.1080/09297049.2012.696603

Brunkhorst-Kanaan

Verdenhalven

Kittel-Schneider

Vainieri

Reif

Grimm

(2020). The Quantified behavioral test: A confirmatory test in the diagnostic process of adult ADHD? Frontiers in Psychiatry, 11, 216. https://doi.org/10.3389/fpsyt.2020.00216

Castellanos

F. X.

Sonuga-Barke

E. J.

Scheres

Di Martino

Hyde

Walters

J. R.

(2005). Varieties of Attention-Deficit/Hyperactivity Disorder-related intra-individual variability. Biological Psychiatry, 57(11), 1416–1423. https://doi.org/10.1016/j.biopsych.2004.12.005

Christakou

Murphy

C. M.

Chantiluke

Cubillo

A. I.

Smith

A. B.

Giampietro

Daly

Ecker

Robertson

Murphy

D. G.

Rubia

.; MRC AIMS Consortium. (2013). Disorder-specific functional abnormalities during sustained attention in youth with attention deficit hyperactivity disorder (ADHD) and with autism. Molecular Psychiatry, 18(2), 236–244. https://doi.org/10.1038/mp.2011.185

Conners

C. K.

(1997). Conners’ rating scales – revised: Technical manual. Multi-Health Systems Inc.

10.

Conners

C. K.

(2014). Conners’ rating scales (3rd ed.). Multi-Health Systems Inc.

11.

Corkum

P. V.

Siegel

L. S.

(1993). Is the continuous performance task a valuable research tool for use with children with attention-deficit-hyperactivity disorder? Journal of Child Psychology and Psychiatry, 34(7), 1217–1239. https://doi.org/10.1111/j.1469-7610.1993.tb01784.x

12.

Epstein

J. N.

Langberg

J. M.

Rosen

P. J.

Graham

Narad

M. E.

Antonini

T. N.

Brinkman

W. B.

Froehlich

Simon

J. O.

Altaye

(2011). Evidence for higher reaction time variability for children with ADHD on a range of cognitive tasks including reward and event rate manipulations. Neuropsychology, 25(4), 427–441. https://doi.org/10.1037/a0022155

13.

Feige

Biscaldi

Saville

C. W.

Kluckert

Bender

Ebner-Priemer

Hennighausen

Rauh

Fleischhaker

Klein

(2013). On the temporal characteristics of performance variability in Attention Deficit Hyperactivity Disorder (ADHD). PLoS One, 8(10), e69674. https://doi.org/10.1371/journal.pone.0069674

14.

Geurts

H. M.

Grasman

R. P.

Verté

Oosterlaan

Roeyers

van Kammen

S. M.

Sergeant

J. A.

(2008). Intra-individual variability in ADHD, autism spectrum disorders and Tourette’s syndrome. Neuropsychologia, 46(13), 3030–3041. https://doi.org/10.1016/j.neuropsychologia.2008.06.013

15.

S. L.

Gau

S. S.

Tzang

S. W.

Hsu

W. Y.

(2013). The ex-Gaussian distribution of reaction times in adolescents with attention-deficit/hyperactivity disorder. Research in Developmental Disabilities, 34(11), 3709–3719. https://doi.org/10.1016/j.ridd.2013.07.025

16.

Hancock

P. A.

(2017). On the nature of vigilance. Human Factors, 59(1), 35–43. https://doi.org/10.1177/0018720816655240

17.

Heinrich

Moll

G. H.

Dickhaus

Kolev

Yordanova

Rothenberger

(2001). Time-on-task analysis using wavelet networks in an event-related potential study on attention-deficit hyperactivity disorder. Clinical Neurophysiology, 112(7), 1280–1287. https://doi.org/10.1016/s1388-2457(01)00541-7

18.

Hervey

A. S.

Epstein

J. N.

Curry

J. F.

Tonev

Eugene Arnold

Keith Conners

Hinshaw

S. P.

Swanson

J. M.

Hechtman

(2006). Reaction time distribution analysis of neuropsychological performance in an ADHD sample. Child Neuropsychology, 12, 125–140. https://doi.org/10.1080/09297040500499081

19.

Huang-Pollock

C. L.

Karalunas

S. L.

Tam

Moore

A. N.

(2012). Evaluating vigilance deficits in ADHD: A meta-analysis of CPT performance. Journal of Abnormal Psychology, 121(2), 360–371. https://doi.org/10.1037/a0027205

20.

Huang-Pollock

C. L.

Nigg

J. T.

Halperin

J. M.

(2006). Single dissociation findings of ADHD deficits in vigilance but not anterior or posterior attention systems. Neuropsychology, 20(4), 420–429. https://doi.org/10.1037/0894-4105.20.4.420

21.

Hwang-Gu

S. L.

Lin

H. Y.

Chen

Y. C.

Tseng

Y. H.

Hsu

W. Y.

Chou

M. C.

Chou

W.-J.

Y.-Y.

Gau

S. S.-F.

(2019). Symptoms of ADHD affect intrasubject variability in youths with Autism Spectrum Disorder: An Ex-Gaussian analysis. Journal of Clinical Child & Adolescent Psychology, 48(3), 455–468. https://doi.org/10.1080/15374416.2018.1452151

22.

Johnson

K. A.

Healy

Dooley

Kelly

S. P.

McNicholas

(2015). Children born with very low birth weight show difficulties with sustained attention but not response inhibition. Child Neuropsychology, 21(5), 629–647. https://doi.org/10.1080/09297049.2014.964193

23.

Johnson

K. A.

Kelly

S. P.

Bellgrove

M. A.

Barry

Cox

Gill

Robertson

I. H.

(2007). Response variability in Attention Deficit Hyperactivity Disorder: Evidence for neuropsychological heterogeneity. Neuropsychologia, 45(4), 630–638. https://doi.org/10.1016/j.neuropsychologia.2006.03.034

24.

Johnson

K. A.

Kelly

S. P.

Robertson

I. H.

Barry

Mulligan

Daly

Lambert

McDonnell

Connor

T. J.

Hawi

Gill

Bellgrove

M. A.

(2008). Absence of the 7-repeat variant of the DRD4 VNTR is associated with drifting sustained attention in children with ADHD but not in controls. American Journal of Medical Genetics Part B Neuropsychiatric Genetics, 147, 927–937. https://doi.org/10.1002/ajmg.b.30718

25.

Karalunas

S. L.

Geurts

H. M.

Konrad

Bender

Nigg

J. T.

(2014). Annual Research Review: Reaction time variability in ADHD and autism spectrum disorders: Measurement and mechanisms of a proposed trans-diagnostic phenotype. Journal of Child Psychology and Psychiatry, 55(6), 685–710. https://doi.org/10.1111/jcpp.12217

26.

Karalunas

S. L.

Huang-Pollock

C. L.

Nigg

J. T.

(2013). Is reaction time variability in ADHD mainly at low frequencies? Journal of Child Psychology and Psychiatry, 54(5), 536–544. https://doi.org/10.1111/jcpp.12028

27.

Kelly

S. P.

Gomez-Ramirez

Foxe

J. J.

(2009). The strength of anticipatory spatial biasing predicts target discrimination at attended locations: a high-density EEG study. European Journal of Neuroscience, 30(11), 2224–2234. https://doi.org/10.1111/j.1460-9568.2009.06980.x

28.

Kofler

M. J.

Rapport

M. D.

Sarver

D. E.

Raiker

J. S.

Orban

S. A.

Friedman

L. M.

Kolomeyer

E. G.

(2013). Reaction time variability in ADHD: a meta-analytic review of 319 studies. Clinical Psychology Review, 33(6), 795–811. https://doi.org/10.1016/j.cpr.2013.06.001

29.

Lacouture

Cousineau

(2008). How to use MATLAB to fit the ex-Gaussian and other probability functions to a distribution of response times. Tutorials in Quantitative Methods for Psychology, 4(1), 35–45. https://doi.org/10.20982/tqmp.04.1.p035

30.

Lee

R. W.

Jacobson

L. A.

Pritchard

A. E.

Ryan

M. S.

Denckla

M. B.

Mostofsky

Mahone

E. M.

(2015). Jitter reduces response-time variability in ADHD: An ex-Gaussian analysis. Journal of Attention Disorders, 19(9), 794–804. https://doi.org/10.1177/1087054712464269

31.

Leth-Steensen

King Elbaz

Douglas

V. I.

(2000). Mean response times, variability, and skew in the responding of ADHD children: a response time distributional approach. Acta Psychologica, 104, 167–190. https://doi.org/10.1016/s0001-6918(00)00019-6

32.

Lewis

F. C.

Reeve

R. A.

Kelly

S. P.

Johnson

K. A.

(2017). Sustained attention to a predictable, unengaging Go/No-Go task shows ongoing development between 6 and 11 years. Attention, Perception and Psychophysics, 79(6), 1726–1741. https://doi.org/10.3758/s13414-017-1351-4

33.

Losier

B. J.

McGrath

P. J.

Klein

R. M.

(1996). Error patterns on the Continuous Performance Test in non-medicated and medicated samples of children with and without ADHD: A meta-analytic review. Journal of Child Psychology and Psychiatry, 37(8), 971–987. https://doi.org/10.1111/j.1469-7610.1996.tb01494.x

34.

Luce

R. D.

(1986). Response Times. Their role in inferring elementary mental organization (Vol. 8). Oxford University Press.

35.

Machida

Murias

Johnson

K. A.

(2019). Electro-physiological correlates of response time variability during a sustained attention task. Frontiers in Human Neuroscience, 13, 363. https://doi.org/10.3389/fnhum.2019.00363

36.

Macmillan

N. A.

Creelman

C. D.

(1991). Detection Theory: A User’’s Guide. Cambridge University Press.

37.

Manly

Owen

A. M.

McAvinue

Datta

Lewis

G. H.

Scott

S. K.

Rorden

Pickard

Robertson

I. H.

(2003). Enhancing the sensitivity of a sustained attention task to frontal damage: convergent clinical and functional imaging evidence. Neurocase, 9(4), 340–349. https://doi.org/10.1076/neur.9.4.340.15553

38.

Merrill

B. M.

Raiker

J. S.

Evans

S. W.

Gnagy

E. M.

Pelham

Jr. (2021). Cognitive mechanisms of methylphenidate in ADHD: Do improvements in sustained attention mediate behavioral improvements in the natural environment? Child Neuropsychology, 27, 425–446. https://doi.org/10.1080/09297049.2020.1862074

39.

Norman

L. J.

Carlisi

C. O.

Christakou

Cubillo

Murphy

C. M.

Chantiluke

Simmons

Giampietro

Brammer

Mataix-Cols

Rubia

(2017). Shared and disorder-specific task-positive and default mode network dysfunctions during sustained attention in paediatric attention-deficit/hyperactivity disorder and obsessive/compulsive disorder. NeuroImage Clinical, 15, 181–193. https://doi.org/10.1016/j.nicl.2017.04.013

40.

O’Connell

R. G.

Dockree

P. M.

Bellgrove

M. A.

Turin

Ward

Foxe

J. J.

Robertson

I. H

. (2009). Two types of action error: Electrophysiological evidence for separable inhibitory and sustained attention neural mechanisms producing error on Go/No-go tasks. Journal of Cognitive Neuroscience, 21(1), 93–104. https://doi.org/10.1162/jocn.2009.21008

41.

Parasuraman

(1979). Memory load and event rate control sensitivity decrements in sustained attention. Science, 205(4409), 924–927. https://doi.org/10.1126/science.472714

42.

Pievsky

M. A.

McGrath

R. E.

(2018). The neurocognitive profile of Attention-deficit/hyperactivity disorder: A review of meta-analyses. Archives of Clinical Neuropsychology, 33(2), 143–157. https://doi.org/10.1093/arclin/acx055

43.

Polanczyk

de Lima

M. S.

Horta

B. L.

Biederman

Rohde

L. A.

(2007). The worldwide prevalence of ADHD: A systematic review and metaregression analysis. American Journal of Psychiatry, 164, 942–948. https://doi.org/10.1176/ajp.2007.164.6.942

44.

Rapport

M. D.

Kofler

M. J.

Alderson

R. M.

Timko

Jr Dupaul

G. J.

(2009). Variability of attention processes in ADHD: Observations from the classroom. Journal of Attention Disorders, 12(6), 563–573. https://doi.org/10.1177/1087054708322990

45.

Rice

M. E.

Harris

G. T.

(2005). Comparing effect sizes in follow-up studies: ROC Area, Cohen’s d, and r. Law and Human Behavior, 29(5), 615–620. https://doi.org/10.1007/s10979-005-6832-7

46.

Robertson

I. H.

Manly

Andrade

Baddeley

B. T.

Yiend

(1997). ’Oops!’: performance correlates of everyday attentional failures in traumatic brain injured and normal subjects. Neuropsychologia, 35(6), 747–758. https://doi.org/10.1016/S0028-3932(97)00015-8

47.

Robin

Turck

Hainard

Tiberti

Lisacek

Sanchez

J.-C.

Müller

(2011). pROC: an open-source package for R and S+ to analyze and compare ROC curves. BMC Bioinformatics, 12(1), 1–8. https://doi.org/10.1186/1471-2105-12-77

48.

Robin

Turck

Hainard

Tiberti

Lisacek

Sanchez

J.-C.

Müller

Siegert

Doering

(2021). pROC. Version 1.17.0.1.

49.

Seymour

K. E.

Mostofsky

S. H.

Rosch

K. S.

(2016). Cognitive load differentially impacts response control in girls and boys with ADHD. Journal of Abnormal Child Psychology, 44(1), 141–154. https://doi.org/10.1007/s10802-015-9976-z

50.

Stanislaw

Todorov

(1999). Calculation of signal detection theory measures. Behavior Research Methods Instruments & Computers, 31(1), 137–149. https://doi.org/10.3758/bf03207704

51.

Sykes

D. H.

Douglas

V. I.

Weiss

Minde

K. K.

(1971). Attention in hyperactive children and the effect of methylphenidate (ritalin). Journal of Child Psychology and Psychiatry, 12(2), 129–139. https://doi.org/10.1111/j.1469-7610.1971.tb01056.x

52.

Taylor

Schachar

Thorley

Wieselberg

(1986). Conduct disorder and hyperactivity: I. Separation of hyperactivity and antisocial conduct in British child psychiatric patients. The British Journal of Psychiatry, 149, 760–767. https://doi.org/10.1192/bjp.149.6.760

53.

Thomson

Vijayakumar

Johnson

K. A.

Malpas

C. B.

Sciberras

Efron

Hazell

Silk

T. J.

(2020). Longitudinal trajectories of sustained attention development in children and adolescents with ADHD. Journal of Abnormal Child Psychology, 48, 1529–1542. https://doi.org/10.1007/s10802-020-00698-5

54.

Vainieri

Michelini

Adamo

Cheung

C. H. M.

Asherson

Kuntsi

(2022). Event-related brain-oscillatory and ex-Gaussian markers of remission and persistence of ADHD. Psychological Medicine, 52, 352–361. https://doi.org/10.1017/S0033291720002056

55.

van der Meere

Wekking

Sergeant

. (1991). Sustained attention and pervasive hyperactivity. Journal of Child Psychology and Psychiatry, 32(2), 275–284. https://doi.org/10.1111/j.1469-7610.1991.tb00306.x

56.

Vaurio

R. G.

Simmonds

D. J.

Mostofsky

S. H.

(2009). Increased intra-individual reaction time variability in attention-deficit/hyperactivity disorder across response inhibition tasks with different cognitive demands. Neuropsychologia, 47, 2389–2396. https://doi.org/10.1016/j.neuropsychologia.2009.01.022

57.

Wechsler

(1992). WISC-III UK. Harcourt Brace & Company.

58.

Wechsler

(2004). Wechsler Intelligence Scale for Children-IV. Harcourt Brace & Company.

59.

Wechsler

(2011). Wechsler Abbreviated Scale of Intelligence (2nd ed.). Pearson.

60.

Willcutt

E. G.

Doyle

A. E.

Nigg

J. T.

Faraone

S. V.

Pennington

B. F.

(2005). Validity of the executive function theory of Attention-Deficit/Hyperactivity Disorder: A meta-analytic review. Biological Psychiatry, 57(11), 1336–1346. https://doi.org/10.1016/j.biopsych.2005.02.006

61.

Wolfers

Onnink

A. M.

Zwiers

M. P.

Arias-Vasquez

Hoogman

Mostert

J. C.

Kan

C. C.

Slaats-Willemse

Buitelaar

J. K.

Franke

(2015). Lower white matter microstructure in the superior longitudinal fasciculus is associated with increased response time variability in adults with attention-deficit/hyperactivity disorder. Journal of Psychiatry and Neuroscience, 40(5), 344–351. https://doi.org/10.1503/jpn.140154

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.03 MB