ADHD neuropsychologia.pdf

2006, Vol. 20, No. 4, 420–429

Single Dissociation Findings of ADHD Deﬁcits in Vigilance but Not

Anterior or Posterior Attention Systems

Cynthia L. Huang-Pollock

Pennsylvania State University

Joel T. Nigg

Michigan State University

Jeffrey M. Halperin

Queens College of the City University of New York

Following a distributed network model of visuospatial attention, the authors used an A – X version of the

Continuous Performance Test and a covert orienting paradigm to examine the vigilance, anterior, and

posterior attention systems. Compared with control participants without attention-deﬁcit/hyperactivity

disorder (ADHD), children with the predominantly inattentive (ADHD-I) and combined (ADHD-C)

subtypes had lower sensitivity ( d ) to detect targets from nontargets. Children with ADHD-C, but not

ADHD-I, additionally had a highly activated response style (ln ). Performance for both subtypes

decreased to a greater extent over time in a manner consistent with problems in sustained attention.

Together, these results suggest the presence of vigilance system deﬁcits in participants with both ADHD

subtypes. However, consistent with previous meta-analytic work, there was no evidence for anterior or

posterior system orienting dysfunctions in either subtype.

Keywords: attention-deﬁcit/hyperactivity disorder (ADHD), visuospatial orienting, Continuous Perfor-

mance Test

Questions remain about the nature of cognitive or neuropsycho-

logical mechanisms that may be involved in the development of

attention-deﬁcit/hyperactivity disorder (ADHD) and its subtypes.

Although recent theories emphasize a range of processes other

than attention, attentional functioning remains an area in need of

further clariﬁcation (Douglas, 1999). In this article, we evaluate

three interconnected attentional systems in children with the two

most common subtypes of ADHD: ADHD combined type

(ADHD-C) and ADHD predominantly inattentive type (ADHD-I).

The relationship between the two subtypes and the status of their

attentional function have been issues of extensive concern and

dispute in the literature (Milich, Balentine, & Lynam, 2001). To

examine this question, we adopted as a conceptual framework a

distributed network model of attention that posits three attentional

systems (Posner & Petersen, 1990). We also drew on the concep-

tual model of Sergeant, Oosterlaan, and van der Meere (1999) and

their analysis of likely ADHD-functional deﬁcits.

The distributed network model is among the best-known models

of visual attention. It speciﬁes three distributed systems: (a) an

automatic posterior attention system (PAS), which involves the

superior parietal cortex, pulvinar, and superior colliculus; (b) a

voluntary anterior attention system (AAS), which involves the

anterior cingulate gyrus and supplementary motor cortex; and (c)

a vigilance system (composed of both sustained attention and

phasic alerting processes), which involves the locus coeruleus

(LC), cholinergic system of the basal forebrain, intralaminar tha-

lamic nuclei, and right prefrontal cortex. These systems are con-

ceptually separable, although as one of their many functions, they

work together to direct attention to locations in space. The strong

empirical support behind the model from lesion and imaging

studies (Posner, Choate, Rafal, & Vaughan, 1985; Posner, Walker,

Friederich, & Rafal, 1984, 1987; Rafal & Posner, 1987) combined

with ease of evaluation has made this model as well as its asso-

ciated measurement paradigm for examining attentional orienting

among the most widely used in the examination of attention

deﬁcits in childhood ADHD (for a review, see Huang-Pollock &

Nigg, 2003).

The measurement paradigm based on reaction time (RT) is a

computer-generated target detection task in which participants are

either peripherally or centrally cued to send their attention to the

right visual ﬁeld (RVF) or left visual ﬁeld (LVF) in anticipation of

a target. Central cues (e.g., arrows) require voluntary cognitive

direction of attention and so probe AAS functionality. Peripheral

cues (e.g., onset of a bright light) activate reﬂexive orienting and

so probe the PAS. By varying the validity of the cues, it is possible

to isolate component processes such as engaging, disengaging, and

shifting of attention.

However, despite intriguing single-study ﬁndings of weaknesses

in the PAS disengage function in ADHD-C, meta-analytic effect

sizes across extant studies using this orienting paradigm have been

small to nonexistent (Huang-Pollock & Nigg, 2003). Thus, any

deﬁcits in the PAS as conceptualized by Posner and Petersen’s

(1990) model are unlikely to be of clinical signiﬁcance in

ADHD-C. Evidence for an AAS deﬁcit is likewise weak and

mixed for ADHD-C (Huang-Pollock & Nigg, 2003), which may be

surprising in light of extensive evidence implicating anterior cor-

Cynthia L. Huang-Pollock, Department of Psychology, Pennsylvania

State University; Joel T. Nigg, Department of Psychology, Michigan State

University; Jeffrey M. Halperin, Department of Psychology, Queens Col-

lege of the City University of New York.

This work was supported in part by National Institute of Mental Health

Fellowship F31-MH12333 and Grant R01-MH59105 awarded to Cynthia

L. Huang-Pollock and Joel T. Nigg, respectively.

Correspondence concerning this article should be addressed to Cynthia

L. Huang-Pollock, 545 Moore Building, Psychology Department, The

Pennsylvania State University, University Park, PA 16802. E-mail:

clh39@psu.edu

420

Neuropsychology

0894-4105/06/$12.00 DOI: 10.1037/0894-4105.20.4.420

ADHD DEFICITS AND THE VIGILANCE SYSTEM

421

tical networks in ADHD. However, most evidence has indicated

that those anterior neural circuits are involved in multiple, perhaps

modularized, control operations of which attentional orienting in

space is only one (Fuster, 1997; Lyon & Krasnegor, 1996; Pen-

nington & Ozonoff, 1996). It is quite possible that ADHD-C

involves breakdowns in executive component functions such as

response suppression (Nigg, 2001; Schachar, Tannock, & Logan,

1993) or working memory (Martinussen, Hayden, Hogg-Johnson,

& Tannock, 2005) but not in executive spatial attention.

Even so, two main issues remain unresolved concerning atten-

tional functioning in ADHD. First, of the studies examining visuo-

spatial orienting with the aforementioned orienting paradigm, only

one included ADHD-I. Thus, the preceding conclusions mainly

pertain to ADHD-C; relatively little is known about attentional

orienting in ADHD-I. This is a key gap in the ﬁeld in light of

long-standing hypotheses that the inattentive subtype may involve

PAS abnormalities (Hynd et al., 1991; Milich et al., 2001).

Second, although previous studies have examined the function-

ing of one or two attentional networks, no studies of ADHD have

examined all three networks within the same sample. This ap-

proach, because it controls for between-sample variance, would

provide a more deﬁnitive answer to the question of whether

children with ADHD demonstrate weaknesses in attentional pro-

cesses. Posner and Petersen’s (1990) model includes both phasic

alerting and sustained attention processes within the rubric of a

vigilance system. Although phasic alerting can be evaluated by

comparing RTs to uncued versus neutrally cued targets (Fan,

McCandliss, Sommer, Raz, & Posner, 2002; Rueda et al., 2004),

the orienting paradigm is not as well suited to the measurement of

sustained attention.

Therefore, to fully assess all three components of the model, one

must include a second measurement paradigm. Among the most

commonly used tests of sustained attention is the Continuous

Performance Test (CPT). Many variations of this paradigm exist

(Corkum & Siegel, 1993; Riccio, Reynolds, Lowe, & Moore,

2002), but nearly all require participants to detect an infrequently

occurring target among rapidly presented nontargets over an ex-

tended period of time. The CPT task in this study is an A – X task

modeled on those used by Rosvold, Mirsky, Sarason, Bransome,

and Beck (1956), Halperin, Sharma, Greenblatt, and Schwartz

(1991), and Halperin, Wolf, Greenblatt, and Young (1991). In

other words, children are instructed to respond when the letter X is

preceded by the letter A . The CPT has itself been extensively

applied to childhood ADHD, especially ADHD-C, with meta-

analyses indicating ADHD weakness on key parameters (Losier,

McGrath, & Klein, 1996).

Experimentally, a sustained attention deﬁcit would lead to per-

formance worsening more rapidly over time (typically measured

by an increase in RT, error rate, or variability in RT over time)

compared with the performance of nonaffected individuals. Mirsky

and Duncan (2001) distinguished sustained attention from a sta-

bilize function that is similar in concept to phasic alertness and is

indexed by overall variability of RT. We therefore conceived of

RT variability as a separate index of alertness that is distinct from

RT change over time, which we consider an index of sustained

attention.

Methods to measure sustained attention on the CPT are not well

standardized. In particular, besides simple error rates, other argu-

ably more sophisticated methods have also been used to identify

performance deﬁcits on the CPT. One of these entails the empir-

refers to an individual’s ability to discriminate between signal and

noise, whereas refers to an individual’s response biases (i.e., to

respond “target present” vs. “target absent”). The cognitive–ener-

getic model (Sergeant et al., 1999) posits that the primary cogni-

tive mechanisms involved in ADHD could involve one or more of

three energetic pools (arousal, activation, and effort) as well as the

executive management of such pools. Sergeant et al. (1999) argued

that cognitive deﬁcits in ADHD may primarily be associated with

arousal or with activation, which they indicated would be indexed

respectively by d and . Under the rubric of this model, d is

believed to be modulated by the arousal pool that is responsible for

phasic alertness, and is considered a measure of activation, or a

more general and tonic state of readiness to respond (Sergeant et

al., 1999). Although activation was not a primary focus of our

study (it is conceptualized as distinct from arousal and vigilance

and may be conceived as part of a broad executive function rubric;

Berger & Posner, 2000), we viewed activation as an additional

state-regulation parameter of theoretical interest in ADHD that we

were able to assess. Therefore, with the CPT, we were able to

assess activation as well as the two components of the vigilance

system: alertness/arousal and sustained attention.

Yet further methodological issues require consideration. For all

their advantages, signal detection indices may be vulnerable to

effects of temporal properties of different error types and partic-

ularly to speed–accuracy trade-off (SATO) effects (Sergeant &

Scholten, 1985). Halperin and colleagues (Halperin, Sharma, et al.,

1991; Halperin, Wolf, et al., 1991; Halperin et al., 1988; Marks,

Himelstein, Newcorn, & Halperin, 1999) developed an alternate

approach on the basis of empirically derived error scores from an

A – X CPT that took into consideration both type of error and speed

of response. They found, in conjunction with RTs, that distinct

CPT error subtypes form meaningful measures of inattention and

impulsivity. Such indices have good test–retest reliability over

durations of several months and have good construct and concur-

rent validity. In their approach, inattentive and impulsive errors

likely index vigilance and activation, respectively. Dyscontrol

errors, which do not meet context or RT criteria for inattention or

impulsive errors, are posited to be nonspeciﬁc in nature. This

approach is described in more detail below. Like other CPT

measures, these error indices have adequate speciﬁcity but low

sensitivity to detect participants with ADHD (as deﬁned in the

Diagnostic and Statistical Manual of Mental Disorders ; 3rd. ed.,

rev.) from control participants without ADHD (Matier-Sharma,

Perachio, Newcorn, Sharma, & Halperin, 1995; Nigg, Hinshaw, &

Halperin, 1996), and they are able to discriminate between chil-

dren with “pure” ADHD and other children, such as children with

anxiety disorders, children with a disruptive behavior disorder

other than ADHD, and control participants (Halperin et al., 1993;

Newcorn et al., 2001). We reasoned that an examination of vigi-

lance system responding can beneﬁt from considering both the

signal detection and error type approaches. Table 1 describes the

cognitive processes and attentional systems under investigation in

this study as well as the indices that in our conceptualization

respectively probe each system.

In all, we sought to evaluate multiple attentional systems in the

two most common Diagnostic and Statistical Manual of Mental

Disorders–IV ( DSM–IV ; 4th ed., American Psychiatric Associa-

tion, 1994) subtypes of ADHD (ADHD-C and ADHD-I) with two

(Green & Swets, 1974). The index d

ically validated and derived variables associated with Signal De-

tection Theory— d

and

422

HUANG-POLLOCK, NIGG, AND HALPERIN

Table 1

Cognitive Processes and Systems Measured by Orienting

Paradigm (OP) and Continuous Performance Test (CPT)

present. Children with ﬁve symptoms of inattention or hyperactivity–

impulsivity by the or algorithm were excluded because it is difﬁcult to

accurately determine subtype in that situation (Lahey et al., 1994). Comor-

bid oppositional deﬁant, conduct, anxiety, and depressive disorders were

identiﬁed with the DISC-IV, and those rates are described later. Children

were deﬁned as having a Reading Disability if (a) their Wechsler Individ-

ual Achievement Test (WIAT; Wechsler, 1992) Basic Reading subtest

score was below 85 and (b) if the score was signiﬁcantly below what would

be predicted given their estimated IQ (as determined by a four-subtest short

form of the Wechsler Intelligence Scale for Children (3rd edition; Wech-

sler, 1991). With this deﬁnition, 1 non-ADHD control participant and 4

participants with ADHD-C met criteria for reading disability. Reading

disability was infrequent because it was an exclusionary criterion during

the initial years of data collection. The simple difference method identiﬁed

the same children as the predicted-achievement method.

Because of a computer failure, we did not have data for the same number

of children on both tasks. Of the participants, 53% ( n

System

Index

Anterior attention

RT to centrally cued targets (OP) a

Posterior attention

RT to peripherally cued targets (OP) a

Vigilance

Sustained attention

Performance over time (CPT) b

Inattentive errors (CPT) c

Phasic alertness

RT to neutral/uncued targets (OP) d

Standard deviation overall (CPT) e

Arousal: d (CPT) f

Activation

ln (CPT) f

Impulsive errors (CPT) c

Note. RT reaction time; d signal detection accuracy; ln re-

sponse bias as deﬁned by Hochhaus (1972).

a See Posner & Petersen (1990)

b See van der Meere & Sergeant (1988)

c See Halperin, Sharma, et al. (1991) and Halperin, Wolf, et al. (1991)

d See Fan et al. (2002) and Rueda et al. (2004)

e See Mirsky & Duncan (2001)

f See Sergeant et al. (1999)

72) completed both

the CPT task and the orienting procedures, 40% ( n

54) had only CPT

9) had only orienting data. Table 2 describes the sample

size for each analysis. Children who completed both the CPT and orienting

task and those who were able to complete only one of the tasks did not

differ in the number or severity of parent- or teacher-reported ADHD

symptoms, IQ, reading, or mathematics achievement (all p s

.20). Results

did not change when analyses were restricted to children who had complete

data on both tasks. Efforts to evaluate the effects of missing data with

maximum likelihood imputation methods (the expectation–maximization

algorithm) yielded no change in conclusions, with one exception that we

detail below. However, the amount of missing data was judged too great to

rely solely on the imputed data. We therefore report all available data

(without imputation) here, adding results for the imputed data set only

when meaningful to clarify results further.

Children were excluded if they had a sensorimotor handicap, neurolog-

ical disorder, autistic disorder, mental retardation, or psychosis by parent

report, if they met criteria for Tourette’s syndrome or bipolar disorder on

the DISC-IV, if they obtained an estimated Full Scale IQ

widely used attention measures. We aimed to clarify whether

either subtype was associated with deﬁcits in the anterior, poste-

rior, or vigilance systems, in addition to activation. On the basis of

our prior meta-analysis (Huang-Pollock & Nigg, 2003), we did not

expect to ﬁnd orienting deﬁcits in the ADHD-C type, but we did

plan to evaluate the hypothesis that the ADHD-I type would be

associated with posterior orienting deﬁcits. Finally, on the basis of

Sergeant et al.’s (1999) study, we hypothesized that the ADHD-C

and ADHD-I types might be differentially associated with alert-

ness/arousal versus activation, if they are distinct conditions as has

been hypothesized (Milich et al., 2001).

75, or if they

were currently taking long-acting psychotropic (e.g., antidepressant) med-

ication. Children taking psychostimulant medication (45% of children with

ADHD, 0% of control participants) washed out for at least 24 hr prior to

their testing (average time off medication

Method

108 hr, range

24–552 hr).

Participants

Procedures

Participants were children aged 7–12 years diagnosed (as deﬁned later)

as either ADHD-C ( n 68), ADHD-I ( n 21), or non-ADHD (control;

n 46). Children were recruited through advertisements in the commu-

nity, schools, clinics, and newspapers to reach as broad and representative

of a sample as possible. The sample thus was representative of the region

with regard to ethnicity: 64% Caucasian, 9% African American, 6%

Hispanic, 2% American Indian, 17% other/mixed, and 2% undisclosed. In

identifying eligible participants, we used a multistage screening procedure

to evaluate DSM–IV criteria on the basis of parent and teacher reports of

symptoms. Children were initially screened as possibly diagnosable with

ADHD if they (a) met or exceeded screening cutoffs on parent–teacher

ratings of the Behavioral Assessment Scale for Children (Reynolds &

Kamphaus, 1992), the Conners’ (1997) Rating Scale, and/or the ADHD

Rating Scale (DuPaul, Power, Anastopoulos, & Reid, 1998) or (b) had been

previously diagnosed by a clinician who had considered parent and teacher

reports. The Diagnostic Interview Schedule for Children, 4th edition

(DISC-IV; Shaffer, Fisher, Lucas, Dulcan, & Schwab-Stone, 2000), a

computer guided, structured diagnostic interview, was then completed by

the primary caregiver, usually the mother.

Following DSM–IV ﬁeld trial data (Lahey et al., 1994), an or algorithm

was used to establish ﬁnal ADHD diagnosis. If children met age of onset,

severity, duration, and cross-situational impairment criteria, a symptom

was counted as present if either the parent (on DISC-IV) or teacher (on the

ADHD Rating Scale asa2ora3onthe0–3scale) endorsed it as being

CPT. To assess vigilance, we administered a computerized A – X CPT

that was modeled after that used by Rosvold et al. (1956) and was identical

to that used by Nigg et al. (1996) and Halperin, Sharma, et al. (1991;

Halperin, Wolf, et al., 1991). On the computer screen, 11 different letters

sequentially appear in a quasi-random order over 400 trials. Total running

time was approximately 12 min that was divided into four 3-min epochs for

data analysis purposes. Each letter was presented for 200 ms, with

a 1,500-ms interstimulus interval. Children pressed a space bar as quickly

Table 2

Sample Size Available for Either the Continuous Performance

Test (CPT) or Covert Orienting Paradigm by Diagnostic

Group

Test

Control ADHD-C ADHD-I Total

Both CPT and orienting

CPT only

Orienting only

Total

135

Note. ADHD-C attention-deﬁcit/hyperactivity disorder–combined

type; ADHD-I attention-deﬁcit/hyperactivity disorder–inattentive type.

data, and 7% ( n

ADHD DEFICITS AND THE VIGILANCE SYSTEM

423

as they could once they detected a target (i.e., A followed by an X ). Targets

occurred in 10% of the trials; false cues (i.e., A not followed by X ) occurred

in 17% of the trials. After instructions were given to the child, the trained

graduate or undergraduate research assistant sat quietly out of view of the

child for the duration of the task.

Examination of error type in combination with error speed allows for the

identiﬁcation of three types of scores that can be conceptualized as inat-

tention , impulsivity , and dyscontrol errors (Halperin, Sharma, et al., 1991;

Halperin, Wolf, et al., 1991; Halperin et al., 1988). Inattention scores were

formed by summing the total number of (a) omissions, (b) slow hits

(correct hits 1700 ms), and (c) slow X -only commission errors (i.e.,

slower than the child’s mean RT, or MRT). Impulsivity scores were formed

by summing the total number of (a) fast A -not- X commission errors (a

keypress RT that is faster than the child’s MRT and is made after the

appearance of an A that is not followed by an X ) and (b) slow A -only

commission errors (a keypress RT made 1,250 ms after the appearance

of an A but prior to the appearance of the next letter; conceptualized as

“jumping the gun”). Dyscontrol scores were formed by summing the total

number of remaining commission errors not already included in the inat-

tention or impulsivity scores, which therefore included (a) responses to any

letter other than A or X , (b) slow A -not- X responses, (c) fast X -only

commission errors, and (d) fast A -only commission errors. Measures of

perceptual sensitivity or signal detection accuracy ( d ) and response bias

(ln ) were calculated with the formulas and table provided in Hochhaus’s

(1972) article. 1

Our index of arousal was d . This index is a standardized measure of

how distinctly a target is perceived from a nontarget. Thus, the smaller the

value of d , the harder it is for a perceiver to distinguish between noise and

targets and thus the weaker is presumed to be the participants’ alertness, or

vigilance system (Pribram & McGuinness, 1976; Sergeant et al., 1999).

When d 0, the hit rate (HR) and false alarm rate (FAR) are equal and

indicate chance performance.

To assess activation, we calculated response bias (ln ), which is an

index of whether a child tends to answer “yes,” resulting in a high HR and

FAR, or whether he or she tends to answer “no,” resulting in a high rate of

correct rejects and omission errors. If the perceiver is unbiased, then

ln 0. If ln is negative, then the perceiver is biased toward “yes”

responses (highly activated response style). If ln is positive, then the

perceiver favors “no” responses (underactivated response style). A positive

score, which suggests a reduced readiness to respond, is taken as an index

of a deﬁcient activation system (Sergeant et al., 1999).

Visuospatial orienting. To assess attentional orienting, we administered a

standard cue-detection task (Posner & Petersen, 1990). To probe the posterior,

reﬂexive orienting system, we administered a fast (150-ms cue–target delay)

exogenous cue condition. To probe the anterior, strategic orienting system, we

administered a slow (800 ms) endogenous cue condition. The exogenous

condition always preceded the endogenous condition to prevent contamination

of data by any carryover strategy effect that might be used during endogenous

cueing. Children completed a total of 120 trials over ﬁve blocks for the

exogenous condition, and 180 trials over ﬁve blocks for the endogenous

conditions. Total running time was 40 min. The child’s head was centered and

stabilized with a chin rest. A digital video camera, mounted behind the

computer monitor and facing the child, provided a close-up of the child’s right

eye. The experimenter observed eye movements on the video camera for each

trial and recorded eye movements that occurred within a trial by using a

response box that was synchronized with the computerized trial-by-trial ad-

ministration. Each trial was experimenter initiated once the experimenter saw

that the child’s gaze was at ﬁxation.

In the exogenous cueing condition, the initial display consisted of a central

ﬁxation cross, ﬂanked 5° to the right and left by boxes, each subtending 1°

(horizontally), which appeared for 1,000 ms. A second, larger box then

appeared around one of the original boxes, which gave the appearance of a

sudden brightening to that location in space. This was the spatial cue, and it

remained present for the duration of the trial. Left-sided, right-sided, and

directionally neutral (brightening of both sides) cue conditions occurred ran-

domly and with equal probability. Single-sided cues validly predicted the

target location 50% of the time so that strategy mechanisms could not be

usefully applied in this condition. Because children cannot voluntarily allocate

spatial attention in less than 300 ms (Munoz, Hampton, Moore, & Goldring,

1999; Ross, Hommer, Breiger, Varley, & Radant, 1994; Rothlind, Posner, &

Schaughency, 1991), the 150-ms cue–target delay is not susceptible to strate-

gic or anterior system responding.

In the endogenous cueing condition, the same initial display appeared

for 1,000 ms. The cue was a centrally appearing arrow subtending 1°

(horizontally) that pointed to the right, the left, or in both directions

(neutral cues). Single-sided cues validly predicted target location 71% of

the time to maximize the beneﬁt of strategic or planned orienting re-

sponses. A single 800-ms cue–target delay was used to allow enough time

for attention to be voluntarily allocated to the target.

In both conditions, targets were arrows subtending 0.5° (vertically) that

randomly and equiprobably pointed up or down. Targets remained present

until children made a forced-choice response indicating whether the target

arrow was pointing up or down. If target arrows were facing up, they

pressed the numeral 9 on a right-sided number pad, and if it was pointing

down, they pressed the 6. Stickers with up and down arrows were placed

over the appropriate numerals. Children were instructed to always keep

their eyes on the central ﬁxation cross. In the exogenous cueing condition,

they were told that sometimes one of the boxes would light up to let them

know where the arrow would appear, sometimes the wrong box would light

up, and sometimes both boxes would light up. In the endogenous condition,

they were told that most of the time the central arrow would direct them to

where the target would appear. Children were told to press the appropriate

up or down button on a response box as fast as they could without making

mistakes to indicate the direction the target arrow was pointing.

Data analysis. We used analysis of variance to decompose the factorial

design on both tasks. To convey effect size, we report the statistic partial

eta squared (

Results

Preliminary Description of Groups

Table 3 provides a description of groups. Consistent with diag-

nostic groupings, compared with control participants, children

with ADHD-I and ADHD-C had more inattentive and more hy-

peractive–impulsive symptoms. Children with ADHD-C had sig-

niﬁcantly more hyperactive–impulsive symptoms and more inat-

tentive symptoms than did children with ADHD-I. Children with

ADHD-C also had a lower estimated IQ than did control partici-

pants or children with ADHD-I. Groups were well matched on age

and did not differ signiﬁcantly in WIAT Basic Reading score.

There were no signiﬁcant differences in sex distribution among

control participants and participants with ADHD subtypes. All of

these features of the sample held when restricted to the subset that

completed only the CPT or only the orienting task. Overall, be-

havioral ratings reﬂected signiﬁcant impairment and supported the

diagnostic groupings assigned.

1 For cases in which HR 1orFAR 0, d and cannot be calculated.

Therefore, following Davies & Parasuraman (1982), when HR 1or

FAR 0, HR was recalculated as 2 -1/number of targets and FAR was recal-

culated as1–2 -1/number of non-targets . Forty-nine children were able to

complete the task with either an HR of 1 or FAR of 0. There were no group

differences in the number of children who were able to maintain an HR

of 1,

2 (1, N 126) 6.13, p .01.

2 ), which is interpreted similar to r squared for a factor.

Results were unchanged when IQ, reading disorder, or conduct disorder

were covaried; we therefore report results without covariates for simplicity.

2 (1, N 126) 3.14, p .08, although 29% of control participants

versus 8% of children with ADHD were able to maintain an FAR of 0,

424

HUANG-POLLOCK, NIGG, AND HALPERIN

Table 3

Demographic, IQ, Achievement, and Symptom Severity Information by Diagnostic Group

Control ( n

46)

ADHD-C ( n

68)

ADHD-I ( n

21)

Variable

Age in months

115.54

13.85

115.20

13.05

114.97

13.96

.99

Estimated FSIQ

111.96

16.03

102.85

14.12

106.05

15.90

.008

WIAT Basic Reading

107.46

12.31

101.82

14.28

103.19

10.50

.08

Total no. of attention symptoms

0.87

1.15

8.47

0.94

7.90

1.00

.001

Total no. of hyper/impulsive symptoms

0.65

1.08

7.99

1.07

2.00

1.45

.001

Parent ratings (T-scores)

Conners’s ADHD index

46.85

4.17

72.85

9.11

71.05

7.98

.001

BASC Hyperactivity

41.96

7.38

71.95

15.71

47.62

9.49

.001

BASC Attention

44.93

7.15

69.28

8.58

68.95

6.24

.001

Teacher ratings (T-scores)

Conners’s ADHD index

44.43

2.67

68.00

9.20

65.48

10.75

.001

BASC Hyperactivity

43.89

5.38

63.35

12.19

52.00

10.54

.001

BASC Attention

43.16

5.42

64.56

7.99

63.71

8.86

.001

Ratio of boys:girls

27:19

48:20

10:11

.12

ODD diagnosis

CD diagnosis

Note. ADHD-C attention-deﬁcit/hyperactivity disorder–combined type; ADHD-I attention-deﬁcit/ hyperactivity disorder–inattentive type; FSIQ

Full Scale IQ; WIAT Wechsler Individual Achievement Test; ADHD attention-deﬁcit/hyperactivity disorder; BASC Behavioral Assessment Scale

for Children; ODD Oppositional Deﬁant Disorder; CD Conduct Disorder; RD Reading Disorder.

Question 1. Vigilance: CPT Analysis

outcome according to some models of sustained attention perfor-

mance (Halperin et al., 1988). SATO effects likely contributed to

the lack of signiﬁcant group differences in MRT on this task, F (2,

123) 1.82 (

Preliminary data check. For control participants, MRT to cor-

rect hits was not signiﬁcantly correlated with the number of false

alarms ( r .16, p .30). Thus, there were no signiﬁcant SATO

effects that could interfere with interpretation of RT scores for the

control group. This was not the case for children with ADHD-C,

however. MRT to correct hits was signiﬁcantly correlated with

number of false alarms ( r .47, p .001). For ADHD-I, MRT

to correct hits was moderately (but nonsigniﬁcantly) correlated

with number of false alarms ( r 0.30, p .23).

The negative correlation between RT to correct hits and number

of errors suggests that children with ADHD-C (and to a lesser

extent, ADHD-I) placed greater emphasis on speed over accuracy,

which is consistent with both behavioral observations of greater

impulsivity and hyperactivity in this group and is an expected

0.03, p .17), but statistically equivalent RTs

among groups does not imply equivalent cognitive efﬁciency. That

is, children with both subtypes of ADHD were still on average

slower than control participants (see Table 4), suggesting that even

a SATO strategy cannot fully account for the greater number of

errors committed.

Concerns regarding SATO effects impacting the interpretability

of RT data are further mitigated by the fact that for both ADHD

groups, MRT to correct hits was signiﬁcantly and positively cor-

related with RT to false alarms (ADHD-C, r .48, p .001;

ADHD-I, r .67, p .001). That is, if children with ADHD

routinely intended to make anticipatory responses, according to the

Table 4

Continuous Performance Test Results by Group: Error Scores and Signal Detection Indices

Control

ADHD-C

ADHD-I

Index

Error scores

Inattention

2.39

2.66

5.08

4.03

6.17

6.19

.001

Impulsivity

1.27

2.33

4.52

7.36

2.00

3.48

.01

Dyscontrol

1.05

1.84

4.86

8.94

3.56

5.11

.02

RT to correct hits

510.00

90.20

540.86

105.52

556.72

110.79

.17

SD of RT

135.89

54.47

173.44

60.18

173.56

59.98

.003

Signal Detection

4.36

0.60

3.68

0.80

3.69

0.81

.001

1.61

1.02

0.98

1.08

1.35

1.26

.01

Note. ADHD-C attention-deﬁcit/hyperactivity disorder–combined type; ADHD-I attention-deﬁcit/hyper-

activity disorder–inattentive type; RT reaction time.

Plik z chomika:

Inne pliki z tego folderu:

Inne foldery tego chomika: