See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/256480986

Use of the Children's Communication Checklist—2 for Classification of

Language Impairment Risk in Young School-Age Children With Attention-
Deficit/Hyperactivity Disorder

Article in American Journal of Speech-Language Pathology · February 2014

DOI: 10.1044/1058-0360(2013/12-0164) · Source: PubMed

CITATIONS READS

32 7,611

1 author:

Geralyn R Timler
James Madison University, Harrisonburg, Virginia, United States
50 PUBLICATIONS 950 CITATIONS

SEE PROFILE

All content following this page was uploaded by Geralyn R Timler on 18 January 2015.

The user has requested enhancement of the downloaded file.

AJSLP

Research Article

Use of the Children’s Communication

Checklist—2 for Classification of Language
Impairment Risk in Young School-Age
Children With Attention-Deficit/
Hyperactivity Disorder
Geralyn R. Timlera

Purpose: Children with attention-deficit/hyperactivity Edition, or on the Test of Narrative Language. Five of these
disorder (ADHD) are at elevated risk for language impairment participants demonstrated 1 or more clinical markers of LI
(LI). This study examined the feasibility of using the Children’s in language samples. Three additional participants, who
Communication Checklist—2 (CCC–2; Bishop, 2006) to received a General Communication Composite score ≤ 85 yet
classify risk for LI in young children, ages 5–8 years, with scored above 85 on the language tests, demonstrated
ADHD. CCC–2 profiles suggestive of pragmatic impairment.
Method: Parents of 32 children with ADHD and 12 typically Sensitivity and specificity rates were 100% and 85.29%,
developing peers completed the CCC–2. The Clinical respectively. CCC–2 scores and most measures were
Evaluation of Language Fundamentals, Fourth Edition significantly correlated.
(Semel, Wiig, & Secord, 2003) and the Test of Narrative Conclusion: The results support the feasibility of using the
Language (Gillam & Pearson, 2004) were administered to CCC–2 as a screener to identify children with ADHD who are
diagnose LI. Language samples were collected to examine at elevated risk for LI and need referral for comprehensive
clinical markers of LI. assessment.
Results: CCC–2 General Communication Composite scores
≤ 85 correctly classified 10 participants with ADHD Key Words: attention-deficit/hyperactivity disorder,
diagnosed with LI as defined by composite scores ≤ 85 on the language impairment, language assessment, school-age
Clinical Evaluation of Language Fundamentals, Fourth children

A
ttention-deficit/hyperactivity disorder (ADHD) for language impairment (LI) in syntactic/semantic (i.e., lan-
affects 9.5% of children ages 4–17 in the United guage structure) and pragmatic (i.e., language use) skills. LIs
States and is the most common neurobehavioral have been reported in approximately 35%–50% of children
disorder of childhood (Centers for Disease Control and who have a diagnosis of ADHD (see, e.g., Cohen et al., 2000;
Prevention, 2010). The diagnosis is based on the presence of Jonsdottir, Bouma, Sergeant, & Scherder, 2005; Tannock &
developmentally inappropriate levels of inattention, overac- Schachar, 1996; Tirosh & Cohen, 1998). Reported rates vary
tivity, and impulsivity (American Psychiatric Association, by referral source, makeup of the assessment protocol, and
2000). The functional impairments of ADHD include com- the criteria for the LI diagnosis (Mueller & Tomblin, 2012).
promised academic achievement and well-being as well as Children with comorbid ADHD and LI have more pro-
difficulty in social interactions. Of interest to speech-language nounced academic and social challenges because LI is also a
pathologists is that children with ADHD are at elevated risk risk factor for compromised social interactions and reading
difficulties (Cohen et al., 2000; Cohen, Barwick, Horodezky,
a Vallance, & Im, 1998; Vallance, Im, & Cohen, 1999). Early
Miami University, Oxford, OH
identification of LI in young children newly diagnosed with
Correspondence to Geralyn R. Timler: [email protected]
ADHD could help facilitate more appropriate educational
Editor: Carol Scheffner Hammer
planning and support services.
Associate Editor: Sean Redmond
Received December 17, 2012
Revision received April 24, 2013
Accepted July 18, 2013 Disclosure: The author has declared that no competing interests existed at the time
DOI: 10.1044/1058-0360(2013/12-0164) of publication.

American Journal of Speech-Language Pathology • Vol. 23 • 73–83 • February 2014 • A American Speech-Language-Hearing Association 73

Downloaded From: http://ajslp.pubs.asha.org/ by Miami University of Ohio - Library, Geralyn Timler on 05/13/2014
The diagnosis of ADHD involves comprehensive and planning and formulation difficulties, including reduced length
multiple parent report, teacher report, and child assessments (Zentall, 1988), poor organization (Tannock, Purvis, &
from several disciplines, including medicine and psychology. Schachar, 1993), and incomplete or confusing story episodes
Current best practice guidelines also recommend assessment (Milch-Reich, Campbell, Pelham, Connelly, & Geva, 1999;
for other conditions that might coexist with ADHD, including Purvis & Tannock, 1997). An important caveat to these de-
developmental disorders such as language and learning dis- scriptions is that these studies generally did not control for
orders (American Academy of Pediatrics, Subcommittee on children’s language status (i.e., with or without LI). It is likely
Attention-Deficit/Hyperactivity Disorder, Steering Committee that some participants had undiagnosed LI because LI is
on Quality Improvement and Management, 2011). The chal- underdiagnosed in children with ADHD (Cohen, Barwick,
lenge for service providers is to identify those children who et al., 1998; Cohen, Davine, Horodezky, Lipsett, & Isaacson,
are most at risk for LI because comprehensive language as- 1993; Cohen, Davine, & Meloche-Kelly, 1989; Mueller &
sessment or referral for comprehensive assessment of all chil- Tomblin, 2012). To determine whether LI status influenced
dren newly diagnosed with ADHD is not an efficient use of narrative production skills in children with ADHD, Luo and
family, hospital, and school personnel resources. One com- Timler (2008) administered the Core Language subtests from
plement to comprehensive assessment is to implement a the Clinical Evaluation of Language Fundamentals, Fourth
screening procedure that could be used to identify children Edition (CELF–4; Semel, Wiig, & Secord, 2003). Children who
who are most at risk for LI. The purpose of this study was obtained a composite standard score of 85 or below were
to examine the feasibility of using a parent report measure, the classified as having an LI (Semel et al., 2003). Four groups of
Children’s Communication Checklist—2 (CCC–2; Bishop, children were identified: (a) ADHD–LI, (b) ADHD+LI, (c) LI
2006), for screening and identification of LI risk in young only, and (d) a typically developing group. Narrative organi-
children with ADHD. Before describing the research questions zation skills were examined from two narratives elicited via
addressed in this study, an overview of language skills in administration of the Test of Narrative Language (TNL;
children with ADHD is provided. Gillam & Pearson, 2004). Language status was found to be
associated with narrative organization skills. The two groups
with LI (i.e., ADHD+LI and LI) demonstrated poorer nar-
Language Abilities in Children With ADHD
rative organization than the ADHD–LI and typically devel-
(With and Without LI) oping groups, who did not differ from each other. These results
Previous investigations into the language skills of underscore the need to examine language status when inter-
children with ADHD have focused on descriptions of lan- preting narrative production skills in children with ADHD.
guage performance on standardized tests and in extended Redmond et al. (2011) also used the TNL to examine narrative
discourse genres such as narration and conversation. Potential performance among three groups of children: (a) those with
differences and areas of overlap in the language abilities of SLI, (b) those with ADHD, and (c) a typically developing
children with ADHD and children with specific language group. As expected, the SLI group’s mean Narrative Language
impairment (SLI) also have been examined (McInnes, Ability Index (NLAI) score was significantly lower than the
Humphries, Hogg-Johnson, & Tannock, 2003; Redmond, ADHD and typically developing groups’ scores, although
2004, 2005; Redmond, Thompson, & Goldstein, 2011). some of the children in the ADHD group did score below the
Standardized test performance. Several studies have recommended clinical cutoff for LI (i.e., a score of 85 or lower
revealed that children with ADHD who do not have LI on the TNL NLAI).
perform slightly below typically developing peers on stan- Conversation skills. Conversation skills, including
dardized language tests, in particular on subtests that tap syntactic, semantic, and pragmatic language use, have been
into verbal working memory, such as nonword and sentence examined through language sample analyses and parent
repetition subtests (Cohen et al., 2000; Jonsdottir et al., 2005; report measures. Redmond (2004) examined conversational
Kim & Kaiser, 2000). For example, children with ADHD language samples in children with ADHD for the presence of
performed below typically developing peers on the Sentence clinical syntactic and semantic markers of SLI, including
Recall and Sentence Formulation subtests of the Clinical mean length of utterance per morpheme (MLUm), verb tense
Evaluation of Language Fundamentals, Third Edition (Semel, (Rice, Wexler, & Hershberger, 1998), and lexical diversity
Wiig, & Secord, 1995), even though children’s composite scores as measured by number of different words in the sample.
were within the average range, confirming absence of LI Conversational language samples were elicited from SLI,
(Kim & Kaiser, 2000; Oram, Fine, Okamoto, & Tannock, ADHD, and typically developing groups. The SLI group
1999). In comprehension tasks, children with ADHD show demonstrated significantly more clinical markers than the
subtle deficits in making inferences and monitoring compre- ADHD and typically developing groups, who did not differ
hension of task directions (McInnes et al., 2003). Children with from each other. Because the group with SLI demonstrated
ADHD and comorbid LI (ADHD+LI) have more pronounced shorter sentences, more verb tense errors, and a fewer
difficulties in receptive language tasks and perform on stan- number of different vocabulary words than the ADHD
dardized tests more like children with SLI than children with and typically developing groups, Redmond concluded that
only ADHD (Cohen et al., 2000; McInnes et al., 2003). these clinical markers could be used to differentiate the
Narrative skills. Perhaps not surprisingly, descriptions of expressive language skills of children with ADHD from
narrative production in children with ADHD have reflected those of children with LI.

74 American Journal of Speech-Language Pathology • Vol. 23 • 73–83 • February 2014

Downloaded From: http://ajslp.pubs.asha.org/ by Miami University of Ohio - Library, Geralyn Timler on 05/13/2014
Whereas syntactic and semantic deficits are usually simultaneously report the findings of a parent report measure
observed only in children with ADHD who also have LI, and child performance on language measures used by cli-
many children with ADHD with and without LI exhibit nicians to diagnose syntactic/semantic LI. As such, the
pragmatic language deficits (Bishop & Baird, 2001; Bruce, CCC–2 and direct child language measures were adminis-
Thernlund, & Nettelbladt, 2006; Geurts et al., 2004; Geurts tered to address the following research questions. First, do
& Embrechts, 2008; Helland, Biringer, Helland, & Heimann, CCC–2 scores differ among children with ADHD (with and
2012; Helland & Heimann, 2007; see also the review by without comorbid LI) and typically developing children?
Tannock & Schachar, 1996). Analyses of conversational Second, what is the relationship among CCC–2 scores and
language samples have revealed more frequent episodes of children’s performance on standardized language tests (the
mazes (Redmond, 2004), turn-taking difficulties (e.g., talking CELF–4 and TNL) and functional language measures,
too much, interrupting others), and difficulty adjusting including clinical markers of LI obtained from a conversa-
language to the listener’s needs within specific contexts tion sample? Third, and last, what is the classification ac-
(Tannock & Schachar, 1994; Vallence et al., 1999). Parents curacy of the CCC–2 composite score for identification of LI
of children with ADHD have also identified pragmatic in children with ADHD?
language deficits as a concern. Examinations of parent report
measures have revealed that children with ADHD have
more difficulty carrying on a conversation than children Method
without ADHD (Bruce et al., 2006). In fact, several studies Participants
that have used the CCC–2 and earlier versions of this mea- All participants were monolingual English speakers.
sure (Bishop, 1998, 2006) have suggested that the pragmatic Inclusionary criteria included parent report of the participant
language skills of children with ADHD are as impaired passing a hearing screen or passing of an audiometric screening
as those of children with autism spectrum disorders (Bishop at the time of study entry. Participants were also required
& Baird; 2001; Geurts et al., 2004; Geurts & Embrechts, to pass the phonological screener from the Test of Early
2008; Helland et al., 2012; Helland & Heimann, 2007). One Grammatical Impairment (Rice & Wexler, 2001). Partici-
limitation of these studies is that children’s language skills pant characteristics are displayed in Table 1.
were not tested as part of the study protocols. It is likely that Typically developing group. Typically developing chil-
some children in the ADHD samples had LI because pre- dren were recruited via flyers posted in community settings.
viously unidentified LI has been found in children with The typically developing group included 12 children between
ADHD when comprehensive language testing is adminis- ages 5;1 and 8;4 (years;months). There were six boys and
tered (e.g., Cohen, Barwick, et al., 1998; McInnes et al., six girls in this group. Eleven children were Caucasian, and
2003; see also the review by Mueller & Tomblin, 2012). one was African American. None of the children were
Therefore, syntactic and semantic language deficits in some receiving special education services. The Kaufmann Brief
of the children could have contributed to compromised Intelligence Test (Kaufman & Kaufman, 2004) nonverbal
pragmatic language ratings. subtests were administered to obtain an estimated nonverbal
Summary. Children with ADHD demonstrate diverse IQ (range = 82–130). Maternal education levels were reported
language profiles. Some children score slightly below typ- as follows: three with graduate degrees, five college graduates,
ically developing peers on language subtests that tap into
verbal working memory, including expressive sentence
formulation tasks and receptive inferencing tasks, even when Table 1. Participant characteristics.
comorbid LI has been ruled out. Narrative production
is compromised for some children, but it is important to ac- TD (n = 12) ADHD (n = 32)
count for children’s language abilities because the narratives Measure M SD M SD F
of children with ADHD+LI are more compromised than
those of children with ADHD only (Cohen et al., 2000; Luo Age (mos) 80.0 13.56 80.88 7.89 0.791
& Timler, 2008). In conversation samples, clinical syntactic Maternal educationa 2.83 0.94 2.84 0.88 0.001
and semantic markers of SLI are observed usually only in Nonverbal IQb 100.18 15.8 9.52 3.44
Verbal IQc 10.74 2.71
children with comorbid LI (Redmond, 2004). Parent report Full Scale IQd 100.33 14.56
measures reveal pragmatic language deficits in children
with ADHD, but no study has yet examined how the impact Note. TD = typically developing; ADHD = attention-deficit/hyper-
of syntactic/semantic impairment affects these measures. activity disorder.
a
1 = high school diploma, 2 = some college, 3 = college degree,
4 = graduate degree. bTD group scores are from the Kaufmann Brief
Purpose Intelligence Test Nonverbal subtest (M = 100, SD = 15), and ADHD
group scores are from the Wechsler Intelligence Scale for Children—
The purpose of the current study was to examine Fourth Edition (WISC–IV) Block Design subtest (M = 10, SD = 3);
the feasibility of using the CCC–2 as a screening tool to group differences were not examined because the same nonverbal
scale was not administered to both groups. cWISC–IV Vocabulary
identify young children with a diagnosis of ADHD who are subtest (M = 10, SD = 3). dEstimated IQ from two WISC–IV subtests
at risk for comorbid LI. To the author’s knowledge, no (M = 100, SD = 15).
studies of children with ADHD have been published that

Timler: Language Impairment Risk and Children With ADHD 75

Downloaded From: http://ajslp.pubs.asha.org/ by Miami University of Ohio - Library, Geralyn Timler on 05/13/2014
three who attended some college, and one with a high school a speech and language impairment and thus a need for
diploma. further evaluation, (b) to identify children with pragmatic LI,
ADHD group. Participants with ADHD were recruited and (c) to assist in identifying children who may have an
from a clinical research sample whose families were about autism spectrum disorder. The CCC–2 is intended for
to enter an intervention study conducted by a university-based children between the ages of 4;0 to 16;11 who “speak in
research clinic. The clinic’s ADHD evaluation protocol in- sentences and whose primary language is English” (Bishop,
cluded independent confirmation of ADHD (any subtype) 2006, p. 1). The composition of the CCC–2 includes 70 items
by a doctoral level clinician and a developmental pediatrician. that caregivers rate using a 4-point numeric frequency scale
Assessment procedures comprised a clinical interview with ranging from 0 (less than once a week [or never]) to 3 (several
the primary caregiver(s) and a review of multiple caregiver times [more than twice] a day [or always]). The 70 items
and teacher report protocols. The clinical interview used the are divided into 10 scales with seven items each: (A) Speech
National Institute of Mental Health Diagnostic Interview (“leaves off beginning or endings of words”), (B) Syntax
Schedule for Children, Version IV (NIMH DISC–IV; Shaffer, (“leaves out ‘is’”), (C) Semantics (“forgets words he or she
Fisher, Lucas, Dulcan, & Schwab-Stone, 2000). Using the knows”), (D) Coherence (“confuses the sequence of events
Diagnostic and Statistical Manual of Mental Health Disorders when trying to tell a story”), (E) Initiation (“is difficult to
(4th ed., text rev.; American Psychiatric Association, 2000), stop him or her from talking”), (F) Scripted Language
the NIMH DISC–IV assesses the symptoms of ADHD, as (“provides over-precise information in his or her talk”), (G)
well as of affective disorders (e.g., depression, anxiety), and Context (“misses the point of jokes and puns”), (H)
screens for exclusionary categories, including pervasive Nonverbal Communication (“does not look at the person
developmental disorders (e.g., autism), seizure disorders (or he or she is talking to”), (I) Social Relations (“appears
medications for seizure disorders), and intellectual disability anxious in the company of other children”), and (J) Interests
(defined as a full scale IQ below 70). The ADHD group (“shows interest in things or activities that most people
included 32 children between 5;7 and 7;9. There were 27 boys would find unusual”).
and five girls. Twenty-two children were Caucasian, five were The CCC–2 yields one composite score—the General
African American, two were Hispanic, two were biracial, Communication Composite (GCC)—and one index score,
and one was Asian. Six children had been prescribed stimulant the Social Interaction Difference Index (SIDI). The GCC is a
medication. Although none of the children were enrolled in norm-referenced standard score (M = 100, SD = 15) that
full-time special education classes, three were receiving reflects overall communication skills; this score is used to
support services under the special education category “Other identify clinically significant communication problems and is
Health Impairment,” and two more were receiving services calculated by summing Scales A through H. The SIDI is a
under the category “Speech and Language Impairment.” Two difference index that reflects the summed difference between
subtests from the Wechsler Intelligence Scale for Children— the structural language scales (A, B, C, and D) and the
Fourth Edition (WISC–IV; Wechsler, 2003) were administered pragmatic language scales (E, H, I, and J). SIDI scores
as part of the university research clinic’s intake process, and a ranging from –10 to 10 are considered typical; scores within
full scale IQ was estimated from the two subtests (Kaplan, this range were obtained by 90% of the CCC–2 normative
Crawford, Dewey, & Fisher, 2000; Sattler, 1988). WISC–IV sample. Scores ≥ 11 suggest syntactic/semantic skills are
estimated IQ range was 73–135. Maternal education levels deficient and relatively poorer than pragmatic skills, whereas
were reported as eight with graduate degrees, 13 college scores ≤ –11 suggest pragmatic language skills are deficient and
graduates, nine who attended some college, and two with high relatively poorer than syntactic/semantic skills; this profile is
school diplomas. associated with autism spectrum disorders (Bishop, 2006).
In this study, the GCC score was used to examine
Procedure group differences as well as the accuracy of the CCC–2 for
identification of LI risk in the ADHD group. The CCC–2
Parents completed the CCC–2 and a case history form.
examiner’s manual reports that a GCC score of 85 or lower is
Child participants completed two to three 60- to 90-min
associated with 70% sensitivity (i.e., 70% of children with SLI
testing sessions (five children in the ADHD group required
were identified) and 85% specificity (i.e., 85% of children
a third testing session due to child fatigue). Two language
who did not have SLI were accurately identified as not
tests and a 12-min language sample (described below)
having a language disorder). Classification accuracy is also
were administered. The author, a certified speech-language
affected by base rates, or the prevalence of the condition
pathologist, conducted or supervised the test sessions. All test
that is being tested. Positive predictive power (PPP) is the
sessions were audio-recorded for data reduction purposes.
likelihood that a person with the disorder actually has the
The CCC–2 and language tests and samples were scored
disorder. Negative predictive power (NPP) is the likelihood
and transcribed after the test sessions by student research
that someone with a negative test result does not have the
assistants who were blind to the group status of the children.
disorder. The CCC–2 manual reports that a 50% prevalence
rate yields PPP and NPP values of .83 and .74, respectively,
Measures for GCC scores of 85 or lower.
CCC–2. The CCC–2 (Bishop, 2006) was developed for The SIDI was used to describe the pragmatic language
three clinical purposes: (a) to identify children who may have profiles of individual participants with GCC scores below 86.

76 American Journal of Speech-Language Pathology • Vol. 23 • 73–83 • February 2014

Downloaded From: http://ajslp.pubs.asha.org/ by Miami University of Ohio - Library, Geralyn Timler on 05/13/2014
The GCC is the primary score that is to be used to identify T-units (Hunt, 1970). Mazes were set apart by parentheses
communication impairment, and, as such, it is recommended and included whole and part word repetitions, revisions, and
that the SIDI be interpreted only when the GCC score is filled pauses (e.g., “um”). Morpheme errors were marked
clinically significant (Bishop, 2013). To examine group dif- at the word level for bound morpheme errors and at the
ferences in pragmatic language skills and the relationship sentence level for free morpheme errors.
among these skills and the child assessments used in the
current study, an unstandardized Pragmatic Composite (PC)
score was calculated as in previous research (Leonard, Data Analyses
Milich, & Lorch, 2011). The PC was calculated by averaging Data preparation. Data from the CELF–4 and TNL
the six pragmatic language scales (E, F, G, H, I, and J), included subtest and composite standard scores. Before
yielding M = 60 and SD = 18. Previous use of this PC in data entry, the scoring of all test protocols was checked by
a community sample of 54 children yielded an acceptable a second research assistant. Scoring disagreements were
Cronbach’s alpha of .88 (Leonard et al., 2011). discussed between the first and second assistant and settled
CELF–4. The CELF–4 Core Language score for by consensus. Language sample transcription accuracy
children ages 5 to 8 comprises four subtests: (a) Concepts & was checked independently by two research assistants who
Following Directions, (b) Word Structure, (c) Recalling had not completed the original transcription. These assis-
Sentences, and (d) Formulated Sentences. The clinical cutoff tants listened to the language samples while reading the
for LI in the present study was a Core Language score of original transcription. Transcript errors were highlighted,
85 or lower. Sensitivity and specificity rates for a Core and potential changes were noted. The errors and changes
Language score ≤ –1 SD (i.e., 85 or lower) are reported as were discussed, and transcripts were corrected as needed
100% and 82%, respectively (CELF–4 examiner’s manual; after a mutual consensus by three research assistants, in-
Semel et al., 2003, p. 272). PPP and NPP values are reported cluding the assistant who had completed the original tran-
for various base rates in the test manual. A base rate of scription. Language sample data were obtained from the
50% (i.e., half of children referred are usually identified as Systematic Analysis of Language Transcripts software using
having a language disorder) was reported to have a PPP of the first 100 complete and intelligible utterances from
.85 and an NPP of 1.00. This means that 85% of the children 43 participants. One participant in the typical group only had
referred who scored ≤ –1 SD actually had a language dis- 99 utterances; her data are included in the analyses. Lan-
order, and the remaining 15% were considered false positives guage sample data from a second participant in the typically
who did not have a language disorder. The reported NPP developing group were excluded from analysis because of
means that 100% of children with a score of 86 or above a recording failure 6 min into sample collection, yielding only
actually did not have a language disorder. The CELF–4 56 complete and intelligible utterances.
standardization sample included more than 4,500 children; Language sample analyses focused on three measures:
approximately 9.5% of the sample received special education (a) MLUm; (b) number of different words; and (c) correct
services. No information is provided in the manual about composite verb tense use, calculated from the mean percentage
whether children with ADHD were included in the stan- of correct use of regular past tense, third person singular, and
dardization sample. the copula and auxiliary BE forms (i.e., correct use/correct
TNL. The TNL’s NLAI score is composed of the use + errors). Language sample data from the 43 participants
Narrative Comprehension and Oral Narration subtests. were transformed into z scores with clinically significant scores
NLAI scores ≤ –1 SD are reported to have a sensitivity rate defined as ≤ –1.00.
of 92% and a specificity rate of 87% from a matched sample Analysis plan. Analyses examined group differences,
of children with and without previously identified language the relationship among CCC–2 scores and language mea-
disorders (TNL examiner’s manual; Gillian & Pearson, 2004, sures, and the accuracy of the CCC–2 GCC score for clas-
pp. 62–63). PPP was reported as 88% and a positive likelihood sification of LI. Group differences on all measures were
ratio (LR+) of 7.0 (i.e., a child is seven times likely to actually examined in two ways. First, differences between the ADHD
have a language disorder with an NLAI score of 85 or less). and typically developing groups on the CCC–2 and direct
Less than 1% of the standardization sample included children language measures were explored with analyses of variance.
with ADHD (Gillian & Pearson, 2004, p. 38). Effect sizes (i.e., h2) were calculated, and effect sizes of .01,
Language sample. A 12-min conversational and nar- .06, and .14 were interpreted as small, medium, and large,
rative language sample was elicited using procedures adapted respectively (Green & Salkind, 2005). The ADHD group was
from Hadley (1998). The sample included a personal retell then divided into two groups, ADHD (–LI) and ADHD+LI,
segment (e.g., “Tell me about your family”), an expository based on LI status (i.e., LI was defined as a score ≤ 85 on
segment (e.g., “Tell me about a game that you like to play”), the CELF–4 or the TNL). Group differences among the
and a narrative retell (e.g., “Tell me about your favorite book CCC–2 GCC and PC scores were then examined.
or movie”). Samples were transcribed and entered into the The relationship among the GCC and PC scores and
Systematic Analysis of Language Transcripts (Miller & direct language measures were explored via bivariate cor-
Iglesias, 2008) by one of four undergraduate and graduate relations. Correlation coefficients of .10, .30, and .50 were
student research assistants. These assistants were blind to the considered small, medium, and large, respectively (Green &
group status of the participants. Utterances were divided into Salkind, 2005).

Timler: Language Impairment Risk and Children With ADHD 77

Downloaded From: http://ajslp.pubs.asha.org/ by Miami University of Ohio - Library, Geralyn Timler on 05/13/2014
Finally, the accuracy of the CCC–2 GCC for identi- Relationship Among CCC–2 Scores and
fication of LI in this sample was examined. The language Language Measures
profiles of the individual participants who received GCC
scores ≤ 85 were reviewed. Next, the accuracy of the GCC Bivariate correlations between the direct language
score was analyzed by two methods. First, empirical measures and the CCC–2 GCC and PC scores across the
classification rates, including sensitivity, specificity, and typically developing and ADHD groups are presented in
overall classification accuracy rate, were performed using Table 4. Significant correlations ranging from .445 to .660
MedCalc for Windows (MedCalc Software, 1993). An were detected among the GCC and the CELF–4 scores, the
empirical receiver operating characteristic (ROC) curve was TNL Narrative Comprehension subtest and composite
generated by comparing the classification of the CCC–2 NLAI scores, and the language sample measures MLUm and
GCC score against the classification of LI. The ROC curve number of different words in 100 utterances. Significant
plots the sensitivity against 100 minus the specificity for correlations were detected for the PC and most language
all obtained values. The area under the ROC curve (AUC) measures.
is an estimate of the overall accuracy of the measure.
The Youden (1950) index (J) is the optimal cutoff point
on an ROC curve; it is defined as the maximum vertical
Accuracy of CCC–2 GCC
distance between the ROC curve and the diagonal line A descriptive summary of the participants whose GCC
and represents the sensitivity of the measure. Second, LRs scores were ≤ 85 is presented in Table 5. Scores from the
were calculated. Unlike sensitivity and specificity, LRs 10 participants classified as ADHD+LI by the CELF–4
remain relatively stable across variations in sample sizes, or TNL are given in the top panel of Table 5; all 10 received
including samples with unequal numbers of participants in a GCC score of ≤ 85. Five of these participants demonstrated
affected and unaffected groups (Dollaghan, 2004). LR+s one or more clinical markers of LI in language samples.
reflect the degree of confidence that a person who scores Note that four of the five participants receiving services
above the criterion truly has an impairment. LR+s ≥ 4.0 but for Speech and Language Impairment or Other Health
< 10.0 are interpreted as intermediate values and are con- Impairment were classified as having an LI by this study’s
sidered insufficient to determine an individual’s diagnostic definition. One child receiving services for Speech and
status, indicating that further testing is needed (Dollaghan, Language Impairment scored above the clinical cutoff on all
2004). tests and measures. A review of his chart and a follow-up
phone conference with his caregiver revealed that his school
services focused on remediation of speech sound disorders.
Results Data from the four participants with ADHD who scored
above 85 on both the CELF–4 and TNL, yet received
Group Differences GCC scores of ≤ 85, are listed in the bottom panel of Table 5.
The group means, SDs, F tests, and effect size indexes Three of these participants (Nos. 44, 29, and 4) received SIDI
(i.e., h2) for the CCC–2 and the direct language measures scores less than –10, a profile suggestive of pragmatic LI.
in the typically developing and ADHD groups are presented The fourth participant (No. 14) had a SIDI score in the
in Table 2. Significant between-group differences were typical range; however, all three of this participant’s lan-
observed for the CCC–2 GCC, the CCC–2 PC, the Recalling guage sample measures were more than 1 SD below this
Sentences and Formulated Sentences subtests from the sample’s mean, which is suggestive of clinical markers of
CELF–4, and TNL subtests and composite TNL NLAI SLI. Finally, one participant in the typically developing
scores. Effect sizes for significant differences ranged from .15 group (No. 45) received a GCC score of ≤ 85; this partic-
to .27. No between-group differences were detected in the ipant’s performance was within the typical range for all
three language sample measures. direct child measures.
CCC–2 mean scores (i.e., the GCC and PC) for the Accuracy was also examined via empirical classifica-
typically developing and ADHD groups, with the ADHD tion rates and LR+. The empirical ROC curve is presented in
group subdivided by LI status, are presented in Table 3. Figure 1. The AUC was calculated as .956 (with a confidence
Recall that the reference standard for classification of LI interval [CI] of 95% [0.847, 0.995], p ≤ .0001). AUCs from
in the current study was a CELF–4 core composite score 0.7 to 0.9 are considered moderately accurate; AUCs above
or TNL NLAI score ≤ 85. This reference standard yielded 0.9 are considered highly accurate (Swets, 1988). The Youden
10 participants in the ADHD group with comorbid LI index was calculated to be 0.8529, with an associated CCC–2
(ADHD+LI), which represents 31% of the ADHD sample. GCC criterion score of ≤ 85. For this study’s sample of
As expected, none of the participants in the typically 44 participants, the criterion cutoff score of 85 or less yielded
developing group had LI. The ADHD+LI group’s GCC 100% sensitivity (95% CI [69.2, 100]) and 85.29% specificity
score was significantly lower than those of the typically (95% CI [68.9, 95.0]). In other words, a CCC–2 GCC score
developing and ADHD groups, which did not differ from of 85 or lower correctly classified 10 out of 10 participants
each other. All three groups differed on the PC score such with ADHD and comorbid LI. As depicted in Figure 2, the
that the typically developing group > ADHD group > specificity rate was lower because five participants with a GCC
ADHD+LI group. score ≤ 85 appeared to be misclassified as having LI; however,

78 American Journal of Speech-Language Pathology • Vol. 23 • 73–83 • February 2014

Downloaded From: http://ajslp.pubs.asha.org/ by Miami University of Ohio - Library, Geralyn Timler on 05/13/2014
Table 2. CCC–2 and language measures in the TD and ADHD groups.

Measure TD groupa ADHD groupb F h2

CCC–2 GCC 103.17 12.85 89.0 13.69 9.65** .187

CCC–2 PC 60.83 9.91 46.19 13.04 12.39** .238
CELF–4 Core Language 109.33 11.13 97.34 14.89 6.40 .132
C & FD 11.42 2.02 9.59 2.87 4.05 .088
RS 11.50 2.11 8.66 3.25 7.88** .158
FS 12.50 2.28 9.81 2.72 9.26** .181
WS 11.00 2.17 10.13 2.97 0.86 .020
TNL NLAI 111.50 14.91 95.69 10.42 15.78*** .273
Narr. Comp. 12.33 1.37 10.34 2.40 7.26** .147
Oral Narr. 11.50 4.08 8.23 1.93 13.22*** .239
Language sample
MLUm 6.96 1.15 6.28 1.16 2.81 .064
NDW-100 211.64 28.69 195.91 26.79 2.72 .106
Composite tense use 0.98 0.03 0.97 0.05 0.06 .001

Note. CCC–2 GCC = Children’s Communication Checklist—2 General Communication Composite; CCC–2 PC = unstandardized Pragmatic
Composite, calculated from the average of the six pragmatic language scales (M = 60, SD = 18); CELF–4 Core Language = Clinical Evaluation of
Language Fundamentals, Fourth Edition, Core Language Score, which comprised Concepts & Following Directions (C & FD), Recalling Sentences
(RS), Formulated Sentences (FS), and Word Structure (WS); TNL NLAI = Test of Narrative Language Narrative Language Ability Index, which
comprised Narrative Comprehension (Narr. Comp.) and Oral Narration (Oral Narr.); MLUm = mean length of utterance per morpheme in 100
utterances; NDW-100 = number of different words in 100 utterances; Composite tense use = mean percentage of correct use of regular past tense,
third person singular, and the copula and auxiliary BE forms.
a
n = 12; for the language sample, n = 11. bn = 32.
**p ≤ .01. ***p ≤ .001.

as described in the preceding paragraph, three of these par- Farinella, 2006). High sensitivity rates suggest that a measure
ticipants had profiles of pragmatic LI, and one had multiple is most useful for ruling out participants as having a specific
clinical markers of SLI in the language sample. condition (Sackett, Straus, Richardson, Rosenberg, & Haynes,
The calculated LR+ was 6.8 (95% CI [3.0, 15.3]), 2000). In this case, the results suggest that if a child’s CCC–2
indicating that the odds of having an LI increased 6.8 times composite score is above 85, the child is less likely to have an
when the GCC score was ≤ 85. LI at that particular point in the child’s development.
One dilemma encountered by speech-language patho-
logists is how best to identify LIs in young children because
Discussion a gold standard diagnostic test does not exist. In lieu of
The results of this study provide preliminary support for one gold standard, speech-language pathologists use a
use of the CCC–2 as a screening tool to identify young chil- combination of standardized tests, criterion-referenced
dren with ADHD who need referral for more comprehensive measures, language samples, observation scales, and care-
language evaluation. A CCC–2 GCC score of 85 or less giver report measures. In the current study, two well-known
correctly classified 100% of the participants who were standardized language tests, the CELF–4 and the TNL, were
identified as having an LI by the language tests used in this selected to identify LI. That these tests tap different areas
study’s protocol. Sensitivity rates above 80% are considered of language abilities was evident. Of the 10 children diag-
acceptable for identification of LI (Spaulding, Plante, & nosed with LI, only five scored below the clinical cutoff on

Table 3. CCC–2 scores in the TD, ADHD (–LI), and ADHD+LI groups.

ADHD (–LI) ADHD+LI

TD (n = 12) (n = 22) (n = 10)
Measure M SD M SD M SD F h2 Post hoc contrasts

GCCa 103.17 12.85 94.82 11.64 76.2 8.01 16.17*** .441 1, 2 > 3
PCb 60.83 9.91 51.27 11.60 35.0 8.33 16.72*** .449 1>2>3

Note. ADHD (–LI) = children with ADHD but without comorbid language impairment; ADHD + LI = children with ADHD and comorbid language
impairment.
a
M = 100, SD = 15. bUnstandardized Pragmatic Composite calculated from the average of the six CCC–2 pragmatic language scales (Initiation,
Scripted Language, Context, Nonverbal Communication, Social Relations, and Interests; M = 60, SD = 18).
***p ≤ .001.

Timler: Language Impairment Risk and Children With ADHD 79

Downloaded From: http://ajslp.pubs.asha.org/ by Miami University of Ohio - Library, Geralyn Timler on 05/13/2014
Table 4. Bivariate correlations between direct language measures findings underscore the importance of using comprehensive
and CCC–2 GCC scores.
assessment protocols, including parent report measures and
observational data, rather than reliance on language tests
Measure CCC–2 GCC CCC–2 PC
alone, when diagnosing LI (Bishop & McDonald, 2009).
CELF–4 Core Language .641*** .496**
C & FD .500*** .469** Limitations
RS .660*** .484**
FS .574*** .500** Several limitations reduce the generalizability of these
WS .461** .267 results, including the participants’ age and the recruitment
TNL NLAI .457** .523**
Narr. Comp. .457** .488** method. This sample was limited to children between the
Oral Narr. .335 .409** ages of 5 and 8 years. The accuracy of the CCC–2 for
Language sample identification of LI in older school-age children with ADHD
MLUm .445** .426** has yet to be examined. Some school-age children and
NDW-100 .447** .369**
Composite tense use .245 .064
adolescents with ADHD demonstrate deficits in organiza-
tional and written language (Bruce et al., 2006); these deficits
**p ≤ .01. ***p ≤ .001. are not represented in CCC–2 items, and they were not tested
in the current study. It may be that some of the participants,
in particular those with borderline GCC scores, will need
both tests. Moreover, only five demonstrated one or more language intervention in the future to address academic
clinical markers of LI in conversational language samples. It needs. School-based speech-language pathologists who serve
should be noted that these measures focus on syntactic and students with ADHD in upper elementary and middle school
semantic language skills. No direct child assessment of settings will need to monitor higher level language functions,
pragmatic language skills was completed. It is interesting including literacy skills, as curriculum demands increase.
that four participants with ADHD, who scored within the Participants with ADHD were recruited from a
typical range on the language tests, had GCC scores below university research clinic, and, as such, parents of these
the clinical cutoff. Three of these participants had SIDI participants were already concerned about one or more
scores suggestive of pragmatic LI. The fourth participant had aspects of their children’s development. Parents reporting
clinical markers of LI in the language sample. These results about a clinical sample, such as the one recruited for
suggest that the CCC–2 may have identified impairment this study, are more likely to identify deficits (in this case,
not identified by the CELF–4 and TNL. Certainly, these LI) than parents who are reporting about children from

Table 5. Performance on direct language measures by the 14 participants in the ADHD group and one participant in the TD group with a GCC
score ≤ 85.

CCC–2 CELF–4 TNL

a a
Group GCC SIDI Core Lang. NLAIa MLUmb NDW-100b Composite: Tense useb

ADHD+LI
30LI 85 –19 88 76 0.26 0.15 0.55
25LI 84 –4 75 76 –0.83 –0.82 –0.23
42LI 83 –9 81 82 –0.12 –0.07 0.55
9LI (S/L)c 81 –6 91 79 –0.64 –0.61 0.55
5LI (OHI)d 78 –17 91 85 –0.84 –0.93 0.55
31LI (OHI)d 77 17 62 97 –0.85 –1.47 –1.91
16LI 74 0 84 82 –0.89 –0.79 –1.17
6LI 71 –4 85 79 –1.26 –1.00 –0.86
41LI (OHI)d 70 3 78 97 –1.05 0.55
10LI 59 2 72 82 –0.02 –1.18 0.02
ADHD
44 83 –15 111 94 0.13 0.76 0.55
29 83 –12 118 109 –0.97 –1.33 0.55
4 79 –13 97 100 –0.67 0.43 0.55
14 78 –5 106 97 –1.87 –1.40 –1.68
TD
45 84 4 102 118 0.79 0.61 –0.15

Note. Numbers in boldface type represent performance ≤1 SD on relevant measures or scores in a clinical range. Digits in the far left column are
participant identification numbers. SIDI = Social Interaction Difference Index (scores between –10 and 10 are obtained by 90% of the normative
population; scores ≥ 11 suggest syntactic/semantic impairment; scores ≤ –11 suggest pragmatic language impairment); Lang. = Language.
a
M = 100, SD = 15. bz scores. cChild receives school services under the category Speech or Language Impairment (S/L). dChild receives school
services under the category Other Health Impairment (OHI).

80 American Journal of Speech-Language Pathology • Vol. 23 • 73–83 • February 2014

Downloaded From: http://ajslp.pubs.asha.org/ by Miami University of Ohio - Library, Geralyn Timler on 05/13/2014
Figure 1. Receiver operating characteristic (ROC) curve associated omitted in clinic practice settings (Hill, 2000; Mueller &
with CCC–2 GCC discrimination of LI according to a reference
standard of 85 or lower on the CELF–4 or the TNL. CCC–2 =
Tomblin, 2012). Use of a parent report measure, such as
Children’s Communication Checklist—2; GCC = General Communi- the CCC–2, which requires approximately 10 to 15 min
cation Composite; CELF–4 = Clinical Evaluation of Language from a caregiver and 5 min from an examiner (if scoring
Fundamentals, Fourth Edition; TNL = Test of Narrative Language. software is used), is a time-efficient method of classifying
young children’s risk status. The CCC–2 requires relatively
little time to complete; however, professionals who diag-
nose ADHD will need to determine an optimal time to ask
parents to complete this measure. Parents complete a
number of report measures related to social skills and
problem behaviors during the ADHD diagnostic process,
and thus it may be prudent to send parents the CCC–2
Caregiver Response form before the first diagnostic
appointment or schedule a follow-up feedback appoint-
ment. The examiner’s manual suggests that if a form is sent
by mail it is helpful to include a cover letter to describe the
purpose of the measure, including why the information will
be important to the assessment process (Bishop, 2006).
Although it is likely that use of the CCC–2 will over-
identify the number of children at risk for LI, the risk of
underidentification through lack of a systematic screening
procedure is more problematic. Undetected LI could nega-
tively influence and impede the potential success of common
community samples. Nevertheless, the proportion of chil- behavioral treatments for children with ADHD because
dren in this study identified with LI in syntactic and semantic many of these treatments rely on talking and listening skills
domains—31% of the sample (nearly 44% if one considers (McInnes et al., 2003). Moreover, early identification of LI
the three children with CCC–2 profiles suggestive of prag- and receipt of appropriate speech and language services
matic LI)—is consistent with previous research (Mueller & could reduce the academic and social consequences of having
Tomblin, 2012). Certainly, replication of this study with a both ADHD and LI. Further development and examination
larger sample is needed to evaluate the sensitivity and spec- of accurate and efficient language screening protocols are
ificity of the CCC–2 for identification of LI in children with sorely needed.
ADHD.

Clinical and Research Implications Acknowledgments

Despite this study’s limitations, the findings have This study was supported by a New Century Scholars Research
important clinical and research implications. Although best Grant from the ASHFoundation. Appreciation is extended to
practice guidelines call for assessment of language and William Pelham and his staff at the University at Buffalo’s Center
for Children and Families (now located at Florida International
learning disorders in children receiving evaluation for
University) for assistance with participant recruitment as well as to
attention and behavior issues, such testing is frequently Dawn Vogler-Elias and Nikki Harty for assistance in data reduction
and supervision of student research assistants.
Figure 2. Interactive dot diagram demonstrating cases correctly and
incorrectly classified by CCC–2 GCC scores ≤ 85. Sens = sensitivity;
Spec = specificity. References
American Academy of Pediatrics, Subcommittee on Attention-
Deficit/Hyperactivity Disorder, Steering Committee on Quality
Improvement and Management. (2011). ADHD: Clinical practice
guideline for the diagnosis, evaluation, and treatment of
attention-deficit/hyperactivity disorder in children and adoles-
cents. Pediatrics, 128, 1007–1022. doi:10.1542/peds.2011-2654
American Psychiatric Association. (2000). Diagnostic and statistical
manual of mental disorders (4th ed., text rev.). Washington, DC:
Author.
Bishop, D. (1998). Development of the Children’s Communication
Checklist (CCC): A method for assessing qualitative aspects
of communicative impairment in children. Journal of Child
Psychology and Psychiatry, 39, 879–891. doi:10.1017/
S0021963098002832
Bishop, D. (2006). Children’s Communication Checklist—2 (U.S. ed.).
San Antonio, TX: Harcourt Assessment.

Timler: Language Impairment Risk and Children With ADHD 81

Downloaded From: http://ajslp.pubs.asha.org/ by Miami University of Ohio - Library, Geralyn Timler on 05/13/2014
Bishop, D. (2013). Children’s Communication Checklist (CCC–2). A pilot study of the Children’s Communication Checklist
In F. Volkmar (Ed.), Encyclopedia of autism spectrum disorders in a Norwegian sample. Logopedics Phoniatrics Vocology, 32,
(pp. 614–618). Berlin, Germany: Springer-Verlag. 23–30.
Bishop, D., & Baird, G. (2001). Parent and teacher report of Hill, G. (2000). A role for the speech-language pathologist in
pragmatic aspects of communication: Use of the Children’s multidisciplinary assessment and treatment of attention-deficit/
Communication Checklist in a clinical setting. Developmental hyperactivity disorder. Journal of Attention Disorders, 4,
Medicine & Child Neurology, 43, 809–819. 69–79.
Bishop, D., & McDonald, D. (2009). Identifying language im- Hunt, K. (1970). Syntactic maturity in school children and adults.
pairment in children: Combining language test scores with Monographs of the Society for Research in Child Development,
parental report. International Journal of Language & Com- 35(Serial No. 134).
munication Disorders, 44, 600–615. doi:10.1080/ Jonsdottir, S., Bouma, A., Sergeant, J., & Scherder, E. (2005). The
13682820802259662 impact of specific language impairment on working memory in
Bruce, B., Thernlund, G., & Nettelbladt, U. (2006). ADHD and children with ADHD combined subtype. Archives of Clinical
language impairment: A study of the Parent Questionnaire FTF Neuropsychology, 20, 443–456. doi:2005-06391-00410.1016/
(Five to Fifteen). European Child and Adolescent Psychiatry, 15, j.acn.2004.10.004
52–60. Kaplan, B., Crawford, S., Dewey, D., & Fisher, G. (2000). The IQs of
Centers for Disease Control and Prevention. (2010). Attention-deficit/ children with ADHD are normally distributed. Journal of
hyperactivity disorder (ADHD): Data and statistics. Retrieved Learning Disabilities, 33, 425–432.
from www.cdc.gov/NCBDDD/adhd/data.html Kaufman, A., & Kaufman, N. (2004). Kaufman Brief Intelligence
Cohen, N., Barwick, M., Horodezky, M., Vallance, D., & Im, N. Test (2nd ed.). Minneapolis, MN: NCS Pearson.
(1998). Language, achievement, and cognitive processing of Kim, O., & Kaiser, A. (2000). Language characteristics of children
psychiatrically disturbed children with previously unidentified with ADHD. Communication Disorders Quarterly, 21, 154–165.
and unsuspected language impairments. Journal of Child Leonard, M., Milich, R., & Lorch, P. (2011). The role of pragmatic
Psychology and Psychiatry, 39, 865–877. language use in mediating the relation between hyperactivity and
Cohen, N., Davine, M., Horodezky, N., Lipsett, L., & Isaacson, L. inattention and social skills problems. Journal of Speech,
(1993). Unsuspected language impairment in psychiatrically Language, and Hearing Research, 54, 567–579. doi:10.1044/
disturbed children: Prevalence and language and behavioral 1092-4388(2010/10-0058)
characteristics. Journal of the American Academy of Child & Luo, F., & Timler, G. (2008). Narrative organization skills in
Adolescent Psychiatry, 32, 595–603. children with attention deficit hyperactivity disorder and
Cohen, N., Davine, M., & Meloche-Kelly, M. (1989). Prevalence of language impairment: Application of the causal network model.
unsuspected language disorders in a child psychiatric popula- Clinical Linguistics & Phonetics, 22, 25–46.
tion. Journal of the American Academy of Child & Adolescent McInnes, A., Humphries, T., Hogg-Johnson, S., & Tannock, R.
Psychiatry, 28, 107–111. (2003). Listening comprehension and working memory are
Cohen, N., Vallance, D., Barwick, M., Im, N., Menna, R., impaired in attention-deficit hyperactivity disorder irrespective
Horodezky, N., & Isaacson, L. (2000). The interface between of language impairment. Journal of Abnormal Child Psychology,
ADHD and language impairment: An examination of language, 31, 427–443.
achievement, and cognitive processing. Journal of Child MedCalc Software. (1993). MedCalc for Windows (Version 12.3.0)
Psychology and Psychiatry, 41, 353–362. [Computer software]. Mariakerke, Belgium: Author.
Dollaghan, C. (2004). Evidence-based practice in communication Milch-Reich, S., Campbell, S., Pelham, W., Connelly, L., & Geva, D.
disorders: What do we know, and when do we know it? Journal (1999). Developmental and individual differences in children’s
of Communication Disorders, 37, 391–400. doi:10.1016/j.jcomdis. on-line representations of dynamic social events. Child Devel-
2004.04.002 opment, 70, 413–431.
Geurts, H., & Embrechts, M. (2008). Language profiles in ASD, Miller, J., & Iglesias, A. (2008). Systematic Analysis of Language
SLI, and ADHD. Journal of Autism and Developmental Transcripts (Research Version 9.1) [Computer software]. Mid-
Disorders, 38, 1931–1943. dleton, WI: SALT Software.
Geurts, H., Verte, S., Oosterlaan, J., Roeyers, H., Hartman, C., Mueller, K., & Tomblin, B. (2012). Examining the comorbidity of
Mulder, E., & Sergeant, J. (2004). Can the Children’s Commu- language impairment and attention-deficit/hyperactivity disor-
nication Checklist differentiate between children with autism, der. Topics in Language Disorders, 32, 228–246.
children with ADHD, and normal controls? Journal of Child Oram, J., Fine, J., Okamoto, C., & Tannock, R. (1999). Assessing
Psychology and Psychiatry, 45, 1437–1453. the language of children with attention deficit hyperactivity dis-
Gillam, R., & Pearson, N. (2004). Test of Narrative Language. order. American Journal of Speech-Language Pathology, 8,
Austin, TX: Pro-Ed. 72–80.
Green, S., & Salkind, N. (2005). Using SPSS for Windows and Purvis, K., & Tannock, R. (1997). Language abilities in children with
Macintosh: Analyzing and understanding data (4th ed.). Upper attention deficit hyperactivity disorder, reading disabilities, and
Saddle River, NJ: Pearson Prentice Hall. normal controls. Journal of Abnormal Child Psychology, 25,
Hadley, P. (1998). Language sampling protocols for eliciting text- 133–144.
level discourse. Language, Speech, and Hearing Services in Redmond, S. (2004). Conversational profiles of children with
Schools, 29, 132–147. ADHD, SLI and typical development. Clinical Linguistics &
Helland, W., Biringer, E., Helland, T., & Heimann, M. (2012). Phonetics, 18, 107–125.
Exploring language profiles for children with ADHD and Redmond, S. (2005). Differentiating SLI from ADHD using
children with Asperger syndrome. Journal of Attention Dis- children’s sentence recall and production of past tense
orders, 16, 34–43. morphology. Clinical Linguistics & Phonetics, 19, 109–127.
Helland, W., & Heimann, M. (2007). Assessment of pragmatic Redmond, S., Thompson, H., & Goldstein, S. (2011). Psycholin-
language impairment in children referred to psychiatric services: guistic profiling differentiates specific language impairment from

82 American Journal of Speech-Language Pathology • Vol. 23 • 73–83 • February 2014

Downloaded From: http://ajslp.pubs.asha.org/ by Miami University of Ohio - Library, Geralyn Timler on 05/13/2014
typical development and from attention-deficit/hyperactivity appropriate? Language, Speech, and Hearing Services in Schools,
disorder. Journal of Speech, Language, and Hearing Research, 37, 61–72.
54, 99–117. Swets, J. (1988, June 3). Measuring the accuracy of diagnostic
Rice, M., & Wexler, K. (2001). Test of Early Grammatical systems. Science, 240, 1285–1293.
Impairment. San Antonio, TX: The Psychological Corporation. Tannock, R., Purvis, K., & Schachar, R. (1993). Narrative abilities
Rice, M., Wexler, K., & Hershberger, S. (1998). Tense over time: in children with attention deficit hyperactivity disorder and normal
The longitudinal course of tense acquisition in children with peers. Journal of Abnormal Child Psychology, 21, 103–117.
specific language impairment. Journal of Speech, Language, and Tannock, R., & Schachar, R. (1996). Executive dysfunction as an
Hearing Research, 41, 1412–1431. underlying mechanism of behaviour and language problems
Sackett, D., Straus, S., Richardson, W., Rosenberg, W., & Haynes, R. in attention deficit hyperactivity disorder. In J. H. Beitchman,
(2000). Evidence-based medicine: How to practice and teach it. N. J. Cohen, M. M. Konstantareas, & R. Tannock (Eds.),
Edinburgh, United Kingdom: Churchill Livingstone. Language, learning, and behaviour disorders: Developmental,
Sattler, J. (1988). Assessment of children (3rd ed.). San Diego, CA: biological, and clinical perspectives (pp. 128–155). Cambridge,
Author. United Kingdom: Cambridge University Press.
Semel, E., Wiig, E., & Secord, W. (1995). Clinical Evaluation of Tirosh, E., & Cohen, A. (1998). Language deficit with attention-
Language Fundamentals, Third Edition. San Antonio, TX: deficit disorder: A prevalent comorbidity. Journal of Child
The Psychological Corporation. Neurology, 13, 493–497. doi:10.1177/088307389801301005
Semel, E., Wiig, E., & Secord, W. (2003). Clinical Evaluation of Vallence, D., Im, N., & Cohen, N. (1999). Discourse deficits asso-
Language Fundamentals, Fourth Edition. San Antonio, TX: The ciated with psychiatric disorders and with language impair-
Psychological Corporation. ments children. Journal of Child Psychology and Psychiatry, 40,
Shaffer, D., Fisher, P., Lucas, C., Dulcan, M., & Schwab-Stone, M. 693–704.
(2000). National Institute of Mental Health Diagnostic Interview Wechsler, D. (2003). Wechsler Intelligence Scale for Children—
Schedule for Children Version IV: Description, differences from Fourth Edition. San Antonio, TX: Harcourt Assessment.
previous versions, and reliability of some common diagnoses. Youden, W. (1950). An index for rating diagnostic tests. Cancer, 3,
Journal of the American Academy of Child & Adolescent Psychiatry, 32–35.
39, 28–38. Zentall, S. (1988). Production deficiencies in elicited language but
Spaulding, T., Plante, E., & Farinella, K. (2006). Eligibility criteria not in the spontaneous verbalizations of hyperactive children.
for language impairment: Is the low end of normal always Journal of Abnormal Child Psychology, 16, 657–673.

Timler: Language Impairment Risk and Children With ADHD 83

DownloadedViewFrom:
publicationhttp://ajslp.pubs.asha.org/
stats by Miami University of Ohio - Library, Geralyn Timler on 05/13/2014