Age 17 Language and Reading

Outcomes in Late-Talking Toddlers:

Support for a Dimensional Perspective
on Language Delay

Leslie Rescorla
Bryn Mawr College,
Bryn Mawr, PA Purpose: This study examined whether late talkers identified at 2431 months
continued to have weaker language and reading skills at 17 years of age than
typically developing peers.
Method: Language and reading outcomes at 17 years of age were examined in
26 children identified as late talkers with normal nonverbal ability and normal
receptive language at intake and in 23 typically developing children matched at
intake on age, socioeconomic status (SES), and nonverbal ability.
Results: Although late talkers performed in the average range on all language and
reading tasks at 17 years of age, they obtained significantly lower Vocabulary/
Grammar and Verbal Memory factor scores than SES-matched peers. The age 17
Vocabulary/Grammar factor had large correlations with the age 17 Verbal Memory
and Reading/Writing factors. The age 17 Vocabulary/Grammar and Reading/
Writing factors were strongly predicted by comparable factors at 13 years of age.
Age 2 Language Development Survey ( L. Rescorla, 1989) vocabulary score
explained 17% of the variance in the age 17 Vocabulary/Grammar and Verbal
Memory factors.
Conclusions: Results suggest that slow language development at 2431 months is
associated with a weakness in language-related skills into adolescence relative to
skills manifested by typically developing peersfindings that are consistent with a
dimensional perspective on language delay.
KEY WORDS: late talkers, adolescent outcomes, early language delay

C
hildren with normal nonverbal cognitive ability, adequate hear-
ing, and typical personality development who have delayed
receptive and /or expressive language skills are generally
referred to as having specific language impairment (SLI; Bishop, 2006;
Leonard, 1998.). When such children are identified at 23 years of age,
they have often been referred to as late talkers, a term that is more de-
scriptive and less diagnostic in nature. One reason this appellation de-
veloped was because of a belief held by many that SLI cannot be reliably
diagnosed at 2 years of age. However, the same criteria are used for
diagnosing a late talker and a preschooler with SLInamely a sig-
nificant language delay in the absence of a more primary condition. A
major justification used for making a diagnostic distinction between
these two groups has been that late talkers appear to manifest a better
outcome than preschoolers with SLI (Paul, 1996; Rescorla & Lee, 2000;
Whitehurst & Fischel, 1994). Evaluation of the merits of this diagnostic

16 Journal of Speech, Language, and Hearing Research Vol. 52 1630 February 2009 D American Speech-Language-Hearing Association
1092-4388/09/5201-0016
distinction may depend on whether one espouses a cat-
egorical or a dimensional theoretical perspective on
language delay.
Contrasting Theoretical Perspectives
on Language Delay
Viewing late talkers and preschoolers with SLI as
manifesting two distinct disorders suggests a categorical
perspective toward language delay whereby the two con-
ditions represent qualitatively different disorders hav-
ing different etiologies and outcomes. The alternative
dimensional theoretical formulation postulates a spec-
trum of language ability whereby late talkers and pre-
schoolers with SLI differ along a quantitative dimension,
with both groups having weaker language ability than
typically developing children.
Researchers who take the categorical approach to
language impairment (Gopnik & Crago, 1991; Rice, 2003;
Rice & Wexler, 1996) emphasize as central to defining SLI
such discrete and focal grammatical deficits as the ex-
tended optional infinitive. Research based in this categor-
ical approach tends to look for clinical markers and specific
genetic causes for what is conceptualized as a discrete
disorder (Bishop, 2006). As Dollaghan (2004) noted, the-
orizing that SLI constitutes a discrete category promotes
research aimed at specifying a unique phenotype,
etiology, base rate, and treatment regimen (p. 464).
In contrast, the dimensional theoretical formulation
postulates a spectrum of language ability. Early ac-
counts of this dimensional perspective were offered by
Bishop and Edmundson (1987), who argued for a spec-
trum of impairment for language problems, and by
Leonard (1987, 1991), who argued that SLI represents
the tail of a normal distribution of language abilities.
Dollaghan (2004, p. 464) noted that the dimensional the-
oretical approach implies the potential for more het-
erogeneity in symptoms, origins, and causal influences
than does the categorical approach. Bishop (2006, p. 220)
used findingssuch as the finding that one monozygotic
twin may be diagnosable with SLI, whereas her co-twin
may have only subclinical language problemsto argue
that investigations of SLI should be looking for dimen-
sions of impairment rather than discrete subtypes.
The dimensional perspective has also been outlined
in Rescorla (2002, 2005), in which it was proposed that
late talkers have below average endowment in a set of
intercorrelated yet diverse language-related abilities,
analogous to differential endowments for intelligence.
The term endowment conveys the assumption that in-
dividual differences in language ability are to some ex-
tent constitutionally basedan assumption supported
by findings from genetic studies, such as those by Dale
et al. (1998) and Bishop, Price, Dale, and Plomin (2003).
Children with more severe and prolonged language de-
lays have weaker skills in a wider variety of language-
related abilities and hence fall further from the mean on
the impaired side of the distribution than children
whose delays are milder or more short-lived. Toddlers
who have receptive as well as expressive delays and tod-
dlers whose expressive delays last long enough for them
to be diagnosed with SLI at 4 years of age are both pre-
sumed to fall further away from the mean on this ability
spectrum than late talkers with no receptive problems
whose delays resolve by 4 years of age.
Family environment factors and speech-language
therapy clearly affect language skills (Girolametto, Pearce,
& Weitzman, 1996; Hart & Risley, 1995). Furthermore,
changes in language input in the home or provision of
speech-language therapy (in addition to the effects of
time) may result in many preschoolers with SLI as well
as many late-talking toddlers moving into the average
range of language skills as they get older. Nonetheless,
the dimensional approach predicts that preschoolers with
SLI or late talkers are very likely to continue to have
weaker language skills than comparison peers who never
manifested language delay because of what is presumed
to be their more compromised language endowment. Out-
come studies of early language delay therefore provide a
crucial test for determining whether this prediction from
the dimensional approach is empirically supported by
longitudinal evidence.
Outcomes for Children With SLI
Although many children diagnosed with SLI as pre-
schoolers have persisting language impairment and later
learning problems, many do not (Aram, Ekelman, &
Nation, 1984; Aram & Hall, 1989; Bishop & Adams, 1990;
Hall & Tomblin, 1978). Despite the fact that a significant
percentage of children with SLI move into the average
range over time, many outcome studies of children with
SLI indicate considerable stability in individual differ-
ences in language skills. Some representative studies
are summarized below.
The study of preschool SLI with the most compre-
hensive follow-up data is that of Bishop and colleagues
(Bishop & Edmundson, 1987). When the children were
seen at age 52 years, 44% of the sample of 68 four-year-olds
with SLI were judged to have good outcome (i.e., above
the 10th percentile in expressive language skills; Bishop
& Edmundson, 1987). At age 82, years, this age 5 good
outcome group performed at an average level on all lan-
guage and reading measures (Bishop & Adams, 1990). In
contrast, the 56% of the sample with a poor outcome at
age 5 years (i.e., scores below the 10% percentile) per-
formed significantly worse than controls on 9 out of
11 measures at age 82 years. Not surprisingly, at age
15 years (Stothard, Snowling, Bishop, Chipchase, &
Rescorla: Age 17 Language and Reading Outcomes 17
Kaplan, 1998), the age 5 poor outcome children were
worse than controls on virtually all language and aca-
demic measures. More surprisingly, the age 5 good
outcome children were significantly different from con-
trols on sentence and nonword repetition, spoonerisms,
and reading/spelling, despite their average scores. This
finding indicates that even though the good outcome
group did not continue to have a significant language
delay, their skills at age 15 years were weaker than those
of peers with typical language histories.
Tomblin, Zhang, Buckwalter, and OBrien (2003)
reported outcomes at 910 years of age (during fourth
grade) for 180 children who were diagnosed with lan-
guage disorder at 56 years of age (during kindergarten),
116 of whom met criteria for SLI and 64 of whom had
performance IQs below 87 and hence were diagnosed with
nonspecific language impairment (NLI). For the full sam-
ple of children with SLI, NLI, and normal language
history, composite language scores in kindergarten were
correlated at r = .77 with composite language scores in
fourth grade, indicating strong continuity in rank or-
dering of language skills over time. Although only 52%
of kindergarten children with SLI/NLI continued to meet
the criterion for language impairment in fourth grade,
their mean composite language score was 1.20 SDs
below the local mean (or well below average). NLI out-
comes were consistently worse than SLI outcomes,
whether determined dichotomously (77% vs. 44% im-
paired in fourth grade) or continuously (1.45 vs. 1.06
SDs). These findings suggest that the children with NLI
fell further away from the mean on the impaired tail of
the language ability spectrum than the children with
SLI but that both groups occupied places on this spec-
trum below those occupied by children with typical lan-
guage histories.
Outcomes for Late Talkers
Several studies of late talkers have reported out-
comes to age 5 years. For example, Girolametto, Wiigs,
Smyth, Weitzman, and Pearce (2001) reported age 5 out-
comes for 21 children identified as late talkers at age
2433 months, all of whom scored below the 5th percen-
tile on the Communicative Development Inventory (CDI;
Fenson et al., 1993). These children had participated in
an 11-week parent-based intervention program at age
2 years; 13 of the 21 children received subsequent speech-
language services. At age 5 follow-up, most of the late
talkers scored in the normal range on various language
measures. However, when compared at age 5 years
with children who had typical language histories, the
late talkers scored significantly lower on most language
measuresparticularly those tapping more complex
skills, such as narrating a story. Of the 10 significant ef-
fects, 9 had Cohens (1988) d values > 0.50, with the largest
18 Journal of Speech, Language, and Hearing Research Vol. 52 1630 February 2009
effect size (ES) on the Test of Language Development
Primary, Second Edition ( TOLD-P:2; Newcomer &
Hammill, 1988b) Oral Vocabulary subtest (1.33).
Moyle, Ellis Weismer, Lindstrom, and Evans (2007)
reported age 5 outcomes for 30 late talkers first iden-
tified at age 2 years, all of whom scored 10th percentile
on the CDI (Fenson et al., 1993). Children in the typ-
ically developing group had all scored 20th percentile
on the CDI at age 2 years and were matched with the late
talkers on age, socioeconomic status (SES), gender, and
nonverbal cognitive ability. At age 5 years, late talkers
had significantly lower scores on the Test of Language
DevelopmentPrimary, Third Edition ( TOLD-P:3;
Newcomer & Hammill, 1997) Oral Vocabulary, Grammatic
Completion, and Sentence Imitation subtests. Although
ESs for these group differences were not reported, in-
spection of the means and standard deviations provided
indicates that the Cohens d values were very large: Oral
Vocabulary, d = 0.97; Grammatic Completion, d = 1.46;
and Sentence Imitation, d = 1.52.
Only a few researchers have reported follow-up data
for late-talking toddlers past age 6 years. Whitehurst
and Fischel (1994) reported age 7 school-administered
standardized test scores for 22 of their late talkers.
Although the group was above average in both reading
and math, performance was significantly better in math
than in reading, suggesting some enduring weakness in
language-related skills.
Paul and colleagues studied a group of about 30 late
talkers and a matched comparison group (Paul, 1993;
Paul, Murray, Clancy, & Andrews, 1997). In Pauls (1993)
cohort, 25% of the late talkers were delayed in receptive
as well as expressive language. At age 7 years, 84% of
Pauls late talkers scored above the 10th percentile in
Developmental Sentence Score (Lee, 1974; i.e., had
recovered). The recovered group was similar to con-
trols on receptive language, reading, spelling, IQ, and
phonological skills. However, they had significantly lower
scores than controls on the Test of Language Development
(TOLD; Newcomer & Hammill, 1988a) Expressive Lan-
guage scale. The age 7 still delayed late talkers were
worse than comparison children on everything except
receptive language and reading/spelling.
Armstrong, Marchman, and Owen (2007) recently
reported findings through fifth grade for late talkers
from the National Institute of Child Health and Human
Development Early Child Care Research Network data
set. On the basis of CDI scores collected at 24 months,
the sample of 689 children was divided into (a) late talkers,
131 toddlers who scored 10th percentile on the CDI at
24 months and who scored <85 on the Reynell Expres-
sive Language Scale (Reynell & Gruber, 1990) at 36 and
54 months; ( b) late bloomers, 39 children who scored
10th percentile on the CDI at 24 months and who
scored 85 on the Reynell Expressive Language Scale
at 54 months; and (c) typically developing children,
558 children who scored 10th percentile on the CDI at
24 months and who scored 85 on the Reynell Expres-
sive Language Scale at 54 months. Differences between
the three groups persisted through fifth grade on the
WoodcockJohnsonRevised ( WJ-R; Woodcock &
Johnson, 1989) Picture Vocabulary, LetterWord Iden-
tification, and Memory for Sentences subtests, although
the late-talker group scored in the average range on the
first two subtests at all time points. On Memory for Sen-
tences, the late talkers scored below the average range
at 54 months and at first grade. For all measures, the
late-talker group performed worst, the typically devel-
oping group performed best, and the late-bloomer group
performed in between the other two groups. The gaps
between the three groups evident at 54 months did not
change significantly through fifth grade. These data
provide strong support for the dimensional perspective
of early language delay, with the late bloomers having a
mild delay and the late talkers having more severe lan-
guage weaknesses, both of which persisted over time.
Rescorla (2002) reported school-age outcome data
for 34 late talkers and 25 comparison children matched
at intake on age, SES, and norverbal cognitive ability. By
age 6 years (Rescorla, 2002), only 6% of the late talkers
had scores on at least two TOLD (Newcomer & Hammill,
1988a) subtests below the 10th percentile benchmark.
Nevertheless, the group means on language measures
were significantly lower than those for typically develop-
ing comparison children. Significant group differences
were found on aggregate Vocabulary score at ages 6, 7,
and 8, with Cohens ds of 0.85, 1.53, and 1.05, respec-
tively. Significant group differences on the Grammar
aggregate were found at 6 and 8 years of age, with
d values of 0.64 and 0.94. The group difference on the
Phonological aggregate was significant at age 6 years
(d = 0.91), as was the group difference on TOLD Sentence
Imitation at age 6 years (d = 1.26). Although negligible
numbers of children in the late-talker group scored below
the 10th percentile on reading measures, significant group
differences for the Reading aggregate were found at
ages 8 and 9 (ds = 0.84 and 0.72, respectively).
Rescorla (2002) reported high correlations among
the various language and reading outcome measures
given from ages 6 to 9 years. The correlations that were
significant at p < .01 (66 of 78) ranged from .33 to .99,
with 32 being large effects (.50) according to Cohen
(1988). These correlations indicate some shared variance
among diverse language and reading measures. Rescorla
also presented regression results, with predictors entered
in chronological order. Vocabulary score on the Language
Development Survey ( LDS; Rescorla, 1989) at 2 years of
age, grammar skills at 3 years of age as measured by
mean length of utterance ( MLU) and scores on the Index
of Productive Syntax ( IPSyn; Scarborough, 1990), and
grammar skills at age 5 years as measured by the Pat-
terned Elicitation Syntax Test (Young & Perachio, 1983)
collectively explained 35% of the variance in age 8 scores
on the Clinical Evaluation of Language Fundamentals
Revised (CELF-R; Semel, Wiig, Secord, & Sabers, 1987).
Age 2 LDS score significantly predicted age 9 reading out-
come on the WJR (Woodcock & Johnson, 1989), but its
beta weight declined once age 5 reading and phonology
scores were entered as predictors. These three predictors
explained 54% of the variance in age 9 reading.
Rescorla (2005) reported language and reading out-
comes at age 13 years for 28 late talkers and 25 typically
developing children from the initial intake sample, with
the two groups matched at intake on age, SES, and non-
verbal ability. As a group, late talkers performed in the
average range on all standardized language and reading
tasks at age 13 years. However, they scored significantly
lower than SES-matched peers on aggregate measures
of Vocabulary, Grammar, and Verbal Memory, as well as
on Reading Comprehension. They were similar to compar-
ison peers in Reading Mechanics and Writing. Correlations
among outcome measures were moderately high, suggest-
ing some shared variance. Regression analyses indicated
that age 2 LDS vocabulary score was a significant predictor
of age 13 Vocabulary, Grammar, Verbal Memory, and Read-
ing Comprehension. Thus, findings suggested that slow
language development at age 2 to 22 years was asso-
ciated with a weakness in language-related skills into
early adolescence relative to typically developing peers.
In summary, Rescorlas (2005) age 13 findings are
consistent with Stothard et al. (1998), Tomblin et al.
(2003), Girolametto et al. (2001), Moyle et al. (2007),
Armstrong et al. (2007), and Paul et al. (1997). Findings
from all of these studies support the dimensional ac-
count of early language delay. According to this dimen-
sional view, children vary widely in their language
ability, from seriously impaired at the lower tail to ex-
tremely gifted at the upper tail. For clinical purposes, a
cutpoint on what is essentially a continuous dimension
may be needed to select children most in need of inter-
vention. However, proponents of a dimensional view
would argue that drawing such a cutpoint imposes a
dichotomy on what is essentially a continuum. The stud-
ies reviewed above indicated that many children with
clinically significant early language delay had moved
into the normal range at follow-up. Nonetheless, consis-
tent with the dimensional view of language delay, children
in these studies who had early delays continued at follow-
up to have lower scores on a variety of different language
measures than children with typical language histories.
Children whose follow-up language status was still
significantly impaired had even lower language scores
than children who had recovered, suggesting that their
impairment was more long lasting and severe.
Rescorla: Age 17 Language and Reading Outcomes 19
Rationale and Goals of the Current Study
The research reported here examined language and
reading outcomes at age 17 years for 26 toddlers iden-
tified as late talkers when they were between 24 and
31 months of age and for 23 typically developing chil-
dren matched at intake on age, SES, and nonverbal
ability. With the exception of Rescorla (2002, 2005), who
reported age 9 and age 13 outcomes for virtually these
same cohorts, no other studies of late talkers identified
before age 3 years have followed children past age 10 years;
the only major follow-up study of preschoolers with SLI
into adolescence is that by Stothard et al. (1998). The
late talkers in the present study, all of whom had normal
language, were identified as delayed when they were
toddlers because they had very small vocabularies and /
or were not combining words into phrases, with most meet-
ing both criteria. Based on the dimensional view of lan-
guage delay articulated above, as well as on the results
of previous outcome studies, it was predicted that the
late talkers would perform close to the average range in
a wide variety of language skills at age 17 years but that
they would continue to manifest weaker skills than de-
mographically matched children with typical language
histories.
The current study tested the following specific
hypotheses: (a) Late talkers at age 17 years would con-
tinue to have weaker language and reading skills than
demographically matched typically developing compar-
ison children, despite scoring in the average range on
most age 17 measures, consistent with the dimensional
view outlined previously; (b) age 17 language and reading
outcome measures would be strongly intercorrelated, con-
sistent with the dimensional view that language ability
can be viewed as a set of interrelated yet diverse language-
related skills; (c) convergent and discriminant predictive
validity would be demonstrated when using regression
analysis to predict age 17 outcomes from age 13 predictors,
with measures tapping the same basic skills being more
highly associated over time than measures tapping related
but different skills; and (d) age 2 language predictors
would significantly predict age 17 outcomes, consistent
with the dimensional view that individual differences in
language skills show substantial consistency over time.
Method
Participants
Participants in this study were 26 late talkers and
23 comparison children matched at intake (2431 months)
on age, SES, and nonverbal ability. The original late-talker
cohort in this longitudinal study consisted of 40 late
talkers (36 boys and 4 girls), whereas the comparison
group consisted of 39 toddlers with normal language
20 Journal of Speech, Language, and Hearing Research Vol. 52 1630 February 2009
histories (38 boys and 1 girl). Follow-up data at age
13 years for 28 of the late talkers and 25 of the compar-
ison children were reported in Rescorla (2005). The
2 late talkers lost to follow-up between age 13 and age
17 years (a set of twins) had moved away and were not
able to return to the area for testing. The 2 comparison
children lost to follow-up between age 13 and age 17 years
had also apparently moved, as the families could not be
located. To test for the possibility of selective attrition,
t tests were used to compare the 26 late talkers seen at
age 17 years with the 14 not seen at 17 years of age on
intake measures for children 2431 months of age. Sim-
ilarly, the 23 comparison children seen at age 17 years
were compared with the 16 not seen. No differences even
approaching significance were found, indicating that
attrition was not selective in either group.
Late talkers were recruited through newspaper ad-
vertisements, notices to pediatricians, and a local infant
laboratory. All but 1 of the children in the original cohort
came from a two-parent, middle-to-upper-class White
family (the mother of 1 late talker was divorced). The
first inclusionary criterion was a score of greater than 85
on the Bayley Mental Development Scale (Bayley, 1969).
Late talkers also had to score within 34 months of
chronological age (CA) on the Reynell Receptive Lan-
guage Scale (Reynell, 1977); 25 of the 26 late talkers
scored within 3 months of CA, and 1 late talker scored
within 4 months of CA. Finally, late talkers had to score
at least 6 months below CA on the Reynell Expressive
Language Scale. Comparison children met the same cri-
teria, except that they had to have a Reynell Expressive
Language Scale score within 34 months of CA; 22 out of
23 comparison children scored within 3 months of CA,
and 1 scored within 4 months of CA.
Demographic information and test scores at intake
(2431 months) for the two age 17 participant groups
appear in Table 1. As toddlers, the late talkers and the
comparison children did not differ in age, Hollingshead
(1975) SES score, or total score on the 19 Bayley Non-
verbal items above the basal level for all children (e.g.,
towering blocks, doing puzzles, drawing, and inserting
pegs)as has been the case in previous reports of these
cohorts (Rescorla, 2002, 2005; Rescorla, Dahlsgaard, &
Roberts, 2000). Intake Reynell performance is reported
in Table 1 using z scores based on British norms, which
were the only norms available for the Reynell measures
at the time these intake data were collected; z scores
rather than raw scores are reported because they not
only allow for group comparisons but also indicate per-
formance of the two groups relative to children of the
same age in the standardization sample (albeit a British
one). As in previous publications, the groups differed sig-
nificantly in receptive language as measured by Reynell
Receptive Language Scale z score ( p < .001), even though
the late talkers had normal receptive skills for their age.
 Table 1. Means, standard deviations, and ranges for intake measures by group.

Late talkers (n = 26) Comparison (n = 23)

Variable M SD Range M SD Range

Intake age 26.62 2.30 24, 30 25.83 1.85 24, 31

Hollingshead total score 53.69 13.18 27, 66 55.52 10.47 35, 66
Bayley Nonverbal items score 14.15 2.57 10, 18 15.17 2.34 10, 19
Reynell Receptive Language Scale z score 0.15 0.49 0.6, 1.4 1.04 0.66 0.4, 2.0***
Reynell Expressive Language Scale z score 1.62 0.34 2.3, 1.2 0.36 0.50 0.7, 1.1***
LDS vocabulary score 24.54 23.62 5, 131 235.17 73.67 27, 319***

Note. LDS = Language Development Survey.

***p < .001 (by independent means t test).

When Cohens d (Cohen, 1988) was used to calculate the
ES for Reynell Expressive Language Scale z score (using
the difference between 1.62 and 0.36 divided by the
standard deviation of the typically developing group
only, because of the extreme restriction in range in the
late-talker group), the ES was large (3.96), as would be
expected given the use of Reynell Expressive Language
Scale score to define the groups. Of the 26 late talkers, 4
were 6 months delayed, and 22 were 713 months delayed
relative to their CA on the Reynell Expressive Language
Scale at intake, with all late talkers scoring at least
1.2 SDs below age expectations (<10th percentile), and
15 of the 26 scoring 1.5 SDs or more below age ex-
pectations (<7th percentile). Finally, the late talkers had
a mean vocabulary of 25 words on Rescorlas (1989) LDS,
in contrast to a mean vocabulary of 235 words for the
comparison group. Mothers of all late talkers reported
fewer than 50 words or no word combinations on the LDS.
By age 4 years, 29% of the late talkers had MLU
scores, and 71% had IPSyn (Scarborough, 1990) scores at
least 1.25 SDs below age expectations (Rescorla et al.,
2000). This indicates that the percentage of the sample
diagnosable with SLI at age 4 years varied with which of
these two measures was used. By age 6 years, about 17%
of the late talkers were judged to have SLI on the basis of
continuing significant delays in expressive language
based on their TOLD (Newcomer & Hammill, 1988a) per-
formance (Rescorla, 2002). Independent of their partici-
pation in this longitudinal study, about one third of the
late talkers received some speech-language therapy dur-
ing the preschool period, and a few participated in pre-
school programs for children with language impairments;
2 late talkers were in special education classes for chil-
dren with language impairments in the early elementary
school years. Some of these children received diagnoses of
SLI from their educational agencies, but diagnostic prac-
tices varied across school districts, and full diagnostic
information was not available for all children. A few
comparison children also received speech and language
therapy or reading remediation. Because degree of re-
mediation was so varied and because information about
interventions was incomplete, intervention was not an-
alyzed for this study.
Procedure
Age 17 participants were seen for a 2-hr session.
Numerous standardized tests were used to assess vocab-
ulary, grammar, verbal memory, and reading/writing
skills. Every child was tested by the author, a licensed
and certified school psychologist. The examiner was not
blind to the childrens language history, having seen
them at regular intervals since age 2 years.
Measures
Wechsler Adult Intelligence ScaleThird Edition
(WAIS-III; Wechsler, 1997a). Participants ability to de-
fine words on the WAIS-III Vocabulary subtest was used
to assess verbal conceptual ability. In addition, the Digit
Span subtest was used to assess recall of strings of digits
in forward and backward order. Finally, the Block De-
sign subtest was given to assess nonverbal ability.
Woodcock Johnson Psychoeducational Battery Tests
of AchievementThird Edition ( WJ-III; Woodcock,
McGrew, & Mather, 2001). The LetterWord Identifica-
tion, Reading Fluency, and Writing Fluency subtests were
administered to assess, respectively, ability to decode
words on a list, to read and comprehend short statements
under time pressure, and to write short statements using
target words under time pressure. In addition, to obtain a
second nonverbal measure, the WJ-III Math Fluency
subtest was given to assess the ability to perform arith-
metic calculations under time pressure.
Comprehensive Assessment of Spoken Language
(CASL; Carrow-Woolfolk, 1999). The Syntax Construc-
tion, Sentence Comprehension, Grammatical Judgment,

Rescorla: Age 17 Language and Reading Outcomes 21

and Ambiguous Sentences subtests were used to assess
grammatical skills. On Syntax Construction, partici-
pants had to formulate sentences given various pictorial
and verbal prompts. On Sentence Comprehension, par-
ticipants had to select which two of three orally pre-
sented sentences had the same meaning. On Grammatical
Judgment, participants had to decide whether orally
presented sentences were grammatically correct and, if
deemed incorrect, to fix them by changing one word. Fi-
nally, on Ambiguous Sentences, participants had to pro-
vide two different meanings for each of a series of orally
presented sentences.
Wechsler Memory ScaleThird Edition ( WMS-III;
Wechsler, 1997b). The Logical Memory subtest was given
to assess recall of an orally presented narrative, whereas
the Verbal Paired Associates subtest was given to assess
recall of arbitrary word pairs.
Data Reduction and Analysis
Eleven language, verbal memory, and reading/writing
subtests constituted the primary dependent variables in
this study (see Table 2). To reduce the risk of Type I
errors arising from performing many statistical tests,
these 11 subtests were aggregated into three composites.
The aggregation was based on a theoretical conceptual-
ization of the skills that the subtests tap, the known
properties of the subtests as presented in their respec-
tive test manuals, observed intercorrelations among the
subtests in this sample, and aggregation procedures
followed in Rescorla (2002, 2005). Principal components
analysis was then used to determine the percentage of
variance in each factor accounted for by the assigned
measures, to obtain loadings of each measure on its
factor, and to obtain factor scores for use in subsequent
analyses. Factor scores effectively weight the components
within a factor according to their degree of association
with the overall factor score. The principal components
analysis results for each factor are summarized below.
Five measures were assigned to a Vocabulary/Grammar
factor (WAIS-III Vocabulary plus CASL Syntax Construc-
tion, Sentence Comprehension, Grammatical Judgment,
and Ambiguous Sentences). These five subtests assigned
to the Vocabulary/Grammar factor accounted for 63%
of the variance in that factor, with the following fac-
tor loadings: Sentence Comprehension = 0.67, Syntax
Construction = 0.78, WAIS-III Vocabulary = 0.79, Grammat-
ical Judgment = 0.82, and Ambiguous Sentences = 0.87.
Cronbachs alpha was large for the five-item Vocabulary/
Grammar factor (.80).
Three measures were assigned to the Verbal Mem-
ory factor ( WAIS-III Digit Span, WMS-III Logical Mem-
ory, and Verbal Paired Associates). The three subtests
assigned to the Verbal Memory factor accounted for 54%

Table 2. Age 17 language and reading test scores: Means, standard deviations, ranges, and percentages below the 10th percentile by group,
plus Cohens d between groups.

Late talkers (n = 26) Comparison (n = 23)

% below % below
Measure M SD Range 10th percentilea M SD Range 10th percentilea d

Vocabulary/Grammar factor
WAIS-III Vocabulary subtest 13.50 2.64 9, 19 0 15.48 2.33 10, 19 0 0.80
CASL Syntax Construction subtest 103.77 14.84 76, 133 4 111.13 10.09 92, 129 0 0.59
CASL Sentence Comprehension subtest 102.88 11.60 84, 122 0 108.13 10.82 87, 127 0 0.47
CASL Grammatical Judgment subtest 103.36 14.06 74, 143 4 114.78 14.58 82, 138 0 0.80
CASL Ambiguous Sentences subtest 111.65 13.29 88, 129 0 119.43 9.33 101, 137 0 0.69
Verbal Memory factor
WAIS-III Digit Span subtest 10.35 3.14 4, 19 0 11.70 2.98 6, 18 0 0.44
WMS-III Logical Memory subtest 9.36 2.60 5, 14 17 11.87 2.05 9, 17 0 1.08
WMS-III Verbal Paired Associates subtest 10.31 3.72 1, 16 15 11.91 2.95 7, 16 0 0.48
Reading/Writing factor
WJ-III LetterWord Identification subtest 104.35 11.60 87, 126 0 110.39 12.32 86, 131 0 0.51
WJ-III Reading Fluency subtest 99.88 11.52 75, 120 12 104.17 14.21 72, 124 9 0.33
WJ-III Writing Fluency subtest 112.12 16.24 64, 133 4 112.96 10.67 93, 128 0 0.06

Note. WAIS-III = Wechsler Adult Intelligence ScaleThird Edition; CASL = Comprehensive Assessment of Spoken Language; WMS-III = Wechsler
Memory ScaleThird Edition; WJ-III = WoodcockJohnson Psychoeducational Battery Tests of AchievementThird Edition.
a
Refers to the percentage of children in each group scoring below the 10th percentile on each measure (<6 or <80).

22 Journal of Speech, Language, and Hearing Research Vol. 52 1630 February 2009
of the variance in that factor, with the following load-
ings: WAIS-III Digit Span = 0.59, WMS-III Logical Mem-
ory = 0.77, WMS-III Verbal Paired Associates = 0.83.
Cronbachs alpha was only moderate for the three-item
Verbal Memory factor (.57).
Finally, three measures were assigned to the
Reading/Writing factor ( WJ-III LetterWord Identifi-
cation, Reading Fluency, and Writing Fluency). The
three subtests assigned to the Reading/Writing factor
accounted for 55% of the variance in that factor, with the
following factor loadings: WJ-III LetterWord Identifi-
cation = 0.69, WJ-III Reading Fluency = 0.76, and WJ-III
Writing Fluency = 0.78. As with the Verbal Memory fac-
tor, Cronbachs alpha was only moderate for the three-
item Reading/Writing factor (.60).
Before conducting the main analyses, t tests on
WAIS-III Block Design and WJ-III Math Fluency scores
were used to determine whether the groups differed
significantly on nonverbal measures. To address the
first hypothesis, differences between the late-talker
and comparison groups were tested using t tests on the
Vocabulary/Grammar, Verbal Memory, and Reading/
Writing factors. Because 2 late talkers were not given
the WMS-III Logical Memory subtest, they would have
been excluded from all follow-up analyses had a multi-
variate analysis of variance been used. Because five sep-
arate t tests were run, p < .01 was used as the significance
level.
To address the second hypothesis, correlations among
the age 17 Vocabulary/Grammar, Verbal Memory, and
Reading/Writing factorsas well as among all 11 age 17
language-related measureswere obtained.
To address the third hypothesis, four of the age 13
aggregates analyzed by Rescorla (2005) were used as
predictors of the three age 17 outcomes in hierarchical
regressions. As described in Rescorlas study, the age 13
Vocabulary aggregate (Cronbachs a = .77) was computed
by averaging scores on the Wechsler Intelligence Scale
for ChildrenThird Edition ( WISC-III; Wechsler, 1991)
Vocabulary subtest and the Test of Adolescent and Adult
LanguageThird Edition ( TOAL-3; Hammill, Brown,
Larsen, & Wiederholt, 1994) Listening Vocabulary and
Reading Vocabulary subtests. The age 13 Grammar ag-
gregate (a = .71) was computed by averaging scores for the
TOAL-3 Listening Grammar and Reading Grammar sub-
tests, the Clinical Evaluation of Language Fundamentals
Third Edition (CELF-III; Semel, Wiig, & Secord, 1995)
Formulated Sentences subtest, and the Test of Language
Competence (TLC; Wiig & Secord, 1989) Ambiguous Sen-
tences subtest. The Verbal Memory aggregate (a = .74)
was composed of the WISC-III Digit Span subtest, the
CELF-III Recalling Sentences subtest, and a Pseudoword
Repetition task modified by Scarborough (1990) from a
procedure reported by Taylor, Lean, and Schwartz (1989).
The Reading Mechanics aggregate (a = .86) was com-
puted by averaging Wechsler Individual Achievement
Test (Wechsler, 1992) Reading and Spelling scaled scores
with scores from the Test of Word Reading Efficiency
(Torgesen, Wagner, & Rashotte, 1999) and with standard
scores for miscues and words per minute computed from
the Qualitative Reading InventoryII (Leslie & Caldwell,
1995).
To address the fourth hypothesis, correlations were
calculated between LDS score, Reynell Expressive Lan-
guage Scale and Reynell Receptive Language Scale scores,
and the three age 17 factor score outcomes. In addi-
tion, Bayley Nonverbal score was tested as a predictor
to examine whether age 2 nonverbal cognitive ability
also predicted age 17 outcomes. Follow-up regressions
were then run to examine how much unique variance in
age 17 outcomes were accounted for by these four age 2
predictors.
Results
Group Differences at Age 17 Years
Late talkers and comparison children did not differ
significantly on either WAIS-III Block Design or WJ-III
Math Fluency. Both groups scored somewhat above
average on Block Design, with the late talkers having
slightly lower scores ( late talkers: M = 12.04, SD = 2.63;
typically developing: M = 13.39, SD = 3.35), t(47) = 1.58,
p = .12. On Math Fluency, both groups scored in the
average range ( late talkers: M = 102.73, SD = 16.24;
typically developing: M = 101.91, SD = 12.31), t(47) = 0.19,
p = .85. Thus, as had been the case 15 years earlier, these
two demographically matched groups of children were
comparable in their nonverbal cognitive abilities.
As can be seen in Table 2, the late talkers generally
obtained scores in the average range or above (i.e., scores
8 or 90) on the language and reading/writing mea-
sures administered. However, on many of these mea-
sures, late talkers scored somewhat below comparison
children. To measure the magnitude of these differences,
Cohens (1988) d was calculated by subtracting the mean
score for the late-talker group from the mean score for
the comparison group and dividing by the pooled SD for
the sample. As shown in Table 2, the largest ds were
found for WAIS-III Vocabulary (d = 0.80), CASL Gram-
matical Judgment (d = 0.80), and WMS-III Logical Mem-
ory (d = 1.08). Table 2 also shows the percentage of
children in each group obtaining scores below the 10th
percentile (scores 6 or 80). These percentages indicate
that few late talkers scored below the 10th percentile on
any of the age 17 language and reading outcome mea-
sures. The most challenging tasks for the late talkers
appeared to be WMS-III Logical Memory and WMS-III
Rescorla: Age 17 Language and Reading Outcomes 23
Verbal Paired Associates, but even on these challenging
verbal memory tasks most late talkers scored in the
average range or above.
Late talkers obtained significantly lower scores than
comparison children on the age 17 Vocabulary/Grammar
factor, t(45.9) = 3.17, p < .01, d = 0.92, and the age 17
Verbal Memory factor, t(42.6) = 3.18, p < . 01, d = 0.93.
The group difference on the age 17 Reading/Writing fac-
tor was not significant, t(46.6) = 1.36, p = .18, d = 0.38.
Thus, despite scoring in the average range on all out-
come measures, late talkers scored significantly less
well than comparison children on factors tapping their
ability to provide word definitions and to perform a
variety of tasks tapping lexical and grammatical skills,
as well as on their ability to recall digit strings, nar-
ratives, and word pairs. The groups did not differ sig-
nificantly in their scores on Reading/ Writing.
Correlations Among Outcome Measures
The Vocabulary/Grammar factor was correlated at
.65 ( p < .001) with the Verbal Memory factor and at .64
( p < .001) with the Reading/Writing factor. The Verbal
Memory factor was correlated at .37 with the Reading/
Writing factor ( p < .05). Thus, the age 17 Vocabulary/
Grammar factor had a large r with the two other age 17
outcome factors, but the r between the other two factors
was only medium in size.
Correlations among the 11 language, memory, and
reading/writing outcome measures are displayed in
Table 3. The 10 rs among the five subtests comprising
the Vocabulary/Grammar factor were all significant, with
9 of the 10 reaching p < .01. The 10 rs ranged from .32
( WAIS-III Vocabulary with CASL Sentence Comprehen-
sion) to .69 (CASL Grammatical Judgment with CASL
Ambiguous Sentences), with five of the rs above the
.50 criterion for a large ES (Cohen, 1988). Two of the
three rs among the three subtests comprising the Verbal
Memory factor were significant at p < .05, with one being
small and two being medium in size. Similarly, two of
the three rs among the three subtests comprising the
Reading/Writing factor were significant at p < .05, with
all three being medium in size.
Consistent with the rs among the three age 17 fac-
tors, the five subtests comprising the Vocabulary/Grammar
factor generally had medium-to-large rs with the sub-
tests comprising the other two factors, whereas the rs
across factors for the other six subtests were generally
more modest. Overall, this pattern of results suggests
that vocabulary and grammar skills are significantly
associated with both verbal memory and reading/writing
skills but that these two latter sets of skills are not highly
associated with each other, as least as instantiated by
the subtests used in this study.
Convergent and Discriminant
Predictive Validity
To examine convergent and discriminant predictive
validity, regression analysis was conducted using the Vo-
cabulary, Grammar, Verbal Memory, and Reading Com-
prehension aggregates analyzed at age 13 years (Rescorla,
2005) as simultaneous predictors of each of the three
age 17 outcomes. As displayed in Table 4, the four age 13
predictors accounted for 67% of the variance in the age 17
Vocabulary/Grammar factor (p < .001). Only the age 13
Vocabulary aggregate and Grammar aggregate had sig-
nificant standardized betas (p < .01, p < .05, respectively),
indicating good convergent and discriminant predictive
validity.

Table 3. Correlations among age 17 language, memory, and reading/writing outcome measures.

Measure 1 2 3 4 5 6 7 8 9 10 11

1. WAIS-III Vocabulary subtest .54** .32* .56** .64** .30* .18 .41** .68** .39** .36*
2. CASL Syntax Construction subtest .46** .49** .57** .41** .25 .49** .37** .36** .43**
3. CASL Sentence Comprehension subtest .47** .48** .36* .26 .43** .31* .05 .21
4. CASL Grammatical Judgment subtest .69** .51** .31* .47** .49** .19 .28
5. CASL Ambiguous Sentences subtest .54** .35* .56** .52** .21 .41*
6. WAIS-III Digit Span subtest .18 .30* .44** .03 .19
7. WMS-III Logical Memory subtest .46** .02 .35* .20
8. WMS-III Verbal Paired Associates subtest .25 .05 .35*
9. WJ-III LetterWord Identification subtest .28 .32*
10. WJ-III Reading Fluency subtest .40*
11. WJ-III Writing Fluency subtest

Note. Values in bold grouped together indicate correlations among tests from the same factor.
*p < .05. **p < .01.

24 Journal of Speech, Language, and Hearing Research Vol. 52 1630 February 2009
Table 4. Predicting age 17 Vocabulary/Grammar factor scores from
age 13 Vocabulary, Grammar, Verbal Memory, and Reading
Comprehension scores.
Predictor B SE B b p
Vocabulary 13 0.17 0.05 .45 .001
Grammar 13 0.16 0.07 .33 .026
Verbal Memory 13 0.08 0.05 .18 .155
Reading 13 0.00 0.01 .05 .733
2 dict the age 17 Reading/Writing factor, the model ac-
Note. R = .67, p < .001.
When the four age 13 aggregates were used to pre-
dict the age 17 Verbal Memory factor, the model ac-
counted for only 42% of the variance (p < .001; see Table 5).
Contrary to prediction, only the age 13 Grammar aggre-
gate had a significant standardized beta ( p < .05), in-
dicating poor convergent and discriminant predictive
validity. To further explore prediction of age 17 Verbal
Memory, correlations between the individual tests mak-
ing up this factor and age 13 subtests were examined. All
four measures comprising the age 13 Grammar aggre-
gate as well as all three tests comprising the age 13
WISC-III Verbal Memory aggregate were significantly
correlated with age 17 WAIS-III Digit Span: TOAL-3
Listening Grammar (r = .55, p < .001); TOAL-3 Reading
Grammar (r = .61, p < .001); CELF-III Formulated Sen-
tences (r = .41, p < .05); TLC Ambiguous Sentences (r = .45,
p < .01); WISC-III Digit Span (r = .72, p < .001); CELF-III
Recalling Sentences (r = .49, p < .01); and Pseudoword
Repetition (r = .59, p < .001). Age 13 TOAL-3 Listening
Grammar and age 13 CELF-III Formulated Sentences,
both in the age 13 Grammar aggregate, were also sig-
nificantly correlated with age 17 WMS-III Logical Memory
(rs = .58 and .39, ps = .001 and .03) and WMS-III Verbal
Paired Associates (rs = .52 and .41, ps < .01 and < .05). On
the other hand, age 13 CELF-III Recalling Sentences and
Pseudoword Repetition, both in the age 13 Verbal Memory
aggregate, were not significantly correlated with WMS-III
Logical Memory or WMS-III Verbal Paired Associates (all
rs < .15). In summary, these results indicate that the skills
tapped by age 13 CELF-III Recalling Sentences and
Pseudoword Repetition differed from those tapped by
age 17 WMS-III Logical Memory and WMS-III Verbal
Paired Associates, but these latter age 17 Verbal Memory
tests were significantly associated with age 13 TOAL-3
Listening Grammar and age 13 CELF-III Formulated
Sentences.
When the four age 13 aggregates were used to pre-
counted for 67% of the variance ( p < .001; see Table 6).
Only the age 13 Reading Comprehension aggregate had
a significant standardized beta (p < .001), indicating good
convergent and discriminant predictive validity.
Prediction From Age 2 to Age 17 Years
Correlations among the three age 2 language mea-
sures were all large by Cohens (1988) standards: LDS
Reynell Expressive (r = .89, p < .001), LDS Reynell Re-
ceptive (r = .59, p < .001), and Reynell Expressive
Reynell Receptive (r = .69, p < .001). None of the three
intake language measures were significantly correlated
with Bayley Nonverbal score (all rs < .10). None of the
four intake measures were significantly correlated with
Hollingshead SES scores (which had restricted range),
and SES was not correlated with any age 17 outcome, so
SES was not further analyzed.
All four age 2 predictors were significantly correlated
with the age 17 Vocabulary/Grammar factor: LDS (r = .43,
p < .01), Reynell Expressive (r = .45, p < .01), Reynell
Receptive (r = .38, p < .01), and Bayley Nonverbal (r = .42,
p < .01). They were also all significantly correlated with
the age 17 Verbal Memory factor: LDS (r = .41, p < .01),
Reynell Expressive (r = .37, p <.05), Reynell Receptive
(r = .36, p < .05), and Bayley Nonverbal (r = .41, p < .01).
These predictive associations were further analyzed using
stepwise multiple regressions. For both regressions, the
order of entry was LDS score, Reynell Expressive Lan-
guage score, Reynell Receptive Language score, and Bayley
Nonverbal score. Because none of the correlations between

Table 5. Predicting age 17 Verbal Memory factor scores from Table 6. Predicting age 17 Reading/Writing factor scores from
age 13 Vocabulary, Grammar, Verbal Memory, and Reading age 13 Vocabulary, Grammar, Verbal Memory, and Reading
Comprehension scores. Comprehension scores.

Predictor B SE B b p Predictor B SE B b p

Vocabulary 13 0.03 0.06 .07 .694 Vocabulary 13 0.07 0.05 .19 .143
Grammar 13 0.24 0.09 .51 .011 Grammar 13 0.05 0.07 .10 .492
Verbal Memory 13 0.05 0.07 .11 .509 Verbal Memory 13 0.03 0.05 .08 .501
Reading 13 0.00 0.02 .03 .857 Reading 13 0.05 0.01 .58 .000

Note. R2 = .42, p < .001. Note. R2 = .67, p < .001.

Rescorla: Age 17 Language and Reading Outcomes 25

age 2 predictors and the age 17 Reading/Writing factor
were significant, regression analysis was not applied to
this outcome.
Only LDS scores and Bayley Nonverbal scores were
significant predictors of age 17 Vocabulary/Grammar
score. The LDS, which was entered first, accounted for 17%
of the variance in age 17 Vocabulary/Grammar (p < .01).
In subsequent steps, Reynell Expressive added 3%
( p = .21), Reynell Receptive added 1% ( p = .49), and
Bayley Nonverbal added 13% ( p < .01). The full model
accounted for 34% of the variance in outcome.
Quite similar results were obtained in the regres-
sion predicting age 17 Verbal Memory scores. The LDS
accounted for 17% of the variance in age 17 Verbal
Memory ( p < .01). Reynell Expressive added 0% ( p = .94),
Reynell Receptive added 2% ( p = .27), and Bayley
Nonverbal added 11% ( p < .05). The full model accounted
for 31% of the variance in outcome.
Discussion
The present research is the only study to report late
adolescent outcomes for late talkers first identified as
such when they were toddlers. With the exception of the
earlier follow-ups from this same cohort (Rescorla, 2002,
2005), no other studies of late-talking toddlers have
followed the children past age 10 years. Furthermore,
there is only one published study reporting follow-up
into adolescence for preschoolers diagnosed with SLI
namely Stothard et al. (1998). The results of the present
study are consistent with those reported by Stothard
et al., but they make an important addition to the lon-
gitudinal literature on language impairment because of
the younger age of the children at intake (2 rather than
4 years of age) and because of the length of the follow-up
period (15 rather than 11 years).
Outcome of Late-Talker and Comparison
Groups at Age 17 Years
The results obtained in this study generally support
the four hypotheses tested. Although late talkers con-
tinued to be comparable with SES-matched comparison
children on nonverbal tasks ( WAIS-III Block Design and
WJ-III Math Fluency), their scores on most of the age 17
language measures were lower than those of the com-
parison children, despite being in the average range. Four
of the ESs were of medium size (Cohen, 1988; d > 0.50),
and one was large (d = 0.80). Late talkers obtained sig-
nificantly lower scores than comparison peers on the age 17
Vocabulary/Grammar and Verbal Memory factors, with
large ESs (ds = 0.92 and 0.93, respectively). Thus, children
identified with delayed expressive language as toddlers
26 Journal of Speech, Language, and Hearing Research Vol. 52 1630 February 2009
generally scored in the average range at age 17 years
and did not have significant language impairment. Fur-
thermore, according to parental report, as well as
reports from the adolescents themselves, they all were
making good progress in high school and were on track to
graduate. Nonetheless, they continued to have weaker
language skills at age 17 years than peers with typical
language historiesa pattern of results supporting a
dimensional view of language ability. This pattern of
findings is consistent with Rescorla (2002, 2005) as well
as with Stothard et al. (1998), Tomblin et al. (2003),
Girolametto et al. (2001), Moyle et al. (2007), Armstrong
et al. (2007), and Paul et al. (1997).
Among the various age 17 subtests comprising the
Vocabulary/Grammar factor, rs were medium-to-large.
These language skills also had moderately strong asso-
ciations with measures comprising the Verbal Memory
and Reading/Writing factors. Thus, the correlations
among the various age 17 language and reading/writing
measures, as well as those among the three age 17 fac-
tors, support the notion that a set of interrelated but
distinct skills comprise language ability, which can be
viewed as a dimension along which individuals differ.
For the age 17 Vocabulary/Grammar and Reading/
Writing factors, strong convergent and discriminant
predictive validity were found on the basis of age 13 scores,
consistent with predictions. Children who had higher
scores on the age 13 Vocabulary and Grammar aggre-
gates tended to have higher scores on the Vocabulary/
Grammar factor 4 years later, but other age 13 scores
were not predictive. Similarly, higher scores on age 13
Reading Comprehension were predictive of higher scores
on age 17 Reading/Writing, but other age 13 scores were
not predictive.
The age 17 Verbal Memory factor was not signifi-
cantly predicted by the age 13 Verbal Memory aggregate,
but it was significantly predicted by the age 13 Grammar
aggregate. Analyses of subtest correlations suggested
that the skills tapped by age 13 CELF-III Recalling Sen-
tences and Pseudoword Repetition differed from those
tapped by WMS-III Logical Memory and WMS-III
Verbal Paired Associates, but these two age 17 Verbal
Memory tests were significantly associated with age 13
TOAL-3 Listening Grammar. One possible reason for
this pattern of results may be that age 17 WMS-III
Logical Memory and WMS-III Verbal Paired Associates
required processing and recall of large amounts of in-
formation on each trial, as did age 13 TOAL-3 Listening
Grammar. In contrast, age 13 CELF-III Recalling Sen-
tences and Pseudoword Repetition both involved rela-
tively short stimuli. Although this explanation is only
speculative, it can be concluded from these results that
different skills were being assessed by the Verbal Mem-
ory measures administered at the two ages, except for
Digit Span, which was highly correlated from age 13 to
age 17 years.
Finally, parent-reported vocabulary score on the
LDS significantly predicted Vocabulary/Grammar and
Verbal Memory skills 15 years later, accounting for 17%
of the variance in both factors. Once the variance as-
sociated with the LDS had been taken into account, the
Reynell Expressive and Receptive Language Scales did
not account for a significant amount of variance in either
of these age 17 outcomes. Thus, significant continuity in
language functioning over 15 years was found in this
study, even when age 2 language skills were measured
by a simple parent-report checklist that takes about
10 min to complete.
An unexpected finding that emerged from these re-
gression analyses was that age 2 Bayley Nonverbal score
was also a significant predictor of age 17 Vocabulary/
Grammar and Verbal Memory skills, accounting for 13%
and 11% of the variance, respectively, beyond that ac-
counted for by the LDS. This suggests that some of the
variance in age 17 Vocabulary/Grammar and Verbal
Memory skills was due to nonlanguage skills measur-
able at age 2 years. These nonlanguage skills include the
various nonverbal cognitive abilities tapped by Bayley
items involving blocks, puzzles, pegs, drawing, and ob-
ject hiding, as well more general test-taking skills, such
as attention, cooperation, and persistence.
In summary, the results of this study are generally
consistent with the notion that language ability consti-
tutes a dimensional spectrum along which individuals
differ. Although the methods of the present study did not
allow disentangling nature versus nurture factors, the
general argument proposed here is that these individual
differences in language ability derive in large part from
constitutionally based variation in endowment. On the
basis of the analogy with intelligence, it can be argued
that this constitutionally based variation in endowment
derives from variation in many discrete skills subserving
language. Although the present study does not directly
indicate what this set of distinct yet interrelated skills
might be, it might be conjectured that likely skills are au-
ditory perception, word retrieval, verbal working memory,
motor planning, phonological discrimination, and gram-
matical rule learning. This notion of a spectrum of lan-
guage ability is consistent with Leonards (1991) argument
that SLI represents the tail of a normal distribution of
language abilities, as well as with Bishop and Edmundsons
(1987) formulation of a spectrum of impairment.
The argument that children are born with differing
endowments for language is similar to the argument
that children are born with differential endowments for
intelligence. Although environmental factors clearly af-
fect both language development and IQ (and environ-
mental factors also covary to some degree with genetic
endowment), the argument advanced here is that dif-
ferential endowment accounts in large part for the con-
siderable stability of individual differences in language
skills as well as IQ across development.
According to the argument advanced here, children
who are very well endowed on most of these skills will
have superior or advanced language ability, whereas chil-
dren with weaker endowment of many of these skills will
have delayed language acquisition initially and weaker
language skills than their typically developing peers as
they mature. Children with the weakest endowment on
these skills will be delayed in language acquisition and
will continue to manifest clinical deficits in language
functioning into adolescence and probably beyond. Fur-
thermore, it is postulated that late talkers have some-
what below-average ability on these constitutionally
based language-related abilities. Toddlers who have re-
ceptive as well as expressive delays and toddlers whose
expressive delays last long enough for them to be diag-
nosed with SLI at age 4 years are both further away from
the mean on this ability spectrum than late talkers with
no receptive problems whose delays resolve by age 4 years.
However, even these less impaired late talkers manifest
significant differences in language skills up to age 17 years
relative to peers with typical language histories.
The fact that the LDS was a significant predictor
of age 17 Vocabulary/Grammar and Verbal Memory fac-
tor scores supports the notion that the LDS at age
2431 months is an excellent early measure of the hy-
pothesized general language ability. However, contrary
to findings reported at 13 years of age, intake Bayley
Nonverbal score was also a significant predictor for both
of these age 17 factors. This suggests that by the time the
participants in this study reached late adolescence,
some nonverbal ability factors that the Bayley tapped at
2431 months also began to account for a significant
percentage of variance in the skills tapped by the age 17
Vocabulary/Grammar and Verbal Memory measures.
The age 2 language measures used in this study
were parent-report or clinical instruments tapping major
language features, such as vocabulary size, use of word
combinations, and comprehension of words and phrases.
They were not experimental measures designed to tap
the language-related skills presumed to underlie language
ability, such as auditory perception, word retrieval, verbal
working memory, motor planning, phonological discrimi-
nation, or grammatical rule learning. However, such
abilities are increasingly being measured in toddlers
and tested as predictors of later language outcomes. For
example, Newman, Ratner, Jusczyk, and Jusczyk (2006)
reported that the ability to segment speech in laboratory
tasks administered before 12 months of age was as-
sociated with vocabulary size at 24 months and with
scores on language tests at ages 46 years.
Rescorla: Age 17 Language and Reading Outcomes 27

Limitations
The sample used in this study was very homoge-
neous in SES and ethnicity. Had it been more diverse,
both the late-talker group and the comparison group
might have had lower scores, and the late talkers might
have fallen below age norms. In addition, the late-talker
group might have received fewer support services, which
might have increased group differences. An additional
limitation of the study is that effects of intervention
could not be studied. Furthermore, not all participants
could be seen at age 17 years. However, the age 17 sam-
ple was reasonably large and was representative of the
intake sample. A final limitation is that testing at age
17 years could not be done by an examiner blind to the
childrens group status. However, all testing was done in
accordance with standardized procedures by an experi-
enced certified school psychologist.
Clinical Implications
This study suggests that most late talkers with
normal receptive language skills will perform in the av-
erage range by national standards in most language
skills by the time they enter school and will continue to
do so through adolescence. On the other hand, the find-
ings suggest that early expressive language delay may
indicate some subclinical weakness in the component
skills that serve language. Continuing weaknesses in
vocabulary, grammar, and verbal memory may be evi-
dent in adolescence, relative to peers with typical lan-
guage histories from the same backgrounds. Although
this study did not assess intervention effects, it is pos-
sible that providing late talkers with activities to im-
prove language processing, phonological discrimination,
verbal memory, and word retrieval might help to pro-
mote stronger language skills and thereby reduce or
eliminate the gap between them and peers with typical
language histories.
Acknowledgments
This research was supported by grants from the Bryn Mawr
College Faculty Research Fund, National Institute of Child
Health and Human Development Grant 1-R15-HD22355-01,
and National Institute on Deafness and Other Communication
Disorders Grant R01-DC00807. I wish to thank the parents
and children whose participation made this research possible.
References
Aram, D., Ekelman, B., & Nation, J. (1984). Preschoolers
with language disorders: 10 years later. Journal of Speech
and Hearing Research, 27, 232244.
Aram, D., & Hall, N. E. (1989). Longitudinal follow-up of
children with study of the relationship between specific
28 Journal of Speech, Language, and Hearing Research Vol. 52 1630 February 2009
language impairment, phonological disorders and reading
retardation. Journal of Child Psychology and Psychiatry, 31,
10271050.
Armstrong, E. S., Marchman, V. A., & Owen, M. T. (2007,
November). School age cognitive and achievement out-
comes for late talkers and late bloomers: Do late bloomers
really bloom? Poster presented at the annual meeting of
the American Speech-Language-Hearing Association,
Boston.
Bayley, N. (1969). The Bayley Scales of Infant Development.
San Antonio, TX: The Psychological Corporation.
Bishop, D. V. M. (2006). What causes specific language im-
pairment? Current Directions in Psychological Science, 15,
217221.
Bishop, D. V. M., & Adams, C. (1990). A prospective study
of the relationship between specific language impairment,
phonological disorders and reading retardation. Journal of
Child Psychology and Psychiatry, 31, 10271050.
Bishop, D. V. M., & Edmundson, A. (1987). Language im-
paired 4-year-olds: Distinguishing transient from persistent
impairment. Journal of Speech and Hearing Disorders, 52,
156173.
Bishop, D. V. M., Price, T. S., Dale, P. S., & Plomin, R.
(2003). Outcomes of early language delay: II. Etiology of
transient and persistent language difficulties. Journal of
Speech, Language, and Hearing Research, 46, 561575.
Carrow-Woolfolk, E. (1999). Comprehensive Assessment of
Spoken Language. Circle Pines, MN: American Guidance
Service.
Cohen, J. (1988). Statistical power analysis for the behavioral
sciences (2nd ed.). New York: Academic Press.
Dale, P. S., Simonoff, E., Dishop, D. V. M., Eley, T. C.,
Oliver, B., Price, T. S., et al. (1998). Genetic influences
on language delay in two-year-old children. Nature Neuro-
science, 1, 324328.
Dollaghan, C. A. (2004). Taxometric analyses of specific
language impairment in 3- and 4-year-old children. Jour-
nal of Speech, Language, and Hearing Research, 47,
464475.
Fenson, L., Reznick, S., Thal, D., Bates, E., Hartung, J.,
Pethick, S., & Reilly, J. (1993). The MacArthur Commu-
nicative Developmental Inventory. San Diego, CA: Singular
Publishing Group.
Girolametto, L., Pearce, P. S., & Weitzman, E. (1996). In-
teractive focused stimulation for toddlers with expressive
vocabulary delays. Journal of Speech and Hearing Research,
39, 12741283.
Girolametto, L., Wiigs, M., Smyth, R., Weitzman, E., &
Pearce, P. S. (2001). Children with a history of expressive
language delay: Outcomes at 5 years of age. American
Journal of Speech-Language Pathology, 10, 358369.
Gopnik, M., & Crago, M. B. (1991). Familial aggregation of
a developmental language disorder. Cognition, 39, 150.
Hall, K., & Tomblin, J. B. (1978). A follow-up study of chil-
dren with articulation and language disorders. Journal of
Speech and Hearing Disorders, 43, 227241.
Hammill, D. D., Brown, V. L., Larsen, S. C., & Wiederholt,
J. L. (1994). Test of Adolescent and Adult LanguageThird
Edition. Austin, TX: Pro-Ed.
Hart, B., & Risley, T. R. (1995). Meaningful differences in the
everyday experiences of young American children. Baltimore:
Brookes.
Hollingshead, A. (1975). Four-Factor Index of Social Status
[ Working paper]. New Haven, CT: Author.
Lee, L. (1974). Developmental sentence analysis: A grammat-
ical assessment procedure for speech and language clini-
cians. Evanston, IL: Northwestern University Press.
Leonard, L. B. (1987). Is specific language impairment a use-
ful construct? In S. Rosenberg ( Ed.), Advances in applied
psycholinguistics: Disorders of first language development
(Vol. 1, pp. 139). New York: Cambridge University Press.
Leonard, L. B. (1991). Specific language impairment as a
clinical category. Language, Speech, and Hearing Services
in Schools, 22, 6668.
Leonard, L. B. (1998). Children with specific language im-
pairment. Cambridge, MA: MIT Press.
Leslie, L., & Caldwell, J. (1995). Qualitative Reading
InventoryII. New York: Harper Collins.
Moyle, M. J., Ellis Weismer, S., Lindstrom, M., & Evans,
J. (2007). Longitudinal relationships between lexical and
grammatical development in typical and late talking chil-
dren. Journal of Speech, Language, and Hearing Research,
50, 508528.
Newcomer, P. L., & Hammill, D. D. (1988a). Test of Lan-
guage Development. Austin, TX: Pro-Ed.
Newcomer, P. L., & Hammill, D. D. (1988b). Test of Lan-
guage DevelopmentPrimary, Second Edition. Austin,
TX: Pro-Ed.
Newcomer, P. L., & Hammill, D. D. (1997). Test of Language
DevelopmentPrimary, Third Edition. Austin, TX: Pro-Ed.
Newman, R., Ratner, N. B., Jusczyk, A. M., & Jusczyk,
P. W. (2006). Infants early ability to segment the conver-
sational speech signal predicts later language development:
A retrospective analysis. Developmental Psychology, 42,
643655.
Paul, R. (1993). Outcomes of early expressive language
delay. Journal of Childhood Communication Disorders,
15, 714.
Paul, R. (1996). Clinical implications of the natural history
of slow expressive language development. American Journal
of Speech-Language Pathology, 5, 521.
Paul, R., Murray, C., Clancy, K., & Andrews, D. (1997).
Reading and metaphonological outcomes in late talkers.
Journal of Speech, Language, and Hearing Research, 40,
10371047.
Rescorla, L. (1989). The Language Development Survey: A
screening tool for delayed language in toddlers. Journal of
Speech and Hearing Disorders, 54, 587599.
Rescorla, L. (2002). Language and reading outcomes to age 9
in late-talking toddlers. Journal of Speech, Language, and
Hearing Research, 45, 360371.
Rescorla, L. (2005). Age 13 language and reading outcomes in
late-talking toddlers. Journal of Speech, Language, and
Hearing Research, 48, 459472.
Rescorla, L., Dahlsgaard, K., & Roberts, J. (2000). Late-
talking toddlers: MLU and IPSyn outcomes at 3;0 and 4;0.
Journal of Child Language, 27, 643664.
Rescorla, L., & Lee, E. C. (2000). Language impairments in
young children. In T. Layton & L. Watson ( Eds.), Handbook
of early language impairment in children: Volume 1: Nature
( pp. 138). New York: Delmar Publishing Company.
Reynell, J. K. (1977). Reynell Developmental Language
Scales. Windsor, England: NferNelson.
Reynell, J. K., & Gruber, C. P. (1990). Reynell Developmental
Language Scales, U.S. edition. Los Angeles: Western Psy-
chological Services.
Rice, M. L. (2003). A unified model of specific and general
language delay: Grammatical tense as a clinical marker
for unexpected variation. In Y. Levt & J. Schaeffer ( Eds.),
Language competence across populations: Toward a defini-
tion of specific language impairment ( pp. 6395). Mahwah,
NJ: Erlbaum.
Rice, M. L., & Wexler, K. (1996). Toward tense as a clinical
marker of specific language impairment in English-speaking
children. Journal of Speech and Hearing Research, 39,
12391257.
Scarborough, H. S. (1990). Index of Productive Syntax. Ap-
plied Psycholinguistics, 11, 112.
Semel, E., Wiig, E., & Secord, W. (1995). Clinical Evaluation
of Language FundamentalsThird Edition. San Antonio,
TX: The Psychological Corporation.
Semel, E., Wiig, E., Secord, W., & Sabers, D. (1987).
Clinical Evaluation of Language FundamentalsRevised.
San Antonio, TX: The Psychological Corporation.
Stothard, S. E., Snowling, M. J., Bishop, D. V. M., Chipchase,
B. B., & Kaplan, C. A. (1998). Language-impaired pre-
schoolers: A follow-up into adolescence. Journal of Speech,
Language, and Hearing Research, 41, 407418.
Taylor, H. G., Lean, D., & Schwartz, S. (1989). Pseudoword
repetition ability in learning-disabled children. Applied
Psycholinguistics, 10, 203219.
Tomblin, J. B., Zhang, X., Buckwalter, P., & OBrien, M.
(2003). The stability of primary language disorder: Four
years after kindergarten diagnosis. Journal of Speech,
Language, and Hearing Research, 46, 12831296.
Torgesen, J., Wagner, R., & Rashotte, C. (1999). Test of
Word Reading Efficiency ( TOWRE). Austin, TX: Pro-Ed.
Wechsler, D. (1991). Wechsler Intelligence Scale for Children
Third Edition. San Antonio, TX: The Psychological
Corporation.
Wechsler, D. (1992). Wechsler Individual Achievement Test.
San Antonio, TX: The Psychological Corporation.
Wechsler, D. (1997a). Wechsler Adult Intelligence Scale
Third Edition. San Antonio, TX: The Psychological
Corporation.
Wechsler, D. (1997b). Wechsler Memory ScaleThird Edition.
San Antonio, TX: The Psychological Corporation.
Whitehurst, G. J., & Fischel, J. E. (1994). Early develop-
mental language delay: What, if anything, should the
clinician do about it? Journal of Child Psychology and
Psychiatry, 35, 613648.
Wiig, W. H., & Secord, W. (1989). Test of Language Competence
Expanded Edition. San Antonio, TX: The Psychological
Corporation.
Rescorla: Age 17 Language and Reading Outcomes 29
Woodcock, R. W., & Johnson, M. B. (1989). Woodcock
Received July 21, 2007
Johnson Psychoeducational BatteryRevised. Chicago:
Riverside Publishing. Revision received January 3, 2008
Woodcock, R. W., McGrew, K. S., & Mather, N. (2001). Accepted May 27, 2008
WoodcockJohnson Psychoeducational BatteryThird Edi- DOI: 10.1044/1092-4388(2008/07-0171)
tion. Chicago: Riverside Publishing.
Contact author: Leslie Rescorla, Department of Psychology,
Young, E. C., & Perachio, J. J. (1983). Patterned Elicitation Bryn Mawr College, 101 North Merion Avenue, Bryn Mawr,
Syntax Test, Revised. Tucson, AZ: Communication Skill
PA 19010. E-mail: lrescorl@brynmawr.edu.
Builders.

30 Journal of Speech, Language, and Hearing Research Vol. 52 1630 February 2009