Language Assessment and Development in Toddlers

1
Language Assessment and Development in Toddlers
with Autism Spectrum Disorder R. Luyster,
University of Michigan M.B. Kadlec, C. Connolly,
A. Carter H. Tager-Flusberg Boston University
School of Medicine
Introduction
Results

Results
Table 3. Hierarchical regression for receptive
language Table 4. Hierarchical
regression for expressive language

• Part 1. Correlational and Descriptive Analyses
  of the MSEL, VABS and MCDI
• Intercorrelations of language measures
• Table 1. Receptive Language Table 2.
  Expressive Language
• Descriptives
• Receptive age equivalents differed across
  measures ?2 (61, 2) 27.79, p lt .001 but
  expressive age equivalents did not ?2 (56, 2)
  4.39, p .11.
• Within the MSEL subdomains, expressive language
  age equivalent was significantly greater than
  receptive (Z 3.54, p lt .001).
• Within the VABS, receptive language age was
  significantly higher than expressive (Z 6.12, p
  lt .001).
• No significant difference in the receptive and
  expressive age equivalents on the MCDI Words and
  Gestures.
• However, the MCDI measures expressive language as
  a necessary subset of receptive language. When
  compared to norms, expressive language was
  relatively in advance of receptive language.
• Figure 1. Expressive language relative to
  receptive language on the MCDI Words and Gestures

• A diagnosis of autism spectrum disorder (ASD)
  is based partly on the presence of a language
  delay or impairment. Children with ASD are being
  identified at earlier ages in recent years, and
  previous findings have consistently suggested
  that early language in this population is both
  heterogeneous and important for later outcome
  (Lord Ventner, 1992). As a result, being able
  to accurately assess and interpret early
  linguistic profiles, as well as the skills which
  may predict early language, is a valuable
  endeavor.
• Language in children with ASD is usually
  measured using standardized tests and parent
  report, which have been found to have fair to
  good agreement (Charman, 2004 Condouris, Meyer,
  Tager-Flusberg, 2003 Stone Yoder, 2001).
  However, analyses of agreement between measures
  generally focuses on production, and
  comprehension has not been throroughly addressed.
  
• Research has indicated that concurrent and
  future language level is predicted by
• joint attention skills (Charman et al., 2003a
  Dawson et al., 2004 Mundy, Sigman, Kasari,
  1990 Sigman McGovern, 2005)
• imitation (Charman et al., 2003a Stone, Ousley,
  Littleford, 1997)
• play (Mundy et al., 1987 Sigman McGovern,
  2005).
• However, most studies on language in children
  with ASD focused on one or two predictors with
  relatively small samples of older children with
  ASD. It remains unclear which among these
  various factors are the most significant
  predictors of language.

Conclusions

• Results indicated close agreement between the
  standardized assessment of language and parent
  report measures, indicating that parents of
  children with ASD provide valid and reliable data
  on their childrens language skills
• Consistent with findings that agreement between
  observational and parent report measures is
  higher for expressive language than for receptive
  (Charman, 2004).
• Our correlations between measures for expressive
  language in particular (.87 to .92) were
  surprisingly high, exceeding those previously
  reported (Charman, 2004 Fenson et al., 1993
  Stone Yoder, 2001).
• It is of note that the discrepancy in receptive
  and expressive age equivalents differed across
  measures.
• Although nonverbal abilities predicted both
  language composites, receptive and expressive
  language were predicted by a different set of
  concurrent skills.
• Interestingly, the most robust predictor of
  receptive and expressive language was gesture
  use, which emerged first and most consistently in
  all models.
• Receptive language was predicted by concurrent
  gesture use, motor skills and response to joint
  attention.
• Expressive language was predicted by concurrent
  gesture use, play and imitation.
• Most of the predictors supported in the present
  study are consistent with those implicated in
  typical development (Carpenter et al., 1998
  Morales et al., 2000 Charman et al., 2000),
  suggesting that language development may be
  similar to that of typically developing children
  in terms of its developmental correlates.

Sponsor NIH (U54 MH 66398)
National Center for Research Resources (M01
RR00533 GCRC)