Living on the Edge of Communicative Competence: Speed of Processing

1
Living on the Edge of Communicative
CompetenceSpeed of Processing

• ASHA Convention
• November 22 2008

2
Project 2 Processing Capacity and Speed of
Processing

• Laurence Leonard
• Carol Miller
• Denise Finneran

• Susan Ellis Weismer
• Christine Weber-Fox
• Robert Kail

3
Starting points

• Children with SLI often perform tasks more slowly
• Not just language tasks
• Example
• Johnston Ellis Weismer (1983)
• slower to perform mental rotation task with
  similar accuracy

4
Starting points

• Kail (1994)
• RTSLI m x RTNLD
• analysis of archived data from 5 studies
• m 1.32
• Children with SLI 32 slower across tasks

5
Two views of response time

• RT as an index of global cognitive speed
• Generalized slowing hypothesis (Kail 1994)
• since then debate about domain-general vs.
  domain-specific slowing
• both
• RT as an index of hidden or residual difficulties
  with specific tasks

6
Preview

• Assessing overall processing speed
• RT battery
• ages 9 14
• Assessing residual difficulties with tense and
  agreement
• sentence processing
• age 16

7
Assessing overall processing speed at age 9

• Sample
• only children with consistent diagnosis
  kindergarten and grade 2
• SLI and NLD matched on PIQ
• 29 NLD
• 29 SLI
• 19 NLI

8
Assessing overall processing speedat age 9

• Nonlinguistic
• Motor
• tapping
• simple RT
• Cognitive
• visual search
• mental rotation

• Linguistic
• Lexical
• picture naming
• picture matching
• Syntactic
• judging grammaticality
• judging truth value
• Phonological
• judging rhymes
• judging initial consonants

9
(No Transcript)
10
(No Transcript)
11
Assessing overall processing speed at age 9

• Results
• SLI group 14 slower than NLD
• NLI group 28 slower than NLD
• For both sub-groups similar degree of slowing
  for linguistic and nonlinguistic tasks
• Individual regression
• 72 SLI and 79 NLI with m gt 1
• 31 SLI and 47 NLI with m larger than all NLD

12
RTSLI (1.14)RTNLD R2 .97
RTNLI (1.28)RTNLD R2 .94
13
Assessing overall processing speed at age 14

• Sample (same as age 9)
• 31 NLD
• 20 SLI
• 15 NLI
• Same tasks as age 9

14
Assessing overall processing speed at age 14

• In the interim evidence that slowing is not
  general (e.g. Windsor et al. 2001)
• Different analytic approach
• individual regressions for all vs. NLD
• ANOVA on slope estimates

15
Assessing overall processing speed at age 14

• Mean slope by group
• NLD 1.00 (0.17)
• SLI 1.22 (0.24)
• NLI 1.29 (0.25)
• Significant group effect
• NLD lt SLI NLI

16
Assessing overall processing speed at age 14

• Significant correlations between mean RTs at 9
  and 14
• NLD r .54
• SLI r .52
• NLI r .71

17
New analyses

• Does RT predict language growth
• Standardized RTs at age 9
• linguistic and nonlinguistic tasks
• Scaled ability scores for vocabulary and grammar
• ages 10 14 16
• Growth curve modeling (linear)

18
Preliminary results

• Both linguistic and nonlinguistic RTs predict
  growth in grammar but not in vocabulary
• Nature of tasks
• Nonlinear component
• Leonard et al. (2007) found that working memory
  and to a lesser extent RT predicted concurrent
  language at age 14

19
Assessing residual difficulties with tense and
agreement at age 16

• Rice and colleagues tense and agreement are
  optionally omitted by children with SLI for an
  extended period
• Demonstrated in Iowa sample (Rice et al. 2004)
• by 4th grade children with SLI were leveling out
  near ceiling for elicited production of -ed and
  3S
• children with NLI lagged but were reaching
  equivalent levels by 4th grade

20
Assessing residual difficulties

• Will sentence processing tasks show problems with
  tense/agreement not seen in production
• grammaticality judgment
• word monitoring

21
Assessing residual difficulties

• Nontense omission
• The children are hide(ing) so they can surprise
  their friend.
• I like Lucy(s) dentist better than my dentist.
• Tense intrusion
• Sue helped to hang(s) a picture last night.
• It is impolite to stare(d) at people at the mall.
• Tense omission
• Every day he tell(s) a joke at lunch.
• Yesterday the friends look(ed) at the magazine
  together.

22
Assessing residual difficulties grammaticality
judgment

• Sample (more inclusive than RT studies)
• Categorized according to age 14 testing
• 108 NLD
• 46 SLI
• 21 NLI

23
Assessing residual difficulties grammaticality
judgment

• Results
• SLI and NLI groups less sensitive (accuracy) to
  violations overall
• F(2 172) 33.3 p lt .001 partial 2 .28
• No interaction of group and sentence type
• Tense omission was not more vulnerable than other
  violations

24
Assessing residual difficulties word monitoring

• Same types of sentences
• Respond to target word should be slower
  following grammatical violation
• Sample
• 106 NLD
• 47 SLI
• 25 NLI

25
Assessing residual difficulties word monitoring

• Results
• Interaction of group with grammaticality of
  sentence was significant only for Tense Omission
• SLI and NLI groups were not slowed by tense
  omission but were slowed by other violations

26
New analysis

• No significant group RT effects for either
  sentence processing task
• Has overall processing speed deficit resolved
• Combined data for both experiments

27
Preliminary results

• Collapsing group means in a quick dirty
  manner
• Effect of group is significant but small
• F(2 172) 3.1 p lt .05 partial 2 .04

28
Summary

• Children and adolescents with language impairment
• show evidence of overall slow processing
• but we dont know if it is truly generalized
• processing speed may predict grammar growth
• adolescents with LI are slower and less accurate
  to process sentences
• Detection of tense omission may continue to be
  compromised even when production is accurate

29
References

• Johnston J. Ellis-Weismer S. (1983). Mental
  rotation abilities in language-disordered
  children. Journal of Speech and Hearing
  Research 26 397-403.
• Kail R. (1994). A method for studying the
  generalized slowing hypothesis in children with
  specific language impairment. Journal of Speech
  and Hearing Research 37 418-421.
• Leonard L.B. Ellis Weismer S. Miller C.A.
  Francis D.J. Tomblin J.B. Kail R.V.
  (2007). Speed of processing working memory and
  language impairment in children. Journal of
  Speech Language and Hearing Research 50
  408-428.
• Leonard L.B. Miller C.A. Finneran D.A. (in
  press). Grammatical morpheme effects on sentence
  processing by school-aged adolescents with
  specific language impairment. Language and
  Cognitive Processes.
• Miller C. A. Kail R. Leonard L. B.
  Tomblin J. B. (2001). Speed of processing in
  children with specific language impairment.
  Journal of Speech Language and Hearing
  Research 44 416-433.
• Miller C.A. Leonard L.B. Kail R.V. Zhang
  X. Tomblin J.B. Francis D. (2006). Response
  time in 14-year-olds with language impairment.
  Journal of Speech Language and Hearing
  Research 49(4) 712-728.
• Miller C.A. Leonard L.B. and Finneran D.
  (2008). Grammaticality judgments in adolescents
  with and without language impairment.
  International Journal of Language and
  Communication Disorders 43(3) 346-360.
• Rice M.L. Tomblin J.B. Hoffman L. Richman
  W.A. Marquis J. (2004). Grammatical tense
  deficits in children with SLI and nonspecific
  language impairment Relationships with
  nonverbal IQ over time. Journal of Speech
  Language and Hearing Research 47(4) 816-834.