Method Data Analysis

1

Semantic and phonological verbal-fluency tasks
in children with cochlear implants Deena
Wechsler-Kashi1, Richard G. Schwartz1, Miranda
Cleary1 Jane R. Madell2 1Program in Speech and
Hearing Sciences, City University of New York,
Graduate Center, New York, N.Y., USA 2 Beth
Israel-New York Eye and Ear Cochlear Implant
Center, New York, N.Y., USA
ASHA, Miami Beach, 2006

• Method- Data Analysis
• Based on Troyer et al. (1997), clusters were
  defined as groups of two or more successively
  produced words that share phonological or
  semantic characteristics. Phonological cluster
  included words beginning with the same two
  initial segments (CC or CV), words sharing first
  and last sounds or words that rhyme (example
  free-fry fat-foot, like-life).
• Semantic cluster consisted of words that are
  related in meaning or that belong to the same
  subcategory (example coffee-cake
  dolphin-whale).
• Number of switches were counted as transitions
  between clusters or words. Counts of clusters and
  switches included repetitions and errors.
• Paired samples T-tests were used for statistical
  analysis.
• Examples
• Phonological VF task - words that begin with /f/
  
• fight, flight, fun, fear, fat, foot, feet,
  federal, flame
• Semantic VF task - words that belong to the
  animal category
• seal, dolphin, whale, fish, sheep, ant, lion,
  giraffe, monkey, dog, goose
• Results
• Table 1 Summary of qualitative and quantitative
  analysis -VF tasks (SD in parentheses)

• Background
• Many hearing impaired children using cochlear
  implants (CIs) have language impairments
  (Svirsky, Robbins, Iler-Kirk, Pisoni Miyamoto,
  2000 Geers, Nicholas, Sedey, 2003). The
  underlying bases of these impairments are not yet
  fully understood.
• Verbal-Fluency (VF) naming tasks have been used
  to examine lexical processing in typically and
  atypically developing children and adults but
  have not yet been applied to the CI population.
  These tasks can shed light on the processes
  underlying the observed variability in language
  outcome
• Goal
• To identify differences in word retrieval
  processes that elucidate the organization and
  representations of words in the mental lexicons
  of children with CIs
• Participants
• 12 NH children 9 females, 3 males Age 72 to
  1011 (M8.92 years)
• 12 CI children 6 females, 6 males Age 70 to
  114 (M8.53 years)
• CI and NH participants were matched according to
  age and Nonverbal IQ scores (Average standardized
  scores were 111.9 for the NH group and 107 for
  the CI group)
• NH inclusion criteria
• Passed hearing screening ( .25K, .50K, 1K, 2K,
  4K Hz)
• TONI-III Nonverbal IQ standardized measuregt80
  (Brown, Sherbenou, Johnsen, 1997)
• Passed CELF-3 language screening measure (Semel,
  Wiig, Secord, 1995)
• No academic, cognitive, emotional or other
  impairments, according to parental report

• Results-continued
• Summary of Results
• CI group generated significantly fewer words
  than NH group on phonological and semantic VF
  tasks
• CI group exhibited fewer switches than the NH
  group on the semantic fluency task, and a similar
  trend was found for the phonological VF task
• CI group tended to form fewer phonological
  clusters than the NH group on the phonological VF
  task, but formed a similar number of semantic
  clusters on the semantic VF task
• No group differences were observed in the mean
  cluster size on both phonological and semantic VF
  tasks
• Discussion

Figure 3 Average number of switches on VF tasks
t(11) 1.8, p0.09
t(11) 2.1, p0.05
t(11) 2.7, plt0.05
t(11) 2.9, plt0.05
Acknowledgements Research supported by NIH-NIDCD
Grant 5R01DC003885 NIH-NIDCD Grant F32 DC006786
Graduate Research Grant, Graduate Center, CUNY
t(11) 1.8, p0.09
t(11) 0.7, pgt0.05