ANKARA D˙ILB˙IL˙IM C¸EVRES˙I/ANKARA LINGUISTIC CIRCLE

8 Ocak, Cuma, 16:40/January 8th, Friday, 16:40 ¨ Enformatik Enstitus ¨ u, ¨ S03 ODTU

¨ TODIL Project: Assessment of Specific Language Impairment in Monolingual and Bilingual Preschool and Elementary School Children (2:0-9:0) Speaking Turkish as Their Native Language Prof. Dr. Seyhun TOPBAS¸ ˙ Anadolu University, DILKOM Researchers have given special emphasis to identify children at an early age who are at risk for later language difficulties, and to describe the nature and severity of their problems so that appropriate interventions can be designed. Identification of language impairment and subsequent qualification for services require comparison of a child’s performance to a normative group (Spaulding et al 2006; Merrell & Plante 1997). Although there are sufficient studies to give a picture of typical language acquisition in Turkish, most of these studies are conducted with small groups of children. On the other hand, the picture is incomplete because there are no norms available indicating which structures are acquired at which ages. A logical and timely first step in normative assessment has led to a new trend, i.e. translating and adapting tests from English to Turkish. As a first step, a project began in 2005 to adapt the Test of Early Language Development (TELD-3) (Hresko, Reid & Hammill, 1999) to Turkish (Topbas¸ & G¨uven 2007; by permission from PROED). The TELD-3 was developed for English children and adapted to Spanish in the US. It was chosen mainly because it is widely used, is easy and short to administer, and provides a good picture of a child’s language performance. The Turkish translation (TELD-3:T) is both a translation and an adaptation. Therefore, performance on the original and the adapted tests cannot be compared directly. Although the items may appear similar, there are inherent linguistic differences in vocabulary, sentence structure, morphology, linguistic cues, word order, subject-verb agreement, verb morphology, and so on. The TELD-3:T pilot sample comprised typically developing and disabled children (n=359), aged 2:0 to 7:0, residing in 14 cities from 5 geographical regions in Turkey. All the children’s first language was Turkish. We included 24 children in a disability group, of which 18 had language delay and/or language learning difficulties and 6 had a diagnosis of mental retardation. The Turkish translation (TELD-3:T) is both a translation and an adaptation. Therefore, performance on the original and the adapted tests cannot be compared directly. Although the items may appear similar, there are inherent linguistic differences in vocabulary, sentence structure, morphology, linguistic cues, word order, subject-verb agreement, verb morphology, and so on. The findings of this study seemed particularly noteworthy because the TELD-3:T has a significant potential to

correctly identify children with impairments. Following this, in a second study, we aimed to focus on quantitative analyses of elicited sentence imitation or repetition tasks (SR). SR data from 30 SLI and 30 normally developing agematched (NDAM) children from the pilot sample were compared. 385 sentences were analyzed statistically. It is assumed that if SR tasks are able to identify the differences between ND and SLI children’s capacity in repeating sentences of syntactic complexity, then the validity of the TELD-3:T would be strengthened. Furthermore, would give invaluable clues for cross-linguistic comparisons. Over the past decade, (SR) have been shown as a reliable and powerful method for assessing grammatical knowledge (Thornton 1995, Crain and Thornton 1998, Lust et al. 1996). More recent studies have suggested that SR skills are potential key markers for language impairments and for SLI specifically, in measuring syntactic and/or morpho-syntactic abilities of children (ContiRamsden et al. 2001). SR measures are easy to administer and permit the assessment of specific target structures in controlled situations (Devescovi & Caselli 2007). SR as a clinical marker has also been investigated in identifying children with normal development and children with different developmental disorders (Botting and Conti-Ramsden 2003; Eadie, Fey, Douglas, and Parsons 2002); as well as in children with SLI speaking such languages as Cantonese (Stokes & Fletcher, 2003), Italian (Vicari, Caselli, Gagliardi, Tonucci, & Volterra, 2002; Volterra, Caselli, Capirci, Tonucci, & Vicari, 2003), and Dutch (Rispens, 2004). Importantly, special SR tests have been developed for this purpose in Italian (Devescovi & Caselli 2007), English (van der Lely, 1997; Gardner 2006 et al 2006), and Dutch (Rispens, 2004). A SR task was also used in determining the language abilities of ND Turkish children during a project for the educational needs of 5-8 old children (preschool, primary 1st and 2nd graders) living in the eastern part of Turkey (Aksu-Koc¸ et al, 2002). The results showed significant differences between the SLI and NDAM children. ¨ ˙ITAK-COST No: 109K001 Project support TUB

http://groups.google.com/group/ankara-linguistic-circle

2