Introduction: Narratives might be more appropriate for assessing bilinguals than standardized tests because their macrostructure is less language dependent than microstructure. In our study, we examined macrostructure for production as suggested in the Multilingual Assessment Instrument for Narratives (MAIN) (Gagarina et al., 2012), with its components including story structure, story complexity and internal state language. The aim of the study was to identify whether Turkish-German speaking bilingual children demonstrate any special difficulty with comprehension or the production of macro structure components in narration in Turkish.

The participants: The participants of the first and the second conditions were 36 (2:11 to 7:11) and 13 children (5:5 to 9:4) respectively. The participants of the study were Turkish-German speaking simultaneous bilingual Turkish children who were born in Germany and have been living in that country since their birth. The younger children were gathered from the BIVEM-study (15 children including 6 boys and 9 girls with an age range 2;11 - 3;10 yrs).

Procedure: The second author in DİLKOM/Turkey was assigned to Berlin to do the experiment; she met the children in Berlin at kindergarten and/or other educational settings. They evaluated children' language proficiency in both languages; in Turkish, with a standardized Turkish test (TEDİL) adopted from TELD (Test of Early Language Development). They asked parents to fill in a form giving information about child's bilingual background, as well.

Two elicitation conditions were compared in the study. In the first condition, we applied “the model and tell story” (that is, after the clinician models Cat or Dog followed by comprehension questions, the child tells Birds or Goats followed by comprehension questions). In the second the “tell and retell story” tasks were applied (that is, child tells Birds or Goats and answered comprehension questions. Then s/he retells what the clinician tells her/him (Cat or Dog) and answers the comprehension questions).

Results: Results have shown that age does not affect macrostructure components except complexity but is more influential on comprehension; that is, in either tasks, the younger children provided the lower responses to the questions. In addition, biological sex does not have an impact either on production of macrostructure components nor the comprehension. In the comparison between the ‘model & tell' story, comprehension was affected in favor of model but not affected in other conditions.