2.2.3. Expressive Language Sampling

Participants' expressive syntactic and lexical levels were assessed using a narration task in which participants were asked to narrate a story using a wordless picture book [23,24]. The task begins with the participant familiarizing themselves with the book by examining each page spread for approximately 10 s before narrating the story depicted in the book. The examiner facilitates the narration by controlling the book and waiting for the participant to finish their description before turning the page. In order to standardize the task, participants received one of two books from the Mercer Mayer's "Frog" series ("Frog Goes to Dinner" or "Frog on His Own"), and administrators relied on a standardized set of prompts and responses to ensure minimal and consistent scaffolding across participants. Participants' speech was transcribed, segmented into C-units (communication units), with a C-unit is defined as an independent clause with associated modifiers, including dependent clauses, and analyzed using the software program Systematic Analysis of Language Transcripts 18 Research (SALT) [30]. Inter-transcriber agreement data were computed and averaged as follows: 87% for utterance segmentation, 87% for identification of partly or fully unintelligible C-units, 84% for identification of the exact lexical and morphemic content of each C-unit, 76% for identification of the exact number of morphemes in each C-unit and 80% for the exact number of words in each C-unit [24]. Construct validity has been established for the ELS narration task such that medium to strong convergent validity was found with directly administered and informant report measures for similar constructs measuring syntactic complexity and lexical diversity in individuals with DS as well as other forms of ID [23,24].

*Syntactic complexity.* Participants' syntactic maturity was assessed by calculating the mean number of morphemes per C-unit. Only complete and fully intelligible C-units were used to calculate this variable.

*Lexical Diversity.* The size of each participant's expressive vocabulary was computed by calculating the number of different word roots in the participant's first 50 complete and fully intelligible C-units. In the event that the participant produced less than 50 complete and fully intelligible C-units, the full sample was used.
