LSA from retell and comprehension questions: an option for teens

Nippold et al., aim to “create a language sampling task appropriate for adolescents.” They use Greek fables (don’t worry—they’re very short!) as a tool to elicit language samples from typically-developing eighth graders (n = 30), with specific retell and comprehension questions to accompany it.

Procedure Details:

  • Time: Sampling, transcription, and analysis took around 1 ½ hours to complete, per student.
  • Sampling procedures: The children retell and answer questions about four short fables. The authors provide the exact questions to ask after each one (see article appendices).
  • Analyses:

c-units (total number): to measure productivity

mean length of c-unit (MLCU): to measure complexity

clausal density: to measure complexity

Why these? The authors identify previous research showing that these measures are sensitive to teens’ language growth and tend to be low in those with language disorder.

What will I need to know how to measure? First, you’ll identify c-units (“an utterance that contains one main clause and any subordinate clauses that are attached to it”). Then you’ll start counting clauses. The authors also highly suggest coding the clauses (main clauses vs. subordinate, and six subordinate types, e.g. adverbial, relative, nominal…), so you have more specific information about the student’s language skills, and thus where to start with therapy (clause coding is included in the 1 ½ hours to complete). The authors also describe how to do the analyses in SALT, rather than by hand.

The authors provide everything you’d need to replicate—the fables, the questions, and directions on which clause types to code. You can then compare the results you obtain with your students to the expected performance of the typically-developing peers from their study, “…with the caveat that the results are preliminary” (because it’s not a huge sample, and kids with language disorders aren’t included in it).

The authors suggest that language sampling is an essential tool in the diagnostic process. They state, “Although norm-referenced, standardized tests are useful for identifying language disorders in adolescents (Nippold et al., 2008, 2009; Strothard et al., 1998), they provide little information about how adolescents actually express themselves for genuine purposes, and they may be subject to cultural bias (Heilmann & Malone, 2014).” Also, most language sampling guidelines were developed for younger children, so having something that can be used on the teen population is needed.

Nippold, M.A., Vigeland, L.M., Frantz-Kaspar, M.W., & and Ward-Lonergan, J.M. (2017). Language Sampling With Adolescents: Building a Normative Database With Fables. American Journal of Speech-Language Pathology. Advance online publication. doi:10.1044/2017_AJSLP-16-0181