| |
|
| |
|
|
|
| |
|
Statistical learning in children with developmental language disorder: Novel insights from an online measure |
|
|
|
|
|
|
|
|
|
|
|
|
|
Imme Lammertink - University of Amsterdam, Amsterdam Center for Language and Communication
|
Frank Wijnen - University of Utrecht, Utrecht Institute of Linguistics
|
Paul Boersma - University of Amsterdam, Amsterdam Center for Language and Communication
|
Judith Rispens - University of Amsterdam, Amsterdam Center for Language and Communication
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| SRCLD Year: |
2018 |
| Presentation Type: |
Special Session |
| Presentation Time: |
(na) |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Rationale: A statistical learning (SL) deficit has been proposed to explain the linguistic problems seen in children with developmental language disorder (DLD). SL is often assessed via artificial language learning tasks, with post-exposure grammaticality judgments as outcome measure. We examined whether 1) a SL-deficit is observed in children with DLD when a novel, online measure of SL is used and 2) whether individual differences in this SL-measure are associated with severity of problems in morphosyntax and morphology.
Methods: our task battery included measures of morphosyntax, morphology, verbal memory (combined forward and backward digit span) and an auditory SL-task.
Results: Results of 33 children with DLD and 33 typically-developing children (M = 9.3y) show that both groups are sensitive to regularities in the input, but that large individual differences exist. The individual differences correlate with morphosyntax when controlling for verbal memory (r=0.34, p<0.01).
Conclusions: Our results challenge the hypothesis that DLD is associated with poor SL across the board. However, the observed association between SL and morphosyntax (in natural language), does suggest that SL is a relevant factor in language development.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|