First-Order and Higher Order Sequence Learning in Specific Language Impairment

Objective: A core claim of the procedural deficit hypothesis of specific language impairment (SLI) is that the disorder is associated with poor implicit sequence learning. This study investigated whether implicit sequence learning problems in SLI are present for first-order conditional (FOC) and higher order conditional (HOC) sequences. Method: Twenty-five children with SLI and 27 age-matched, nonlanguage-impaired children completed 2 serial reaction time tasks. On 1 version, the sequence to be implicitly learnt comprised a FOC sequence and on the other a HOC sequence. Results: Results showed that the SLI group learned the HOC sequence (eta(2)(p) = .285, p = .005) but not the FOC sequence (eta(2)(p) = .099, p = .118). The control group learned both sequences (FOC eta(2)(p) = .497, HOC eta(2)(p) = .465, ps < .001). Conclusions: The SLI group's difficulty learning the FOC sequence is consistent with the procedural deficit hypothesis. However, the study provides new evidence that multiple mechanisms may underpin the learning of FOC and HOC sequences.