Determining Implicit and Explicit Contributions to Sequence Learning in ASC

Background:

There does not appear to be a general ASC-deficit in implicit learning, as evidenced by preserved performance on a variety of implicit learning tasks (Brown et al., 2008; Barnes et al., 2008; Kourkoulou et al., 2008; Travers et al., 2008). However, the fact remains that ASC-deficits have been occasionally observed on different versions of implicit learning tasks (Klinger et al., 2007; Mostofsky et al., 2000).

The authors hypothesise that ASC-deficits on implicit learning tasks arise from atypical explicit learning contributions. For example, Mostofsky et al., (2000) found an ASC-deficit on the Serial Reaction Time (SRT) task using conditions that have been subsequently shown to involve the use of explicit learning processes (completely reliable sequences and long inter-stimulus-intervals - Destrebecqz & Cleeremans, 2001). In contrast, equivalent ASC performance has been identified (e.g. Brown et al., 2008) using conditions that tend to prevent explicit learning (Destrebecqz & Cleeremans, 2001). We examine our hypothesis by comparing ASC and TD children on a SRT-task using completely reliable (deterministic) sequences.

An additional feature of the design allows the mechanism of the atypical explicit contribution to be explored: the sequence information will be embedded within another implicit learning design, Contextual Cueing (CC). Vazquez (2008) developed this hybrid (SRT-CC) design and found that for most sequences, both sequential and contextual learning emerge and remain entirely unaffected by the presence of the other. However, when the sequence is deterministic, contextual cueing is reduced. It is assumed that the reduction of the CC-effect is contingent upon explicit knowledge of the sequence. Concordantly, there is evidence of explicit knowledge from the SRT generation task only when the sequence is deterministic. It has previously been demonstrated that there is intact performance in ASC individuals on the classic (non-hybrid) CC-task (e.g. Barnes et al., 2008). Therefore, if we observe the predicted ASC-deficit on this deterministic-SRT, then we can make inferences about the explicit contribution to SRT-learning from ASC performance on the CC-component of the task. If the CC-effect is superior in the ASC-group, it suggests that the ASC-group acquired less explicit knowledge of the sequence and the deterministic-SRT ASC-deficit stems from the fact that they are worse at learning such tasks explicitly. However, if the CC-effect is reduced equivalently in both groups, then this would suggest that the ASC-group acquired an equivalent amount of explicit sequence knowledge. This would imply that the previously unaffected implicit contribution to SRT-performance in ASC-individuals becomes relatively impaired only in the presence of explicit knowledge.

Objectives:

To test our hypothesis that ASC-deficits on implicit learning tasks arise when there is an atypical explicit learning contribution. Insofar that our hypothesis is true, the design will elaborate the mechanism of the atypical explicit contribution in ASC.

Methods:

The performance of children with ASC is compared with that of TD children matched for IQ and chronological-age (11-14 years-old) on a deterministic SRT-task embedded within a CC-design. Participants also complete post-test generation tasks.

Results:

At the time of writing, data-collection for ASC-participants was incomplete.

Conclusions:

As above, so not possible.