IMPLICIT COVARIATION LEARNING

Lewicki and colleagues have demonstrated that people can implicitly le arn hidden covariations between elements of presented stimuli. Under c onditions of tight experimental control, we obtained little evidence f or such Hidden Covariation Detection (HCD) in both conceptual and exac t replications of Lewicki's studies. The results of that HCD research are summarised and reviewed in this paper. These results cast doubt on the generality and robustness of HCD and suggest that HCD is subjecte d to various boundary conditions. Other paradigms more easily result i n implicit covariation detection and provide a good starting point for more process-related investigations into implicit covariation learnin g. More specifically, we investigated whether the classical conditioni ng phenomenon of 'blocking' can also be observed in implicit learning. No evidence for blocking was obtained suggesting that blocking does n ot necessarily represent an acquisition failure but may be the result of a performance deficit. The consequences of this finding for the imp licit/explicit distinction are discussed.