Original citation. Pacton, S., & Perruchet, P. (2008). An attention-based associative account of adjacent and nonadjacent dependency learning. Journal of Experimental Psychology: Learning, Memory, and Cognition, 34, 80-96.
Target of replication. Pacton and Perruchet tested whether attending to non-adjacent elements is sufficient and necessary for associative learning. They used sequences of digits and a subtraction task that required attention either to adjacent (spatially contiguous) or to non-adjacent digits. Learning of adjacent and non-adjacent dependencies was tested with a two-choice recognition test. Here, the last experiment in the reported series (Experiment 4b) is replicated, in which attention was directed to non-adjacent digits only. In the recognition test, the percentage of correct responses to items with nonadjacent dependencies (M = 74.48, SE = 2.86) was higher than to items with adjacent dependencies (M = 49.48, SE = 4.02), This is the target finding for replication.
A priori replication criteria. A successful replication woud find higher recognition performance for items with nonadjacent dependencies than for items with adjacent dependencies. Performance for items with adjacent dependencies should not differ from the chance level of 50%.
Materials, Data, and Report. Study materials can be found in the materials component of this project. Raw data can be found in the dataset node. The full report appears in the files section of this node.
Conclusions. The target finding – better recognition performance for items with nonadjacent dependencies to which attention was directed in the study phase - was successfully replicated with a similar effect size (partial eta-squared of .46 as compared to .51 in the original experiment). As in the original experiment, only performance for items with nonadjacent dependencies was above chance level. This held for familiar and transfer items.