Abstract
Seven experiments, all focused on serial position phenomena, showed the usual primacyrecency pattern to be easily disrupted. The theoretical assumption that learning operations lose power as a function of continuous use was refuted on empirical grounds. Repeated observations revealed how difficult it is to represent spatial series on the basis of randomized temporal inputs. A two‐stage interpretation of serial learning was claimed to be compatible with the experimental findings: Stage I involves item differentiation, Stage II integration of the differentiated items.

This publication has 7 references indexed in Scilit: