Development and Validation of the Word-Pseudoword Associative Learning task: an Italian web-based parallel-forms tool for declarative memory assessment

The measurement of associative memory often relies on verbal paired-associate learning tasks, yet existing implementations remain highly heterogeneous, lack standardized parallel forms, and frequently suffer from uncontrolled semantic and psycholinguistic variability. To address these limitations, we developed and validated the Word-Pseudoword Associative Learning (WPAL) task, a fully self-administered, web-based paradigm designed to provide a controlled and replicable assessment of declarative associative learning in Italian populations. Two parallel 40-item forms were constructed through a rigorous multistep procedure balancing real nouns and pseudowords across lexical, affective, and neighborhood metrics. Ninety healthy adults completed the two parallel task forms one week apart in a counterbalanced within-subject design, with a 12-hour retention interval including either nocturnal sleep (SLEEP group, n = 52) or daytime wakefulness (WAKE group, n = 38). Psychometric properties, item characteristics, parallel-forms reliability, and sensitivity to time- and sleep-dependent changes in memory were evaluated.Convergent evidence supported statistical interchangeability between parallel forms. Learning trajectories and immediate recall performance were virtually identical across forms, and internal consistency was excellent for both versions. As expected, recall declined over the 12-hour interval in the overall sample. However, the reduction was significantly attenuated in the SLEEP compared with the WAKE group, demonstrating robust task sensitivity to time- and sleep-dependent memory stabilization.Overall, the WPAL offers a psychometrically validated, remotely deployable, open-access tool for assessing verbal associative memory in Italian adults. Its standardized construction, parallel-form equivalence, and demonstrated sensitivity to consolidation processes make it suitable for reliable and replicable within-subjects investigations in real-world settings.