Durations as part of equivalence classes

This study investigated whether durations can serve as elements of equivalence classes in humans. Eight adult participants underwent conditional discrimination training using a one-to-many (OTM) matching-to-sample (MTS) procedure. Sample stimuli consisted of two durations (400 ms and 1200 ms), and comparison stimuli were abstract colored figures. Five baseline relations (AB, AC, AD, AE, AF) were trained. Equivalence class formation was assessed through post-training sorting and tests of emergent relations under extinction. Four participants met the training criterion. Among them, three showed class-consistent responding in the MTS test, and three showed consistent sorting. A time-estimation task followed, requiring participants to reproduce the durations associated with each comparison stimulus via keypress. Generalized linear mixed-effects modeling indicated a significant effect of stimulus class (Short vs. Long) on estimated durations ( z = − 16 . 04 , p < . 001 ), with average estimates of 0.80 s and 1.76 s, respectively. These findings suggest that durations may function as class elements and that their control may transfer to novel response topographies, although, due to between-subject variability, the generality of these findings remains limited. The results contribute to theoretical models of symbolic behavior and support the inclusion of temporal properties in the analysis of equivalence class formation. • We tested whether durations can be included in stimulus equivalence classes. • A cross-modal MTS procedure trained relations between durations and figures. • Emergent relations indicated class-consistent responding in trained participants. • Time estimation distinguished between short and long stimulus classes. • Results suggest that durations may be part of equivalence classes.