A hallmark of human memory is the ability to integrate discrete experiences into cognitive maps. A fundamental form of this integration is transitive inference (TI), in which overlapping premises (A < B, B < C) are integrated into a unified representation of a relational hierarchy (A < B < C). Few existing theories provide a mechanistic account of this construction of relational knowledge and how it is shaped by different training conditions. This study builds on recent findings that TI is facilitated by chaining of overlapping premises, with a new behavioral experiment confirming an advantage over non-overlapping sequences matched for premise frequency and spacing. A subsequent simulation study shows that the chaining effect is captured by a particle filter which performs approximate Bayesian inference about the latent hierarchy. These results provide a better understanding of how chaining shapes the construction of relational knowledge in the face of uncertainty and forgetting.

Prior knowledge of relational structure allows people toquickly make sense of and respond to new experiences. Whenawareness of such structure is not necessary to support learn-ing, however, it is unclear when and why individuals “spon-taneously discover” an underlying relational schema. Thepresent study examines the determinants of such discovery indiscrimination-based transitive inference (TI), whereby peo-ple learn about a hierarchy of interrelated premises and aretested on their ability to draw inferences that bridge studiedassociations. Experiencing “chained” sequences of overlap-ping premises during training was predicted to facilitate thediscovery of relational structure. Among individuals withoutprior knowledge of the hierarchy, chaining improved relationallearning and was most likely to result in explicit awareness ofthe underlying relations between items. These findings addto growing evidence that the temporal dynamics of training,including successive presentation of overlapping associations,are key to understanding spontaneous relational discovery dur-ing learning.

A growing body of research indicates that “active learning” im-proves episodic memory for material experienced during study.It is less clear how active learning impacts the integration ofthose experiences into flexible, generalizable knowledge. Thisstudy used a novel active transitive inference task to investi-gate how people learn a relational hierarchy through activeselection of premise pairs. Active control improved memoryfor studied premises as well as transitive inferences involv-ing items that were never experienced together during study.Active learners also exhibited a systematic search preference,generating sequences of overlapping premises that may fa-cilitate relational integration. Critically, however, advantagesfrom active control were not universal: Only participants withhigher working memory capacity benefited from the opportu-nity to select premise pairs during learning. These findingssuggest that active control enhances integrative encoding ofstudied material, but only among individuals with sufficientcognitive resources.

This paper reports an experiment testing whether volitionalcontrol over the presentation of stimuli leads to enhancedrecognition memory in 6- to 8-year-old children. Childrenwere presented with a simple memory game on an iPad.During the study phase, for half of the materials childrencould decide the order and pacing of stimuli presentation(active condition). For the other half of the materials, childrenobserved the study choices of another child (yokedcondition). We found that recognition performance was betterfor the objects studied in the active condition as compared tothe yoked condition. Furthermore, we found that the memoryadvantages of active learning persisted over a one-week delaybetween study and test. Our results support pedagogicalapproaches that emphasize self-guided learning and show thateven young children benefit from being able to control howthey learn.

Previous research with typically developing (TD) childrenand adults show an advantage of active control for episodicmemory as compared to conditions lacking this control. Thepresent study attempts to replicate this effect in autisticchildren. Six- to 12-year-old autistic children (n = 30) wereinstructed to remember as many of 64 presented objects aspossible. For half of the materials presented, participantscould decide the order and pacing of study (Active condition).For the other half, they passively observed the study decisionsof a previous participant (Yoked condition). We found thatrecognition memory was more accurate for objects studied inthe active as compared to the yoked condition, even after aweek-long delay. The magnitude of the effect was comparableto that obtained in previous studies with TD children andadults, suggesting a strong robustness for the benefits ofactive learning. We discuss how pedagogical approaches maybe encouraged to utilize self-directed learning strategies topromote inclusive learning.

In decisions from experience (DFE), people sample from two or more lotteries prior to making a consequential choice. Although existing models can account for how sampled experiences relate to choice, they don’t explain decisions about how to search (in particular, when to stop sampling information). We propose that both choice and search behavior in this context can be understood as a sequential sampling process whereby decision makers sequentially accumulate outcome information from each option to form a preference for one alternative over the other. We formalize this process in a new model, Choice from Accumulated Samples of Experience (CHASE). The model provides a good account of choice behavior and goes beyond existing models by explaining variations in sample size under different task conditions. This approach offers a process-level framework for understanding how interactions between the choice environment and properties of the decision maker give rise to decisions from experience.