Maria Robinson¹ (), Timothy Brady²; ¹University of Warwick, ²University of California, San Diego

Can time-based memory effects (like recency) and strength-based effects (like repetition) be unified under a single explanatory framework? We addressed this question by testing whether the Target Competition Confusability (TCC) framework (e.g., Robinson & Brady, 2023) could predict performance in repetition conditions using model parameters estimated solely from temporal order effect conditions. We developed a summative model that conceptualizes repetition-based effects as the cumulative activation strengths elicited by individual items over time. The model treats repeated items as the sum of their individual presentation activations, using parameters that are independently estimated from a sequential task condition where items were probed in specific orders. For instance, in a six-item sequence with the first two items repeated, the model predicted performance for the repeated item by summing the activations from single-presentation trials that probed the first and second items in the sequence. Our results showed that this summative model successfully predicted the entire distribution of memory errors in repetition conditions with no free parameters used to fit these distributions. Furthermore, it did so across different numbers of repetitions and for different combinations of temporal orders. While prediction accuracy did not match test-retest reliability, the model captured both the shapes of error distributions and the ordinal effects of repetition-based manipulations. These findings demonstrate that a single modeling framework can unify seemingly distinct temporal order and repetition effects on memory, and we discuss how this elucidates the underlying principles that govern memory performance more broadly.