Heightened familiarity drives the negative retrieval bias in depression: Evidence from the PRISM task

Major Depressive Disorder (MDD) is associated with emotional memory deficits with substantial downstream consequences, but treatment is limited by a poor understanding of the upstream mechanisms that drive such behavior. Our previous work linked depression to a negative retrieval bias rooted in abnormal evidence accumulation (Cataldo et al., 2023). Computational modeling with the Drift Diffusion Model can account for such a bias in two ways: increased familiarity, in which depression strengthens evidence for all negative memories—even false ones; or motivated retrieval, in which depression increases the propensity to judge all evidence as “old”—even if it is weak. Thus, it is unclear whether depression affects the quality of negative memories or the way they are acted upon, limiting both basic and applied depression research. The current work distinguishes the familiarity vs. motivated retrieval accounts via the Parceling Recognition Into Strength and Motivation (PRISM) task, which isolates memory strength from decision processes by extending single-item recognition behavior to forced choices between targets and lures (Starns et al., 2018). In a sample of 53 community adults ranging in depressive severity, we found that the negative retrieval bias extended across both single-item and forced-choice recognition, thus supporting the false familiarity account. A qualitative analysis of participants’ self-reported strategies further indicated that increased schema use may be an important mechanism. In sum, the current results provide critical evidence that the negative retrieval bias results from disrupted memory representations, and point to schemas as a promising context for studying such representations.