Elsevier, Journal of Mathematical Psychology, (75), p. 86-95
DOI: 10.1016/j.jmp.2016.03.001
Full text: Unavailable
Recent years have seen an increased interest in models of recognition memory’s decision stage. However, only a relatively narrow set of candidate models has been considered thus far, with comparisons being typically restricted to signal detection and high-threshold models. Here, we consider a third alternative, Luce’s (1963) low threshold model (LTM). We evaluated the LTM’s predictions for existing Yes–No receiver-operating characteristic data (Dube et al., 2012) as well as data from K alternative ranking tasks (Kellen and Klauer, 2014). The LTM, which to this point has been largely ignored in the recognition memory literature, turns out to perform at least as well as the most popular model in this domain, the Gaussian signal detection model. These results suggest future work concerning the decision stage of recognition should consider the LTM in addition to the continuous and discrete-state models that have dominated the literature so far.