“…These include both the simple random walk models, where a decision maker has a discrete set of states (e.g., 11 confidence levels, 0/10/20/.../100) that they move through over time, shown in Figure 2, and the more common diffusion models where the "states" are a continuously-valued level of evidence (such as 0-100, including all numbers in between). Since the discrete-state random walks approach a diffusion process as the number of states gets very large, we group these two approaches together under the umbrella of Markov process, which have been used to model choices and response times (Emerson, 1970;Luce, 1986;Stone, 1960) as well as probability judgments (Edwards et al, 1963;Wald & Wolfowitz, 1949, 1948Kvam & Pleskac, 2016;Moran et al, 2015;Ratcliff & Starns, 2009;Yu et al, 2015) in domains such as memory (Ratcliff, 1978), categorization (Nosofsky & Palmeri, 1997), and inference (Pleskac & Busemeyer, 2010).…”