In memory research, concepts of forgetting and the rate of forgetting are different. To estimate the rate of forgetting,w e need to construct a mathematical function at first, such as power function. Different rate of forgetting can be viewed as an evidence for different memory systems, but the problem for us is how to estimate such parameter. The same parameter estimated from a single-power function by least-squares regression often changes at different recall period in a whole range. Then, which recall period's parameters can be used to theory testing is a question. In light of nonlinear science, we put forward an intermediate asymptotics approach to capture the key characteristics of different memory. Using new method, we find that different types memory do have different rate of forgetting, which is inconsistent with McBride and Doshner (1999), but support multiple memory systems view (e.g., Schacter & Tulving, 1994).
展开▼