Overview
Rational analysis, developed by @1990andersonAdaptive, is a paradigm for modeling human cognition that assumes that cognition is a rational adaptation to environmental structure and goals pursued within it. Rational process models of human behavior can then be derived from assumptions of the environment. This is similar to ecological explanations of animal behavior based on evolutionary fitness.
Resource-rational analysis, after @2020liederResource, extends this paradigm to account for limited computational resources using the principle of bounded optimality, a normative standard in which the cognitive process an agent executes performs as well or better than any other process the agent could perform given their constraints.
Rational process models help to connect computational and algorithmic levels of analysis.
Problems of indiscriminate confirmation
The classical inductive scientific method is epistemically flawed in an immature science like psychology, which lacks an established body of facts: confirmation of a hypothesis by an observation is a logical fallacy, affirming the consequent (); on the other hand, failure to confirm a hypothesis could be attributed to both the main hypothesis as well as any number of other auxiliary assumptions.
The resource-rational method is deductive: $$ \text{environment} \ \wedge \ \text{rationality assumption} \ \wedge \ \text{constraints} \ \implies \text{behavior}.
The environment consists of observables that are true by definition, and the rationality assumption is intended to be trivially true. Thus, in the case that observed behavior does not confirm a hypothesis, we know to modify our model of constraints. --- # Steps of rational vs. resource-rational analysis >[!example] Steps of rational analysis ([[@1990andersonAdaptive]]): > >1. Precisely specify what the goals of the cognitive system are. >2. Develop a formal model of the environment to which the system is adapted. >3. Make the minimal assumptions about computational limitations. >4. Derive the optimal behavioral function given (1) to (3). >5. Examine the empirical literature to see if the predictions of the behavioral function are confirmed. >6. If predictions are off, iterate. >[!example] Six steps of resource-rational analysis ([[@2020liederResource]]): > >1. Begin by giving an aspect of cognition with a [[Classical levels of cognitive theory|computational-level]] (i.e., [[Classical theories of mind|functional]]) description in terms of a problem and its solution. >2. Determine which class of [[Classical levels of cognitive theory|algorithms]] the mind’s computational architecture may be used to solve these this problem in a **satisficing** way, then determine the resource cost of these algorithms and the utility of more accurately approximating the correct solution. >3. Find the particular algorithm that optimally trades off resource cost and utility. >4. Evaluate the resulting **rational process model** against empirical data. >5. Refine computational-level theory from step (1) or assumptions about computational architecture and constraints (2) to make a new resource-rational model, then iterate until model’s predictions are sufficiently realistic.