Modeling and Diagnostics of Biased Judgments

The paper presents a heuristic modeling and solution of cognitive biases for subjective understanding and misjudgment in an ambiguous and comparative situation on the example of Ellsberg's two-color paradox.It applies an original symptom-based context evaluation procedure for a clear and rational interpretation of situational awareness in the decision-making process.By sequentially and simultaneously adopting and using two overlapping types of cognitive process, additive and subtractive, judgmental unreliability is assessed for conflicting and violating contexts.The main idea to overcome these biases is to use the assumption of the dual involvement of the symptom as delay waves in the intuitive conscious processing of the information.To account for the causes of misjudgments in different erroneous actions one must dynamically track and identify the possible trajectories of the actual recognized context.Judgment in an unambiguous context is a monotonic process with decreasing information entropy and contextual probability of erroneous action, i.e. increasing probability of successful cognition as a sum of increasing discrete probabilistic amplitudes of contextual symptom recognition.But judgment in an ambiguous context is a wave-like alternating process with successively alternating decreasing-increasing (concave) information entropy (context probability) and alternating increasing-decreasing (convex) orderliness (cognitive probability) as the corresponding sum of varying discrete probabilistic amplitudes of all trajectories.These probabilistic amplitudes result from a combination of the shifts between numbers of existed objectively and imagined subjectively symptoms of context with their weights in a given situation.By distinguishing these at least two overlapping additive and subtractive types of cognition consisting of different numbers of stages, we can resolve some of the paradoxes of expected utility theory by predicting possible alternatives for context development, possibly quantifiable and comparable probability trajectories that are applicable to explaining known cognitive biases.