Open AccessProfiling dynamic decision-makers
Author(s) -
Konrad Grabiszewski,
Alex R. Horenstein
Publication year - 2022
Publication title -
plos one
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.99
H-Index - 332
ISSN - 1932-6203
DOI - 10.1371/journal.pone.0266366
Subject(s) - profiling (computer programming) , predictive power , computer science , robustness (evolution) , construct (python library) , machine learning , data science , artificial intelligence , biology , philosophy , biochemistry , epistemology , gene , programming language , operating system
From CEOs confronting competition to children playing board games, our professional and personal lives are full of dynamic decisions. Naturally, while playing the role of a decision-maker, people differ. To comprehend and analyze how they differ, first it is necessary to construct a profiling method that classifies dynamic decision-makers. Developing such a method is the main objective of our article. We equate dynamic decision-making with backward inducting. We rely on response times to construct the profiles. Our method has both descriptive power and predictive power: a subject’s profile resembles her reasoning process and forecasts the likelihood of her correctly backward inducting. To test the proposed profiling method, we use data generated by 22 different finite dynamic scenarios from the mobile app Blues and Reds . Our sample consists of 35,826 observations from 6,463 subjects located in 141 countries. We construct the profiles of our subjects, and, in a variety of exercises supported by an array of robustness checks, we successfully establish the predictive power of our profiling method.