The choices we make not only influence how we remember the past but also how we will perceive the future. In both cases, our judgments are biased in favor of our preceding choice. Such choice-induced biases have been known for centuries. They affect our judgments in daily life including in situations where judgment errors are critical (e.g., medical diagnoses, scientific hypothesis testing, or political decisions). Why the human brain generates these biases and tends towards self-confirmation has long remained elusive. Supported by the Collaborative Research in Computational Neuroscience program, the proposed research will set out to identify the underlying computational and neural mechanisms of choice-induced biases in human decision-making. The proposed work employs a highly interdisciplinary approach that combines measures of behavior and brain activity with theory and computational modeling. It perfectly draws from the complimentary experimental and computational expertise of each PI and their laboratories. The goal is to develop and validate a novel theory of decision-making under uncertainty, in which choice-induced biases play a central role. Unraveling the source and function of choice-induced biases is not only critical for understanding the limits of human rationality but has also immediate and important implications for society at large. For example, the insights can help training of physicians to reduce judgment errors in clinical diagnostics, as well as that of other decision-makers in the legal and corporate sectors. Finally, the experimental methods and results of the proposed research may help to shed light on the biological underpinnings of important brain disorders, in particular schizophrenia, for which confirmation biases are aggravated.
The central focus of the proposed work is to test the hypothesis that choices generate subjective expectations that influence the subsequent evaluation process of both past and future evidence. The proposed research will record psychophysical and functional neuroimaging (magnetoencephalography (MEG)) measurements of healthy human subjects while they are performing a range of novel behavioral tasks specifically developed to assess choice-induced biases in low-level perception as well has high-level cognition. Theory and computational modeling will be crucial to interpret these signals and ultimately to help infer the underlying cognitive operations. The results of this research will not only advance our understanding of choice-induced bias effects in decision-making, but of decision-making in general. The combined approach of simultaneously using behavioral, neural, and theoretical/computational measurements and methods imposes strong constraints on the expected results and explanations while at the same time lending also extraordinarily strong confidence to the findings as they are supported on all levels. To the degree that they demonstrate how choices influence subsequent evidence evaluation, our results have the potential to drastically change the traditional understanding of decision-making as a simple feedforward accumulation-to-bound process.
A companion project is being funded by the Federal Ministry of Education and Research, Germany (BMBF).
This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.
