Have a personal or library account? Click to login
The Value of Non-Instrumental Information in Anxiety: Insights from a Resource-Rational Model of Planning Cover

The Value of Non-Instrumental Information in Anxiety: Insights from a Resource-Rational Model of Planning

Open Access
|Feb 2025

Figures & Tables

cpsy-9-1-124-g1.png
Figure 1

Task instructions and general findings.

A) Task instructions.

B) Example scenario.

C,D,E) Average willingness to pay as a function of outcome at 1 hour (C), probability (D), and cost to know now (E) for the N = 227 subjects across 30 scenarios each. Willingness to pay is the Likert response for each scenario, scored from –2 for ‘definitely not’ to +2 for ‘definitely yes.’

cpsy-9-1-124-g2.png
Figure 2

Willingness to pay as a function of somatic anxiety and negative affect.

A,B,C) For somatic anxiety, willingness to pay as a function of outcome at 1 hour (A), probability of outcome (B), and cost to know now (C).

D,E,F) For negative affect, willingness to pay as a function of outcome at 1 hour (D), probability of outcome (E), and cost to know now (F).

Solid lines model simulations. For visualization only, data are median split into high anxiety (bright colors) and low anxiety (dull colors).

cpsy-9-1-124-g3.png
Figure 3

Example calculation of the value of information.

A) Expected utilities of information (Uinfo), planning in uncertainty (Uplan), and living in denial (Uleisure) as a function of probability of the outcome. For this simulation, we used s = 1, Uoutcome = 1.5, Cplan = 0.8, Cinfo = 0.1, c = 1, and Uleisure = 0.25.

B) The value of information as a function of probability. Here, there is an inverted-U relationship between the value of information and the probability of the outcome.

cpsy-9-1-124-g4.png
Figure 4

Relationship between anxiety factor scores and cognitive model variables and parameters.

A,B,C) Standardized regression coefficients between anxiety factors and model variables. The model variables listed here are the value of information (A), the utility of information (B), and utility of planning (C).

D,E,F,G,H) Standardized regression coefficients between anxiety factors and model parameters. The model parameters listed here are the utility of outcome (D), cost of planning (E), cost of information (F), utility of leisure (G), and the outcome scaling, represented as the log of the ratio between gains and losses (H).

Error bars are 95% confidence intervals. * denotes p < 0.05.

DOI: https://doi.org/10.5334/cpsy.124 | Journal eISSN: 2379-6227
Language: English
Submitted on: May 29, 2024
Accepted on: Jan 19, 2025
Published on: Feb 12, 2025
Published by: Ubiquity Press
In partnership with: Paradigm Publishing Services
Publication frequency: 1 issue per year

© 2025 Bilal A. Bari, Samuel J. Gershman, published by Ubiquity Press
This work is licensed under the Creative Commons Attribution 4.0 License.