Fig. 6
From: The neural computation of inconsistent choice behavior
The neural random utility model with two alternatives. a Inconsistent binary choice: A draw (red) of utilities from a log normal distribution with mean v1 = 4 and v2 = 2, (log s.d. = 0.7). In this draw, \(\tilde v_2 > \tilde v_1\), so the inconsistent option (with lower mean valuation) is chosen. b Average random utility: A sample of 2,000,000 utilities from the distribution in a. The average utility is higher when the inconsistent option is chosen. c Larger difference: A sample of utilities from distributions with a larger difference in mean valuations (log normal, means v1 = 7 and v2 = 2, (log s.d. = 0.7). The average utility of an inconsistent choice is larger when the choice is more inconsistent (the difference in mean valuations is larger compared to b)
