Table 1.

Results of behavioral model fitting

Participant	Gain			Loss
	α	λ	Log-likelihood	α	λ	Log-likelihood
1	0.12	0.02	−3.60	0.12	0.02	−5.75
2	0.42	0.74	−18.15	0.94	0.77	−12.15
3	0.53	0.74	−16.30	0.71	0.89	−16.96
4	0.74	0.03	−8.38	0.75	0.03	−9.12
5	0.13	0.02	−3.60	0.73	0.03	−3.60
6	0.33	0.02	−5.74	0.54	0.47	−9.39
7	0.12	0.02	−9.56	0.13	0.02	−5.74
8	0.12	0.02	−5.74	0.33	0.02	−5.74
9	0.12	0.02	−8.38	0.13	0.02	−9.56
10	0.17	0.85	−20.71	0.13	0.02	−7.28
11	0.56	0.81	−17.36	1.11	0.03	−0.00
12	0.33	0.03	−9.12	0.47	0.43	−9.38
13	0.33	0.50	−14.04	0.44	0.73	−17.34
14	0.34	0.03	−3.60	0.34	0.03	−3.60
15	0.74	0.03	−5.75	0.76	0.03	−5.74
16	0.21	0.02	−0.00	0.32	0.93	−23.41
17	0.13	1.11	−24.70	0.20	1.14	−26.02
18	0.34	0.03	−7.27	1.07	0.71	−9.39

The ‘envy’ parameter α reflects the degree to which an individual cares about inequality, and the ‘temperature’ parameter λ reflects decision randomness. Log-likelihoods show the goodness of fit (the closer to zero, the better the fit).