Table 5. Donation and π: Nonparametric tests (MTurk).

This table shows that, along most thresholds for π, Mann-Whitney tests yield significant differences in the distribution of donations as π increases.

	Wilcoxon-Mann-Whitney two-sided test (P values)
	(1)	(2)	(3)
Thresholds	Κ-Unknown or 10¢	Κ = 0¢ or 50¢	Κ-Pooled
π = 1 versus π ≤ 0.67	0.91	0.05	0.11
π ≥ 0.67 versus π ≤ 0.33	0.07	1.00	0.20
π ≥ 0.33 versus π ≤ 0.05	0.05	0.10	0.01
π ≥ 0.05 versus π = 0.01	0.05	0.02	0.01
		π-Pooled
K ≥ 10¢ versus K = 0¢		0.040
K = 50¢ versus K ≤ 10¢		0.11