Table 1.

The differences in median and the 95th percentile range of Robinson–Foulds values between the Mk and both parsimony models are greater in the full dataset compared with the realistic homoplasy subsets. mk, Bayesian Mk model; ew, equal-weights parsimony; iw, implied weights parsimony and its attendant K values.

	100 characters	100 characters CI	350 characters	350 characters CI	1000 characters	1000 characters CI
mk	45 (29–64)	40.5 (28.2–62.5)	20 (10–51)	19.5 (10.2–57.3)	19.5 (10.2–57.3)	11 (5–27.8)
ew	61 (31–98)	53 (29–91.8)	27 (12–70)	28 (12–74.8)	28 (12–74.8)	16 (6.2–43.7)
iw k2	89 (39–119)	77 (38.2–117.7)	36 (18–76)	36 (17.2–81.3)	36 (17.2–81.3)	19.5 (10–35.7)
iw k3	76 (38–112)	69 (36.4–108)	32 (16–69)	34 (15.2–70)	34 (15.2–70)	18 (9.2–35.7)
iw k5	68 (36–104)	61 (32.2–102)	30 (14–66)	31.5 (15.2–68)	31.5 (15.2–68)	18 (9–34)
iw k10	63 (34–100)	55.5 (32–98)	28 (13–68)	30 (15.2–69.7)	30 (15.2–69.7)	16 (8–34)
iw k20	64 (34–100)	53 (33–97.8)	28 (14–68)	30 (13.2–71.7)	30 (13.2–71.7)	17 (8–39.3)
iw k200	65 (34–100)	55 (32.2–97.7)	28 (14–72)	30.5 (15–76)	30.5 (15–76)	18 (8–44)