Skip to main content
. 2016 Aug 29;113(35):9751–9756. doi: 10.1073/pnas.1602504113

Table S4.

Fisher’s exact tests for multiple analyses using the complete (nonoptimized) subset

Data subset Test EWS Pass Fail CI low CI high P
Tau for all data (τo) Sampling (Ps) AR (1) 2 7 0.13 201.64 0.55
σ 5 4 0.94 703.78 0.05**
1γ 5 4 0.94 703.78 0.05**
Null model (Pn) AR (1) 0 9 0.00 51.96 1.00
σ 3 6 0.32 313.65 0.27
1γ 1 8 0.02 117.49 1.00

Significance level: **P = 0.10. CI, confidence interval.

Pass/fail at P < 0.1.

Expected ratio used in Fisher test is 1:8, pass:fail.