Semantic and cognitive tools to aid statistical science: replace confidence and significance by compatibility and surprise

Table 2 Reanalysis of the Brown et al. HDPS results [34]^a

Test Hypothesis (H)	P-value (compatibility)	S-value (bits of information)	Maximum-likelihood ratio	Likelihood-ratio statistic
Halving of hazard, HR = 0.5	1.6 × 10^− 6	19.3	1.0 × 10⁵	23.1
No association (null), HR = 1	0.0505	4.31	6.77	3.82
Point estimate, HR = 1.61	1.00	0.00	1.00	0.00
Doubling of hazard, HR = 2	0.373	1.42	1.49	0.79
Tripling of hazard, HR = 3	0.01	6.56	26.2	6.53
Quintupling of hazard, HR = 5	3.3 × 10⁻⁶	18.2	5.0 × 10⁴	21.7

^aComputed from the normal approximations given in the Appendix
P-values, S-values, maximum-likelihood ratios, and likelihood-ratio statistics for several test hypotheses about the hazard ratio (HR) computed from Brown et al. HDPS results [34].

ISSN: 1471-2288