T-test

Options consider:
Confidence intervals
Effect Sizes

Mean 1: 8
Mean 2: 162
N1: 18198
N2: 18325

t (0.025) for 95% CI= 1.9608
declare p larger than alpha=0.05 not significant.

COMPARE TWO
BINOMIAL PROPORTIONS

mean1 eq: 0.0004 (sd= 0.021) (se= 0)
mean2 eq: 0.0088 (sd= 0.094) (se= 0.001)

Difference between means:
M1-M2=0.0004-0.0088=-0.0084
sd=0.0962; se=0.0007
95% CI of difference:
-0.0098 <-0.0084< -0.007 (Wald)
-0.0098 <-0.0084< -0.007 (CCWald)
-0.0099 <-0.0084< -0.0071 (N-W)
t-difference: -11.853
df-t: 18416.3; p= -0
(left p: 1; two sided: 0)

Effect size:
Odds Ratio: 0.0493; s.e: 0.0179
This seems a large effect size.
95% CI: 0.024 >0.0493> 0.1

Risk Ratio: 0.0497=p1/p2=0/0.01
95% CI: 0.024 <0.05< 0.101
Efficacy(VE): 95.03%
95% CI: 89.9% <95.03%< 97.6%
Phi/Cramer's V/R: 0.0617
95% CI: 0.051 <0.0617< 0.072
Yules-Q: -0.906; s.e.= 0.0325
95% CI: -0.97 <-0.906< -0.842
Yules-Y: -0.6366; s.e.= 0.0539
95% CI: -0.742 <-0.6366< -0.531

###### For help go to SISA

More:

Two by Two table analysis
Chi squares, Risk Ratios, Odds Ratios
and other table statistics

Fisher exact test

Calculate the minimum Sample
Size
required to see if the
difference between 8 and 162
is statistically significant

Calculate CI
around Mean1= 8.

Calculate CI
around Mean2= 162.