Table III.
95% Confidence intervals for 1000 resampling with replacement | |||||
---|---|---|---|---|---|
| |||||
Lower interval | Upper intervals | Regression coefficient |
Original OR estimate |
Critical ratio* | |
Sufficient saliva data | −0.052 | 11.77 | 1.78 | 2.88 | 1.78/2.83=0.63 |
| |||||
Missing tubes data | −0.012 | 11.81 | 1.87 | 2.52 | 1.87/3.05=0.61 |
Note: Logistic regression bootstrap CI intervals for the comparison of age group (i.e., AYA vs. C) on dependent variables (i.e., Sufficiency and Missing tubes) with C being the referent group.
Regression coefficient divided by standard deviation yields the critical ratio with absolute numbers < 1.96 equating to non-significance.