Rotor-Gene Q MDx CE User Manual 02/2022
202
For estimation of standards, a tighter confidence interval can be obtained:
The implication of this formula is that adding replicates to an standard individual concentration
reduces the width of the interval for all estimates, as n is increased. Adding a large number of
replicates to an unknown reduces its uncertainty to that of a single standard. The extra replicates
reduce the uncertainty due to the unknown not forming part of the linear model.
15.1.2
Confidence intervals for CT values
We assume that error in replicate CT values is linear and normally distributed.
We therefore use the One-Sample t Confidence Interval. Let µ be the mean value for a replicate's
CT values
. Then, a 100(1–
α
)% confidence interval for a CT value µ is:
We would like to thank Peter Cook from the Mathematics Department of the University of NSW,
Sydney, Australia, whose help was invaluable in verifying the mathematical approaches used.