Confidence intervals wider than R version

[jamalsenouci]

breaking out the wider confidence intervals from the rest of the issues raised in #7. This is potentially due to the R version placing an upper limit on the standard deviation