How to Find 90% or 95% conf Interval , and a suitable length L

linear_regression
r
data_science
statistics

#1

I am trying to implement an model that finds a 90% or 95% conf Interval , and a suitable length L ( ie number of measures in sample space) to narrow down the confidence interval to 1.3 units, in a bi-variate normal distribution independent forming a linear regression model ? Could you please provide some guidance and examples.