Hi - I did a linear regression on log_turbidity (explanatory variable) vs log_ecoli (response variable) and am now trying to compute, and more importantly, understand intuitively, what my 95% confidence interval is in arithmetic space. Specifically, I need error bars in arithmetic space to go along with my back transformed predicted ecoli values, but I'm not sure how to do and interpret this step. (Note that my log-residuals are normally distributed.) Thanks in advance!
