Dear all, I try to test behavioural data durations whether they follow an exponential distribution. My data have the particularity that there are no values below a certain threshold, and the only test adapted to this case I could find is the Shapiro-Wilk test for exponentiality (Shapiro & Wilk 1972, Technometrics 14, 355-370). But it requires to compute Shapiro and Wilks critical W values. Is there any R function that lets me compute these values? I checked out the shapiro.test source which seems to compute these values in swilk.c, but I could not find an R function wrapper that would compute either these critical values for a given alpha and n (number of values) or a function that comptues directly the p-value for a given value of the W statistic and n. Any idea where I could find that? Or are there specific packages that let me test for exponentiality (tried package "survival", but my stat level seems to be too low to understand application of this package beyond simple survival curve visualisation). Thanks for any help, Christian.