Censoring and edge effects in areal and line transect sampling of rock joint traces

This paper reviews techniques for correcting for bias incurred when incomplete observations form part of a data set. The similarities and differences between biases due to censoring and edge effects are emphasized, the latter being of importance when the distribution of joint traces is estimated from an exposure. A new technique for analyzing line transect data is presented, and a new analysis of published data on joint trace lengths given.