
Spacing 285
8.11 Theory for edge corrections
∗
The estimation o f summary f unctions is hampered by ed ge effects, arising when we c a nnot see
beyond the boundary of our study region. If no correction is made for edge effects, th en estimates
of the functions F, G, and K are negatively biased, with the magnitude of bias increasing with r.
Edge corrections are routinely used to avoid this problem, and are strongly recommended, unless
the data are huge (when th e inter e stin g values of r will be small, and the edge effect bias also small).
This section gives essential details about edge corrections for the distan c e distributions F and G.
Readers