p. 6, line 7 - "time-to-EVENT" ?
p. 7, line 14 - "time-to-EVENT".
p. 8, line 3 - I prefer "ESTIMATE field effects" to "PREDICT field effects".
p. 8, line 12 - you might consider mentioning R along with Splus.  R is
freeware and is popular among statistically trained researchers.
p. 16, line 22 - you might just say "assess the fit of the Cox regression
model."
p. 17, line 1 - you might just say the KM and the Cox model appear to
be in reasonable agreement in their estimation of survival probability,
except for slight differences for spot at 25 degrees C.
p. 23, line 23, (Figure 7) instead of (Fig 7).
Figure 7 is still not monotone.  It takes little jumps upward.