Accurate prediction of the timing of death in patients with a terminal illness is important for their care. The following graph compares the survival times of terminally ill cancer patients with the clinical prediction of their survival times (data from Glare et al. 2003).

a. Describe in words what features most of the frequency distributions of actual survival times have in common, based on the box plots for each group.

b. Describe the differences in shape of actual survival time distributions between those for 1 to 5 months predicted survival times and those for 6 to 24 months.

c. Describe the trend in median actual survival time with increasing predicted number.

d. The predicted survival times of terminally ill cancer patients tend to overestimate the medians of actual survival times. Are the means of actual survival times likely to be closer to, further from, or no different from the predicted times than the medians? Explain.

Transcribed Image Text:

Actual survival (months) 15 10- 5 0 I I I ||||||||||||| I I I I |||| 1 2 3 4 5 II 11 6 7-9 10-12 13-24 Clinical prediction (months)