Abstract
In this paper the author tried to calculate sample sizes for clinical trials in lung cancer treatment results that the survival-time patterns of patients have lognormality using the Owen's noncentral t-distribution and given following tables.
The table la and lb are given the relation between the expected difference Δμ in logmean and the required sample sizes that should be included in each group of a clinical trial in lung cancer treatment results in oder to have a 90% (β=0.90), 70% (β=0.70) and 50% (β=0.50) probability detecting significant difference (!α=0.05, α=0.01) in results and covered any value of the difference ranging from 0.03 to 0.30 and any value of ranging from 0.30 to 0.50.
