I was reading "Sample Size Calculations in Clinical Research", a book by Chow, Shao, and Wang. On p. 6 there is a table of the probability that a finding will be reproduced in a second trial. For example, the probability that a group difference with significance p<.05 will be reproduced is about .50, for p<.01 it is .73, and for p<.001 it is .90. I find this quite interesting but can't locate any further information about this type of calculation/statistic. Does anyone have any insights?