“…Similar with the previously published study on κσ 2 (C 4 /C 2 ) [25], we randomly divide the total number of events into 450, 225, 90, 45, 30, 15, 9, and 5 sub-samples. Correspondingly, the number of events for each kind of sub-sample are 1M, 2M, 5M, 10M, 15M, 30M, 50M, and 90M, respectively.…”