marap
Computer
- May 15, 2015
- 1
Hi All,
I'm trying to reconcile a discrepancy between two sources of critical values of Cramer-von Mises (CVM) statistic for the CVM goodness-of-fit test. On the one hand, the table on page 229 in shows the values x such that P(omega <= x) = p for various values of p, where omega is the CVM statistic. On the other hand, the table lists the critical values by significance level, e.g., for alpha = 0.2, the entry x in the table satisfies P(omega > x) = 0.2. Why do the two tables disagree? For example, the entry for p=0.95 and n=20 in the first table is 0.45778 and the corresponding value in the second table (for alpha=0.05 and n=20) is 0.217. I would expect to see the same numbers here but they are different. Any ideas?
Thanks in advance!
Maxim
I'm trying to reconcile a discrepancy between two sources of critical values of Cramer-von Mises (CVM) statistic for the CVM goodness-of-fit test. On the one hand, the table on page 229 in shows the values x such that P(omega <= x) = p for various values of p, where omega is the CVM statistic. On the other hand, the table lists the critical values by significance level, e.g., for alpha = 0.2, the entry x in the table satisfies P(omega > x) = 0.2. Why do the two tables disagree? For example, the entry for p=0.95 and n=20 in the first table is 0.45778 and the corresponding value in the second table (for alpha=0.05 and n=20) is 0.217. I would expect to see the same numbers here but they are different. Any ideas?
Thanks in advance!
Maxim
