> That is a matter of opinion. Another view is that perceived randomness in > any sphere, biological, physical or whatever, is merely a symptom of our > ignorance of, or inability to cope with, the underlying deterministic > details. Fascinating as this discussion is, the real question of interest to statisticians should be whether the random numbers are good enough for practical use in statistics. This is an important question because while, as Daniel Bebber states, Excel is not in itself designed to do complex statistics, it is an excellent tool for programming statisical techniques and simulating data, e.g. for use in stats training. My own experience suggests that the Excel random number generator is adequate for most practical purposes. However, I have no hard data to back this up - does anyone else have evidence on this point? Steve Langton ******************************************************************* Steve Langton E-mail: [log in to unmask] MAFF Central Science Lab Phone: 01 904 462521 Sand Hutton Fax: 01 904 462111 York, YO41 1LZ, UK. ******************************************************************* %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%